Physics 2000

E. R. Huggins
Dartmouth College

physics2000.com
MKS Units (link to CGS Units)
m = meters kg = kilograms s = seconds
N = newtons J = joules C = coulombs
T = tesla F = farads H = henrys Powers of 10
A = amperes K = kelvins mol = mole
Power Prefix Symbol
speed of light c 3.00 × 10 8 m / s
10 12 tera T
gravitational constant G 6.67 × 10 –11N⋅m2 / kg 2
10 9 giga G
permittivity constant ε0 8.85 × 10 – 12F / m
10 6 mega M
permeability constant µ0 1.26 × 10 –6
H/m
10 3 kilo k
elementary charge e 1.60 × 10 –19
C 10 2 hecto h
electron volt eV 1.60 × 10 –19
J 10 – 1 deci d
electron rest mass me 9.11 × 10 – 31
kg 10 – 2 centi c
proton rest mass mp 1.67 × 10 – 27kg 10 – 3 milli m
Planck constant h 6.63 × 10 – 34
J⋅ s 10 –6
micro µ
Planck constant / 2 π h 1.06 × 10 – 34 J⋅ s 10 – 9 nano n
Bohr radius rb 5.29 × 10 – 11m 10 – 12 pico p
10 – 15 femto f
Bohr magneton µb 9.27 × 10 – 24J / T
Boltzmann constant k 1.38 × 10 –23J / K
Avogadro constant NA 6.02 × 10 23mol – 1
universal gas constant R 8.31 J /mol⋅ K

Dimensions
Quantity Unit Equivalents
2
Force newton N J/m kg • m/ s
2 2
Energy joule J N• m kg • m /s
2 3
Power watt W J/s kg • m /s
Pressure pascal Pa N/m 2 kg/m• s2
Frequency hertz Hz cycle/s s–1
Electric charge coulomb C A•s
Electric potential volt V J/C kg • m2/A • s3
Electric resistance ohm Ω V/A kg • m2/A2• s3
Capacitance farad F C/V A2• s4/kg • m2
Magnetic field tesla T N •s/C • m kg/A • s2
Magnetic flux weber Wb T• m2 kg • m2/A• s2
Inductance henry H V• s/A kg • m2/A2• s2

Copyright © 2000 Moose Mountain Digital Press

Etna, New Hampshire 03750
All rights reserved
 Preface & TOC-i

Physics2000

Student project by Bob Piela

explaining the hydrogen
molecule ion.

by E. R. Huggins
Department of Physics
Dartmouth College
Hanover, New Hampshire

i
 Preface & TOC-iii

Preface

ABOUT THE COURSE Most students experience difficulty when they first
Physics2000 is a calculus based, college level introduc- encounter abstract concepts like vector fields and Gauss’
tory physics course that is designed to include twentieth law. To provide a familiar model for a vector field, we
century physics throughout. This is made possible by begin the section on electricity and magnetism with a
introducing Einstein’s special theory of relativity in the chapter on fluid dynamics. It is easy to visualize the
first chapter. This way, students start off with a modern velocity field of a fluid, and Gauss’ law is simply the
picture of how space and time behave, and are prepared statement that the fluid is incompressible. We then show
to approach topics such as mass and energy from a that the electric field has mathematical properties simi-
modern point of view. lar to those of the velocity field.

The course, which was developed during 30 plus years The format of the standard calculus based introductory
working with premedical students, makes very gentle physics text is to put a chapter on special relativity
assumptions about the student’s mathematical back- following Maxwell’s equations, and then put modern
ground. All the calculus needed for studying Phys- physics after that, usually in an extended edition. This
ics2000 is contained in a supplementary chapter which format suggests that the mathematics required to under-
is the first chapter of a physics based calculus text. We stand special relativity may be even more difficult than
can cover all the necessary calculus in one reasonable the integral-differential equations encountered in
length chapter because the concepts are introduced in Maxwell’s theory. Such fears are enhanced by the
the physics text and the calculus text only needs to strangeness of the concepts in special relativity, and are
handle the formalism. (The remaining chapters of the driven home by the fact that relativity appears at the end
calculus text introduce the mathematical tools and con- of the course where there is no time to comprehend it.
cepts used in advanced introductory courses for physics This format is a disaster.
and engineering majors. These chapters will appear on Special relativity does involve strange ideas, but the
a later version of the Physics2000 CD, hopefully next mathematics required is only the Pythagorean theorem.
year.) By placing relativity at the beginning of the course you
In the physics text, the concepts of velocity and accelera- let the students know that the mathematics is not diffi-
tion are introduced through the use of strobe photo- cult, and that there will be plenty of time to become
graphs in Chapter 3. How these definitions can be used familiar with the strange ideas. By the time students
to predict motion is discussed in Chapter 4 on calculus have gone through Maxwell’s equations in Physics2000,
and Chapter 5 on the use of the computer. they are thoroughly familiar with special relativity, and
are well prepared to study the particle-wave nature of
Students themselves have made major contributions to matter and the foundations of quantum mechanics. This
the organization and content of the text. Student’s material is not in an extended edition because there is of
enthusiasm for the use of Fourier analysis to study time to cover it in a comfortably paced course.
musical instruments led to the development of the
MacScope™ program. The program makes it easy to
use Fourier analysis to study such topics as the normal
modes of a coupled aircart system and how the energy-
time form of the uncertainty principle arises from the
particle-wave nature of matter.

iii
Preface & TOC-iv

ABOUT THE PHYSICS2000 CD ABOUT THE AUTHOR

The Physics2000 CD contains the complete Physics2000 E. R. Huggins has taught physics at Dartmouth College
text in Acrobat™ form along with a supplementary since 1961. He was an undergraduate at MIT and got his
chapter covering all the calculus needed for the text. Ph.D. at Caltech. His Ph.D. thesis under Richard
Included on the CD is a motion picture on the time Feynman was on aspects of the quantum theory of
dilation of the Muon lifetime, and short movie segments gravity and the non uniqueness of energy momentum
of various physics demonstrations. Also a short cook- tensors. Since then most of his research has been on
book on several basic dishes of Caribbean cooking. The superfluid dynamics and the development of new teach-
CD is available at the web site ing tools like the student built electron gun and
MacScope™. He wrote the non calculus introductory
www.physics2000.com physics text Physics1 in 1968 and the computer based
text Graphical Mechanics in 1973. The Physics2000
The cost is $10.00 postpaid.
text, which summarizes over thirty years of experiment-
Also available is a black and white printed copy of the ing with ways to teach physics, was written and class
text, including the calculus chapter and the CD, at a cost tested over the period from 1990 to 1998. All the work
of $ 39 plus shipping. of producing the text was done by the author, and his
wife, Anne Huggins. The text layout and design was
The supplementary calculus chapter is the first chapter done by the author’s daughter Cleo Huggins who de-
of a physics based calculus text which will appear on a signed eWorld™ for Apple Computer and the Sonata™
later edition of the Physics2000 CD. As the chapters are music font for Adobe Systems.
ready, they will be made available on the web site.
The author’s eMail address is
Use of the Text Material [email protected]
Because we are trying to change the way physics is
taught, Chapter 1 on special relativity, although copy- The author is glad to receive any comments.
righted, may be used freely (except for the copyrighted
photograph of Andromeda and frame of the muon film).
All chapters may be printed and distributed to a class on
a non profit basis.

iv
 Preface & TOC-i

Table of Contents
PART 1
CHAPTER 1 PRINCIPLE OF RELATIVITY
Front Cover
MKS Units ............................................... Front cover-2 The Principle of Relativity ............................................. 1-2
Dimensions ............................................. Front cover-2 A Thought Experiment ........................................... 1-3
Powers of 10 ........................................... Front cover-2 Statement of the Principle of Relativity .................... 1-4
Basic Law of Physics ............................................. 1-4
Wave Motion ............................................................... 1-6
Preface Measurement of the Speed of Waves .................... 1-7
About the Course ........................................................... iii Michaelson-Morley Experiment ............................ 1-11
About the Physics2000 CD ............................................. iv Einstein’s Principle of Relativity .................................. 1-12
Use of the Text Material ............................................ iv The Special Theory of Relativity ........................... 1-13
About the Author ............................................................ iv Moving Clocks ..................................................... 1-13
Other Clocks ........................................................ 1-18
INTRODUCTION—AN OVERVIEW OF PHYSICS Real Clocks .......................................................... 1-20
Space And Time ......................................................... int-2 Time Dilation ........................................................ 1-22
The Expanding Universe ....................................... int-3 Space Travel ........................................................ 1-22
The Lorentz Contraction ....................................... 1-24
Structure of Matter ...................................................... int-5
Relativistic Calculations ....................................... 1-28
Atoms ................................................................... int-5
Approximation Formulas ...................................... 1-30
Light ..................................................................... int-7
Photons ................................................................. int-8 A Consistent Theory .................................................. 1-32
The Bohr Model .................................................... int-8 Lack of Simultaneity .................................................. 1-32
Particle-Wave Nature of Matter ................................. int-10 Causality ................................................................... 1-36
Conservation of Energy ............................................ int-11 Appendix A ............................................................... 1-39
Anti-Matter ................................................................ int-12 Class Handout ..................................................... 1-39
Particle Nature of Forces .......................................... int-13
Renormalization .................................................. int-14 CHAPTER 2 VECTORS
Gravity ................................................................ int-15 Vectors ........................................................................ 2-2
A Summary .............................................................. int-16 Displacement Vectors ............................................ 2-2
The Nucleus ............................................................. int-17 Arithmetic of Vectors .............................................. 2-3
Stellar Evolution ........................................................ int-19 Rules for Number Arithmetic .................................. 2-4
The Weak Interaction .......................................... int-20 Rules for Vector Arithmetic ..................................... 2-4
Leptons ............................................................... int-21 Multiplication of a Vector by a Number .................. 2-5
Nuclear Structure ................................................ int-22 Magnitude of a Vector ............................................ 2-6
Vector Equations .................................................... 2-6
A Confusing Picture .................................................. int-22
Graphical Work ...................................................... 2-6
Quarks ..................................................................... int-24
Components ................................................................ 2-8
The Electroweak Theory ........................................... int-26 Vector Equations in Component Form .................. 2-10
The Early Universe ................................................... int-27 Vector Multiplication .................................................. 2-11
The Thermal Photons .......................................... int-29 The Scalar or Dot Product .................................... 2-12
Interpretation of the Dot Product .......................... 2-14
Vector Cross Product ........................................... 2-15
Magnitude of the Cross Product .......................... 2-17
Component Formula for the Cross Product .......... 2-17
Right Handed Coordinate System ............................. 2-18

i
Preface & TOC-ii

CHAPTER 3 DESCRIPTION OF MOTION CHAPTER 5 COMPUTER PREDICTION OF

Displacement Vectors ................................................. 3-5 MOTION
A Coordinate System ............................................. 3-7 Step-By-Step Calculations ........................................... 5-1
Computer Calculations ................................................ 5-2
Manipulation of Vectors .......................................... 3-8 Calculating and Plotting a Circle ............................ 5-2
Measuring the Length of a Vector .......................... 3-9 Program for Calculation ............................................... 5-4
Coordinate System and Coordinate Vectors ........ 3-11 The DO LOOP ........................................................ 5-4
Analysis of Strobe Photographs ................................ 3-11 The LET Statement ................................................. 5-5
Velocity ................................................................ 3-11 Variable Names ..................................................... 5-6
Acceleration ......................................................... 3-13 Multiplication .......................................................... 5-6
Determining Acceleration Plotting a Point ....................................................... 5-6
from a Strobe Photograph .................................... 3-15 Comment Lines ...................................................... 5-7
The Acceleration Vector ....................................... 3-15 Plotting Window ..................................................... 5-7
Projectile Motion ........................................................ 3-16 Practice ................................................................. 5-8
Uniform Circular Motion ............................................. 3-17 Selected Printing (MOD Command) ..................... 5-10
Magnitude of the Acceleration for Circular Motion 3-18 Prediction of Motion ................................................... 5-12
An Intuitive Discussion of Acceleration ...................... 3-20 Time Step and Initial Conditions ................................ 5-14
Acceleration Due to Gravity ................................. 3-21 An English Program for Projectile Motion ................... 5-16
Projectile Motion with Air Resistance .................... 3-22 A BASIC Program for Projectile Motion ...................... 5-18
Instantaneous Velocity .............................................. 3-24 Projectile Motion with Air Resistance ......................... 5-22
Instantaneous Velocity from a Strobe Photograph 3-26 Air Resistance Program ....................................... 5-24

CHAPTER 4 CALCULUS IN PHYSICS CHAPTER 6 MASS

Limiting Process .......................................................... 4-1 Definition of Mass ........................................................ 6-2
The Uncertainty Principle ....................................... 4-1 Recoil Experiments ................................................ 6-2
Calculus Definition of Velocity...................................... 4-3 Properties of Mass ................................................. 6-3
Acceleration ................................................................ 4-5 Standard Mass ...................................................... 6-3
Components .......................................................... 4-6 Addition of Mass .................................................... 6-4
Distance, Velocity and A Simpler Way to Measure Mass ........................... 6-4
Acceleration versus Time Graphs .......................... 4-7 Inertial and Gravitational Mass ............................... 6-5
The Constant Acceleration Formulas ........................... 4-9 Mass of a Moving Object ....................................... 6-5
Three Dimensions ................................................ 4-11 Relativistic Mass .......................................................... 6-6
Projectile Motion with Air Resistance ......................... 4-12 Beta (β ) Decay ....................................................... 6-6
Differential Equations ................................................ 4-14 Electron Mass in β Decay ...................................... 6-7
Solving the Differential Equation ........................... 4-14 Plutonium 246 ........................................................ 6-8
Solving Projectile Motion Problems ............................ 4-16 Protactinium 236 .................................................... 6-9
Checking Units .................................................... 4-19 The Einstein Mass Formula ........................................ 6-10
Nature’s Speed Limit ............................................ 6-11
Zero Rest Mass Particles ........................................... 6-11
Neutrinos ................................................................... 6-13
Solar Neutrinos .................................................... 6-13
Neutrino Astronomy ............................................. 6-14

ii
 Preface & TOC-iii

CHAPTER 7 CONSERVATION OF LINEAR & CHAPTER 9 APPLICATIONS OF NEWTON’S

ANGULAR MOMENTUM SECOND LAW
Conservation of Linear Momentum ............................. 7- 2 Addition of Forces ....................................................... 9-2
Collision Experiments ................................................. 7- 4 Spring Forces .............................................................. 9-3
Subatomic Collisions ............................................. 7- 7 The Spring Pendulum ............................................ 9-4
Example 1 Rifle and Bullet .................................... 7- 7 Computer Analysis of the Ball Spring Pendulum .... 9-8
Example 2 ............................................................ 7- 8 The Inclined Plane ..................................................... 9-10
Conservation of Angular Momentum .......................... 7- 9 Friction ...................................................................... 9-12
A More General Definition of Angular Momentum ..... 7- 12 Inclined Plane with Friction ................................... 9-12
Angular Momentum as a Vector ............................... 7- 14 Coefficient of Friction ........................................... 9-13
Formation of Planets ........................................... 7- 17 String Forces ............................................................. 9-15
The Atwood’s Machine .............................................. 9-16
CHAPTER 8 NEWTONIAN MECHANICS The Conical Pendulum .............................................. 9-18
Force ........................................................................... 8-2 Appendix: The ball spring Program ........................... 9-20
The Role of Mass ......................................................... 8-3
Newton’s Second Law ................................................. 8-4 CHAPTER 10 ENERGY
Newton’s Law of Gravity .............................................. 8-5 ` ................................................................................. 10-1
Big Objects ............................................................ 8-5 Conservation of Energy ............................................. 10-2
Galileo’s Observation ............................................. 8-6 Mass Energy ............................................................. 10-3
The Cavendish Experiment ......................................... 8-7 Ergs and Joules ................................................... 10-4
"Weighing” the Earth .............................................. 8-8 Kinetic Energy ........................................................... 10-5
Inertial and Gravitational Mass ............................... 8-8 Example 1 ............................................................ 10-5
Satellite Motion ............................................................ 8-8 Slowly Moving Particles ........................................ 10-6
Other Satellites ..................................................... 8-10 Gravitational Potential Energy .................................... 10-8
Weight ................................................................. 8-11 Example 2 .......................................................... 10-10
Earth Tides ........................................................... 8-12 Example 3 .......................................................... 10-11
Planetary Units ..................................................... 8-14 Work ........................................................................ 10-12
Table 1 Planetary Units ....................................... 8-14 The Dot Product ................................................. 10-13
Computer Prediction of Satellite Orbits ...................... 8-16 Work and Potential Energy ................................. 10-14
New Calculational Loop ....................................... 8-17 Non-Constant Forces ......................................... 10-14
Unit Vectors ......................................................... 8-18 Potential Energy Stored in a Spring .................... 10-16
Calculational Loop for Satellite Motion ................. 8-19 Work Energy Theorem ............................................. 10-18
Summary ............................................................. 8-20 Several Forces ................................................... 10-19
Working Orbit Program ........................................ 8-20 Conservation of Energy ...................................... 10-20
Projectile Motion Program .................................... 8-21 Conservative and Non-Conservative Forces ...... 10-21
Orbit-1 Program .................................................. 8-21
Gravitational Potential Energy on a Large Scale ...... 10-22
Satellite Motion Laboratory ................................... 8-23
Zero of Potential Energy ..................................... 10-22
Kepler's Laws ............................................................ 8-24 Gravitational PotentialEnergy in a Room ............ 10-25
Kepler's First Law ................................................. 8-26
Satellite Motion and Total Energy ............................ 10-26
Kepler's Second Law ........................................... 8-27
Example 4 Escape Velocity .............................. 10-28
Kepler's Third Law ............................................... 8-28
Black Holes ............................................................. 10-29
Modified Gravity and General Relativity ..................... 8-29
A Practical System of Units ................................ 10-31
Conservation of Angular Momentum ......................... 8-32
Conservation of Energy ............................................. 8-35

iii
Preface & TOC-iv

CHAPTER 11 SYSTEMS OF PARTICLES CHAPTER 13 EQUILIBRIUM

Center of Mass .......................................................... 11-2 Equations for equilibrium ........................................... 13-2
Center of Mass Formula ....................................... 11-3 Example 1 Balancing Weights ............................ 13-2
Dynamics of the Center of Mass .......................... 11-4 Gravitational Force acting at the Center of Mass ....... 13-4
Newton’s Third Law ................................................... 11-6 Technique of Solving Equilibrium Problems ............... 13-5
Conservation of Linear Momentum ............................ 11-7 Example 3 Wheel and Curb ................................ 13-5
Momentum Version of Newton’s Second Law ...... 11-8 Example 4 Rod in a Frictionless Bowl .................. 13-7
Collisions ................................................................... 11-9 Example 5 A Bridge Problem .............................. 13-9
Impulse ................................................................ 11-9 Lifting Weights and Muscle Injuries ......................... 13-11
Calibration of the Force Detector ....................... 11-10
The Impulse Measurement ................................. 11-11 CHAPTER 14 OSCILLATIONS AND RESONANCE
Change in Momentum ........................................ 11-12
Oscillatory Motion ...................................................... 14-2
Momentum Conservation during Collisions ........ 11-13
Collisions and Energy Loss ................................ 11-14 The Sine Wave .......................................................... 14-3
Collisions that Conserve Momentum and Energy 11-16 Phase of an Oscillation ......................................... 14-6
Elastic Collisions ................................................ 11-17 Mass on a Spring;Analytic Solution ........................... 14-7
Discovery of the Atomic Nucleus ............................. 11-19 Conservation of Energy ...................................... 14-11
Neutrinos ................................................................. 11-20 The Harmonic Oscillator .......................................... 14-12
Neutrino Astronomy ........................................... 11-21 The Torsion Pendulum ....................................... 14-12
The Simple Pendulum ........................................ 14-15
Small Oscillations ............................................... 14-16
CHAPTER 12 ROTATIONAL MOTION Simple and Conical Pendulums ......................... 14-17
Radian Measure ........................................................ 12-2 Non Linear Restoring Forces ................................... 14-19
Angular Velocity ................................................... 12-2 Molecular Forces ..................................................... 14-20
Angular Acceleration ........................................... 12-3
Damped Harmonic Motion ...................................... 14-21
Angular Analogy .................................................. 12-3
Critical Damping ................................................ 14-23
Tangential Distance, Velocity and Acceleration ... 12-4
Radial Acceleration .............................................. 12-5 Resonance .............................................................. 14-24
Bicycle Wheel ...................................................... 12-5 Resonance Phenomena ..................................... 14-26
Transients .......................................................... 14-27
Angular Momentum ................................................... 12-6
Angular Momentum of a Bicycle Wheel ............... 12-6 Appendix 14–1 Solution of the Differential Equation
Angular Velocity as a Vector ................................ 12-7 for Forced Harmonic Motion .................................. 14-28
Angular Momentum as a Vector ........................... 12-7 Appendix 14-2 Computer analysis
Angular Mass or Moment of Inertia ............................ 12-7 of oscillatory motion ............................................... 14-30
Calculating Moments of Inertia ............................. 12-8 English Program ................................................ 14-31
The BASIC Program ........................................... 14-32
Vector Cross Product ................................................ 12-9
Damped Harmonic Motion ................................. 14-34
Right Hand Rule for Cross Products .................. 12-10
Cross Product Definition of Angular Momentum ...... 12-11
CHAPTER 15 ONE DIMENSIONAL WAVE MOTION
The r × p Definition of Angular Momentum ...... 12-12
Wave Pulses ............................................................. 15-3
Angular Analogy to Newton’s Second Law .............. 12-14
Speed of a Wave Pulse ............................................. 15-4
About Torque .......................................................... 12-15
Dimensional Analysis ................................................ 15-6
Conservation of Angular Momentum ....................... 12-16
Speed of Sound Waves ............................................. 15-8
Gyroscopes ............................................................. 12-18
Linear and nonlinear Wave Motion .......................... 15-10
Start-up .............................................................. 12-18
Precession ......................................................... 12-19 The Principle of Superposition ................................. 15-11
Rotational Kinetic Energy ........................................ 12-22 Sinusoidal Waves .................................................... 15-12
Wavelength, Period, and Frequency .................. 15-13
Combined Translation and Rotation ........................ 12-24
Angular Frequency ω ....................................... 15-14
Example—Objects Rolling
Spacial Frequency k .......................................... 15-14
Down an Inclined Plane ..................................... 12-25
Traveling Wave Formula .................................... 15-16
Proof of the Kinetic Energy Theorem ....................... 12-26 Phase and Amplitude ......................................... 15-17
Standing Waves ...................................................... 15-18
Waves on a Guitar String ......................................... 15-20
Frequency of Guitar String Waves ...................... 15-21
Sound Produced by a Guitar String ................... 15-22

iv
 Preface & TOC-v

CHAPTER 16 FOURIER ANALYSIS, CHAPTER 18 ENTROPY

NORMAL MODES AND SOUND Introduction ............................................................... 18-2
Harmonic Series ........................................................ 16-3 Work Done by an Expanding Gas ............................. 18-5
Normal Modes of Oscillation ...................................... 16-4 Specific Heats CV and Cp ........................................ 18-6
Fourier Analysis ......................................................... 16-6 Isothermal Expansion and PV Diagrams .................... 18-8
Analysis of a Sine Wave ....................................... 16-7 Isothermal Compression ...................................... 18-9
Analysis of a Square Wave .................................. 16-9 Isothermal Expansion of an Ideal Gas .................. 18-9
Repeated Wave Forms ...................................... 16-11
Adiabatic Expansion ................................................. 18-9
Analysis of the Coupled Air Cart System ................. 16-12
The Carnot Cycle .................................................... 18-11
The Human Ear ....................................................... 16-15 Thermal Efficiency of the Carnot Cycle .............. 18-12
Stringed Instruments ............................................... 16-18 Reversible Engines ............................................ 18-13
Wind Instruments .................................................... 16-20 Energy Flow Diagrams ............................................ 18-15
Percussion Instruments ........................................... 16-22 Maximally Efficient Engines ................................ 18-15
Sound Intensity ........................................................ 16-24 Reversibility ....................................................... 18-17
Bells and Decibels ............................................. 16-24 Applications of the Second Law .............................. 18-17
Sound Meters .................................................... 16-26 Electric Cars ...................................................... 18-19
Speaker Curves ................................................. 16-27 The Heat Pump .................................................. 18-19
Appendix A: Fourier Analysis Lecture ...................... 16-28 The Internal Combustion Engine ........................ 18-21
Square Wave ..................................................... 16-28 Entropy .................................................................... 18-22
Calculating Fourier Coefficients ......................... 16-28 The Direction of Time ......................................... 18-25
Amplitude and Phase ......................................... 16-31 Appendix: Calculation of the
Amplitude and Intensity ..................................... 16-33 Efficiency of a Carnot Cycle .................................. 18-26
Appendix B: Inside the Cochlea .............................. 16-34 Isothermal Expansion ......................................... 18-26
Adiabatic Expansion .......................................... 18-26
CHAPTER 17 ATOMS, MOLECULES AND The Carnot Cycle ............................................... 18-28
ATOMIC PROCESSES
Molecules .................................................................. 17-2 CHAPTER 19 THE ELECTRIC INTERACTION
Atomic Processes ..................................................... 17-4 The Four Basic Interactions ....................................... 19-1
Thermal Motion .......................................................... 17-6 Atomic Structure ........................................................ 19-3
Thermal Equilibrium ................................................... 17-8 Isotopes ............................................................... 19-6
Temperature .............................................................. 17-9 The Electric Force Law .............................................. 19-7
Absolute Zero ...................................................... 17-9 Strength of the Electric Interaction ....................... 19-8
Temperature Scales ........................................... 17-10 Electric Charge ......................................................... 19-8
Molecular Forces ..................................................... 17-12 Positive and Negative Charge ............................ 19-10
Evaporation ........................................................ 17-14 Addition of Charge ............................................. 19-10
Pressure .................................................................. 17-16 Conservation of Charge .......................................... 19-13
Stellar Evolution ................................................. 17-17 Stability of Matter ............................................... 19-14
The Ideal Gas Law .................................................. 17-18 Quantization of Electric Charge .......................... 19-14
Ideal Gas Thermometer ..................................... 17-20 Molecular Forces ..................................................... 19-15
The Mercury Barometer Hydrogen Molecule ........................................... 19-16
and Pressure Measurements ............................ 17-22 Molecular Forces—A More Quantitative Look .... 19-18
Avogadro’s Law ...................................................... 17-24 The Bonding Region .......................................... 19-19
Electron Binding Energy .................................... 19-20
Heat Capacity ......................................................... 17-26
Electron Volt as a Unit of Energy ........................ 19-21
Specific Heat ..................................................... 17-26
Electron Energy in the Hydrogen Molecule Ion .. 19-21
Molar Heat Capacity .......................................... 17-26
Molar Specific Heat of Helium Gas .................... 17-27
Other Gases ...................................................... 17-27 CHAPTER 20 NUCLEAR MATTER
Equipartition of Energy ............................................ 17-28 Nuclear Force ........................................................... 20-2
Real Molecules .................................................. 17-30 Range of the Nuclear Force ................................. 20-3
Failure of Classical Physics ..................................... 17-31 Nuclear Fission .......................................................... 20-3
Freezing Out of Degrees of Freedom ................. 17-32 Neutrons and the Weak Interaction ........................... 20-6
Thermal Expansion .................................................. 17-33 Nuclear Structure ...................................................... 20-7
Osmotic Pressure .................................................... 17-34 α (Alpha) Particles .............................................. 20-8
Elasticity of Rubber ................................................. 17-35
Nuclear Binding Energies .......................................... 20-9
A Model of Rubber ............................................. 17-36
Nuclear Fusion ........................................................ 20-12
Stellar Evolution ....................................................... 20-13
Neutron Stars .......................................................... 20-17
Neutron Stars
and Black Holes ...................................................... 20-18
v
PART 2

CHAPTER 23 FLUID DYNAMICS CHAPTER 25 FIELD PLOTS AND

The Current State of Fluid Dynamics .................... 23-1 ELECTRIC POTENTIAL
The Velocity Field ...................................................... 23-2 The Contour Map .......................................................... 25-1
The Vector Field ................................................... 23-3 Equipotential Lines ........................................................ 25-3
Streamlines .......................................................... 23-4 Negative and Positive Potential Energy ................... 25-4
Continuity Equation .............................................. 23-5 Electric Potential of a Point Charge ............................... 25-5
Velocity Field of a Point Source ............................ 23-6 Conservative Forces ..................................................... 25-5
Velocity Field of a Line Source ............................. 23-7 Electric Voltage ............................................................. 25-6
Flux ........................................................................... 23-8 A Field Plot Model ................................................. 25-10
Bernoulli’s Equation ................................................... 23-9 Computer Plots ...................................................... 25-12
Applications of Bernoulli’s Equation ......................... 23-12
Hydrostatics ....................................................... 23-12 CHAPTER 26 ELECTRIC FIELDS AND
Leaky Tank ........................................................ 23-12 CONDUCTORS
Airplane Wing .................................................... 23-13
Electric Field Inside a Conductor .................................. 26-1
Sailboats ............................................................ 23-14 Surface Charges ..................................................... 26-2
The Venturi Meter ............................................... 23-15 Surface Charge Density .......................................... 26-3
The Aspirator ..................................................... 23-16 Example: Field in a Hollow Metal Sphere ................. 26-4
Care in Applying Bernoulli’s Equation ................ 23-16 Van de Graaff generator .............................................. 26-6
Hydrodynamic Voltage ...................................... 23-17 Electric Discharge ................................................... 26-7
Town Water Supply ............................................ 23-18 Grounding ............................................................... 26-8
Viscous Effects .................................................. 23-19
The Electron Gun .......................................................... 26-8
Vortices ................................................................... 23-20 The Filament ............................................................ 26-9
Quantized Vortices in Superfluids ...................... 23-22 Accelerating Field ................................................. 26-10
A Field Plot ............................................................ 26-10
CHAPTER 24 COULOMB'S AND GAUSS' LAW Equipotential Plot ................................................... 26-11
Coulomb's Law ......................................................... 24-1 Electron Volt as a Unit of Energy ................................. 26-12
CGS Units ............................................................ 24-2 Example ................................................................ 26-13
MKS Units ............................................................ 24-2 About Computer Plots ........................................... 26-13
Checking Units in MKS Calculations .................... 24-3 The Parallel Plate Capacitor ........................................ 26-14
Summary ............................................................. 24-3 Deflection Plates .................................................... 26-16
Example 1 Two Charges .................................... 24-3
Example 2 Hydrogen Atom ................................ 24-4 CHAPTER 27 BASIC ELECTRIC CIRCUITS
Force Produced by a Line Charge ............................ 24-6 Electric Current ............................................................ 27- 2
Short Rod ............................................................. 24-9 Positive and Negative Currents .............................. 27- 3
The Electric Field ..................................................... 24-10 A Convention .......................................................... 27- 5
Unit Test Charge ................................................ 24-11 Current and Voltage ..................................................... 27- 6
Electric Field lines ................................................... 24-12 Resistors ................................................................ 27- 6
Mapping the Electric Field ................................. 24-12 A Simple Circuit ...................................................... 27- 8
Field Lines ......................................................... 24-13 The Short Circuit ..................................................... 27- 9
Continuity Equation for Electric Fields ................ 24-14 Power ..................................................................... 27- 9
Flux .................................................................... 24-15 Kirchoff’s Law ............................................................ 27- 10
Negative Charge ................................................ 24-16 Application of Kirchoff’s Law ................................ 27- 11
Flux Tubes ......................................................... 24-17 Series Resistors .................................................... 27- 11
Conserved Field Lines ....................................... 24-17 Parallel Resistors .................................................. 27- 12
A Mapping Convention ...................................... 24-17 Capacitance and Capacitors ..................................... 27- 14
Summary ........................................................... 24-18 Hydrodynamic Analogy ........................................ 27- 14
A Computer Plot ................................................. 24-19 Cylindrical Tank as a Constant Voltage Source .... 27- 15
Electrical Capacitance ......................................... 27- 16
Gauss’ Law ............................................................. 24-20
Electric Field of a Line Charge ........................... 24-21 Energy Storage in Capacitors .................................... 27- 18
Flux Calculations ................................................ 24-22 Energy Density in an Electric Field ....................... 27- 19
Area as a Vector ................................................ 24-22 Capacitors as Circuit Elements .................................. 27- 20
Gauss' Law for the Gravitational Field ..................... 24-23 The RC Circuit ............................................................ 27- 22
Gravitational Field of a Point Mass ..................... 24-23 Exponential Decay ............................................... 27- 23
Gravitational Field The Time Constant RC .......................................... 27- 24
Half-Lives ............................................................. 27- 25
of a Spherical Mass ........................................... 24-24
Initial Slope ........................................................... 27- 25
Gravitational Field Inside the Earth ..................... 24-24
The Exponential Rise ............................................ 27- 26
Solving Gauss' Law Problems ............................ 24-26
The Neon Bulb Oscillator ........................................... 27- 28
Problem Solving ...................................................... 24-29 The Neon Bulb ..................................................... 27- 28
The Neon Oscillator Circuit ................................... 27- 29
Period of Oscillation .............................................. 27- 30
Experimental Setup .............................................. 27- 31
vi
 Preface & TOC-vii

CHAPTER 28 MAGNETISM CHAPTER 30 FARADAY'S LAW

Two Garden Peas ............................................... 28- 2 Electric Field of Static Charges .................................. 30-2
A Thought Experiment .............................................. 28- 4 A Magnetic Force Experiment ................................... 30-3
Charge Density on the Two Rods ........................ 28- 6 Air Cart Speed Detector ............................................ 30-5
A Proposed Experiment ...................................... 28- 7 A Relativity Experiment .............................................. 30-9
Origin of Magnetic Forces ................................... 28- 8
Faraday's Law ......................................................... 30-11
Magnetic Forces ............................................... 28- 10
Magnetic Flux .................................................... 30-11
Magnetic Force Law ............................................... 28- 10 One Form of Faraday's Law ............................... 30-12
The Magnetic Field B ........................................ 28- 10 A Circular Electric Field ...................................... 30-13
Direction of the Magnetic Field .......................... 28- 11
The Right Hand Rule for Currents ..................... 28- 13 Line Integral of E around a Closed Path ............ 30-14
Parallel Currents Attract .................................... 28- 14 Using Faraday's Law ............................................... 30-15
The Magnetic Force Law .................................. 28- 14 Electric Field of an Electromagnet ...................... 30-15
Lorentz Force Law ............................................ 28- 15 Right Hand Rule for Faraday's Law .................... 30-15
Dimensions of the Electric Field of Static Charges .......................... 30-16
Magnetic Field, Tesla and Gauss ...................... 28- 16 The Betatron ............................................................ 30-16
Uniform Magnetic Fields ................................... 28- 16 Two Kinds of Fields ................................................. 30-18
Helmholtz Coils ................................................. 28- 18
Note on our E⋅d meter .................................. 30-20
Motion of Charged Particles in Magnetic Fields ...... 28- 19
Motion in a Uniform Magnetic Field ................... 28- 20 Applications of Faraday’s Law ................................. 30-21
Particle Accelerators ......................................... 28- 22 The AC Voltage Generator ................................. 30-21
Gaussmeter ....................................................... 30-23
Relativistic Energy and Momenta ........................... 28- 24
A Field Mapping Experiment .............................. 30-24
Bubble Chambers .................................................. 28- 26
The Mass Spectrometer .................................... 28- 28
Magnetic Focusing ........................................... 28- 29 CHAPTER 31 INDUCTION AND
Space Physics ....................................................... 28- 31 MAGNETIC MOMENT
The Magnetic Bottle .......................................... 28- 31 The Inductor .............................................................. 31-2
Van Allen Radiation Belts .................................. 28- 32 Direction of the Electric Field ................................ 31-3
Induced Voltage .................................................. 31-4
Inductance ........................................................... 31-5
CHAPTER 29 AMPERE'S LAW
Inductor as a Circuit Element .................................... 31-7
The Surface Integral .................................................. 29-2
The LR Circuit ...................................................... 31-8
Gauss’ Law .......................................................... 29-3
The LC Circuit ......................................................... 31-10
The Line Integral ....................................................... 29-5
Intuitive Picture of the LC Oscillation .................. 31-12
Ampere’s Law ........................................................... 29-7 The LC Circuit Experiment ................................. 31-13
Several Wires ..................................................... 29-10
Measuring the Speed of Light ................................. 31-15
Field of a Straight Wire ....................................... 29-11
Magnetic Moment ................................................... 31-18
Field of a Solenoid ................................................... 29-14
Magnetic Force on a Current ............................. 31-18
Right Hand Rule for Solenoids ........................... 29-14
Torque on a Current Loop .................................. 31-20
Evaluation of the Line Integral ............................ 29-15
Magnetic Moment .............................................. 31-21
Calculation of i encl os ed ....................................... 29-15 Magnetic Energy ................................................ 31-22
Using Ampere's law ........................................... 29-15 Summary of Magnetic Moment Equations .......... 31-24
One More Right Hand Rule ................................ 29-16 Charge q in a Circular Orbit ............................... 31-24
The Toroid .......................................................... 29-17
Iron Magnets ........................................................... 31-26
The Electromagnet ............................................. 31-28
The Iron Core Inductor ....................................... 31-29
Superconducting Magnets ................................. 31-30
Appendix: The LC circuit and Fourier Analysis ........ 31-31

vii
Preface & TOC-viii

CHAPTER 32 MAXWELL'S EQUATIONS CHAPTER 34 PHOTONS

Gauss’ Law for Magnetic Fields ............................... 32- 2 Blackbody Radiation ................................................. 34-2
Maxwell’s Correction to Ampere’s Law ..................... 32- 4 Planck Blackbody Radiation Law ......................... 34-4
Example: Magnetic Field The Photoelectric Effect ............................................. 34-5
between the Capacitor Plates ........................... 32- 6 Planck's Constant h ................................................... 34-8
Maxwell’s Equations ................................................. 32- 8 Photon Energies ........................................................ 34-9
Symmetry of Maxwell’s Equations ............................. 32- 9 Particles and Waves ................................................ 34-11
Maxwell’s Equations in Empty Space ..................... 32- 10 Photon Mass ........................................................... 34-12
A Radiated Electromagnetic Pulse .................... 32- 10 Photon Momentum ............................................. 34-13
A Thought Experiment ...................................... 32- 11 Antimatter ................................................................ 34-16
Speed of an Electromagnetic Pulse .................. 32- 14 Interaction of Photons and Gravity ........................... 34-18
Electromagnetic Waves .......................................... 32- 18 Evolution of the Universe ......................................... 34-21
Electromagnetic Spectrum ..................................... 32- 20 Red Shift and the Expansion of the Universe ..... 34-21
Components of the Electromagnetic Spectrum . 32- 20 Another View of Blackbody Radiation ................ 34-22
Blackbody Radiation ......................................... 32- 22 Models of the universe ............................................ 34-23
UV, X Rays, and Gamma Rays ......................... 32- 22 Powering the Sun ............................................... 34-23
Polarization ............................................................. 32- 23 Abundance of the Elements ............................... 34-24
Polarizers .......................................................... 32- 24 The Steady State Model of the Universe ............ 34-25
Magnetic Field Detector .................................... 32- 26 The Big Bang Model ................................................ 34-26
Radiated Electric Fields .......................................... 32- 28 The Helium Abundance ..................................... 34-26
Field of a Point Charge ...................................... 32- 30 Cosmic Radiation ............................................... 34-27
The Three Degree Radiation ................................... 34-27
CHAPTER 33 LIGHT WAVES Thermal Equilibrium of the Universe ................... 34-28
Superposition of The Early Universe .................................................. 34-29
Circular Wave Patterns .............................................. 33-2 The Early Universe ............................................. 34-29
Huygens Principle ..................................................... 33-4 Excess of Matter over Antimatter ........................ 34-29
Two Slit Interference Pattern ...................................... 33-6 Decoupling (700,000 years) .............................. 34-31
The First Maxima .................................................. 33-8 Guidebooks ....................................................... 34-32
Two Slit Pattern for Light .......................................... 33-10
The Diffraction Grating ............................................ 33-12 CHAPTER 35 BOHR THEORY OF HYDROGEN
More About Diffraction Gratings ......................... 33-14 The Classical Hydrogen Atom ................................... 35-2
The Visible Spectrum ......................................... 33-15 Energy Levels ...................................................... 35-4
Atomic Spectra .................................................. 33-16 The Bohr Model ......................................................... 35-7
The Hydrogen Spectrum ......................................... 33-17 Angular Momentum in the Bohr Model ................. 35-8
The Experiment on Hydrogen Spectra ............... 33-18 De Broglie's Hypothesis .......................................... 35-10
The Balmer Series .............................................. 33-19
CHAPTER 36 SCATTERING OF WAVES
The Doppler Effect .................................................. 33-20 Scattering of a Wave by a Small Object .................... 36-2
Stationary Source and Moving Observer ............ 33-21
Reflection of Light ...................................................... 36-3
Doppler Effect for Light ...................................... 33-22
Doppler Effect in Astronomy .............................. 33-23 X Ray Diffraction ........................................................ 36-4
The Red Shift and theExpanding Universe ......... 33-24 Diffraction by Thin Crystals .................................. 36-6
A Closer Look at Interference Patterns .................... 33-26 The Electron Diffraction Experiment .......................... 36-8
Analysis of the Single Slit Pattern ....................... 33-27 The Graphite Crystal ............................................ 36-8
The Electron Diffraction Tube ............................... 36-9
Recording Diffraction Grating Patterns .................... 33-28
Electron Wavelength ............................................ 36-9
The Diffraction Pattern ........................................ 36-10
Analysis of the Diffraction Pattern ....................... 36-11
Other Sets of Lines ............................................. 36-12
Student Projects ................................................. 36-13
Student project by Gwendylin Chen ................... 36-14

viii
 Preface & TOC-ix

CHAPTER 37 LASERS, A MODEL ATOM CHAPTER 40 QUANTUM MECHANICS

AND ZERO POINT ENERGY Two Slit Experiment ................................................... 40-2
The Laser and Standing Light Waves ........................ 37-2 The Two Slit Experiment
Photon Standing Waves ....................................... 37-3 from a Particle Point of View ................................. 40-3
Photon Energy Levels .......................................... 37-4 Two Slit Experiment—One Particle at a Time ....... 40-3
A Model Atom ........................................................... 37-4 Born’s Interpretation of the Particle Wave ............. 40-6
Zero Point Energy ...................................................... 37-7 Photon Waves ...................................................... 40-6
Definition of Temperature ..................................... 37-8 Reflection and Fluorescence ................................ 40-8
A Closer Look at the Two Slit Experiment ............. 40-9
Two dimensional standing waves .............................. 37-8
The Uncertainty Principle ........................................ 40-14
Position-Momentum Form
CHAPTER 38 ATOMS
of the Uncertainty Principle ...................................... 40-15
Solutions of Schrödinger’s Single Slit Experiment ........................................ 40-16
Equation for Hydrogen .............................................. 38-2
Time-Energy Form of the Uncertainty Principle ........ 40-19
The = 0 Patterns ................................................ 38-4 Probability Interpretation .................................... 40-22
The ≠ 0 Patterns ................................................ 38-5 Measuring Short Times ...................................... 40-22
Intensity at the Origin ........................................... 38-5 Short Lived Elementary Particles ........................ 40-23
Quantized Projections of Angular Momentum ...... 38-5 The Uncertainty Principleand Energy Conservation . 40-24
The Angular Momentum Quantum Number ......... 38-7 Quantum Fluctuations and Empty Space ................ 40-25
Other notation ...................................................... 38-7 Appendix: How a pulse is formed from sine waves 40-27
An Expanded Energy Level Diagram ................... 38-8
Multi Electron Atoms .................................................. 38-9
Pauli Exclusion Principle ...................................... 38-9
Electron Spin ....................................................... 38-9
The Periodic Table .................................................. 38-10
Electron Screening ............................................. 38-10
Effective Nuclear Charge ................................... 38-12
Lithium ............................................................... 38-12
Beryllium ............................................................ 38-13
Boron ................................................................. 38-13
Up to Neon ........................................................ 38-13
Sodium to Argon ................................................ 38-13
Potassium to Krypton ......................................... 38-14
Summary ........................................................... 38-14
Ionic Bonding .......................................................... 38-15

CHAPTER 39 SPIN
The Concept of Spin .................................................. 39-3
Interaction of the Magnetic Field with Spin ................ 39-4
Magnetic Moments and the Bohr Magneton ........ 39-4
Insert 2 here .............................................................. 39-5
Electron Spin Resonance Experiment .................. 39-5
Nuclear Magnetic Moments ................................. 39-6
Sign Conventions ................................................. 39-6
Classical Picture of Magnetic Resonance ............ 39-8
Electron Spin Resonance Experiment ....................... 39-9
Appendix:Classical Picture of Magnetic Interactions 39-14

ix
Preface & TOC-x

CHAPTER ON GEOMETRICAL OPTICS

Reflection from Curved Surfaces ......................... Optics-3
A PHYSICS BASED CALCULUS TEXT
The Parabolic Reflection ................................ Optics-4 CHAPTER 1 INTRODUCTION TO CALCULUS
Mirror Images ...................................................... Optics-6 Limiting Process .................................................... Cal 1-3
The Corner Reflector ...................................... Optics-7 The Uncertainty Principle ................................. Cal 1-3
Motion of Light through a Medium ....................... Optics-8 Calculus Definition of Velocity................................ Cal 1-5
Index of Refraction ......................................... Optics-9 Acceleration .......................................................... Cal 1-7
Cerenkov Radiation ........................................... Optics-10 Components .................................................... Cal 1-7
Snell’s Law ........................................................ Optics-11 Integration ............................................................. Cal 1-8
Derivation of Snell’s Law .............................. Optics-12 Prediction of Motion ......................................... Cal 1-9
Internal Reflection .............................................. Optics-13 Calculating Integrals ...................................... Cal 1-11
Fiber Optics ................................................. Optics-14 The Process of Integrating ............................. Cal 1-13
Medical Imaging .......................................... Optics-15 Indefinite Integrals ......................................... Cal 1-14
Prisms ............................................................... Optics-15 Integration Formulas ...................................... Cal 1-14
Rainbows ..................................................... Optics-16 New Functions .................................................... Cal 1-15
The Green Flash .......................................... Optics-17 Logarithms ..................................................... Cal 1-15
Halos and Sun Dogs .................................... Optics-18 The Exponential Function ............................... Cal 1-16
Lenses .............................................................. Optics-18 Exponents to the Base 10 .............................. Cal 1-16
Spherical Lens Surface ................................ Optics-19 The Exponential Function yx ......................... Cal 1-16
Focal Length of a Spherical Surface ............ Optics-20 Euler's Number e = 2.7183. . . ....................... Cal 1-17
Aberrations .................................................. Optics-21 Differentiation and Integration.............................. Cal 1-18
Thin Lenses ....................................................... Optics-23 A Fast Way to go Back and Forth ................... Cal 1-20
The Lens Equation ....................................... Optics-24 Constant Acceleration Formulas .................... Cal 1-20
Negative Image Distance ............................. Optics-26 Constant Acceleration Formulas
Negative Focal Length & Diverging Lenses . Optics-26 in Three Dimensions ...................................... Cal 1-22
Negative Object Distance ............................ Optics-27 More on Differentiation ........................................ Cal 1-23
Multiple Lens Systems ................................. Optics-28 Series Expansions ......................................... Cal 1-23
Two Lenses Together ................................... Optics-29
Derivative of the Function x n ........................ Cal 1-24
Magnification ............................................... Optics-30
The Human Eye ................................................. Optics-31 The Chain Rule .............................................. Cal 1-25
Nearsightedness and Farsightedness ......... Optics-32 Remembering The Chain Rule ....................... Cal 1-25
Partial Proof of the Chain Rule (optional) ........ Cal 1-26
The Camera ...................................................... Optics-33
Depth of Field .............................................. Optics-34 Integration Formulas ............................................ Cal 1-27
Eye Glasses and a Home Lab Experiment ... Optics-36 Derivative of the Exponential Function ........... Cal 1-28
Integral of the Exponential Function ............... Cal 1-29
The Eyepiece .................................................... Optics-37
The Magnifier ............................................... Optics-38 Derivative as the Slope of a Curve ....................... Cal 1-30
Angular Magnification .................................. Optics-39 Negative Slope .............................................. Cal 1-31
Telescopes ........................................................ Optics-40 The Exponential Decay ....................................... Cal 1-32
Reflecting telescopes .................................. Optics-42 Muon Lifetime ................................................ Cal 1-32
Large Reflecting Telescopes. ...................... Optics-43 Half Life ......................................................... Cal 1-33
Hubbel Space Telescope ............................ Optics-44 Measuring the Time
World’s Largest Optical Telescope .............. Optics-45 Constant from a Graph .................................. Cal 1-34
Infrared Telescopes ..................................... Optics-46 The Sine and Cosine Functions ........................... Cal 1-35
Radio Telescopes ........................................ Optics-48 Radian Measure ............................................. Cal 1-35
The Very Long Baseline Array (VLBA) .......... Optics-49 The Sine Function .......................................... Cal 1-36
Microscopes ..................................................... Optics-50 Amplitude of a Sine Wave .............................. Cal 1-37
Scanning Tunneling Microscope .................. Optics-51 Derivative of the Sine Function ....................... Cal 1-38
Physical Constants in CGS Units ............ Back cover-1
Photograph credits ........................................................... i
Conversion Factors ................................. Back cover-1

x
Physics 2000
E. R. Huggins
Dartmouth College

Part I
Mechanics,
Waves & Particles

physics2000.com
Introduction
An Overview of Physics

INTRODUCTION—AN OVERVIEW OF PHYSICS

With a brass tube and a few pieces of glass, you can as an incredibly small, incredibly hot speck that has
construct either a microscope or a telescope. The expanded to the universe we see today. By looking
difference is essentially where you place the lenses. farther and farther out, astronomers have been
With the microscope, you look down into the world of looking farther and farther back in time, closer to
the small, with the telescope out into the world of the that hot, dense beginning. Physicists, by looking at
large. matter on a smaller and smaller scale with the even
more powerful accelerators, have been studying
In the twentieth century, physicists and astronomers matter that is even hotter and more dense. By the
have constructed ever larger machines to study matter end of the twentieth century, physicists and astrono-
on even smaller or even larger scales of distance. For mers have discovered that they are looking at the
the physicists, the new microscopes are the particle same image.
accelerators that provide views well inside atomic
nuclei. For the astronomers, the machines are radio It is likely that telescopes will end up being the most
and optical telescopes whose large size allows them to powerful microscopes. There is a limit, both finan-
record the faintest signals from space. Particularly cial and physical, to how big and powerful an
effective is the Hubble telescope that sits above the accelerator we can build. Because of this limit, we
obscuring curtain of the earth’s atmosphere. can use accelerators to study matter only up to a
certain temperature and density. To study matter
The new machines do not provide a direct image like that is still hotter and more dense, which is the same
the ones you see through brass microscopes or tele- as looking at still smaller scales of distance, the only
scopes. Instead a good analogy is to the Magnetic “machine” we have available is the universe itself.
Resonance Imaging (MRI) machines that first collect a We have found that the behavior of matter under the
huge amount of data, and then through the use of a extreme conditions of the very early universe have
computer program construct the amazing images show- left an imprint that we can study today with tele-
ing cross sections through the human body. The scopes.
telescopes and particle accelerators collect the vast
amounts of data. Then through the use of the theories In the rest of this introduction we will show you some
of quantum mechanics and relativity, the data is put of the pictures that have resulted from looking at
together to construct meaningful images. matter with the new machines. In the text itself we
will begin to learn how these pictures were con-
Some of the images have been surprising. One of the structed.
greatest surprises is the increasingly clear image of the
universe starting out about fourteen billion years ago
Int-2 An Overview of Physics

SPACE AND TIME has a diameter of 5 light hours (not including the cloud
The images of nature we see are images in both space of comets that lie out beyond the planets.)
and time, for we have learned from the work of Einstein The closest star, Proxima Centauri, is 4.2 light years
that the two cannot be separated. They are connected away. Light from this star, which started out when you
by the speed of light, a quantity we designate by the entered college as a freshman, will arrive at the earth
letter c, which has the value of a billion (1,000,000,000) shortly after you graduate (assuming all goes well).
feet (30 cm) in a second. Einstein’s remarkable discov- Stars in our local area are typically 2 to 4 light years
ery in 1905 was that the speed of light is an absolute apart, except for the so called binary stars which are
speed limit. Nothing in the current universe can travel pairs of stars orbiting each other at distances as small as
faster than the speed c. light days or light hours.
Because the speed of light provides us with an absolute On a still larger scale, we find that stars form island
standard that can be measured accurately, we use the structures called galaxies. We live in a fairly typical
value of c to relate the definitions of time and distance. galaxy called the Milky Way. It is a flat disk of stars
The meter is defined as the distance light travels in an with a slight bulge at the center much like the Sombrero
interval of 1/299,792.458 of a second. The length of a Galaxy seen edge on in Figure (1) and the neighboring
second itself is provided by an atomic standard. It is the spiral galaxy Andromeda seen in Figure (2). Our
time interval occupied by 9,192,631,770 vibrations of Milky Way is a spiral galaxy much like Andromeda,
a particular wavelength of light radiated by a cesium with the sun located about 2/3 of the way out in one of
atom. the spiral arms. If you look at the sky on a dark clear
Using the speed of light for conversion, clocks often night you can see the band of stars that cross the sky
make good meter sticks, especially for measuring called the Milky Way. Looking at these stars you are
astronomical distances. It takes light 1.27 seconds to looking sideways through the disk of the Milky Way
travel from the earth to the moon. We can thus say that galaxy.
the moon is 1.27 light seconds away. This is simpler
than saying that the moon is 1,250,000,000 feet or
382,000 kilometers away. Light takes 8 minutes to
reach us from the sun, thus the earth’s orbit about the
sun has a radius of 8 light minutes. Radio signals,
which also travel at the speed of light, took 2 1/2 hours
to reach the earth when Voyager II passed the planet
Uranus (temporarily the most distant planet). Thus
Uranus is 2 1/2 light hours away and our solar system

Figure 1 Figure 2
The Sombrero galaxy. The Andromeda galaxy.
 Int-3

Our galaxy and the closest similar galaxy, Androm- The Expanding Universe
eda, are both about 100,000 light years (.1 million light In the 1920s, Edwin Hubble made the surprising dis-
years) in diameter, contain about a billion stars, and are covery that, on average, the galaxies are all moving
about one million light years apart. These are more or away from us. The farther away a galaxy is, the faster
less typical numbers for the average size, population it is moving away. Hubble found a simple rule for this
and spacing of galaxies in the universe. recession, a galaxy twice as far away is receding twice
as fast.
To look at the universe over still larger distances, first
imagine that you are aboard a rocket leaving the earth At first you might think that we are at the exact center
at night. As you leave the launch pad, you see the of the universe if the galaxies are all moving directly
individual lights around the launch pad and street lights away from us. But that is not the case. Hubble’s
in neighboring roads. Higher up you start to see the discovery indicates that the universe is expanding
lights from the neighboring city. Still higher you see uniformly. You can see how a uniform expansion
the lights from a number of cities and it becomes harder works by blowing up a balloon part way, and drawing
and harder to see individual street lights. A short while a number of uniformly spaced dots on the balloon.
later all the bright spots you see are cities, and you can Then pick any dot as your own dot, and watch it as you
no longer see individual lights. At this altitude you continue to blow the balloon up. You will see that the
count cities instead of light bulbs. neighboring dots all move away from your dot, and you
will also observe Hubble’s rule that dots twice as far
Similarly on our trip out to larger and larger distances away move away twice as fast.
in the universe, the bright spots are the galaxies for we
can no longer see the individual stars inside. On Hubble’s discovery provided the first indication that
distances ranging from millions up to billions of light there is a limit to how far away we can see things. At
years, we see galaxies populating the universe. On this distances of about fourteen billion light years, the
scale they are small but not quite point like. Instru- recessional speed approaches the speed of light. Re-
ments like the Hubble telescope in space can view cent photographs taken by the Hubble telescope show
structure in the most distant galaxies, like those shown galaxies receding at speeds in excess of 95% the speed
in Figure (3) . of light, galaxies close to the edge of what we call the
visible universe.

The implications of Hubble’s rule are more dramatic if

you imagine that you take a moving picture of the
expanding universe and then run the movie backward
in time. The rule that galaxies twice as far away are
receding twice as fast become the rule that galaxies
twice as far away are approaching you twice as fast. A
more distant galaxy, one at twice the distance but
heading toward you at twice the speed, will get to you
at the same time as a closer galaxy. In fact, all the
galaxies will reach you at the same instant of time.

Now run the movie forward from that instant of time,

and you see all the galaxies flying apart from what
looks like a single explosion. From Hubble’s law you
can figure that the explosion should have occurred
about fourteen billion years ago.
Figure 3
Hubble photograph of the most distant galaxies.
Int-4 An Overview of Physics

Did such an explosion really happen, or are we simply By 1960, it was recognized that, to explain the abun-
misreading the data? Is there some other way of dance of the elements as a result of nuclear reactions
interpreting the expansion without invoking such a inside of stars, you have to start with a mixture of
cataclysmic beginning? Various astronomers thought hydrogen and helium. Where did the helium come
there was. In their continuous creation theory they from? Could it have been created in a Big Bang?
developed a model of the universe that was both
unchanging and expanding at the same time. That As early as 1948, the Russian physicist George Gamov
sounds like an impossible trick because as the universe studied the consequences of the Big Bang model of the
expands and the galaxies move apart, the density of universe. He found that if the conditions in the early
matter has to decrease. To keep the universe from universe were just right, there should be light left over
changing, the model assumed that matter was being from the explosion, light that would now be a faint glow
created throughout space at just the right rate to keep the at radio wave frequencies. Gamov talked about this
average density of matter constant. prediction with several experimental physicists and
was told that the glow would be undetectable. Gamov’s
With this theory one is faced with the question of which prediction was more or less ignored until 1964 when
is harder to accept—the picture of the universe starting the glow was accidently detected as noise in a radio
in an explosion which was derisively called the Big telescope. Satellites have now been used to study this
Bang, or the idea that matter is continuously being glow in detail, and the results leave little doubt about
created everywhere? To provide an explicit test of the the explosive nature of the birth of the universe.
continuous creation model, it was proposed that all
matter was created in the form of hydrogen atoms, and What was the universe like at the beginning? In an
that all the elements we see around us today, the carbon, attempt to find out, physicists have applied the laws of
oxygen, iron, uranium, etc., were made as a result of physics, as we have learned them here on earth, to the
nuclear reactions inside of stars. collapsing universe seen in the time reversed motion
picture of the galaxies. One of the main features that
To test this hypothesis, physicists studied in the labo- emerges as we go back in time and the universe gets
ratory those nuclear reactions which should be relevant smaller and smaller, is that it also becomes hotter and
to the synthesis of the elements. The results were quite hotter. The obvious question in constructing a model
successful. They predicted the correct or nearly correct of the universe is how small and how hot do we allow
abundance of all the elements but one. The holdout was it to get? Do we stop our model, stop our calculations,
helium. There appeared to be more helium in the when the universe is down to the size of a galaxy? a
universe than they could explain. star? a grapefruit? or a proton? Does it make any sense
to apply the laws of physics to something as hot and
dense as the universe condensed into something smaller
than, say, the size of a grapefruit? Surprisingly, it may.
One of the frontiers of physics research is to test the
application of the laws of physics to this model of the
hot early universe.
 Int-5

We will start our disruption of the early universe at a STRUCTURE OF MATTER

time when the universe was about a billionth of a We will start our trip down to small scales with a rather
second old and the temperature was three hundred large, familiar example—the earth in orbit about the
thousand billion ( 3 × 1014 ) degrees. While this sounds sun. The earth is attracted to the sun by a force called
like a preposterously short time and unbelievably high gravity, and its motion can be accurately forecast, using
temperature, it is not the shortest time or highest a set of rules called Newtonian mechanics. The basic
temperature that has been quite carefully considered. concepts involved in Newtonian mechanics are force,
For our overview, we are arbitrarily choosing that time mass, velocity and acceleration, and the rules tell us
because of the series of pictures we can paint which how these concepts are related. (Half of the traditional
show the universe evolving. These pictures all involve introductory physics courses is devoted to learning
the behavior of matter as it has been studied in the these rules.)
laboratory. To go back earlier relies on theories that we
are still formulating and trying to test.
Atoms
To recognize what we see in this evolving picture of the We will avoid much of the complexity we see around
universe, we first need a reasonably good picture of us by next focusing in on a single hydrogen atom. If we
what the matter around us is like. With an understand- increase the magnification so that a garden pea looks as
ing of the building blocks of matter, we can watch the big as the earth, then one of the hydrogen atoms inside
pieces fit together as the universe evolves. Our discus- the pea would be about the size of a basketball. How
sion of these building blocks will begin with atoms we interpret what we see inside the atom depends upon
which appear only late in the universe, and work down our previous experience with physics. With a back-
to smaller particles which play a role at earlier times. ground in Newtonian mechanics, we would see a
To understand what is happening, we also need a miniature solar system with the nucleus at the center
picture of how matter interacts via the basic forces in and an electron in orbit. The nucleus in hydrogen
nature. consists of a single particle called the proton, and the
electron is held in orbit by an electric force. At this
When you look through a microscope and change the
magnification, the proton and electron are tiny points,
magnification, what you see and how you interpret it,
too small to show any detail.
changes, even though you are looking at the same
sample. To get a preliminary idea of what matter is
made from and how it behaves, we will select a
particular sample and magnify it in stages. At each
stage we will provide a brief discussion to help interpret
what we see. As we increase the magnification, the
interpretation of what we see changes to fit and to
explain the new picture. Surprisingly, when we get
down to the smallest scales of distance using the
greatest magnification, we see the entire universe at its
infancy. We have reached the point where studying
matter on the very smallest scale requires an under-
Figure 8-25a
standing of the very largest, and vice versa.
Elliptical orbit of an earth satellite calculated
using Newtonian mechanics.
Int-6 An Overview of Physics

There are similarities and striking differences between protons repel each other and the electrons repel to give
the gravitational force that holds our solar system the two repulsive forces. The net result is that the
together and the electric force that holds the hydrogen attractive and repulsive forces cancel and we end up
atom together. Both forces in these two examples are with essentially no electric force between the atoms.
attractive, and both forces decrease as the square of the
distance between the particles. That means that if you Rather than counting individual forces, it is easier to
double the separation, the force is only one quarter as add up electric charge. Since a proton and an electron
strong. The strength of the gravitational force depends have opposite charges, the total charge in a hydrogen
on the mass of the objects, while the electric force atom adds up to zero. With no net charge on either of
depends upon the charge of the objects. the two hydrogen atoms in our example, there is no net
electric force between them. We say that a complete
One of the major differences between electricity and hydrogen atom is electrically neutral.
gravity is that all gravitational forces are attractive,
while there are both attractive and repulsive electric While complete hydrogen atoms are neutral, they can
forces. To account for the two types of electric force, attract each other if you bring them too close together.
we say that there are two kinds of electric charge, which What happens is that the electron orbits are distorted by
Benjamin Franklin called positive charge and negative the presence of the neighboring atom, the electric
charge. The rule is that like charges repel while forces no longer exactly cancel, and we are left with a
opposite charges attract. Since the electron and the small residual force called a molecular force. It is the
proton have opposite charge they attract each other. If molecular force that can bind the two hydrogen atoms
you tried to put two electrons together, they would repel together to form a hydrogen molecule. These molecu-
because they have like charges. You get the same lar forces are capable of building very complex objects,
repulsion between two protons. By the accident of like people. We are the kind of structure that results
Benjamin Franklin’s choice, protons are positively from electric forces, in much the same way that solar
charged and electrons are negatively charged. systems and galaxies are the kind of structures that
result from gravitational forces.
Another difference between the electric and gravita-
tional forces is their strengths. If you compare the Chemistry deals with reactions between about 100
electric to the gravitational force between the proton different elements, and each element is made out of a
and electron in a hydrogen atom, you find that the different kind of atom. The basic distinction between
electric force is 227000000000000000000000000 atoms of different elements is the number of protons in
0000000000000 times stronger than the gravitational the nucleus. A hydrogen nucleus has one proton, a
force. On an atomic scale, gravity is so weak that it is helium nucleus 2 protons, a lithium nucleus 3 protons,
essentially undetectable. on up to the largest naturally occurring nucleus, ura-
nium with 92 protons.
On a large scale, gravity dominates because of the
cancellation of electric forces. Consider, for example, Complete atoms are electrically neutral, having as
the net electric force between two complete hydrogen many electrons orbiting outside as there are protons in
atoms separated by some small distance. Call them the nucleus. The chemical properties of an atom are
atom A and atom B. Between these two atoms there are determined almost exclusively by the structure of the
four distinct forces, two attractive and two repulsive. orbiting electrons, and their electron structure depends
The attractive forces are between the proton in atom A very much on the number of electrons. For example,
and the electron in atom B, and between the electron in helium with 2 electrons is an inert gas often breathed by
atom A and the proton in atom B. However, the two deep sea divers. Lithium with 3 electrons is a reactive
metal that bursts into flame when exposed to air. We
go from an inert gas to a reactive metal by adding one
electron.
 Int-7

Light
The view of the hydrogen atom as a miniature solar Visible light is only a small part of what we call the
system, a view of the atom seen through the “lens” of electromagnetic spectrum. Our eyes are sensitive to
Newtonian mechanics, fails to explain much of the light waves whose wavelength varies only over a very
atom’s behavior. When you heat hydrogen gas, it narrow range. Shorter wavelengths lie in the ultravio-
glows with a reddish glow that consists of three distinct let or x ray region, while at increasingly longer wave-
colors or so called spectral lines. The colors of the lines lengths are infra red light, microwaves, and radio
are bright red, swimming pool blue, and deep violet. waves. Maxwell’s theory made it clear that these other
You need more than Newtonian mechanics to under- wavelengths should exist, and within a few years, radio
stand why hydrogen emits light, let alone explain these waves were discovered. The broadcast industry is now
three special colors. dependent on Maxwell’s equations for the design of
radio and television transmitters and receivers.
In the middle of the 1800s, Michael Faraday went a (Maxwell’s theory is what is usually taught in the
long way in explaining electric and magnetic phenom- second half of an introductory physics course. That
ena in terms of electric and magnetic fields. These gets you all the way up to 1860.)
fields are essentially maps of electric and magnetic
forces. In 1860 James Clerk Maxwell discovered that While Maxwell’s theory works well for the design of
the four equations governing the behavior of electric radio antennas, it does not do well in explaining the
and magnetic fields could be combined to make up behavior of a hydrogen atom. When we apply
what is called a wave equation. Maxwell could con- Maxwell’s theory to the miniature solar system model
struct his wave equation after making a small but of hydrogen, we do predict that the orbiting electron
crucial correction to one of the underlying equations. will radiate light. But we also predict that the atom will
self destruct. The unambiguous prediction is that the
The importance of Maxwell’s wave equation was that electron will continue to radiate light of shorter and
it predicted that a particular combination of electric and shorter wavelength while spiraling in faster and faster
magnetic fields could travel through space in a wave- toward the nucleus, until it crashes. The combination
like manner. Equally important was the fact that the of Newton’s laws and Maxwell’s theory is known as
wave equation allowed Maxwell to calculate what the Classical Physics. We can easily see that classical
speed of the wave should be, and the answer was about physics fails when applied even to the simplest of
a billion feet per second. Since only light was known atoms.
to travel that fast, Maxwell made the guess that he had
discovered the theory of light, that light consisted of a
wave of electric and magnetic fields of force.

visible ultraviolet
radio, television, radar light rays gamma rays

wavelength, cm

106 5 4 3 2 10 1 10 -1 -2 -3 -4 -5 -6 -7 -8 -9 -10 -11 -12

Figure 32-24 infrared rays X-rays

The electromagnetic spectrum.
Int-8 An Overview of Physics

Photons The Bohr Model

In the late 1890’s, it was discovered that a beam of light In 1913, the year after the nucleus was discovered,
could knock electrons out of a hydrogen atom. The Neils Bohr developed a somewhat ad hoc model that
phenomenon became known as the photoelectric ef- worked surprisingly well in explaining hydrogen. Bohr
fect. You can use Maxwell’s theory to get a rough idea assumed that the electron in hydrogen could travel on
of why a wave of electric and magnetic force might be only certain allowed orbits. There was a smallest,
able to pull electrons out of a surface, but the details all lowest energy orbit that is occupied by an electron in
come out wrong. In 1905, in the same year that he cool hydrogen atoms. The fact that this was the
developed his theory of relativity, Einstein explained smallest allowed orbit meant that the electron would
the photoelectric effect by proposing that light con- not spiral in and crush into the nucleus.
sisted of a beam of particles we now call photons.
When a metal surface is struck by a beam of photons, Using Maxwell’s theory, one views the electron as
an electron can be knocked out of the surface if it is radiating light continuously as it goes around the orbit.
struck by an individual photon. A simple formula for In Bohr’s picture the electron does not radiate while in
the energy of the photons led to an accurate explanation one of the allowed orbits. Instead it radiates, it emits a
of all the experimental results related to the photoelec- photon, only when it jumps from one orbit to another.
tric effect.
To see why heated hydrogen radiates light, we need a
Despite its success in explaining the photoelectric picture of thermal energy. A gas, like a bottle of
effect, Einstein’s photon picture of light was in conflict hydrogen or the air around us, consists of molecules
not only with Maxwell’s theory, it conflicted with over flying around, bouncing into each other. Any moving
100 years of experiments which had conclusively object has extra energy due to its motion. If all the parts
demonstrated that light was a wave. This conflict was of the object are moving together, like a car traveling
not to be resolved in any satisfactory way until the down the highway, then we call this energy of motion
middle 1920s. kinetic energy. If the motion is the random motion of
molecules bouncing into each other, we call it thermal
The particle nature of light helps but does not solve the energy.
problems we have encountered in understanding the
behavior of the electron in hydrogen. According to The temperature of a gas is proportional to the average
Einstein’s photoelectric formula, the energy of a pho- thermal energy of the gas molecules. As you heat a gas,
ton is inversely proportional to its wavelength. The the molecules move faster, and their average thermal
longer wavelength red photons have less energy than
the shorter wavelength blue ones. To explain the
special colors of light emitted by hydrogen, we have to
be able to explain why only photons with very special
energies can be emitted.

Lyma
ns
Ba

e
lm
rie

er s
s

Paschen series

r1
eries

r2

r3

Figure 35-6
The allowed orbits of the Bohr Model.
 Int-9

energy and temperature rises. At the increased speed However this is not the case for the Bohr model. Bohr
the collisions between molecules are also stronger. found a simple formula for the electron energies of all
the allowed orbits. This one formula in a sense explains
Consider what happens if we heat a bottle of hydrogen the many spectral lines of hydrogen. A lot more came
gas. At room temperature, before we start heating, the out of Bohr’s model than Bohr had to put in.
electrons in all the atoms are sitting in their lowest
energy orbits. Even at this temperature the atoms are The problem with Bohr’s model is that it is essentially
colliding but the energy involved in a room tempera- based on Newtonian mechanics, but there is no excuse
ture collision is not great enough to knock an electron whatsoever in Newtonian mechanics for identifying
into one of the higher energy orbits. As a result, room any orbit as special. Bohr focused the problem by
temperature hydrogen does not emit light. discovering that the allowed orbits had special values
of a quantity called angular momentum.
When you heat the hydrogen, the collisions between
atoms become stronger. Finally you reach a tempera- Angular momentum is related to rotational motion, and
ture in which enough energy is involved in a collision in Newtonian mechanics angular momentum increases
to knock an electron into one of the higher energy continuously and smoothly as you start to spin an
orbits. The electron then falls back down, from one object. Bohr could explain his allowed orbits by
allowed orbit to another until it reaches the bottom, proposing that there was a special unique value of
lowest energy orbit. The energy that the electron loses angular momentum—call it a unit of angular momen-
in each fall, is carried out by a photon. Since there are tum. Bohr found, using standard Newtonian calcula-
only certain allowed orbits, there are only certain tions, that his lowest energy orbit had one unit of
special amounts of energy that the photon can carry out. angular momentum, orbit 2 had two units, orbit 3 three
units, etc. Bohr could explain his entire model by the
To get a better feeling for how the model works, one assumption that angular momentum was quan-
suppose we number the orbits, starting at orbit 1 for the tized, i.e., came only in units.
lowest energy orbit, orbit 2 for the next lowest energy
orbit, etc. Then it turns out that the photons in the red Bohr’s quantization of angular momentum is counter
spectral line are radiated when the electron falls from intuitive, for it leads to the picture that when we start to
orbit 3 to orbit 2. The red photon’s energy is just equal rotate an object, the rotation increases in a jerky fashion
to the energy the electron loses in falling between these rather than continuously. First the object has no
orbits. The more energetic blue photons carry out the angular momentum, then one unit, then 2 units, and on
energy an electron loses in falling from orbit 4 to orbit up. The reason we do not see this jerky motion when
2, and the still more energetic violet photons corre- we start to rotate something large like a bicycle wheel,
spond to a fall from orbit 5 to orbit 2. All the other jumps is that the basic unit of angular momentum is very
give rise to photons whose energy is too large or too small. We cannot detect the individual steps in angular
small to be visible. Those with too much energy are momentum, it seems continuous. But on the scale of an
ultraviolet photons, while those with too little are in the atom, the steps are big and have a profound effect.
infra red part of the spectrum. The jump down to orbit
1 is the biggest jump with the result that all jumps down With Bohr’s theory of hydrogen and Einstein’s theory
to the lowest energy orbit results in ultraviolet photons. of the photoelectric effect, it was clear that classical
physics was in deep trouble. Einstein’s photons gave
It appears rather ad hoc to propose a theory where you a lumpiness to what should have been a smooth wave
invent a large number of special orbits to explain what in Maxwell’s theory of light and Bohr’s model gave a
we now know as a large number of spectral lines. One jerkiness to what should be a smooth change in angular
criterion for a successful theory in science is that you momentum. The bumps and jerkiness needed a new
get more out of the theory than you put in. If Bohr had picture of the way matter behaves, a picture that was
to invent a new allowed orbit for each spectral line introduced in 1924 by the graduate student Louis de
explained, the theory would be essentially worthless. Broglie.
Int-10 An Overview of Physics

PARTICLE-WAVE NATURE OF this works in detail is the subject of our chapter on

MATTER Quantum Mechanics. One fundamental requirement
Noting the wave and particle nature of light, of our modern interpretation of the particle wave is that,
for the interpretation to be meaningful, all forms of
de Broglie proposed that the electron had both a wave
matter, without exception, must have this particle wave
and a particle nature. While electrons had clearly
exhibited a particle behavior in various experiments, de nature. This general requirement is summarized by a
rule discovered by Werner Heisinberg, a rule known as
Broglie suggested that it was the wave nature of the
the uncertainty principle. How the rule got that name
electron that was responsible for the special allowed
orbits in Bohr’s theory. De Broglie presented a simple is also discussed in our chapter on quantum mechanics.
wave picture where, in the allowed orbits, an integer In 1925, after giving a seminar describing de Broglie’s
number of wavelengths fit around the orbit. Orbit 1 had model of electron waves in hydrogen, Erwin
one wavelength, orbit 2 had two wavelengths, etc. In Schrödinger was chided for presenting such a “child-
De Broglie’s picture, electron waves in non allowed ish” model. A colleague reminded him that waves do
orbits would cancel themselves out. Borrowing some not work that way, and suggested that since Schrödinger
features of Einstein’s photon theory of light waves, de had nothing better to do, he should work out a real wave
Broglie could show that the angular momentum of the equation for the electron waves, and present the results
electron would have the special quantized values when in a couple of weeks.
the electron wave was in one of the special, non
cancelling orbits. It took Schrödinger longer than a couple of weeks, but
he did succeed in constructing a wave equation for the
With his simple wave picture, de Broglie had hit upon electron. In many ways Schrödinger’s wave equation
the fundamental idea that was missing in classical for the electron is analogous to Maxwell’s wave equa-
physics. The idea is that all matter, not just light, has tion for light. Schrödinger’s wave equation for the
a particle wave nature. electron allows one to calculate the behavior of elec-
trons in all kinds of atoms. It allows one to explain and
It took a few years to gain a satisfactory interpretation
predict an atom’s electron structure and chemical
of the dual particle wave nature of matter. The current
properties. Schrödinger’s equation has become the
interpretation is that things like photons are in fact
fundamental equation of chemistry.
particles, but their motion is governed, not by Newto-
nian mechanics, but by the laws of wave motion. How

Figure 35-9 Figure 35-10

De Broglie picture of an electron If the circumference of the orbit is an integer
wave cancelling itself out. number of wavelengths, the electron wave will go
around without any cancellation.
 Int-11

CONSERVATION OF ENERGY One of the big surprises in physics was Einstein’s

Before we go on with our investigation of the hydrogen discovery of the equivalence of mass and energy, a
atom, we will take a short break to discuss the idea of relationship expressed by the famous equation
conservation of energy. This idea, which originated in E = mc 2 . In that equation, E stands for the energy of
Newtonian mechanics, survives more or less intact in an object, m its mass, and c is the speed of light. Since
our modern particle-wave picture of matter. the factor c 2 is a constant, Einstein’s equation is
basically saying that mass is a form of energy. The c 2
Physicists pay attention to the concept of energy only is there because mass and energy were initially thought
because energy is conserved. If energy disappears to be different quantities with different units like kilo-
from one place, it will show up in another. We saw this grams and joules. The c 2 simply converts mass units
in the Bohr model of hydrogen. When the electron lost into energy units.
energy falling down from one allowed orbit to a lower
energy orbit, the energy lost by the electron was carried What is amazing is the amount of energy that is in the
out by a photon. form of mass. If you could convert all the mass of a
pencil eraser into electrical energy, and sell the electri-
You can store energy in an object by doing work on the cal energy at the going rate of 10¢ per kilowatt hour,
object. When you lift a ball off the floor, for example, you would get about 10 million dollars for it. The
the work you did lifting the ball, the energy you problem is converting the mass to another, more useful,
supplied, is stored in a form we call gravitational form of energy. If you can do the conversion, however,
potential energy. Let go of the ball and it falls to the the results can be spectacular or terrible. Atomic and
floor, loosing its gravitational potential energy. But hydrogen bombs get their power from the conversion
just before it hits the floor, it has a lot of energy of of a small fraction of their mass energy into thermal
motion, what we have called kinetic energy. All the energy. The sun gets its energy by “burning” hydrogen
gravitational potential energy the ball had before we nuclei to form helium nuclei. The energy comes from
dropped it has been converted to kinetic energy. the fact that a helium nucleus has slightly less mass than
the hydrogen nuclei out of which it was formed.
After the ball hits the floor and is finally resting there,
it is hard to see where the energy has gone. One place If you have a particle at rest and start it moving, the
it has gone is into thermal energy, the floor and the ball particle gains kinetic energy. In Einstein’s view the
are a tiny bit warmer as a result of your dropping the particle at rest has energy due to its rest mass. When
ball. you start the particle moving, it gains energy, and since
mass is equivalent to energy, it also gains mass. For
Another way to store energy is to compress a spring. most familiar speeds the increase in mass due to kinetic
When you release the spring you can get the energy energy is very small. Even at the speeds travelled by
back. For example, compress a watch spring by rockets and spacecraft, the increase in mass due to
winding up the watch, and the energy released as the kinetic energy is hardly noticeable. Only when a
spring unwinds will run the watch for a day. We could particle’s speed gets up near the speed of light does the
call the energy stored in the compressed spring spring increase in mass become significant.
potential energy. Physicists invent all sorts of names
for the various forms of energy.
Int-12 An Overview of Physics

One of the first things we discussed about the behavior ANTI-MATTER

of matter is that nothing can travel faster than the speed Schrödinger’s equation for electron waves is a nonrela-
of light. You might have wondered if nature had traffic tivistic theory. It accurately describes electrons that are
cops to enforce this speed limit. It does not need one, moving at speeds small compared to the speed of light.
it uses a law of nature instead. As the speed of an object This is fine for most studies in chemistry, where
approaches the speed of light, its mass increases. The chemical energies are much much less than rest mass
closer to the speed of light, the greater increase in mass. energies. You can see the difference for example by
To push a particle up to the speed of light would give comparing the energy released by a conventional chemi-
it an infinite mass and therefore require an infinite cal bomb and an atomic bomb.
amount of energy. Since that much energy is not
available, no particle is going to exceed nature’s speed Schrödinger of course knew Einstein’s theory of rela-
limit. tivity, and initially set out to derive a relativistic wave
equation for the electron. This would be an equation
This raises one question. What about photons? They that would correctly explain the behavior of electrons
are particles of light and therefore travel at the speed of even as the speed of the electrons approached the speed
light. But their energy is not infinite. It depends instead of light and their kinetic energy became comparable to
on the wavelength or color of the photon. Photons or even exceeded their rest mass energy.
escape the rule about mass increasing with speed by
starting out with no rest mass. You stop a photon and Schrödinger did construct a relativistic wave equation.
nothing is left. Photons can only exist by traveling at The problem was that the equation had two solutions,
the speed of light. one representing ordinary electrons, the other an appar-
ently impossible particle with a negative rest mass. In
When a particle is traveling at speeds close enough to physics and mathematics we are often faced with
the speed of light that its kinetic energy approaches its equations with two or more solutions. For example, the
rest mass energy, the particle behaves differently than formula for the hypotenuse c of a right triangle with
slowly moving particles. For example, push on a sides of lengths a and b is
slowly moving particle and you can make the particle
c
move faster. Push on a particle already moving at c2 = a2 + b2 b
nearly the speed of light, and you merely make the a
particle more massive since it cannot move faster.
Since the relationship between mass and energy came This equation has two solutions, namely
out of Einstein’s theory of relativity, we say that c = + a 2 + b 2 and c = – a 2 + b 2 . The negative
particles moving near the speed of light obey relativis- solution does not give us much of a problem, we simply
tic mechanics while those moving slowly are nonrela- ignore it.
tivistic. Light is always relativistic, and all automobiles
Schrödinger could not ignore the negative mass solu-
on the earth are nonrelativistic.
tions in his relativistic wave equation for the following
reason. If he started with just ordinary positive mass
electrons and let them interact, the equation predicted
that the negative mass solutions would be created! The
peculiar solutions could not be ignored if the equation
was to be believed. Only by going to his nonrelativistic
equation could Schrödinger avoid the peculiar solu-
tions.
 Int-13

A couple years later, Dirac tried again to develop a PARTICLE NATURE OF FORCES
relativistic wave equation for the electron. At first it De Broglie got his idea for the wave nature of the
appeared that Dirac’s equation would avoid the nega- electron from the particle-wave nature of light. The
tive mass solutions, but with little further work, Dirac particle of light is the photon which can knock electrons
found that the negative mass solutions were still there. out of a metal surface. The wave nature is the wave of
Rather than giving up on his new equation, Dirac found electric and magnetic force that was predicted by
a new interpretation of these peculiar solutions. Instead Maxwell’s theory. When you combine these two
of viewing them as negatively charged electrons with aspects of light, you are led to the conclusion that
a negative mass, he could interpret them as positive electric and magnetic forces are ultimately caused by
mass particles with a positive electric charge. Accord- photons. We call any force resulting from electric or
ing to Dirac’s equation, positive and negative charged magnetic forces as being due to the electric interaction.
solutions could be created or destroyed in pairs. The The photon is the particle responsible for the electric
pairs could be created any time enough energy was interaction.
available.
Let us see how our picture of the hydrogen atom has
Dirac predicted the existence of this positively charged evolved as we have learned more about the particles
particle in 1929. It was not until 1933 that Carl and forces involved. We started with a miniature solar
Anderson at Caltech, who was studying the elementary system with the heavy proton at the center and an
particles that showered down from the sky (particles electron in orbit. The force was the electric force that
called cosmic rays), observed a positively charged in many ways resembled the gravitational force that
particle whose mass was the same as that of the keeps the earth in orbit around the sun. This picture
electron. Named the positron, this particle was imme- failed, however, when we tried to explain the light
diately identified as the positive particle expected from radiated by heated hydrogen.
Dirac’s equation.
The next real improvement comes with Schrödinger’s
In our current view of matter, all particles are described wave equation describing the behavior of the electron
by relativistic wave equations, and all relativistic wave in hydrogen. Rather than there being allowed orbits as
equations have two kinds of solutions. One solution is in Bohr’s model, the electron in Schrödinger’s picture
for ordinary matter particles like electrons, protons, has allowed standing wave patterns. The chemical
and neutrons. The other solution, which we now call properties of atoms can be deduced from these wave
antimatter, describes anti particles, the antielectron patterns, and Schrödinger’s equation leads to accurate
which is the positron, and the antiproton and the predictions of the wavelengths of light radiated not
antineutron. Since all antiparticles can be created or only by hydrogen but other atoms as well.
destroyed in particle-anti particle pairs, the antiparticle
has to have the opposite conserved property so that the There are two limitations to Schrödinger’s equation.
property will remain conserved. As an example, the One of the limitations we have seen is that it is a non
positron has the opposite charge as the electron so that relativistic equation, an equation that neglects any
electric charge is neither created or destroyed when change in the electron’s mass due to motion. While this
electron-positron pairs appear or disappear. is a very good approximation for describing the slow
speed electron in hydrogen, the wavelengths of light
While all particles have antiparticles, some particles radiated by hydrogen can be measured so accurately
like the photon, have no conserved properties other that tiny relativistic effects can be seen. Dirac’s relativ-
than energy. As a result, these particles are indistin- istic wave equation is required to explain these tiny
guishable from their antiparticles. relativistic corrections.
Int-14 An Overview of Physics

The second limitation is that neither Schrödinger’s or Renormalization

Dirac’s equations take into account the particle nature The early attempts to construct the theory of quantum
of the electric force holding hydrogen together. In the electrodynamics were plagued by infinities. What
hydrogen atom, the particle nature of the electric force would happen is that you would do an initial approxi-
has only the very tiniest effect on the wavelength of the mate calculation and the results would be good. You
radiated light. But even these effects can be measured would then try to improve the results by calculating
and the particle nature must be taken into account. The what were supposed to be tiny corrections, and the
theory that takes into account both the wave nature of corrections turned out to be infinitely large. One of the
the electron and the particle nature of the electric force main accomplishments of Feynman and Schwinger
is called quantum electrodynamics, a theory finally was to develop a mathematical procedure, sort of a
developed in 1947 by Richard Feynman and Julian mathematical slight of hand, that got rid of the infini-
Schwinger. Quantum electrodynamics is the most ties. This mathematical procedure became known as
precisely tested theory in all of science. renormalization.
In our current picture of the hydrogen atom, as de- Feynman always felt that renormalization was simply
scribed by quantum electrodynamics, the force be- a trick to cover up our ignorance of a deeper more
tween the electron and the proton nucleus is caused by accurate picture of the electron. I can still hear him
the continual exchange of photons between the two saying this during several seminars. It turned out
charged particles. While being exchanged, the photon however that renormalization became an important
can do some subtle things like create a positron electron guide in developing theories of other forces. We will
pair which quickly annihilates. These subtle things shortly encounter two new forces as we look down into
have tiny but measurable effects on the radiated wave- the atomic nucleus, forces called the nuclear interac-
lengths, effects that correctly predicted by the theory. tion and the weak interaction. Both of these forces
have a particle-wave nature like the electric interaction,
The development of quantum electrodynamics came and the successful theories of these forces used
nearly 20 years after Dirac’s equation because of renormalization as a guide.
certain mathematical problems the theory had to over-
come. In this theory, the electron is treated as a point
particle with no size. The accuracy of the predictions
of quantum electrodynamics is our best evidence that
this is the correct picture. In other words, we have no
evidence that the electron has a finite size, and a very
accurate theory which assumes that it does not. How-
ever, it is not easy to construct a mathematical theory in
which a finite amount of mass and energy is crammed
into a region of no size. For one thing you are looking
at infinite densities of mass and energy.

Figure 8-33
Einstein’s theory of gravity predicted that Mercury’s
elliptical orbit “precessed” or rotated somewhat like the
rotation seen in the above orbit. Mercury’s precession
is much, much smaller.
 Int-15

Gravity Einstein’s theory predicts that the orbiting stars should

The one holdout, the one force for which we do not radiate gravitational waves and spiral in toward each
have a successful theory, is gravity. We have come a other. This is reminiscent of what we got by applying
long way since Newton’s law of gravity. After Einstein Maxwell’s theory to the electron in hydrogen, but in the
developed his theory of relativity in 1905, he spent the case of the pair of neutron stars the theory worked. The
next 12 years working on a relativistic theory of period of the orbit of these stars is changing in exactly
gravity. The result, known as general relativity is a the way one would expect if the stars were radiating
theory of gravity that is in many ways similar to gravitational waves.
Maxwell’s theory of electricity. Einstein’s theory
predicts, for example, that a planet in orbit about a star If our wave-particle picture of the behavior of matter is
should emit gravitational waves in much the same way correct, then the gravitational waves must have a
that Maxwell’s theory predicts that an electron in orbit particle nature like electromagnetic waves. Physicists
about a nucleus should emit electromagnetic radiation call the gravitational particle the graviton. We think we
or light. know a lot about the graviton even though we have not
yet seen one. The graviton should, like the photon,
One of the difficulties working with Einstein’s theory have no rest mass, travel at the speed of light, and have
of gravity is that Newton’s theory of gravity explains the same relationship between energy and wavelength.
almost everything we see, and you have to look very
hard in places where Newton’s law is wrong and One difference is that because the graviton has energy
Einstein’s theory is right. There is an extremely small and therefore mass, and because gravitons interact with
but measurable correction to the orbit of Mercury that mass, gravitons interact with themselves. This self
Newton’s theory cannot explain and Einstein’s theory interaction significantly complicates the theory of grav-
does. ity. In contrast photons interact with electric charge,
but photons themselves do not carry charge. As a
Einstein’s theory also correctly predicts how much result, photons do not interact with each other which
light will be deflected by the gravitational attraction of considerably simplifies the theory of the electric inter-
a star. You can argue that because light has energy and action.
energy is equivalent to mass, Newton’s law of gravity
should also predict that starlight should be deflected by An important difference between the graviton and the
the gravitational pull of a star. But this Newtonian photon, what has prevented the graviton from being
argument leads to half the deflection predicted by detected, is its fantastically weak interaction with mat-
Einstein’s theory, and the deflection predicted by Ein- ter. You saw that the gravitational force between the
stein is observed. electron and a proton is a thousand billion billion billion
billion times weaker than the electric force. In effect
The gravitational radiation predicted by Einstein’s this makes the graviton a thousand billion billion
theory has not been detected directly, but we have very billion billion times harder to detect. The only reason
good evidence for its existence. In 1974 Joe Taylor we know that this very weak force exists at all is that it
from the University of Massachusetts, working at the gets stronger and stronger as we put more and more
large radio telescope at Arecibo discovered a pair of mass together, to form large objects like planets and
neutron stars in close orbit about each other. We will stars.
have more to say about neutron stars later. The point is
that the period of the orbit of these stars can be
measured with extreme precision.
Int-16 An Overview of Physics

Not only do we have problems thinking of a way to A SUMMARY

detect gravitons, we have run into a surprising amount Up to this point our focus has been on the hydrogen
of difficulty constructing a theory of gravitons. The atom. The physical magnification has not been too
theory would be known as the quantum theory of great, we are still picturing the atom as an object
gravity, but we do not yet have a quantum theory of magnified to the size of a basketball with two particles,
gravity. The problem is that the theory of gravitons the electron and proton, that are too small to see. They
interacting with point particles, the gravitational anal- may or may not have some size, but we cannot tell at
ogy of quantum electrodynamics, does not work. The this scale.
theory is not renormalizable, you cannot get rid of the
infinities. As in the case of the electric interaction the What we have done is change our perception of the
simple calculations work well, and that is why we think atom. We started with a picture that Newton would
we know a lot about the graviton. But when you try to recognize, of a small solar system with the massive
make what should be tiny relativistic corrections, the proton at the center and the lighter electron held in orbit
correction turns out to be infinite. No mathematical by the electric force. When we modernize the picture
slight of hand has gotten rid of the infinities. by including Maxwell’s theory of electricity and mag-
netism, we run into trouble. We end up predicting that
The failure to construct a consistent quantum theory of the electron will lose energy by radiating light, soon
gravity interacting with point particles has suggested to crashing into the proton. Bohr salvaged the picture by
some theoretical physicists that our picture of the introducing his allowed orbits and quantized angular
electron and some other particles being point particles momentum, but the success of Bohr’s theory only
is wrong. In a new approach called string theory, the strengthened the conviction that something was funda-
elementary particles are view not as point particles but mentally wrong with classical physics.
instead as incredibly small one dimensional objects
called strings. The strings vibrate, with different Louis de Broglie pointed the way to a new picture of the
modes of vibration corresponding to different elemen- behavior of matter by proposing that all matter, not just
tary particles. light, had a particle-wave nature. Building on de
Broglie’s idea, Schrödinger developed a wave equa-
String theory is complex. For example, the strings exist tion that not only describes the behavior of the electron
in a world of 10 dimensions, whereas we live in a world in hydrogen, but in larger and more complex atoms as
of 4 dimensions. To make string theory work, you have well.
to explain what happened to the other six dimensions.
While Schrödinger’s non relativistic wave equation
Another problem with string theory is that it has not led adequately explains most classical phenomena, even in
to any predictions that distinguish it from other theo- the hydrogen atom, there are tiny but observable rela-
ries. There are as yet no tests, like the deflection of tivistic effects that Dirac could explain with his relativ-
starlight by the sun, to demonstrate that string theory is istic wave equation for the electron. Dirac handled the
right and other theories are wrong. problem of all relativistic wave equations having two
String theory does, however, have one thing going for solutions by reinterpreting the second solution as rep-
it. By spreading the elementary particles out from zero resenting antimatter.
dimensions (points) to one dimensional objects (strings),
the infinities in the theory of gravity can be avoided.
 Int-17

Dirac’s equation is still not the final theory for hydro- THE NUCLEUS
gen because it does not take into account the fact that To see the nucleus we have to magnify our hydrogen
electric forces are ultimately caused by photons. The atom to a size much larger than a basketball. When the
wave theory of the electron that takes the photon nature atom is enlarged so that it would just fill a football
of the electric force into account is known as quantum stadium, the nucleus, the single proton, would be about
electrodynamics. The predictions of quantum electro- the size of a pencil eraser. The proton is clearly not a
dynamics are in complete agreement with experiment, point particle like the electron. If we enlarge the atom
it is the most precisely tested theory in science. further to get a better view of the nucleus, to the point
The problems resulting from treating the electron as a where the proton looks as big as a grapefruit, the atom
point particle were handled in quantum electrodynam- is about 10 kilometers in diameter. This grapefruit
ics by renormalization. Renormalization does not sized object weighs 1836 times as much as the electron,
work, however, when one tries to formulate a quantum but it is the electron wave that occupies the 10 kilometer
theory of gravity where the gravitational force par- sphere of space surrounding the proton.
ticle—the graviton—interacts with point particles. This Before we look inside the proton, let us take a brief look
has led some theorists to picture the electron not as a at the nuclei of some other atoms. Once in a great while
point but as an incredibly small one dimensional object you will find a hydrogen nucleus with two particles.
called a string. While string theory is renormalizable, One is a proton and the other is the electrically neutral
there have been no experimental tests to show that particles called the neutron. Aside from the electric
string theory is right and the point particle picture is charge, the proton and neutron look very similar. They
wrong. This is as far as we can take our picture of the are about the same size and about the same mass. The
hydrogen atom without taking a closer look at the neutron is a fraction of a percent heavier than the
nucleus. proton, a small mass difference that will turn out to have
some interesting consequences.

As we mentioned, the type of element is determined by

the number of protons in the nucleus. All hydrogen
atoms have one proton, all helium atoms 2 protons, etc.
But for the same element there can be different num-
bers of neutrons in the nucleus. Atoms with the same
p numbers of protons but different numbers of neutrons
Hydrogen-1 are called different isotopes of the element. Another
isotope of hydrogen, one that is unstable and decays in
p roughly 10 years, is a nucleus with one proton and two
p
n n p neutrons called tritium.
Hydrogen-2 Helium-3
(Deuterium) The most stable isotope of helium is helium 4, with 2
p n protons and 2 neutrons. Helium 3 with 2 protons and
p n p one neutron is stable but very rare. Once we get beyond
n n Helium-4 hydrogen we name the different isotopes by adding a
Hydrogen-3 number after the name, a number representing the total
(Tritium) number of protons and neutrons. For example the
Figure 19-2 heaviest, naturally occurring atom is the isotope Ura-
Isotopes of hydrogen and helium. nium 238, which has 92 protons and 146 neutrons for
a total of 238 nuclear particles, or nucleons as we
sometimes refer to them.
Int-18 An Overview of Physics

The nucleons in a nucleus pack together much like the One of the important features of the nuclear force
grapes in a bunch, or like a bag of grapefruit. At our between nucleons is that it has a short range. Compared
enlargement where a proton looks as big as a grapefruit, to the longer range electric force, the nuclear force is
the uranium nucleus would be just over half a meter in more like a contact cement. When two protons are next
diameter, just big enough to hold 238 grapefruit. to each other, the attractive nuclear force is stronger
than the electric repulsion. But separate the protons by
When you look at a uranium nucleus with its 92 more than about 4 protons diameters and the electric
positively charged protons mixed in with electrically force is stronger.
neutral neutrons, then you have to wonder, what holds
the thing together? The protons, being all positively If you make nuclei by adding nucleons to a small
charged, all repel each other. And because they are so nucleus, the object becomes more and more stable
close together in the nucleus, the repulsion is extremely because all the nucleons are attracting each other. But
strong. It is much stronger than the attractive force felt when you get to nuclei whose diameter exceeds around
by the distant negative electrons. There must be 4 proton diameters, protons on opposite sides of the
another kind of force, and attractive force, that keeps nucleus start to repel each other. As a result nuclei
the protons from flying apart. larger than that become less stable as you make them
bigger. The isotope Iron 56 with 26 protons and 30
The attractive force is not gravity. Gravity is so weak neutrons, is about 4 proton diameters across and is the
that it is virtually undetectable on an atomic scale. The most stable of all nuclei. When you reach Uranium
attractive force that overpowers the electric repulsion is which is about 6 proton diameters across, the nucleus
called the nuclear force. The nuclear force between has become so unstable that if you jostle it by hitting it
nucleons is attractive, and essentially blind to the with a proton, it will break apart into two roughly equal
difference between a proton and a neutron. To the sized more stable nuclei. Once apart, the smaller nuclei
nuclear force, a proton and a neutron look the same. repel each other electrically and fly apart releasing
The nuclear force has no effect whatsoever on an electric potential energy. This process is called nuclear
electron. fission and is the source of energy in an atomic bomb.

While energy is released when you break apart the

large unstable nuclei, energy is also released when you
add nucleons to build up the smaller, more stable
nuclei. For example, if you start with four protons (four
hydrogen nuclei), turn two of the protons into neutrons
(we will see how to do this shortly) and put them
together to form stable helium 4 nucleus, you get a
considerable release of energy. You can easily figure
out how much energy is released by noting that 4
protons have a mass that is about .7 percent greater than
a helium nucleus. As a result when the protons
combine to form helium, about .7 percent of their mass
is converted to other forms of energy. Our sun is
powered by this energy release as it “burns” hydrogen
to form helium. This process is called nuclear fusion
and is the source of the energy of the powerful hydro-
gen bombs.
Figure 19-1
Styrofoam ball model of the uranium nucleus..
 Int-19

STELLAR EVOLUTION What is left behind of the core of the star depends on
Our sun is about half way through burning up the how massive the star was to begin with. If what remains
hydrogen in its hot, inner core. When the hydrogen is of the core is 1.4 times as massive as our sun, then the
exhausted in another 5 billion years, the sun will gravitational force will be strong enough to cram the
initially cool and start to collapse. But the collapse will electrons into the nuclei, turning all the protons into
release gravitational potential energy that makes the neutrons, and leaving behind a ball of neutrons about
smaller sun even hotter than it was before running out 20 kilometers in diameter. This is called a neutron star.
of hydrogen. The hotter core will emit so much light A neutron star is essentially a gigantic nucleus held
that the pressure of the light will expand the surface of together by gravity instead of the nuclear force.
the sun out beyond the earth’s orbit, and the sun will If you think that squeezing the mass of a star into a ball
become what is known as a red giant star. Soon, over 20 kilometers in diameter is hard to picture (at this
the astronomically short time of a few million years, the density all the people on the earth would fit into the
star will cool off becoming a dying, dark ember about volume of a raindrop), then consider what happens if
the size of the earth. It will become what is known as the remaining core is about six times as massive as the
a black dwarf. sun. With such mass, the gravitational force is so strong
If the sun had been more massive when the hydrogen that the neutrons are crushed and the star becomes
ran out and the star started to collapse, then more smaller and smaller.
gravitational potential energy would have been re- The matter in a neutron star is about as rigid as matter
leased. The core would have become hotter, hot enough can get. The more rigid a substance is, the faster sound
to ignite the helium to form the heavier nucleus carbon. waves travel through the substance. For example,
Higher temperatures are required to burn helium be- sound travels considerably faster through steel than air.
cause the helium nuclei, with two protons, repel each The matter in a neutron star is so rigid, or shall we say
other with four times the electric repulsion than hydro- so incompressible, that the speed of sound approaches
gen nuclei. As a result more thermal energy is required the speed of light.
to slam the helium nuclei close enough for the attractive
nuclear force to take over.

Once the helium is burned up, the star again starts to

cool and contract, releasing more gravitational poten-
tial energy until it becomes hot enough to burn the
carbon to form oxygen nuclei. This cycle keeps
repeating, forming one element after another until we
get to Iron 56. When you have an iron core and the star
starts to collapse and gets hotter, the iron does not burn.
You do not get a release of energy by making nuclei
larger than iron. As a result the collapse continues
resulting in a huge implosion.

Once the center collapses, a strong shock wave races

out through the outer layers of the star, tearing the star
apart. This is called a supernova explosion. It is in
these supernova explosions with their extremely high
temperatures that nuclei larger than iron are formed.
Figure 4
All the elements inside of you that are down the
1987 supernova as seen by the Hubble telescope.
periodic table from iron were created in a supernova.
Part of you has already been through a supernova
explosion.
Int-20 An Overview of Physics

When gravity has crushed the neutrons in a neutron The Weak Interaction
star, it has overcome the strongest resistance any In addition to gravity, the electric interaction and the
known force can possibly resist. But, as the collapse nuclear force, there is one more basic force or interac-
continues, gravity keeps getting stronger. According tion in nature given the rather bland name the weak
to our current picture of the behavior of matter, a rather interaction. While considerably weaker than electric
unclear picture in this case, the collapse continues until or nuclear forces, it is far far stronger than gravity on a
the star becomes a point with no size. Well before it nuclear scale.
reaches that end, gravity has become so strong that light
can no longer escape, with the result that these objects A distinctive feature of the weak interaction is its very
are known as black holes. short range. A range so short that only with the
construction of the large accelerators since 1970 has
We have a fuzzy picture of what lies at the center of a one been able to see the weak interaction behave more
black hole because we do not have a quantum theory of like the other forces. Until then, the weak interaction
gravity. Einstein’s classical theory of gravity predicts was known only by reactions it could cause, like
that the star collapses to a point, but before that happens allowing a proton to turn into a neutron or vice versa.
we should reach a state where the quantum effects of
gravity are important. Perhaps string theory will give Because of the weak interaction, an individual neutron
us a clue as to what is happening. We will not learn by is not stable. Within an average time of about 10
looking because light cannot get out. minutes it decays into a proton and an electron. Some-
times neutrons within an unstable nucleus also decay
The formation of neutron stars and black holes empha- into a proton and electron. This kind of nuclear decay
sizes an important feature of gravity. On an atomic was observed toward the end of the nineteenth century
scale, gravity is the weakest of the forces we have when knowledge of elementary particles was very
discussed so far. The gravitational force between an limited, and the electrons that came out in these nuclear
electron and a proton is a thousand billion billion billion decays were identified as some kind of a ray called a
billion ( 10 39 ) times weaker than the electric force. Yet beta ray. (There were alpha rays which turned out to
because gravity is long range like the electric force, and be helium nuclei, beta rays which were electrons, and
has no cancellation, it ends up dominating all other gamma rays which were photons.) Because the elec-
forces, even crushing matter as we know it, out of trons emitted during a neutron decay were called beta
existence. rays, the process is still known as the beta decay
process.

The electron is emitted when a neutron decays in order

to conerve electric charge. When the neutral neutron
decays into a positive proton, a negatively charged
particle must also be emitted so that the total charge
does not change. The lightest particle available to carry
out the negative charge is the electron.

Early studies of the beta decay process indicated that

while electric charge was conserved, energy was not.
For example, the rest mass of a neutron is nearly 0.14
percent greater than the rest mass of a proton. This
Figure 5 mass difference is about four times larger than the rest
Hubble telescope’s first view of a lone neutron star in mass of the electron, thus there is more than enough
visible light. This star is no greater than 16.8 miles (28
kilometers) across.
 Int-21

mass energy available to create the electron when the Leptons

neutron decays. If energy is conserved, you would We now know that neutrinos are emitted in the beta
expect that the energy left over after the electron is decay process because of another conservation law, the
created would appear as kinetic energy of the electron. conservation of leptons. The leptons are a family of
light particles that include the electron and the neutrino.
Careful studies of the beta decay process showed that When an electron is created, an anti neutrino is also
sometimes the electron carried out the expected amount created so that the number of leptons does not change.
of energy and sometimes it did not. These studies were
carried out in the 1920s, when not too much was known Actually there are three distinct conservation laws for
about nuclear reactions. There was a serious debate leptons. The lepton family consists of six particles, the
about whether energy was actually conserved on the electron, two more particles with rest mass and three
small scale of the nucleus. different kinds of neutrino. The other massive particles
are the muon which is 207 times as massive as the
In 1929, Wolfgang Pauli proposed that energy was electron, and the recently discovered tau particle which
conserved, and that the apparenty missing energy was is 3490 times heavier. The three kinds of neutrino are
carried out by an elusive particle that had not yet been the electron type neutrino, the muon type neutrino
seen. This elusive particle, which became known as the and the tau type neutrino. The names come from the
neutrino or “little neutral one”, had to have some rather fact that each type of particle is separately conserved.
peculiar properties. Aside from being electrically For example when a neutron decays into a proton and
neutral, it had to have essentially no rest mass because an electron is created, it is an anti electron type neu-
in some reactions the electron was seen to carry out all trino that is created at the same time to conserve
the energy, leaving none to create a neutrino rest mass. electron type particles.
The most bizarre property f the neutrino was its In the other common beta decay process, where a
undetectability. It had to pass through matter leaving proton turns into a neutron, a positron is created to
no trace. It was hard to believe such a particle could conserve electric charge. Since the positron is the anti
exist, yet on the other hand, it was hard to believe particle of the electron, its opposite, the electron type
energy was not conserved. The neutrino was finally neutrino, must be created to conserve leptons.
detected thirty years later and we are now quite confi-
dent that energy is conserved on the nuclear scale.

The neutrino is elusive because it interacts with matter

only through the weak interaction (and gravity). Pho-
tons interact via the strong electric interaction and are
quickly stopped when they encounter the electric charges
in matter. Neutrinos can pass through light years of
lead before there is a good chance that they will be
stopped. Only in the collapsing core of an exploding
star or in the very early universe is matter dense enough
to significantly absorb neutrinos. Because neutrinos
have no rest mass, they, like photons, travel at the speed
of light.
Int-22 An Overview of Physics

Nuclear Structure A CONFUSING PICTURE

The light nuclei, like helium, carbon, oxygen, gener- By 1932, the basic picture of matter looked about as
ally have about equal numbers of protons and neutrons. simple as it can possibly get. The elementary particles
As the nuclei become larger we find a growing excess were the proton, neutron, and electron. Protons and
of neutrons over protons. For example when we get up neutrons were held together in the nucleus by the
to Uranium 238, the excess has grown to 146 neutrons nuclear force, electrons were bound to nuclei by the
to 92 protons. electric force to form atoms, a residual of the electric
force held atoms together to form molecules, crystals
The most stable isotope of a given element is the one
and living matter, and gravity held large chunks of
with the lowest possible energy. Because the weak
matter together for form planets, stars and galaxies.
interaction allows protons to change into neutrons and
The rules governing the behavior of all this was quan-
vice versa, the number of protons and neutrons in a
tum mechanics on a small scale, which became New-
nucleus can shift until the lowest energy combination
tonian mechanics on the larger scale of our familiar
is reached.
world. There were a few things still to be straightened
Two forms of energy that play an important role in their out, such as the question as to whether energy was
proess are the extra mass energy of the neutrons, and conserved in beta decays, and in fact why beta decays
the electric potential energy of the protons. It takes a lot occurred at all, but it looked as if these loose ends
f to shove two protons together against their electric should be soon tied up.
repulsion. The work you do in shoving them together
The opposite happened. By 1960, there were well over
is stored as electric potential energy which will be
100 so called elementary particles, all of them unstable
released if you let go and the particles fly apart. This
except for the familiar electron, proton and neutron.
energy will not be released, however, if the protons are
Some lived long enough to travel kilometers down
latched together by the nuclear force. But in that case
through the earth’s atmosphere, others long enough to
the electric potential energy can be released by turning
be observed in particle detectors. Still others had such
one of the protons into a neutron. This will happen if
short lifetimes that, even moving at nearly the speed of
enough electric potential energy is available not only to
light, they could travel only a few proton diameters
create the extra neutron rest mass energy, but also the
before decaying. With few exceptions, these particles
positron required to conserve electric charge.
were unexpected and their behavior difficult to explain.
The reason that the large nuclei have an excess of Where they were expected, they were incorrectly iden-
neutrons over protons is that electric potential energy tified.
increases faster with increasing number of protons than
One place to begin the story of the progression of
neutron mass energy does with increasing numbers of
unexpected particles is with a prediction made in 1933
neutrons. The amount of extra neutron rest mass
by Heidi Yukawa. Yukawa proposed a new theory of
energy is more or less proportional to the number of
the nuclear force. Noting that the electric force was
neutrons. But the increase in electric potential energy
ultimately caused by a particle, Yukawa proposed that
as you add a proton depends on the number of protons
the nuclear force holding the protons and neutrons
already in the nucleus. The more protons already there,
together in the nucleus was also caused by a particle, a
the stronger the electric repulsion when you try to add
particle that became known as the nuclear force me-
another proton, and the greater the potential energy
son. The zero rest mass photon gives rise to the long
stored. As a result of this increasing energy cost of
range electric force. Yukawa developed a wave equa-
adding more protons, the large nuclei find their lowest
tion for the nuclear force meson in which the range of
energy balance having an excess of neutrons.
the force depends on the rest mass of the meson. The
bigger the rest mass of the meson, the shorter the range.
(Later in the text, we will use the uncertainty principle
to explain this relationship between the range of a force
and the rest mass of the particle causing it.)
 Int-23

From the fact that iron is the most stable nucleus, boggling to think of the nuclear force as being caused
Yukawa could estimate that the range of the nuclear by over 100 different kinds of mesons, while the
force is about equal to the diameter of an iron nucleus, electric force had only one particle, the photon.
about four proton diameters. From this, he predicted
that the nuclear force meson should have a rest mass One of the helpful ways of viewing matter at that time
bout 300 times the rest mass of the electron (about 1/6 was to identify each of the particle decays with one of
the rest mass of a proton). the four basic forces. The very fastest decays were
assumed to be caused by the strong nuclear force.
Shortly after Yukawa’s prediction, the muon was Decays that were about 100 times slower were identi-
discovered in the rain of particles that continually strike fied with the slightly weaker electric force. Decays that
the earth called cosmic rays. The rest mass of the muon took as long as a billionth of a second, a relatively long
was found to be about 200 times that of the electron, not lifetime, were found to be caused by the weak interac-
too far off the predicted mass of Yukawa’s particle. For tion. The general scheme was the weaker the force, the
a while the muon was hailed as Yukawa’s nuclear force longer it took to cause a particle decay.
meson. But further studies showed that muons could
travel considerable distances through solid matter. If
the muon were the nuclear force meson, it should
interact strongly with nuclei and be stopped rapidly.
Thus the muon was seen as not being Yukawa’s
particle. Then there was the question of what role the
muon played. Why did nature need it?

In 1947 another particle called the π meson was

discovered. (There were actually three π mesons, one
with a positive charge, the π + , one neutral, the π ° , and
one with a negative charge, the π – .) The π mesons
interacted strongly with nuclei, and had the mass close
to that predicted by Yukawa, 274 electron masses. The
π mesons were then hailed as Yukawa’s nuclear force
meson.

However, at almost the same time, another particle

called the K meson, 3.5 times heavier than the π
meson, was discovered. It also interacted strongly with
nuclei and clearly played a role in the nuclear force.
The nuclear force was becoming more complex than
Yukawa had expected. Figure 6 e–
First bubble chamber
photograph of the Ω
– p+
Experiments designed to study the π and K mesons π–
particle. The Ω ,Ξ ,Λ
– 0 0 +
e
revealed other particles more massive than protons and and p + are all members
neutrons that eventually decayed into protons and of the proton family, the γ e
+

neutrons. It became clear that the proton and neutron K’s and π ’s are mesons, K
+
e–
the γ ’s are photons and
Λ0
were just the lightest members of a family of proton like
particles. The number of particles in the proton family the e– and e+ are γ

was approaching 100 by 1960. During this time it was electrons and positrons. K 0
Ξ0
Here we see two π–
also found that the π and K mesons were just the examples of the creation
Ω
–

lightest members of another family of particles whose of an electron-positron K

–

number exceeded 100 by 1960. It was rather mind pair by a photon.

Int-24 An Overview of Physics

QUARKS proton family or a quark and an antiquark to create a

The mess seen in 1960 was cleaned up, brought into meson. The fast elementary particle reactions were the
focus, primarily by the work of Murray Gell-Mann. In result of the rearrangement of the quarks within the
1961 Gell-Mann and Yval Neuman found a scheme particle, while the slow reactions resulted when the
that allowed one to see symmetric patterns in the weak interaction changed one kind of quark into an-
masses and charges of the various particles. In 1964 other.
Gell-Mann and George Zweig discovered what they A peculiar feature of the quark model is that quarks
thought was the reason for the symmetries. The have a fractional charge. In all studies of all elemen-
symmetries would be the natural result if the proton and tary particles, charge was observed to come in units of
meson families of particles were made up of smaller the amount of charge on the electron. The electron had
particles which Gell-Mann called quarks. (–1) units, and the neutron (0) units. All of the more
Initially Gell-Mann proposed that there were three than 100 “elementary” particles had either +1, 0, or –1
different kinds of quark, but the number has since units of change. Yet in the quark model, quarks had a
grown to six. The lightest pair of the quarks, the so charge of either (+2/3) units like the up quark or (-1/3)
called up quark and down quark are found in protons units like the down quark. (The anti particles have the
and neutrons. If the names “up quark” and “down opposite charge, -2/3 and +1/3 units respectively.) You
quark” seem a bit peculiar, they are not nearly as can see that a proton with two up and one down quark
confusing as the names strange quark, charm quark, has a total charge of (+2/3 +2/3 -1/3) = (+1) units, and
bottom quark and top quark given the other four the neutron with two down and one up quark has a total
members of the quark family. It is too bad that the charge (-1/3 -1/3 +2/3) = (0) units.
Greek letters had been used up naming other particles. The fact that no one had ever detected an individual
In the quark model, all members of the proton family quark, or ever seen a particle with a fractional charge,
consist of three quarks. The proton and neutron, are made the quark model hard to accept at first. When
made from the up and down quarks. The proton Gell-Mann initially proposed the model in 1963, he
consists of two up and one down quark, while the presented it as a mathematical construct to explain the
neutron is made from one up and two down quarks. The symmetries he had earlier observed.
weak interaction, which as we saw can change protons The quark model gained acceptance in the early 1970s
into neutrons, does so by changing one of the proton’s when electrons at the Stanford high energy accelerator
up quarks into a down quark. were used to probe the structure of the proton. This
The π meson type of particles, which were thought to machine had enough energy, could look in sufficient
be Yukawa’s nuclear force particles, turned out instead detail to detect the three quarks inside. The quarks were
to be quark-antiquark pairs. The profusion of what real.
were thought to be elementary particles in 1960 re- In 1995, the last and heaviest of the six quarks, the top
sulted from the fact that there are many ways to quark, was finally detected at the Fermi Lab Accelera-
combine three quarks to produce members of the tor. The top quark was difficult to detect because it is
185 times as massive as a proton. A very high energy
accelerator was needed to create and observe this
massive particle.
 Int-25

With the quark model, our view of matter has become Our picture of the four basic interactions has also
relatively simple again: there are two families of par- become clearer since the early 1930s. The biggest
ticles called quarks and leptons. Each family contains change is in our view of the nuclear force. The basic
six particles. It is not a coincidence that there are the nuclear force is now seen to be the force between
same number of particles in each family. In the current quarks that holds them together to form protons, neu-
theory of matter called the standard model, each pair trons and other particles. What we used to call the
of leptons is intimately connected to a pair of quarks. nuclear force, that short range force binding protons
The electron type leptons are associated with the down and neutrons together in a nucleus, is now seen as a
and up quarks, the muon and muon type neutrino with residual effect of the force between quarks. The old
the strange and charm quarks, and the tau and tau type nuclear force is analogous to the residual electric force
neutrino with the bottom and top quarks. that binds complete atoms together to form molecules.
Are there more than six quarks and six leptons? Are As the electric interaction is caused by a particle, the
there still heavier lepton neutrino pairs associated with photon, the nuclear force is also caused by particles,
still heavier quarks? That the answer is no, that six is eight different ones called gluons. The nuclear force is
the limit, first came not from accelerator experiments, much more complex than the electric force because
but from studies of the early universe. Here we have a gluons not only interact with quarks, they also interact
question concerning the behavior of matter on the very with themselves. This gives rise to a very strange force
smallest of scales of distance, at the level of quarks between quarks. Other forces get weaker as you
inside proton like particles, and we find the answer by separate the interacting particles. The nuclear force
looking at matter on the very largest of scales, the entire between quarks gets stronger! As a result quarks are
universe. The existence of more than six leptons and confined to live inside particles like protons, neutrons
quarks would have altered the relative abundance of and mesons. This is why we have still never seen an
hydrogen, deuterium, and helium remaining after the individual quark or an isolated particle with a fractional
big bang. It would have led to an abundance that is not charge.
consistent with what we see now. Later experiments
with particle accelerators confirmed the results we first
learned from the early universe.

Figure 28-29 Figure 28-28

Fermi Lab accelerator where the Fermi Lab accelerator magnets.
top quark was first observed.
Int-26 An Overview of Physics

THE ELECTROWEAK THEORY cancelled superconducting supercollider was supposed

Another major advance in our understanding of the to allow us to study the behavior of matter at these
nature of the basic interactions came in 1964 when temperatures.
Steven Weinberg, Abdus Salam and Sheldon Glashow One of the major predictions of the electroweak theory
discovered a basic connection between the electric and was that after the electric and weak interactions had
weak interactions. Einstein had spent the latter part of separated, electric forces should be caused by zero rest
his life trying without success to unify, find a common mass photons and the weak interaction should be
basis for, the electric and the gravitational force. It caused by three rather massive particles given the
came somewhat as a surprise that the electric and weak names W + , W – and Z 0 mesons. These mesons were
interactions, which appear so different, had common found, at their predicted mass, in a series of experi-
origins. Their theory of the two forces is known as the ments performed at CERN in the late 1970s.
electroweak theory.
We have discussed Yukawa’s meson theory of forces,
In the electroweak theory, if we heat matter to a a theory in which the range of the force is related to the
temperature higher than 1000 billion degrees, we will rest mass of the particle responsible for the force. As
find that the electric and weak interaction are a single it turns out, Yukawa’s theory does not work for nuclear
force. If we then let the matter cool, this single forces for which it was designed. The gluons have zero
electroweak force splits into the two separate forces, rest mass but because of their interaction, gives rise to
the electric interaction and the weak interaction. This a force unlike any other. What Yukawa’s theory does
splitting of the forces is viewed as a so called phase describe fairly well is the weak interaction. The very
transition, a transition in the state of matter like the one short range of the weak interaction is a consequence of
we see when water turns to ice at a temperature of 0°C. the large masses of the weak interaction mesons W + ,
The temperature of the phase transition for the W – and Z 0 . (The W mesons are 10 times as massive
electroweak force sounds impossibly hot, but it is as a proton, the Z 0 is 11 times as massive.)
attainable if we build a big enough accelerator. The

Figure 28-30
Paths for the large particle accelerators at CERN. The Geneva airport is in the foreground.
 Int-27

THE EARLY UNIVERSE In the next 10 millionths of a second the universe

In the reverse motion picture of the expanding uni- expands and cools to a point where the photons and
verse, the universe becomes smaller and smaller and mesons no longer have enough energy to recreate the
hotter as we approach the big bang that created it. How rapidly annihilating proton and neutron pairs. Soon the
small and how hot are questions we are still studying. protons and neutrons and their antiparticles will have
But it now seems that with reasonable confidence we essentially disappeared from the universe.
can apply the laws of physics to a universe that is about Matter particles survive
one nanosecond old and at a temperature of three The protons and neutrons will have almost disappeared
hundred thousand billion degrees. This is the tempera- but not quite. For some reason, not yet completely
ture of the electroweak transition where the weak and understood, there was a tiny excess of protons
electric interactions become separate distinct forces. over antiprotons and neutrons over
We have some confidence in our knowledge of the antineutrons. The estimate is that there were
behavior of matter at this temperature because this 100,000,000,001 matter particles for every
temperature is being approached in the largest of the 100,000,000,000 antimatter particles. It was
particle accelerators. the tiny excess of matter over antimatter that survived
the proton and neutron annihilation.
3 ×1014 degrees
At three hundred thousand billion degrees the only 3 ×1010 degrees
structures that survive the energetic thermal collisions After this annihilation, nothing much happens until the
are the elementary particles themselves. At this time universe approaches the age of a tenth of a second and
the universe consists of a soup of quarks and anti the temperature has dropped to 30 billion degrees.
quarks, leptons and anti leptons, gravitons and gluons. During this time the particles we see are photons,
Photons and the weak interaction mesons W + , W – neutrinos and antineutrinos and electrons and positrons.
and Z ° are just emerging from the particle that gave These particles exist in roughly equal numbers. The
rise to the electroweak force. The situation may not electron-positron pairs are rapidly annihilating to pro-
actually be that simple. When we get to that tempera- duce photons, but the photons are equally rapidly
ture we may find some of the exotic elementary par- creating electron positron pairs.
ticles suggested by some recent attempts at a quantum
theory of gravity. 38% neutrons
There are still the relatively few protons and neutrons
13
10 degrees that survived the earlier annihilation. The weak inter-
When the universe reaches the ripe old age of a action allows the protons to turn into neutrons and vice
millionth of a second, the time it takes light to travel versa, with the result there are roughly equal numbers
1000 feet, the temperature has dropped to 10 thousand of protons and neutrons. The numbers are not quite
billion degrees. At these temperatures the gluons are equal, however, because at those temperatures there is
able to hold the quarks together to form protons, a slightly greater chance for the heavier neutron to
neutrons, mesons, and their anti particles. It is still decay into a lighter proton than vice versa. It is
much too hot, however, for protons and neutrons to estimated that the ratio of neutrons to protons has
stick together to form nuclei. dropped to 38% by the time the universe is .11 seconds
old. The temperature is still too high for protons and
When we look closely at the soup of particles at 10 neutrons to combine to form nuclei.
thousand billion degrees, there is activity in the form of
the annihilation and creation of particle-antiparticle
pairs. Proton-antiproton pairs, for example, are rapidly
annihilating, turning into photons and mesons. But just
as rapidly photons and mesons are creating proton-
antiproton pairs.
Int-28 An Overview of Physics

Neutrinos escape at one second Positrons annihilated

As we noted, neutrinos are special particles in that the After about three minutes the positrons are gone and
only way they interact with matter is through the weak from then on the universe consists of photons, neutri-
interaction. Neutrinos pass right through the earth with nos, anti neutrinos and the few matter particles. The
only the slightest chance of being stopped. But the neutrinos are not interacting with anything, and the
early universe is so dense that the neutrinos interact matter particles are outnumbered by photons in a ration
readily with all the other particles. of 100,000,000,000 to one. The photons essentially
dominate the universe.
When the universe reaches an age of about one second,
the expansion has reduced the density of matter to the Deuterium bottleneck
point that neutrinos can pass undisturbed through At the time of 13.8 seconds the temperature was 3
matter. We can think of the neutrinos as decoupling billion degrees, cool enough for helium nuclei to
from matter and going on their own independent way. survive. But helium nuclei cannot be made without
From a time of one second on, the only thing that will first making deuterium, and deuterium is not stable at
happen to the neutrinos is that they will continue to cool that temperature. Thus while there are still neutrons
as the universe expands. At an age of 1 second, the around, protons and neutrons still cannot form nuclei
neutrinos were at a temperature of 10 billion degrees. because of this deuterium bottleneck.
By today they have cooled to only a few degrees above Helium created
absolute zero. This is our prediction, but these cool
When the universe reaches an age of three minutes and
neutrinos are too elusive to have been directly ob-
2 seconds, and the ratio of neutrons to protons has
served.
dropped to 13%, finally deuterium is stable. These
24% neutrons surviving neutrons are quickly swallowed up to form
Some other interesting things are also beginning to deuterium which in turn combine to form the very
happen at the time of 1 second. The photons have stable helium nuclei. Since there are equal numbers of
cooled to a point that they just barely have enough protons and neutrons in a helium nucleus, the 13% of
energy to create electron-positron pairs to replace those neutrons combined with an equal number of protons to
that are rapidly annihilating. The result is that the give 26% by weight of helium nuclei and 74% protons
electrons and positrons are beginning to disappear. At or hydrogen nuclei.
these temperatures it is also more favorable for neu-
trons to turn into protons rather than vice versa, with the By the time the helium nuclei form, the universe has
result that the ratio of neutrons to protons has dropped become too cool to burn the helium to form heavier
to 24%. elements. The creation of the heavier elements has to
wait until stars begin to form one third of a million years
3 ×109 degrees (13.8 seconds) later.
When the temperature of the universe has dropped to 3 The formation of elements inside of stars was the basis
billion degrees, at the time of 13.8 seconds, the energy of the continuous creation theory. As we mentioned,
of the photons has dropped below the threshold of one could explain the abundance of all the elements
being able to create electron-positron pairs and the except helium as being a by product of the evolution of
electrons and the positrons begin to vanish from the stars. To explain the helium abundance it was neces-
universe. There was the same tiny excess of electrons sary to abandon his continuous creation theory and
over positrons as there had been of protons over accept that there might have been a big bang after all.
antiprotons. Only the excess of electrons will survive.
 Int-29

The Thermal Photons Think about what it means that the universe became
After the electron positron pairs had vanished, what is transparent at an age of .7 million years. In our
left in the universe are the photons, neutrinos, anti telescopes, as we look at more and more distant galax-
neutrinos, and the few matter particles consisting of ies, the light from these galaxies must have taken more
protons, helium nuclei and a trace of deuterium and and more time to reach us. As we look farther out we
lithium. There are enough electrons to balance the are looking farther back in time. With the Hubble
charge on the hydrogen and helium nuclei, but the telescope we are now looking at galaxies formed when
photons are energetic enough to break up any atoms the universe was less than a billion years old, less than
that might try to form. The neutrinos have stopped 10% of its current age.
interacting with anything and the matter particles are
outnumbered by photons in a ratio of 100 billion to one. Imagine that you could build a telescope even more
At this time the photons dominate the universe. powerful than the Hubble, one that was able to see as
far out, as far back to when the universe was .7 million
One way to understand why the universe cools as it years old. If you could look that far out what would you
expands is to picture the expansion of the universe as see? You would be staring into a wall of heated opaque
stretching the wavelength of the photons. Since the hydrogen. You would see this wall in every direction
energy of a photon is related to its wavelength (the you looked. If you tried to see through the wall, you
longer the wavelength the lower the energy), this would be trying to look at the universe at earlier, hotter
stretching of wavelengths lowers the photon energies. times. It would be as futile as trying to look inside the
Because the photons dominate the young universe, sun with a telescope.
when the photons lose energy and cool down, so does
everything else that the photons are interacting with. Although this wall at .7 million light years consists of
essentially the same heated hydrogen as the surface of
.7 million years the sun, looking at it would not be the same as looking
Until the universe reaches the age of nearly a million at the sun. The light from this wall has been traveling
years, the photons are knocking the matter particles toward us for the last 14 billion years, during which
around, preventing them from forming whole atoms or time the expansion of the universe has stretched the
gravitational structures like stars. But at the age of .7 wavelength and cooled the photons to a temperature of
million years the temperature has dropped to 3000 less than 3 degrees, to a temperature of 2.74 degrees
degrees, and something very special happens at that above absolute zero to be precise.
point. The matter particles are mostly hydrogen, and if
you cool hydrogen below 3000 degrees it becomes Photons at a temperature of 2.74 degrees can be ob-
transparent. The transition in going from above 3000 served, not by optical telescopes but by radio antennas
degrees to below, is like going from inside the surface instead. In 1964 the engineers Arno Penzias and
of the sun to outside. We go from an opaque, glowing Robert Wilson were working with the radio antenna
universe to a transparent one. that was communicating with the Telstar satellite. The
satellite was a large aluminized balloon that was sup-
Transparent universe posed to reflect radio signals back to earth. The radio
When the universe becomes transparent, the photons antenna had to be very sensitive to pick up the weak
no longer have any effect on the matter particles and the reflected signals.
matter can begin to form atoms, stars, and galaxies.
Everything we see today, except for the primordial
hydrogen and helium, was formed after the universe
became transparent.
Int-30 An Overview of Physics

In checking out the antenna, Penzias and Wilson were This photograph shows an extremely uniform uni-
troubled by a faint noise that they could not eliminate. verse. The smoothness shows us that stars and galaxies
Further study showed that the noise was characteristic had not yet begun to form. In fact the universe was so
of a thermal bath of photons whose temperature was smooth that it is difficult to explain how galaxies did
around 3 degrees. After hearing a seminar on the theory form in the time between when the universe went
of the big bang and on the possibility that there might transparent and when we see galaxies in the most
be some light remaining from the explosion, Penzias distant Hubble telescope photographs. The COBE
and Wilson immediately realized that the noise in their (Cosmic Background Explorer) satellite was able to
antenna was that light. Their antenna in effect was detect tiny fluctuations in the temperatures of the
looking at light from the time the universe became background radiation, indicating that there was per-
transparent. At that time, only a few astronomers and haps just enough structure in the early hot universe to
physicists were taking the big bang hypothesis seri- give us the stars, galaxies and clustering of galaxies we
ously. The idea of the universe beginning in an see today.
explosion seemed too preposterous. After Penzias and
Wilson saw the light left over from the hotter universe, One of the questions you may have had reading our
no other view has been acceptable. discussion of the early universe, is how do we know
that the photons, and earlier the particle-anti particle
The fact that the universe became transparent at an age pairs outnumbered the matter particles by a ratio of 100
of .7 million years, means that the photons, now called billion to one? How do we estimate the tiny excess of
the cosmic background radiation, travelled undis- matter over anti matter that left behind all the matter we
turbed by matter. By studying these photons carefully, see today? The answer is that the thermal photons we
which we are now doing in various rocket and satellite see today outnumber protons and neutrons by a factor
experiments, we are in a sense, taking an accurate of 100 billion to one and that ratio should not have
photograph of the universe when the universe was .7 changed since the universe was a few minutes old.
million years old.
We also mentioned that it would be futile to try to look
under the surface of the sun using a telescope. That is
true if we try to use a photon telescope. However we
can, in effect, see to the very core of the sun using
neutrinos. In the burning of hydrogen to form helium,
for each helium nucleus created, two protons are con-
verted to neutrons via the weak interaction. In the
process two neutrinos are emitted. As a result the core
of the sun is a bright source of neutrinos which we can
detect and study here on earth.

While it would be futile to use photons to see farther

back to when the universe was about .7 million years
old, we should be able to see through that barrier using
neutrinos. The universe became transparent to neutri-
nos at the end of the first second. If we could detect
these neutrinos, we would have a snapshot of the
universe as it looked when it was one second old. Thus
far, we have not found a way to detect these cosmic
Figure 34-11 background neutrinos.
Penzias and Wilson, and the Holmdel radio telescope.
Chapter 1
Principle of Relativity
CHAPTER 1 PRINCIPLE OF RELATIVITY

The subject of this book is the behavior of matter—the approach and begin with the development of Newton’s
particles that make up matter, the interactions between theory of mechanics. You do not need a very sophisti-
particles, and the structures that these interactions cated picture of space and time to understand Newto-
create. There is a wondrous variety of activity, as nian mechanics, and this theory explains an enormous
patterns and structures form and dissipate, and all of range of phenomena, more than you can learn in one
this activity takes place in an arena we call space and or several years. There are three main reasons why we
time. The subject of this chapter is that arena —space will not start off with the Newtonian picture. The first
and time itself. is that the simple Newtonian view of space and time is
approximate, and the approximation fails badly in
Initially, one might think that a chapter on space and many examples we will discuss in this text. By starting
time would either be extraordinarily dull, or too eso- with a more accurate picture of space and time, we can
teric to be of any use. From the it’s too dull point of view these examples as successful predictions rather
view, distance is measured by meter sticks, and there than failures of the Newtonian theory.
are relationships like the Pythagorean theorem and
various geometric and trigonometric rules already The second reason is that the more accurate picture of
familiar to the reader. Time appears to be less chal- space and time is based on the simplest, yet perhaps
lenging—it is measured by clocks and seems to march most general law in all of physics—the principle of
inexorably forward. relativity. The principle of relativity not only underlies
all basic theories of physics, it was essential in the
On the too esoteric side are the theories like Einstein’s discovery of many of these theories. Of all possible
General Theory of Relativity which treats gravity as a ways matter could behave, only a very, very few are
distortion of space and time, the Feynman-Wheeler consistent with the principle of relativity, and by con-
picture of antimatter as being matter traveling back- centrating on these few we have been able to make
ward in time, and recent “super symmetry” theories enormous strides in understanding how matter inter-
which assume a ten dimensional space. All of these acts. By beginning the text with the principle of
theories are interesting, and we will briefly discuss relativity, the reader starts off with one of the best
them. We will do that later in the text after we have built examples of a fundamental physical law.
up enough of a background to understand why these
theories were put forth. Our third reason for starting with the principle of
relativity and the nature of space and time, is that it is
What can we say in an introductory chapter about fun. The math required is simple – only the Pythagorean
space and time that is interesting, or useful, or neces- theorem. Yet results like clocks running slow, lengths
sary for a physics text? Why not follow the traditional contracting, the existence of an ultimate speed, and
1-2 Principle of Relativity

questions of causality, are stimulating topics. Many of THE PRINCIPLE OF RELATIVITY

these results are counter intuitive. Your effort will not In this age of jet travel, the principle of relativity is not
be in struggling with mathematical formulas, but in a strange concept. It says that you cannot feel motion
visualizing yourself in new and strange situations. This in a straight line at constant speed. Recall a smooth
visualization starts off slowly, but you will get used to flight where the jet you were in was traveling at perhaps
it and become quite good at it. By the end of the course 500 miles per hour. A moving picture is being shown
the principle of relativity, and the consequences known and all the window shades are closed. As you watch the
as Einstein’s special theory of relativity will be second movie are you aware of the motion of the jet? Do you
nature to you. feel the jet hurtling through the air at 500 miles per
hour? Does everything inside the jet crash to the rear
of the plane because of this immense speed?

No—the only exciting thing going on is the movie. The

smooth motion of the jet causes no excitement whatso-
ever. If you spill a diet Coke, it lands in your lap just as
it would if the plane were sitting on the ground. The
problem with walking around the plane is the food and
drink cart blocking the aisles, not the motion of the
plane. Because the window shades are closed, you
cannot even be sure that the plane is moving. If you
open your window shade and look out, and if it is
daytime and clear, you can look down and see the land
move by. Flying over the Midwestern United States
you will see all those square 40 acre plots of land move
by, and this tells you that you are moving. If someone
suggested to you that maybe the farms were moving
and you were at rest, you would know that was ridicu-
lous, the plane has the jet engines, not the farms.

Despite the dull experience in a jumbo jet, we often are

able to sense motion. There is no problem in feeling
motion when we start, stop, or go around a sharp curve.
But starting, stopping, and going around a curve are not
examples of motion at constant speed in a straight line,
the kind of motion we are talking about. Changes in
speed or in direction of motion are called accelera-
tions, and we can feel accelerations. (Note: In physics
a decrease in speed is referred to as a negative
acceleration.)

Even without accelerations, even when we are moving

at constant speed in a straight line, we can have a strong
sense of motion. Driving down a freeway at 60 miles
per hour in a low-slung, open sports car can be a
notable, if not scary, experience.
 Principle of Relativity 1-3

This sense of motion can be misleading. The first wide For our thought experiment, imagine that we are going
screen moving pictures took the camera along on a to take the Concorde supersonic jet from Boston,
roller coaster ride. Most people in the audience found Massachusetts to San Francisco, California. The jet
watching this ride to be almost as nerve wracking as has been given special permission to fly across the
actually riding a roller coaster. Some even became country at supersonic speeds so that the trip, which is
sick. Yet the audience was just sitting at rest in the scheduled to leave at noon, takes only three hours.
movie theater.
When we arrive in San Francisco we reset our watches
to Pacific Standard Time to make up for the 3 hour
Exercise 1
difference between Boston and San Francisco. We
Throughout this text we will insert various exercises
reset our watches to noon. When we left, it was noon
where we want you to stop and think about or work with
the material. At this point we want you to stop reading
and the sun was overhead. When we arrive it is still
and think about various times you have experienced noon and the sun is still overhead. One might say that
motion. Then eliminate all those that involved accelera- the jet flew fast enough to follow the sun, the 3 hour trip
tions, where you speeded up, slowed down, or went just balancing the 3 hours time difference.
around a curve. What do you have left, and how real
were the sensations? But there is another view of the trip shown in Figure (1).
When we took off at noon, the earth, the airplane and
One of my favorite examples occurred while I was at a sun were lined up as shown in Figure (1a). Three hours
bus station in Boston. A number of busses were lined later the earth, airplane and sun are still lined up as
up side by side waiting for their scheduled departure shown in Figure (1b). The only difference between
times. I recall that after a fairly long wait, I observed that (1a) and (1b) is that the earth has been rotating for three
we were moving past the bus next to us. I was glad that hours so that San Francisco, rather than Boston is now
we were finally leaving. A few seconds later I looked out under the plane. The view in Figure (1) is what an
the window again; the bus next to us had left and we
astronaut approaching the earth in a spacecraft might
were still sitting in the station. I had mistaken that bus’s
see.
motion for our own!

sun sun
A Thought Experiment
Not only can you feel accelerated motion, you can
easily see relative motion. I had no problem seeing the
bus next to us move relative to us. My only difficulty supersonic jet supersonic jet
was in telling whether they were moving or we were
moving. S
Boston Fr an Bo
An example of where it is more obvious who is moving San an sto
ci n
Francisco sc
is the example of the jet flying over the Midwestern ro o
plains. In the daytime the passengers can see the farms ea tatin
rotating rth g
go by; it is easy to detect the relative motion of the plane earth
and the farms. And it is quite obvious that it is the plane
moving and the farms are at rest. Or is it?
a) Supersonic jet over b) Supersonic jet
To deal with this question we will go through what is Boston just after takeoff. three hours later
called a thought experiment where we solve a prob- over San Francisco.
Figure 1
lem by imagining a sometimes contrived situation, and One view of a three hour trip from Boston to San
then figure out what the consequences would be if we Francisco. It is possible, even logical, to think of
were actually in that situation. Galileo is well known the jet as hovering at rest while the earth turns
for his use of thought experiments to explain the underneath.
concepts of the new mechanics he was discovering.
1-4 Principle of Relativity

For someone inside the jet, looking down at the Mid- In the above definition the capsule can use anything you
western farms going by, who is really moving? Are the want as an example—a jet plane, a car, or a room in a
farms really at rest and the plane moving? Or is the plane building. Generally, think of it as a sealed capsule like
at rest and the farms going by? Figure (1) suggests that the jet plane where the moving picture is being shown
the latter point of view may be more accurate, at least and all of the window shades are shut. Of course you can
from the perspective of one who sees the bigger picture look outside, and you may see things going by. But, as
including the earth, airplane, and sun. shown in Figure (1), seeing things outside go by does not
prove that you and the capsule are moving. That cannot
But, you might ask, what about the jet engines and all the be used as evidence of your own uniform motion.
fuel that is being expended to move the jet at 1000 miles/
hour? Doesn’t that prove that it is the jet that is moving? Think about what kind of experiments you might per-
Not necessarily. When the earth rotates, it drags the form in the sealed capsule to detect your uniform mo-
atmosphere around with it creating a 1000 mi hr wind tion. One experiment is to drop a coin on the floor. If you
that the plane has to fly through in order to stand still. are at rest, the coin falls straight down. But if you are in
Without the jet engines and fuel, the plane would be a jet travelling 500 miles per hours and the flight is
dragged back with the land and never reach San Fran- smooth, and you drop a coin, the coin still falls straight
cisco. down. Dropping a coin does not distinguish between
being at rest or moving at 500 miles per hour; this is one
This thought experiment has one purpose. To loosen
experiment that does not violate the principle of relativ-
what may have been a firmly held conviction that when
ity.
you are in a plane or car, you are moving and the land that
you see go by must necessarily be at rest. Perhaps, under There are many other experiments you can perform.
some circumstances it is more logical to think of your- You could use gyroscopes, electronic circuits, nuclear
self at rest and the ground as moving. Or, perhaps it reactions, gravitational wave detectors, anything you
does not make any difference. The principle of relativ- want. The principle of relativity states that none of these
ity allows us to take this last point of view. will allow you to detect your uniform motion.

Statement of the Exercise 2

Principle of Relativity Think about what you might put inside the capsule and
Earlier we defined uniform motion as motion at constant what experiments you might perform to detect the
speed in a straight line. And we mentioned that the motion of the capsule. Discuss your ideas with others
principle of relativity said that you could not feel this and see if you can come up with some way of violating
uniform motion. Since it is not exactly clear what is the principle of relativity.
meant by “feeling” uniform motion, a more precise
statement of the principle of relativity is needed, a
statement that can be tested by experiment. The follow- Basic Law of Physics
ing is the definition we will use in this text. We mentioned that one of the incentives for beginning
the text with the principle of relativity is that it is an
excellent example of a basic law of physics. It is simple
Imagine that you are in a capsule and
and easy to state—there is no experiment that you can
you may have any equipment you
perform that allows you to detect your own uniform
wish inside the capsule. The principle
motion. Yet it is general—there is no experiment that
of relativity states that there is no
can be done at any time, at any place, using anything,
experiment you can perform that will
that can detect your uniform motion. And most impor-
allow you to tell whether or not the
tant, it is completely subject to experimental test on an
capsule is moving with uniform mo-
all-or-nothing basis. Just one verifiable experiment
tion—motion in a straight line at con-
detecting one’s own uniform motion, and the principle
stant speed.
of relativity is no longer a basic law. It may become a
useful approximation, but not a basic law.
 Principle of Relativity 1-5

Once a fundamental law like the principle of relativity The clue was that the speed c of this wave was so fast
is discovered or accepted, it has a profound effect on the that only light had a comparable speed. And more
way we think about things. In this case, if there is no way remarkably the known speed of light, and the speed c
that we can detect our own uniform motion, then we of his wave were very close—to within experimental
might as well ignore our motion and always assume that error they were equal. As a consequence Maxwell
we are at rest. Nature is usually easier to explain if we proposed that he had discovered the theory of light, and
take the point of view that we are at rest and that other that this wave of electric and magnetic force was light
people and things are moving by. It is the principle of itself.
relativity that allows us to take this self-centered point of
view. Maxwell’s theory explained properties of light such as
polarization, and made predictions like the existence of
It is a shock, a lot of excitement is generated, when what radio waves. Many predictions were soon verified, and
was accepted as a basic law of physics is discovered not within a few years there was little doubt that Maxwell
to be exactly true. The discovery usually occurs in some had discovered the theory of light.
obscure corner of science where no one thought to look
before. And it will probably have little effect on most One problem with Maxwell’s theory is that by mea-
practical applications. But the failure of a basic law surements of the speed of light, it appears that one
changes the way we think. should be able to detect one’s own uniform motion. In
the next section we shall see why. This had two
Suppose, for example, that it was discovered that the immediate consequences. One was a change in the
principle of relativity did not apply to the decay of an view of nature to make it easy to see who was moving
esoteric elementary particle created only in the gigantic and who was not. The second was a series of experi-
particle accelerating machines physicists have recently ments to see if the earth were moving or not.
built. This violation of the principle of relativity would
have no practical effect on our daily lives, but it would In the resulting view of nature, all of space was filled
have a profound psychological effect. We would then with an invisible substance called ether. Light was
know that our uniform motion could be detected, and pictured as a wave in the ether medium just as ocean
therefore on a fundamental basis we could no longer waves are waves in the medium of water. The experi-
take the point of view that we are at rest and others are ments, initiated by Michaelson and Morley, were de-
moving. There would be legitimate debates as to who signed to detect the motion of the earth by measuring
was moving and who was at rest. We would search for how fast the earth was moving through the ether
a formulation of the laws of physics that made it medium.
intuitively clear who was moving and who was at rest.
The problem with the ether theory was that all experi-
This is almost what happened in 1860. In that year, ments designed to detect ether, or to detect motion
James Clerk Maxwell summarized the laws of electric- through it, seemed to fail. The more clever the experi-
ity and magnetism in four short equations. He then ment, the more subtle the apparent reason for the
solved these equations to predict the existence of a wave failure. We will not engage in any further discussion of
of electric and magnetic force that should travel at a the ether theory, because ether still has never been
speed of approximately 3 × 10 8 meters per second. detected. But we will take a serious look in the next
The predicted speed, which we will call c, could be section at how the measurement of the speed of a pulse
determined from simple measurements of the behavior of light should allow us to detect our own uniform
of an electric circuit. motion. And then in the rest of the chapter we will
discuss how a young physicist, working in a patent
Before Maxwell, no one had considered the possibility office in 1905, handled the problem.
that electric and magnetic forces could combine in a
wavelike structure that could travel through space. The
first question Maxwell had to answer was what this
wave was. Did it really exist? Or was it some spurious
solution of his equations?
1-6 Principle of Relativity

Figure 2 Figure 3
Rain drops creating circular waves on the surface of a This ocean wave traveled hundreds of miles from
puddle. (Courtesy Bill Jack Rodgers, Los Alamos Hurricane Bertha to the Maine coast (July 31, 1990).
Scientific Laboratory.)

WAVE MOTION Although you cannot see them, sound waves are a more
We do not need to know the details of Maxwell’s theory familiar form of wave motion. Sound moving through
to appreciate how one should be able to use the theory air, waves moving over water, and light, all have certain
to violate the principle of relativity. All we need is an common features and ways of behaving which we
understanding of some of the basic properties of wave classify as wave motion. In later chapters we will study
motion. the subject of wave motion in considerable detail. For
now we will limit our discussion to a few of the features
The most familiar examples of wave motion are the we need to understand the impact of Maxwell’s theory.
waves on the surface of water. We have seen the waves
that spread out in circles when a stone is dropped in a Two examples of wave motion that are easy to study are
pond, or rain hits a puddle in a sidewalk as shown in a wave pulse traveling down a rope as indicated in
Figure (2). And most of us have seen the ocean waves Figure (4) or down a stretched Slinky® (the toy coil
destroying themselves as they crash into the beach. that ‘climbs’ down stairs) as shown in Figure (5). The
The larger ocean waves often originate at a storm far advantage of using a stretched Slinky is that the waves
out to sea, and have traveled hundreds or even a travel so slowly that you can study them as they move.
thousand miles to reach you (see Figure (3)). The very It turns out that the speed of a wave pulse depends upon
largest ocean waves, created by earthquakes or explod- the medium along which, or through which, it is
ing volcanos have been known to travel almost around traveling. For example, the speed of a wave pulse along
the earth. a rope or Slinky is given by the formula

Figure 4 Figure 5
Wave pulse traveling along a rope. Wave pulse traveling along a Slinky.
 Wave Motion
1-7

Speed of = τ A substance like air, which is relatively compressible,

(1)
wave pulse µ has a small rigidity B, while substances like steel and
granite are very rigid and have large values of B. As a
where τ is the tension in the rope or Slinky, and µ the result sound travels much faster in steel and granite
mass per unit length. Do not worry about precise than in air. For air at room temperature and one
definitions of tension or mass, the important point is atmosphere of pressure, the speed of sound is 343
that there is a formula for the speed of the wave pulse, meters or 1125 feet per second. Sound travels about 20
a formula that depends only on the properties of the times faster in steel and granite. Again the important
medium along which the pulse is moving. point is that the speed of a wave depends on the
properties of the medium through which it is moving,
The speed does not depend upon the shape of the pulse
and not on the shape of the wave or the way it was
or how the pulse was created. For example, the Slinky
produced.
pulse travels much more slowly than the pulse on the
rope because the suspended Slinky has very little
tension τ. We can slow the Slinky wave down even Measurement of the Speed of Waves
more by hanging crumpled pieces of lead on each end If you want to know how fast your car is traveling you
of the coils of the Slinky to increase its mass per unit look at the speedometer. Some unknown machinery in
length µ. the car makes the needle of the speedometer point at the
correct speed. Since the wave pulses we are discussing
Another kind of wave we can create in the Slinky is the do not have speedometers, we have to carry out a series
so called compressional wave shown in Figure (6). of measurements in order to determine their speed. In
Here the end of the Slinky was pulled back and this section we wish to discuss precisely how the
released, giving a moving pulse of compressed coils. measurements can be made using meter sticks and
The formula for the speed of the compression wave is clocks so that there will be no ambiguity, no doubt
still given by Equation (1), if we interpret τ as the about precisely what we mean when we talk about the
stiffness (Youngs modulus) of the suspended Slinky. speed of a wave pulse. We will use the Slinky wave
If we use a loudspeaker to produce a compressional pulse as our example, because the wave travels slowly
pulse in air, we get a sound wave that travels out from enough to actually carry out these measurements in a
the loudspeaker at the speed of sound. The formula for classroom demonstration.
the speed of a sound wave is The first experiment, shown in Figure (7), involves two
students and the instructor. One student stands at the
Speed of = B (2) end of the stretched Slinky and releases a wave pulse
sound ρ
like that shown in Figure (6). The instructor holds a
where B is the bulk modulus which can be thought of meter stick up beside the Slinky as shown. The other
as the rigidness of the material, and the mass per unit student has a stopwatch and measures the length of
length µ is replaced by the mass per unit volume ρ. time it takes the pulse to travel from the front to the back
of the stick. (She presses the button once when the
pulse reaches the front of the meter stick, presses it
again when the pulse gets to the back, and reads the
elapsed time T.) The speed of the pulse is then defined
to be

Speed of 1 meter
= (3)
Figure 6 Slinky pulse T seconds
To create a compressional wave on a suspended
Slinky, pull the end back a bit and let go.
1-8 Principle of Relativity

this the time T1. This time T1 is less than T because Bill
and the meter stick are moving toward the pulse.

To Bill, the pulse passes his one meter long stick in a

time T1, therefore the speed of the pulse past him is

1 meter speed of pulse

v1 = = (4a)
T1 seconds relative to Bill
Bill should also have carried the stopwatch so that v1
would truly represent his measurement of the speed of
the pulse. But it is too awkward to hold the meter stick,
and run and observe when the pulse is passing the ends
of the stick.

The speed v1 measured by Bill is not the same as the

Figure 7 speed v measured by the instructor in Figure (7). v1 is
Experiment to measure the speed of a wave greater than v because Bill is moving toward the wave
pulse on a suspended Slinky. Here the pulse. This is not surprising: if you are on a freeway
instructor holds the meter stick at rest.
and everyone is traveling at a speed v = 55 miles per
hour, the oncoming traffic in the opposite lane is
Later in the course, when we have discussed ways of traveling past you at a speed of 110 miles per hour
measuring tension τ and mass per unit length µ, we can because you are moving toward them.
compare the experimental result we get from Equation
(3) with the theoretically predicted result of Equation In Figure (8b) we again have the same situation as in
(1). With a little practice using the stopwatch, it is not Figure (7) except that Bill is now replaced by Joan who
difficult to get reasonable agreement between theory is running away from the student who releases the
and experiment . pulse. Joan is moving in the same direction as the pulse
and it takes a longer time T2 for the pulse to pass her.
In our second experiment, shown in Figure (8a) every- (Assume that Joan is not running faster than the pulse.)
thing is the same except that the instructor has been The speed of the pulse relative to Joan is
replaced by a student, let us say it is Bill, holding the
meter stick and running toward the student who re- v2 = 1 meter = speed of pulse (4b)
leases the wave pulse. Again the second student T2 seconds relative to Joan
measures the length of time it takes the pulse to travel Joan’s speed v2 will be considerably less than the speed
from the front to the back of the meter stick. Let us call v observed by the instructor.

Figure 8a Figure 8b
Bill runs toward the source of the Joan runs away from the source of
pulse while measuring its speed. the pulse while measuring its speed.
 Wave Motion 1-9

In these three experiments, the instructor is special get, and then see what the consequences would be if we
(wouldn’t you know it). Only the instructor measures actually got those results.
the speed v predicted by theory, only for the instructor
is the speed given by v = τ/µ . Both the students Bill The apparatus for our new thought experiment is
and Joan observe different speeds, one larger and one shown in Figure (9). We have a laser which can
smaller than the theoretical value. produce a very short pulse of light – only a few
millimeters long. The meter stick now has photo
What is special about the instructor? In this case the detectors and clocks mounted on each end, so that we
instructor gets the predicted answer because she is at can accurately record the times at which the pulse of
rest relative to the Slinky. If we hadn’t seen the light passed each end. These clocks were synchro-
experiment, but just looked at the answers, we could nized, so the time difference is the length of time T it
tell that the instructor was at rest because her result takes the pulse of light to pass the meter stick.
agreed with the predicted speed of a Slinky wave. Bill
got too high a value because he was moving toward the Before the experiment, the instructor gives a short
pulse; Joan, too low a value because she was moving in lecture to the class. She points out that according to
the same direction. Maxwell’s theory of light, a light wave should travel at
a speed c given by the formula
The above set of experiments is not strikingly pro-
found. In a sense, we have developed a new and rather c = 1
µ0 ε0 (5)
cumbersome way to tell who is not moving relative to
the Slinky. But the same procedures can be applied to
where µ0 and εo are constants in the theory of electric-
a series of experiments that gives more interesting
ity. She says that later on in the year, the students will
results. In the new series of experiments, we will use
perform an experiment in which they measure the
a pulse of light rather than a wave pulse on a Slinky.
value of the product µ0 ε0. This experiment involves
Since the equipment is not likely to be available among measuring the size of coils of wire and plates of
the standard set of demonstration apparatus, and since aluminum, and timing the oscillation of an electric
it will be difficult to run at speeds comparable to the current sloshing back and forth between the plates and
speed of light we will do this as a thought experiment. the coil. The important point is that these measure-
We will imagine that we can measure the time it takes ments do not involve light. It is analogous to the Slinky
a light pulse to go from the front to the back of a meter where the predicted speed τ/µ of a Slinky wave
stick. We will imagine the kind of results we expect to involved measurements of the stiffness τ and mass per
unit length µ, and had nothing to do with observations
of a Slinky wave pulse.
plates
photo detectors electric
laser pulse
with clocks current

laser
meter stick
coil
Figure 9
Apparatus for the thought experiment. Now we wish Figure 10a Figure 10b
to measure the speed of a laser wave pulse, rather Plates and coil for The plates and coil we
than the speed of a Slinky wave pulse. The photo measuring the use in the laboratory.
detectors are used to measure the length of time the experimental value
laser pulse takes to pass by the meter stick. of µ0 ε0.
1-10 Principle of Relativity

laser pulse
laser c = 3 × 10 8 meter 9 feet
sec = 10 sec
(8)
c = 1 –9
foot
10 sec

foot
c = 1
nanosecond

She says that because this is such an easy number to

remember, she will use it throughout the rest of the
course.

The lecture on Maxwell’s theory being over, the in-

Figure 11
structor starts in on the thought experiment. In the first
Experiment to measure the speed of a light
wave pulse from a laser. Here the run she stands still, holding the meter stick, and the
instructor holds the meter stick at rest. student with the laser emits a pulse of light as shown in
Figure (11). The pulse passes the 3.28 foot length of the
Although she is giving out the answer to the lab meter stick in an elapsed time of 3.28 nanoseconds, for
experiment, she points out that the value of c from these a measured speed
measurements is
1 v light pulse = 3.28 feet
c = = 3 × 108 meters/second (6) 3.28 nanoseconds
µ 0 ε0
= 1 foot (9)
which is a well-known but uncomfortably large and nanosecond
hard to remember number. However, she points out,
The teacher notes, with a bit of complacency, that she
3 × 108 meters is almost exactly one billion (109) feet. got the predicted speed of 1 foot/nanosecond. Again,
If you measure time, not in seconds, but in billionths of
the instructor is special.
a second, or nanoseconds, where
Then the instructor invites Bill to hold the meter stick
1 nanosecond ≡ 10-9 seconds (7)
and run toward the laser as shown in Figure (12a).
then since light travels only one foot in a nanosecond, Since this is a thought experiment, she asks Bill to run
the speed of light is simply at nearly the speed of light, so that the time should be
cut in half and Bill should see light pass him at nearly
a speed of 2c.
laser pulse laser pulse
laser laser

Figure 12a Figure 12b

Bill runs toward the source of the Joan runs away from the source of
pulse while measuring its speed. the pulse while measuring its speed.
 Wave Motion 1-11

Then she invites Joan to hold the meter stick and run at Michaelson-Morley Experiment
about half the speed of light in the other direction as The period from 1860 to 1905 saw a number of
shown in Figure (12b). One would expect that the light attempts to detect one’s own or the earth’s motion
would take twice as long to pass Joan as it did the through space by measuring the speed of pulses of
instructor and that Joan should obtain a value of about light. Actually it was easier and far more accurate to
c/2 for the speed of light. compare the speeds of light traveling in different
directions. If you were moving forward through space
Suppose it turned out this way. Suppose that the (like Bill in our thought experiment), you should see
instructor got the predicted answer 1 foot/nanosecond, light coming from in front of you traveling faster than
while Bill who is running toward the pulse got a higher light from behind or even from the side.
value and Joan, running with the pulse got a lower value.
Just as in our Slinky pulse experiment we could say that Michaelson and Morley used a device called a
the instructor was at rest while both Bill and Joan were Michaelson interferometer which compared the speeds
moving. of pulses of light traveling at right angles to each other.
A detailed analysis of their device is not hard, just a bit
But, moving relative to what? In the Slinky experiment, lengthy. But the result was that the device should be
the instructor was at rest relative to the Slinky – the able to detect small differences in speeds, small enough
medium through which a Slinky wave moves. Light differences so that the motion of the earth through
pulses travel through empty space. Light comes to us space should be observable -- even the motion caused
from stars 10 billion light years away, almost across the by the earth orbiting the sun.
entire universe. The medium through which light
moves is empty space. At this point we can summarize volumes of the history
of science by pointing out that no experiment using the
If the experiment came out the way we described, the Michaelson interferometer, or any device based on
instructor would have determined that she was at rest measuring or comparing the speed of light pulses, ever
relative to empty space, while Bill and Joan would have succeeded in detecting the motion of the earth.
determined that they were moving. They would have
violated the principle of relativity, which says that you
cannot detect your own motion relative to empty space. Exercise 3
Units of time we will often use in this course are the
The alert student might argue that the pulses of light millisecond, the microsecond, and the nanosecond,
come out of the laser like bullets from a gun at a definite where
muzzle velocity, and that all the instructor, Bill and Joan 1 millisecond = 10– 3 seconds (one thousandth)
are doing is measuring their speed relative to the laser. 1 microsecond = 10– 6 seconds (one millionth)
Experiments have carefully demonstrated that the speed 1 nanosecond = 10– 9 seconds (one billionth)
of a pulse of light depends in no way on the motion of
the emitter just as the Slinky pulse depended in no way How many feet does light travel in
on how the student started the pulse. Maxwell’s theory a) one millisecond (1ms)?
predicts that light is a wave, and many experiments have b) one microsecond (1µs )?
c) one nanosecond (1ns )?
verified the wave nature of light, including the fact that
its speed does not depend on how it was emitted.

From the logical simplicity of the above thought experi-

ment, from the ease with which we should be able to
violate the principle of relativity (if we could accurately
measure the speed of a pulse of light passing us), it is not
surprising that after Maxwell developed this theory of
light, physicists did not take the principle of relativity
seriously, at least for the next 45 years.
1-12 Principle of Relativity

EINSTEIN’S PRINCIPLE from Maxwell’s theory, they could all safely assume
OF RELATIVITY that they were at rest. At the very least, their measure-
ments of the speed of the light pulse could not be used
In 1905 Albert Einstein provided a new perspective on
to detect their own motion.
the problems we have been discussing. He was appar-
ently unaware of the Michaelson-Morley experiments. As we said, the idea is simple. You always get the
Instead, Einstein was familiar with Maxwell’s equa- answer c whenever you measure the speed of a light
tions for electricity and magnetism, and noted that pulse moving past you. But the idea is horrendous.
these equations had a far simpler form if you took the Einstein went against more than 200 years of physics
point of view that you are at rest. He suggested that and centuries of observation with this suggestion.
these equations took this simple form, not just for some
privileged observer, but for everybody. If the principle Suppose, for example, we heard about a freeway where
of relativity were correct after all, then everyone, no all cars traveled at precisely 55 miles per hour – no
matter how they were moving, could take the point of exceptions. Hearing about this freeway, our three
view that they were at rest and use the simple form of people in the thought experiment decide to test the rule.
Maxwell’s equations. The instructor sets up measuring equipment in the
median strip and observes that the rule is correct. Cars
How did Einstein deal with measurements of the speed in the north bound lane travel north at 55 miles per hour,
of light? We have seen that if someone, like Bill in our and cars in the south bound lane go south at 55 miles per
thought experiment, detects a pulse of light coming at hour.
them at a speed faster than c = 1 foot/nanosecond, then
that person could conclude that they themselves were For his part of the experiment, Bill gets into one of the
moving in the direction from which the light was north bound cars. Since Bill knows about the principle
coming. They would have thereby violated the prin- of relativity he takes the point of view that he is at rest.
ciple of relativity. If the 55 miles per hour speed is truly a fundamental
law, then he, who is at rest, should see the south bound
Einstein’s solution to that problem was simple. He cars pass at 55 miles per hour.
noted that any measurement of the speed of a pulse of
light that gave an answer different from c = 1 foot/ Likewise, Joan, who is in a south bound car, can take
nanosecond could be used to violate the principle of the point of view that she is at rest. She knows that if
relativity. Thus if the principle of relativity were cor- the 55 miles per hour speed limit is a fundamental law,
rect, all measurements of the speed of light must give then north bound cars must pass her at precisely 55
the answer c. miles per hour. If the instructor, Bill and Joan all
observe that every car on the freeway always passes
Let us put this in terms of our thought experiment. them at the same speed of 55 miles per hour, then none
Suppose the instructor observed that the light pulse of them can use this observation to detect their own
passed the 3.24 foot long meter stick in precisely 3.24 motion.
nanoseconds. And suppose that Bill, moving at nearly
the speed of light toward the laser, also observed that Freeways do not work that way. Bill will see south
the light took 3.24 nanoseconds to pass by his meter bound cars passing him at 110 miles per hour. And
stick. And suppose that Joan, moving away from the Joan will see north bound cars passing at 110 miles per
laser at half the speed of light, also observed that the hour. From these observations Bill and Joan will
pulse of light took 3.24 nanoseconds to pass by her conclude that in fact they are moving – at least relative
meter stick. If the instructor, Bill and Joan all got to the freeway.
precisely the same answer for the speed of light, then
none of their results could be used to prove that one was Measurements of the speed of a pulse of light differ,
at rest and the others moving. Since their answer of however, in two significant ways from measurements
3.24 feet in 3.24 nanoseconds or 1 foot/nanosecond is of the speed of a car on a freeway. First of all, light
in agreement with the predicted value c = 1/ µ0 ε0 moves through empty space, not relative to anything.
 Moving Clocks 1-13

Secondly, light moves at enormous speeds, speeds that Moving Clocks

lie completely outside the realm of common experi- Our first thought experiment for Einstein’s special
ence. Perhaps, just perhaps, the rules we have learned relativity will deal with the behavior of clocks. We saw
so well from common experience, do not apply to this that the measurement of the speed of a pulse of light
realm. The great discoveries in physics often came required a timing device, and perhaps the peculiar
when we look in some new realm on the very large results can be explained by the peculiar behavior of the
scale, or the very small scale, or in this case on the scale timing device.
of very large, unfamiliar speeds.
Also the peculiar behavior seems to happen at high
speeds near the speed of light, not down at freeway
The Special Theory of Relativity speeds. Thus the question we would like to ask is what
Einstein developed his special theory of relativity from happens to a clock that is moving at a high speed, near
two assumptions: the speed of light?
1) The principle of relativity is correct. That is a tough question. There are many kinds of
2) Maxwell’s theory of light is correct. clocks, ranging from hour glasses dripping sand, to the
popular digital quartz watches, to the atomic clocks
As we have seen, the only way Maxwell’s theory of used by the National Bureau of Standards. The oldest
light can be correct and not violate the principle of clock, from which we derive our unit of time, is the
relativity, is that every observer who measures the motion of the earth on its axis each 24 hours. We have
speed of light, must get the predicted answer both the problem of deciding which kind of clock we
c = 1/ µ0 ε0 = 1 foot/nanosecond. Temporarily we wish to consider moving at high speeds, and then figure
will use this as the statement of Einstein’s second out how that clock behaves.
postulate:
The secret of working with thought experiments is to
2a) Everyone, no matter how he or she is moving, keep everything as simple as possible and do not try to
must observe that light passes them at do too much at once. If we want to understand what
precisely the speed c. happens to a moving clock, we should start with the
simplest clock we can find. If we cannot understand
Postulates (1) and (2a) salvage both the principle of that one, we will imagine an even simpler one.
relativity and Maxwell’s theory, but what else do they
predict? We have seen that measurements of the speed A clock that is fairly easy to understand is the old
of a pulse of light do not behave in the same way as grandfather’s clock shown in Figure (13), where the
measurements of the speed of cars on a freeway. timing device is the swinging pendulum. There are also
Something peculiar seems to be happening at speeds wheels, gears, and hands, but these merely count
near the speed of light. What are these peculiar things? swings of the pendulum. The pendulum itself is what
How do we find out? is important. If you shorten the pendulum it swings
faster and the hands go around faster.
To determine the consequences of his two postulates,
Einstein borrowed a technique from Galileo and used hands wheels &
a series of thought experiments. Einstein did this so gears
clearly, explained the consequences so well in his 1905
paper, that we will follow essentially the same line of
reasoning. The main difference is that Einstein made swinging
a number of strange predictions that in 1905 were hard pendulum
to believe. But these predictions were not only verified,
they became the cornerstone of much of 20th century
physics. We will be able to cite numerous tests of all the
Figure 13
predictions. Grandfather’s clock.
1-14 Principle of Relativity

We could ask what we would see if we observed a For our thought experiment it is convenient to construct
grandfather's clock moving past us at a high speed, near two identical light pulse clocks as shown in Figure (15).
the speed of light. The answer is likely to be “I don’t We wish to take great care that they are identical, or at
know”. The grandfather’s clock, with its swinging least that they run at precisely the same rate. Once they
pendulum mechanism, is still too complicated. are finished, we adjust them so that the pulses bounce
up and down together for weeks on end.
A simpler timing device was considered by Einstein,
namely a bouncing pulse of light. Suppose, we took the Now we get to the really hypothetical part of our
grandfather’s clock of Figure (13), and replaced the thought experiment. We give one of the clocks to an
pendulum by two mirrors and a pulse of light as shown astronaut, and we keep the other for reference. The
in Figure (14). Space the mirrors 1 foot apart so that the astronaut is instructed to carefully pack his clock,
pulse of light will take precisely one nanosecond to accelerate up to nearly the speed of light, unpack his
bounce either up or down. Leave the rest of the clock, and go by us at a constant speed so that we can
machinery of the grandfather’s clock more or less compare our reference clock to his moving clock.
intact. In other words have the wheels and gears now
count bounces of the pulse of light rather than swings Before we describe what we see, let us take a look at a
of the pendulum. And recalibrate the face of the clock brief summary of the astronaut’s log book of the trip.
so that for each bounce, the hand advances one nano- The astronaut writes, “ I carefully packed the light pulse
second. (The marvelous thing about thought experi- clock because I did not want it damaged during the
ments is that you can get away with this. You do not accelerations. My ship can maintain an acceleration of
have to worry about technical feasibility, only logical 5gs, and even then it took about a month to get up to our
consistency.) final speed of just over half the speed of light.”

The advantage of replacing the pendulum with a bounc- “Once the accelerations were over and I was coasting,
ing light pulse is that, so far, the only thing whose I took the light pulse clock out of its packing and set it
behavior we understand when moving at nearly the up beside the window, so that the class could see the
speed of light is light itself. We know that light always clock as it went by. Before the trip I was worried that
moves at the speed c in all circumstances, to any I might have some trouble getting the light pulse into
observer. If we use a bouncing light pulse as a timing the clock, but it was no problem at all. I couldn't even
device, and can figure out how the pulse behaves, then tell that I was moving! The light pulse went in and the
we can figure out how the clock behaves. clock started ticking just the way it did back in the lab,
before we started the trip.”

“It was not long after I started coasting, that the class
mirror went by. After that, I packed everything up again,
decelerated, and returned to earth.”

pulse of light
1 ft bouncing
between mirrors

mirror

Figure 14
Light pulse clock. We can construct a clock by
having a pulse of light bounce between two mirrors.
If the mirrors are one foot apart, then the time Figure 15
between bounces will be one nanosecond. The face Two identical light pulse clocks.
of the clock displays the number of bounces.
 Moving Clocks 1-15
vastronaut

c c c c
c

our clock astronaut's moving clock

Figure 16
In order to stay in the astronaut’s moving clock, the light pulse must follow a longer, saw-tooth, path.

What we saw as the astronaut went by is illustrated in and the astronaut’s pulse traveling along the saw tooth
the sketch of Figure (16). On the left is our reference path are both traveling at the same speed c. Since the
clock, on the right the astronaut’s clock moving by. astronaut’s pulse travels farther, the astronaut’s clock
You will recall that the astronaut had no difficulty must take longer to tick off a nanosecond. The
getting the light pulse to bounce, and as a result we saw astronaut’s clock must be running slower!
his clock go by with the pulse bouncing inside.
Because there are no budget constraints in a thought
For his pulse to stay in his clock, his pulse had to travel experiment, we are able to get a better understanding of
along the saw-tooth path shown in Figure (16). The how the astronaut’s clock was behaving by having the
saw-tooth path is longer than the up and down path astronaut repeat the trip, this time going faster, about
taken by the pulse in our reference clock. His pulse had .95 c. What we saw is shown in Figure (17). The
to travel farther than our pulse to tick off one nanosec- astronaut’s clock is moving so fast that the saw tooth
ond. path is stretched way out. The astronaut’s pulse takes
a long time to climb from the bottom to the top mirror
Here is what is peculiar. If Einstein’s postulate is right, in his clock, his nanoseconds take a long time, and his
if the speed of a pulse of light is always c under any clock runs very slowly.
circumstances, then our pulse bouncing up and down,

v astronaut

c c

our clock astronaut's moving clock

Figure 17
When the astronaut goes faster, his light pulse has to go farther in order to register a bounce. Since the speed of
light does not change, it takes longer for one bounce to register, and the astronaut's moving clock runs slower.
 1-16 Principle of Relativity

It does not take too much imagination to see that if the distance vT' as shown. This gives us a right triangle
astronaut came by at the speed of light c, the light pulse, whose base is vT' , whose hypotenuse is cT' , and
also traveling at a speed c, would have to go straight whose height, determined from our clock, is cT. Ac-
ahead just to stay in the clock. It would never be able cording to the Pythagorean theorem, these sides are
to get from the bottom to the top mirror, and his clock related by
would never tick off a nanosecond. His clock would
2 2
stop! cT' = vT' + cT 2 (10)
Carrying out the squares, and collecting the terms with
Exercise 4
T' on one side, we get
Discuss what the astronaut should have seen when the
class of students went by. In particular, draw the c2T' 2 - v2 T' 2 = c2 - v2 T' 2 = c 2 T 2
astronaut's version of Figure (16) and describe the
situation from the astronaut's point of view. 2
2 c 2T 2 c 2T 2 × 1/c T2
T' = = =
It is not particularly difficult to calculate the amount by c2 - v2 c 2 - v 2 × 1/c 2 1 - v 2 /c 2
which the astronaut’s clock runs slow. All that is
required is the Pythagorean theorem. In Figure (18), on Taking the square root of both sides gives
the left, we show the path of the light pulse in our
reference clock, and on the right the path in the T
T' = (11)
astronaut’s moving clock. Let T be the length of time 2
1 - v /c 2
it takes our pulse to go from the bottom to the top
mirror, and T' the longer time light takes to travel along Equation (11) gives a precise relationship between the
the diagonal line from his bottom mirror to his top length of our nanosecond T and the astronaut’s longer
mirror. We can think of T as the length of one of our nanosecond T' . We see that the astronaut’s basic time
nanoseconds, and T' as the length of one of the unit T' is longer than our basic time unit T by a factor
astronaut’s longer nanoseconds. 1/ 1 - v2 /c2 .
The distance an object, moving at a speed v, travels in The factor 1/ 1 - v2 /c2 appears in a number of calcu-
a time T, is vT. (If you go 30 miles per hour for 3 hours, lations involving Einstein’s special theory of relativity.
you travel 90 miles.) Thus, in Figure (18), the distance As a result, it is essential to develop an intuitive feeling
our light pulse travels in going from the bottom to the for this number. Let us consider several examples to
top mirror is cT as shown. The astronaut’s light pulse, begin to build this intuition. If v = 0, then
which takes a time T' to travel the diagonal path, must
have gone a distance cT' as shown. T T
T' = =
1 - v 2 /c 2 1-0
During the time T' , while the astronaut’s light pulse is
going along the diagonal path, the astronaut’s clock, T
T' = = T (v = 0) (12)
which is traveling at a speed v, moves forward a 1

time for one

time for one bounce, T'
bounce, T Figure 18
In our clock, the light pulse
c c c travels a distance cT in one
bounce. In the astronaut's
'
cT

cT
clock, the pulse travels a
distance cT' while the clock
moves forward a distance vT'
our clock vT' astronaut's clock during one bounce.
 Moving Clocks 1-17

and we see that a clock at rest keeps the same time as In this case, the astronaut’s seconds would be infinitely
ours. If the astronaut goes by at one tenth the speed of long and the astronaut’s clock would stop. This agrees
light, v = .1 c, and we get with our earlier observation that if the astronaut went
by at the speed of light, the light pulse in his clock would
T T have to go straight ahead just to stay in the clock. It
T' = =
2 1 - .01
1 - .1c /c 2 would not have time to move up or down, and therefore
not be able to tick off any seconds.
T
T' = = 1.005T (v = c/10) (13) So far we have been able to use a pocket calculator to
.99
evaluate 1/ 1 - v2 /c2 . But if the astronaut were flying
In this case the astronaut’s seconds lengthen only by a in a commercial jet plane at a speed of 500 miles per
factor 1.005 which represents only a .5% increase. If hour, you have problems because 1/ 1 - v2 /c2 is so
the astronaut’s speed is increased to half the speed of close to 1 that the calculator cannot tell the difference.
light, we get In a little while we will show you how to do such
calculations, but for now we will just state the answer.
T T
T' = =
2 1 - .25
1 - .5c /c 2 1 for a speed
= 1 + 2.7 × 10 –13 of 500 mi/hr (17)
2 2
T 1 - v /c
T' = = 1.15T (v = c/2) (14)
.75 To put this result in perspective suppose the astronaut
flew on the jet for what we thought was a time T = 1 hour
Now we are getting a 15% increase in the length of the
or 3600 seconds. The astronaut’s light pulse clock
astronaut’s seconds.
would show a longer time T' given by
When we work with atomic or subatomic particles, it is
not difficult to accelerate these particles to speeds close T
T' =
to the speed of light. Shortly we will consider a particle 1 - v 2 /c 2
called a muon, that is traveling at a speed v = .994 c. For
this particle we have = 1 + 2.7 × 10 -13 × 3600 seconds
T T
T' = = = 1 hour + .97 × 10 –9 seconds
2 2 1 - .988
1 - .994c /c
(15) Since .97 x 10-9 seconds is close to a nanosecond, we
T
T' = = 9T (v = .994c) can write
.012
Here we are beginning to see some large effects. If the T' ≈ 1 hour + 1 nanosecond (18)
astronaut were traveling this fast, his seconds would be The astronaut’s clock takes 1 hour plus 1 nanosecond
9 times longer than ours, his clock would be running to move its hand forward 1 hour. We would say that his
only 1/9th as fast. light pulse clock is losing a nanosecond per hour.
If we go all the way to v = c, Equation (11) gives Students have a tendency to memorize formulas, and
T T Equation (11), T' = T/ 1 - v2 /c2 looks like a good
T' = = candidate. But don’t! If you memorize this formula,
2 2 1-1
1 - c /c you will mix up T' and T, forgetting which seconds
T belong to whom. There is a much easier way to always
T' = = ∞ (v = c) (16) get the right answer.
0
1-18 Principle of Relativity

For any speed v less than or equal to c (which is all we the rate of the moving clock is reduced by a factor
will need to consider) the quantity 1 - v2 /c2 is always 1 - v2 /c2 .
a number less than or equal to 1, and 1/ 1 - v2 /c2 is
always greater than or equal to 1. For the examples we The factor 1 - v2 /c2 will appear numerous times
have considered so far, we have throughout the text. But in every case you should have
an intuitive idea of whether the quantity under consid-
eration should increase or decrease. If it increases,
Table 1
divide by the 1 - v2 /c2 , and if it decreases, multiply
by 1 - v2 /c2 . This approach gives the right answer,
v 1 - v2 /c2 1/ 1 - v2 /c2 reduces memorization, and eliminates obscure nota-
0 1 1 tion like T' and T.

500 mi/hr 1 - 2.7 × 10 -13 1 + 2.7 × 10 -13

Other Clocks
c/10 .995 1.005 So far we have an interesting but limited result. We
have predicted that if someone carrying a light pulse
c/2 .87 1.15 clock moves by us at a speed v, we will see that their
light pulse clock runs slow by a factor 1 - v2 /c2 . Up
.994c 1/9 9
until now we have said nothing about any other kind of
c 0 ∞ clock, and we have the problem that no one has actually
constructed a light pulse clock. But we can easily
generalize our result with another thought experiment
You also know intuitively that for the moving light fairly similar to the one we just did.
pulse clock, the light pulse travels a longer path, and
therefore the moving clock’s seconds are longer. For the new thought experiment let us rejoin the
discussion between the astronaut and the class of
If you remember that 1 - v2 /c2 appears somewhere students. We begin just after the students have told the
in the formula, all you have to do is ask yourself what astronaut what they saw. “I was afraid of that,” the
to do with a number less than one to make the answer astronaut replies. “I never did trust that light pulse
bigger; clearly, you have to divide by it. clock. I am not at all surprised that it ran slow. But now
my digital watch, it’s really good. It is based on a quartz
As an example of this way of reasoning, note that if a crystal and keeps really good time. It wouldn’t run
moving clock’s seconds are longer, then the rate of the slow like the light pulse clock.”
clock is slower. The number of ticks per unit time is
less. If we want to talk about the rate of a moving clock,
do we multiply or divide by 1 - v2 /c2 ? To get a
reduced rate, we multiply by 1 - v2 /c2 since that
number is always less than one. Thus we can say that
 Moving Clocks 1-19

“I’ll bet it would,” Bill interrupts. “How much?,” the “Are you insinuating,” the astronaut continued, “that
astronaut responds indignantly. “The cost of one more the reason I did not detect my light pulse running slow
trip,” Bill answers. was because my digital watch was also running slow?”

In the new trip, the astronaut is to place his digital watch “Almost,” replied Bill, “but you have other timing
right next to the light pulse clock so that the astronaut devices in your capsule. You shave once a day because
and the class can see both the digital watch and the light you do not like the feel of a beard. This is a cyclic
pulse clock at the same time. The idea is to compare the process that could be used as the basis of a new kind of
rates of the two timing devices. clock. If your shaving cycle clock did not slow down
just like the light pulse clock, you could time your
“Look what would happen,” Bill continues, “if your shaving cycle with the light pulse clock and detect your
digital watch did not slow down. When you come by, motion. You would notice that you had to shave more
your digital watch would be keeping “God’s time” as times per light pulse month when you were moving
you call it, while your light pulse clock would be than when you were at rest. This would violate the
running slower.” principle of relativity.”
“The important part of this experiment is that because “Wow,” the astronaut exclaimed, “if the principle of
the faces of the two clocks are together, if we see them relativity is correct, and the light pulse clock runs
running at different rates, you will too. You would slow, then every process, all timing devices in my ship
notice that here on earth, when you are at rest, the two have to run slow in precisely the same way so that I
clocks ran at the same rate. But when you were cannot detect the motion of the ship.”
moving at high speed, they would run at different
rates. You could use this difference in rates to detect The astronaut’s observation highlights the power and
your own motion, and therefore violate the principle generality of the principle of relativity. It turns a limited
of relativity.” theory about the behavior of one special kind of clock
into a general theory about the behavior of all possible
The astronaut thought about this for a bit, and then clocks. If the light pulse clock in the astronaut’s
responded, “I’ll grant that you are partly right. On my capsule is running slow by a factor 1 - v2 /c2 , then all
previous trips, after the accelerations ceased and I clocks must run slow by exactly the same factor so that
started coasting toward the class, I did not feel any the astronaut cannot detect his motion.
motion. I had no trouble unpacking the equipment and
setting it up. The light pulse went in just as it had back
in the lab, and I was sure that the light pulse clock was
working just fine. I certainly would have noticed any
difference in the rates of the two clocks.”
1-20 Principle of Relativity

Real Clocks Muons are created when cosmic rays from outer space
Our theory still has a severe limitation. We have to strike the upper atmosphere. Few cosmic rays make it
assume that the light pulse clock runs slow. But no one down to the lower atmosphere, so that most muons are
has yet built a light pulse clock. Thus our theory is still created in the upper atmosphere, several miles up. The
based on thought experiments and conjectures about interesting results, observed in the 1930s was that there
the behavior of light. If we had just one real clock that were almost as many high energy muons striking the
ran slow by a factor 1 - v2 /c2 , then the principle of surface of the earth as there were several miles up. This
relativity would guarantee that all other clocks ran slow indicated that most of the high energy muons seemed
in precisely the same way. Then we would not need any to be surviving the several mile trip down through the
conjectures about the behavior of light. The principle earth’s atmosphere.
of relativity would do it all!
Suppose we have a muon traveling at almost the speed
In 1905 when Einstein proposed the special theory of of light, almost 1 foot per nanosecond. To go a mile,
relativity, he did not have any examples of moving 5280 feet, would take 5,280 nanoseconds or about 5
clocks that were observed to run slow. He had to rely microseconds. Therefore a 2 mile trip takes at least 10
on his intuition and the two postulates. It was not until microseconds, which is 5 half lives. One would expect
the early 1930s, in studies of the behavior of an that in this 2 mile trip, only one out of every 32 muons
elementary particle called the muon, that experimental that started the trip would survive. Yet the evidence
evidence was obtained showing that a real moving was that most of the high energy muons, those traveling
clock actually ran slow. close to the speed of light, survived. How did they do
this?
A muon at rest has a half life of 2.2 microseconds or
2,200 nanoseconds. That means that if we start with We can get an idea of why the muons survive when we
1000 muons, 2.2 microseconds later about half will realize that the muon half life can be used as a timing
have decayed and only about 500 will be left. Wait device for a clock. Imagine that we have a box with a
another 2.2 microseconds and half of the remaining dial on the front as shown in Figure (19). We set the
muons will decay and we will have only about 250 left, hand to 0 and put 1000 muons in the box. We wait until
etc. If we wait 5 half lives, just over 10 microseconds, half the muons decay, whereupon we advance the hand
only one out of 32 of the original particles remain 2.2 microseconds, replace the decayed muons so that
(1/2 × 1/2 × 1/2 × 1/2 × 1/2 = 1/32 ). we again have 1000 muons, and then wait until half
have decayed again. If we keep repeating this process
muons go the hand will advance one muon half life in each cycle.
into box
Here we have a clock based on the muon half life rather
than the swings of a pendulum or the vibrations of a
quartz crystal.

0 The fact that most high energy muons raining down

through the atmosphere survive the trip means that
2.2
11

their half life is in excess of 10 microseconds, much

longer than the 2.2 microsecond half life of a muon at
rest. A clock based on these moving muons would run
4.4
8.8

6.6 much slower than a muon clock at rest. Thus the

experimental observation that the muons survive the
MUON CLOCK
trip down through the atmosphere gives us our first
example of a real clock that runs slow when moving.
Figure 19
In our muon clock, every time half of the muons
inside decay, we replace them and move the
hand on the face forward by 2.2 microseconds.
 1-21

In the early 1960s, a motion picture was made that 500 miles per hour should lose one nanosecond every
carefully studied the decay of muons in the trip down hour. This loss was detected to an accuracy of 1% when
from the top of Mount Washington in New Hampshire physicists at the University of Maryland in the early
to sea level (the sea level measurements were made in 1980s flew an atomic clock for 15 hours over Chesa-
Cambridge, Massachusetts), a trip of about 6000 feet. peake Bay.
Muons traveling at a speed of v = .994c were studied
and from the number surviving the trip, it was deter- In more recent times atomic clocks have become so
mined that the muon half life was lengthened to about accurate that the slowing down of the clock has become
20 microseconds, a factor of 9 times longer than the 2.2 a nuisance. When these clocks are moved from one
microsecond lifetime of muons at rest. Since location to another, they have to be corrected for the
1/ 1 - v2 /c2 = 9 for v = .994c, a result we got back in time that was lost due to their motion. For these clocks,
Equation (16), we see that the moving picture provides even a one nanosecond error is too much.
an explicit example of a moving clock that runs slow by Thus today the slowing down of moving clocks is no
a factor 1 - v2 /c2 . longer a hypothesis but a common observational fact.
At the present time there are two ways to observe the The slowing down by 1 - v2 /c2 has been seen both
slowing down of real clocks. One is to use elementary for clocks moving at the slow speeds of a commercial
particles like the muon, whose lifetimes are lengthened jet and the high speeds travelled by elementary par-
significantly when the particle moves at nearly the ticles. We now have real clocks that run slow by a
speed of light. The second way is to use modern atomic factor 1 - v2 /c2 and no longer need to hypothesize
clocks which are so accurate that one can detect the tiny about the behavior of light pulses. All of our conjec-
slowing down that occurs when the clock rides on a tures in this chapter hinge on the principle of relativ-
commercial jet. We calculated that a clock traveling ity alone.

Movie
To play the movie, click the
cursor in the photo to the
left. Use up or down arrows
on the keyboard to raise or
lower volume. Left and right
arrows step one frame
foreward or back and esc
stops it. The movie is 36
minutes long. The Movie
Time Dilation: An
Experiment with Mu-Mesons
is presented with the
permission of Education
Development Center Inc.,
Newton, Massachusetts.

Figure 19a -- Muon Lifetime Movie

The lifetimes of 568 muons, traveling at a speed of .994c, were plotted as vertical
lines. If the muon’s clocks did not run slow, these lines would show how far the
muons could travel before decaying. One can see that very few of the muons
would survive the trip from the top of Mt. Washington to sea level. Yet the
majority do survive.
1-22 Principle of Relativity

Time Dilation Space Travel

If all moving clocks run slow, does time itself run slow In human terms, time dilation should have its greatest
for the moving observer? That raises the question of effect on space travelers who need to travel long dis-
how we define time. If time is nothing more than what tances and therefore must go at high speeds. To get an
we measure by clocks, and all clocks run slow, we might idea of the distances involved in space travel, we note
as well say that time runs slow. And we can give this that light takes 1.25 seconds to travel from the earth to
effect a name like time dilation, the word dilation the moon (the moon is 1.25 billion feet away), and 8
referring to the stretching out of seconds in a moving minutes to travel from the sun to the earth. We can say
clock. that the moon is 1.25 light seconds away and the sun is
8 light minutes distant.
But time is such a personal concept, it plays such a basic
role in our lives, that it seems almost demeaning that Currently Neptune is the most distant planet (Pluto will
time should be nothing more than what we measure by be the most distant again in a few years). When Voyager
clocks. We have all had the experience that time runs II passed Neptune, the television signals from Voyager,
slow when we are bored, and fast when we are busy. which travel at the speed of light, took 2.5 hours to reach
Time is associated with all aspects of our life, including us. Thus our solar system has a radius of 2.5 light hours.
death. Can such an important concept be abstracted to It takes 4 years for light to reach us from the nearest star
be nothing more than the results of a series of measure- from our sun; stars are typically one to a few light years
ments? apart.

Let us take the following point of view. Let physicists’ If you look up at the sky at night and can see the Milky
time be that which is measured by clocks. Physicists’ Way, you will see part of our galaxy, a spiral structure
time is what runs slow for an object moving by. If your of stars that looks much like our neighboring galaxy
sense of time does not agree with physicists’ time, think Andromeda shown in Figure (20). Galaxies are about
of that as a challenge. Try to devise some experiment to 100,000 light years across, and typically spaced about a
show that your sense of time is measurably different
from physicists’ time. If it is, you might be able to devise
an experiment that violates the principle of relativity.

Figure 20
The Andromeda galaxy, about a million light years
away, and about 1/10 million light years in diameter.
 Space Travel 1-23

million light years apart. As we will see there are even Does Einstein’s theory preclude the possibility that we
larger structures in space; there are interesting things to could visit a distant world in our lifetime; are we
study on an even grander scale. confined to our local neighborhood of stars by Nature’s
speed limit? The behavior of the muons raining down
Could anyone who is reading this text survive a trip to through the atmosphere suggests that we are not con-
explore our neighboring galaxy Andromeda, or just fined. The muons, you will recall, live only 2.2
survive a trip to some neighboring star, say, only 200 microseconds (on the average) when at rest. Yet the
light years away? muons go much farther than the 2200 feet that light
Before Einstein’s theory, one would guess that the best
could travel in a muon half life. They survive the trip
way to get to a distant star would be to go so fast that the
down through the atmosphere because their clocks are
trip would not take very long. But now we have a
running slow.
problem. In Einstein’s theory, the speed of light is a If humans could accompany muons on a trip at a speed
special speed. If we had the astronaut carry our light v = .994c, the human clocks should also run slow, their
pulse clock at a speed greater than the speed of light, the lifetimes should also expand by the same factor of 9. If
light pulse could not remain in the clock. The astronaut the human clocks did not run slow and the muon clocks
would also notice that he could not keep the light pulse did, the difference in rates could be used to detect
in the clock, and could use that fact to detect his own uniform motion in violation of the principle of
uniform motion. In other words, the principle of relativ- relativity.
ity implies that we or the astronaut cannot travel faster
than the speed of light. The survival of the muons suggest that we should be
able to travel to a distant star in our own lifetime.
That the speed of light is a limiting speed is common Suppose, for example, we wish to travel to the star Zeta
knowledge to physicists working with elementary par- (we made up that name) which is 200 light years away.
ticles. Small particle accelerators about a meter in If we traveled at the speed v = .994c, our clocks should
diameter can accelerate electrons up to speeds ap- run slow by a factor 1/ 1 - v2 /c2 = 1/9, and the trip
proaching v = . 9999c. The two mile-long accelerator should only take us 200 × 1 9 = 22.4 years. We
at Stanford University, which holds the speed record would be only 22.4 years older when we get there. A
for accelerating elementary particles here on earth, can healthy, young crew should be able to survive that.
only get electrons up to a speed v = . 999999999c. The
speed of light is Nature’s speed limit, how this speed
limit is enforced is discussed in Chapter 6.
1-24 Principle of Relativity

The Lorentz Contraction The mission control officer then wrote down the fol-
A careful study of this proposed trip to star Zeta lowing calculations that were later verified in detail.
uncovers a consequence of Einstein’s theory that we “The spacecraft is traveling at a speed v = .994c, so that
have not discussed so far. To see what this effect is, to it will take 1/.994 times longer than it takes a pulse of
see that it is just as real as the slowing down of moving light to reach the star. Since the star is 200 light years
clocks, we will treat this proposed trip as a new thought away, the spacecraft should take 200/.994 = 201.2
experiment which will be analyzed from several points years to get there.
of view.
But the passengers inside are also moving at a speed
In this thought experiment, the instructor and the class, v = .994c, their clocks and biological processes run
who participated in the previous thought experiments, slow by a factor 1 - v2 /c2 = 1/9, and the amount of
decide to travel to Zeta at a speed of v = .994c. They time they will age is
have a space ship constructed which on the inside looks amount of time
just like their classroom, so that classroom discussions 1
space travelers = 201.2 years × = 22.4 years
can be continued during the trip. 9
age
On the earth, a permanent government subagency of
NASA is established to record transmissions from the Even the oldest member of the crew, the instructor, will
space capsule and maintain an earth bound log of the be able to survive.”
trip. Since the capsule, traveling at less than the speed The 2406 entry continued; “During the intervening
of light, will take over 200 years to get to Zeta, and since years we maintained communication with the capsule
the transmissions upon arrival will take 200 years to get and everything seemed to go well. There were some
back, the NASA agency has to remain in operation for complaints about our interpretation of what was hap-
over 400 years to complete its assignment. NASA’s pening but that did not matter, everything worked out
summary of the trip, written in the year 2406, reads as just as we had predicted. The spacecraft flew past Zeta
follows: “The spacecraft took off in the year 2001 and in March of the year 2206, and we received the commu-
spent four years accelerating up to a speed of v = .994c. nications of the arrival this past March. The instructor
During this acceleration everything was packed away, said she planned to retire after they decelerated and the
but when they got up to the desired speed, the rocket spacecraft landed on a planet orbiting Zeta. She was
engines were shut off and they started the long coast to not quite sure what her class of middle aged students
the Zeta. This coast started with a close fly-by of the would do.”
earth in late January of the year 2005. The NASA
mission control officer who recorded the fly-by noted
that his great, great, great, grandchildren would be alive
when the spacecraft reached its destination.”
 Lorentz Contraction 1-25

NASA’s predictions may have come true, but from the Very shortly after the engines were shut off, the earth
point of view of the class in the capsule, not everything went by. This was expected, and the students were
worked out the way NASA said it did. ready to measure the speed of the earth as it passed.
There were two windows 100 feet apart on the back
As NASA mentioned, a few years were spent acceler- wall of the classroom, as shown in Figure (21). When
ating the space capsule to the speed v = .994c. The the earth came by, there was an orbiting spacecraft,
orbit was chosen so that just after the engines were shut essentially at rest relative to the earth, that passed close
off and the coast to Zeta began, the spacecraft would to the windows. The students measured the time it took
pass close to the earth for one final good-by. the front edge of this orbiting craft to travel the 100 feet
There was quite a change from the acceleration phase between the windows. They got 100.6 nanoseconds
to the coasting phase. During the acceleration every- and therefore concluded that the orbiting craft and the
thing had to be securely fastened, and there was the earth itself were moving by at a speed
constant vibration of the engines. But when the engines 100 feet
were shut off, you couldn’t feel motion any more; vearth =
100.6 nanoseconds
everything floated as in the TV pictures of the early
astronauts orbiting the earth. = .994 feet = .994 c
nanosecond
When the coasting started, the instructor and class So far so good. That was supposed to be the relative
settled down to the business of monitoring the trip. The speed of the earth.
first step was to test the principle of relativity. Was
there any experiment that they could do inside the In the first communications with earth, NASA mission
capsule that could detect the motion of the capsule? control said that the space capsule passed by the earth
Various experiments were tried, but none demon- at noon, January 17, 2005. Since all the accurate clocks
strated that the capsule itself was moving. As a result had been dismantled to protect them from the accelera-
the students voted to take the point of view that they and tion, and only put back together when the coasting
the capsule were at rest, and the things outside were started, the class was not positive about what time it
moving by. was. They were willing to accept NASA’s statement
that the fly-by occurred on January 17, 2005. From
then on, however, the class had their own clocks in
v earth
earth order—light pulse clocks, digital clocks and an atomic
clock. From then on they would keep their own time.
For the next 22 years the trip went smoothly. There
100 ft were numerous activities, video movies, etc., to keep
the class occupied. Occasionally, about once every
classroom in
other month, a star went by. As each star passed, its
space capsule speed v was measured and the class always got the
answer v = .994c. This confirmed that the earth and the
neighboring stars were all moving together like bright
dots on a huge moving wall.

Figure 21
To measure the speed of the earth as it passes
by, the class measures the time it takes a small
satellite to pass by the windows in the back of
the classroom. The windows are 100 feet apart.
1-26 Principle of Relativity

The big day was June 13, 2027, the 45th birthday of Jill measurements of the distance to that star. We knew that
who was eighteen when the trip was planned. This was it was 200 light years away, and we knew that traveling
the day, 22.4 years after the earth fly-by, that Zeta went at a speed .994c, we could survive the trip in our
by. The students made one more speed measurement lifetime.”
and determined that Zeta went by at a speed v = .994c.
An arrival message was sent to NASA, one day was Bill responded, “I think you entered the room too late
allowed for summary discussions of the trip, and then and missed my argument. Let me summarize it. Point
the deceleration was begun. 1: the earth went by a little over 22 years ago. Point 2:
we actually measured that the earth was traveling by us
After a toast to Jill for her birthday, Bill began the at almost the speed of light. Point 3: even light cannot
conversation. “Over the past few years, the NASA go farther than 22 light years in 22 years. The earth can
communications and even our original plans for the trip be no farther than about 22 light years away. Point 4:
have been bothering me. The star charts say that Zeta Zeta passed by us today, thus the distance from the
is 200 light years from the earth, but that cannot be earth to Zeta is about 22 light years, not 200 light
true.” years!”

“Look at the problem this way.” Bill continues. “The “But what about NASA’s calculations and all their
earth went by us at noon on January 17, 2005, just 22.4 plans,” the instructor said, interrupting a bit nervously.
years ago. When the earth went by, we observed that
it took 100.6 nanoseconds to pass by our 100 foot wide “We do not care what NASA thinks,” responded Bill.
classroom. Thus the earth went by at a speed v = .994 “We have had no acceleration since the earth went by.
feet/nanosecond, or .994c. Where is the earth now, Thus the principle of relativity guarantees that we can
22.4 years later? How far could the earth have gotten, take the point of view that we are at rest and that it is the
traveling at a speed .994c for 22.4 years? My answer earth and NASA that are moving. From our point of
is view, the earth is 22 light years away. What NASA
thinks is their business.”
distance of earth light year
= .994 × 22.4 years Joan interrupts, “Let us not argue on this last day. Let’s
from spaceship year
figure out what is happening. There is something more
= 22 light years important here than just how far away the earth is.”
“You’re right!” Joan interrupted, “Even if the earth had
gone by at the speed of light, it would have gone only Zeta
22.4 light years in the 22.4 years since fly-by. The star 03
chart must be wrong.”

The instructor, who had just entered the room, said, “I

object to that remark. As a graduate student I sat in on
part of a course in astronomy and they described how
the distance to Zeta was measured.” The instructor
drew a sketch, Figure (22), and continued. “Here is the
earth in its orbit about the sun, and two observations, six
months apart, are made of Zeta. You see that the two 01 Sun 02
positions of the earth and the star form a triangle.
Telescopes can accurately measure the two angles I e a rth's o r b i t
labeled θ1 and θ2, and the distance across the earth’s Figure 22
orbit is accurately known to be 16 light minutes. If you Instructor's sketch showing how the distance from the
know two angles and one side of a triangle, then you can earth to the star Zeta was measured. (For a star 200
calculate the other sides from simple geometry. One light years away, θ3 is 4.5 millionths of a degree.)
reason for choosing a trip to Zeta is that we had accurate
 Lorentz Contraction 1-27

“Remember in the old lectures on time dilation where “That’s easy,” replies Bill. Just imagine that we string
the astronaut carried a light pulse clock. We used the high tensile carbon filament meter sticks between the
peculiar behavior of that clock and the principle of sun and Zeta. I estimate (after a short calculation) that
relativity to deduce that time ran slow for a moving it should take only 6 × 1017 of them. As we go on our
observer." trip, it doesn’t make any difference whether the meter
sticks are there or not, everything between the earth and
“Now for us, NASA is the moving observer. More than Zeta passes by in 22 years. We still see 6 × 1017 meter
that, the earth, sun, and the stars, including Zeta, have sticks. But each one must have shortened by a factor
all passed us going in the same direction and the same 1 - v2 /c2 so that all of them fit in the shortened
speed v = .994c. We can think of them as all in the same distance of 22 light years. It does not make any
huge space ship. Or we can think of the earth and the difference what the sticks are made of.”
stars as painted dots on a very long rod. A very long rod
moving past us at a speed v = .994c. See my sketch Jim, who had not said much up until now, said, “OK,
(Figure 23)." from your arguments I see that the length of the meter
sticks, the length in the direction of motion must
“To NASA, and the people on earth, this huge rod, with contract by a factor 1 - v2 /c2 , but what about the
the sun at one end and Zeta at the other, is 200 light width? Do the meter sticks get skinnier too?”
years long. Our instructor showed us how earth people
measured the length of the rod. But as Bill has pointed The class decided that Jim’s question was an excellent
out, to us this huge rod is only 22 light years long. That one, and that a new thought experiment was needed to
moving rod is only 1/9th as long as the earth people decide.
think it is.”
Let’s try this,” suggested Joan. “Imagine that we have
“But,” Bill interrupts, the factor of 1/9 is exactly the a space ship 10 feet in diameter and we build a brick
factor 1 - v2 /c2 by which the earth people thought wall with a circular hole in it 10 feet in diameter (Figure
our clocks were running slow. Everyone sees some- (24)). Let us assume that widths, as well as lengths,
thing peculiar. The earth people see our clocks running contract. To test the hypothesis, we hire an astronaut to
slow by a factor 1 - v2 /c2 , and we see this hypotheti-
Zeta
cal rod stretched from the sun to Zeta contracted by a
factor 1 - v2 /c2 .”

“But I still worry about the peculiar rod of Joan’s,” Bill

continues, “what about real rods, meter sticks, and so
forth? Will they also contract?” hypothetical
measuring
At this point Joan sees the answer to that. “Remember, rod between
Bill, when we first discussed moving clocks, we had our sun and
only the very peculiar light pulse clock that ran slow. the star Zeta
But then we could argue that all clocks, no matter how
they are constructed, had to run slow in exactly the
same way, or we could violate the principle of
relativity.”

“We have just seen that my ‘peculiar’ rod, as you call Sun
it, contracts by a factor 1 - v2 /c2 . We should be able
Figure 23
to show with some thought experiments that all rods, no Joan's sketch of the Sun and Zeta
matter what they are made of, must contract in exactly moving by. This “object” passed by in
the same way as my peculiar one or we could violate the about 22 years, moving at nearly the
principle of relativity.” speed of light. Thus the “object” was
v = .994c about 22 light years long.
1-28 Principle of Relativity

fly the 10 foot diameter capsule through the 10 foot hole Leaving our thought experiment, it is interesting to note
at nearly the speed of light, say at v = .994c. If widths that the discovery of the contraction of moving lengths
contract like lengths, the capsule should contract to 10/ occurred before Einstein put forth the special theory of
9 of a foot; it should be just over 13 inches in diameter relativity. In the 1890s, physicist George Fitzgerald
when it gets to the 10 foot hole. It should have no assumed that the length of one of the arms in
trouble getting through.” Michaelson’s interferometer, the arm along the direc-
tion of motion, contracted by a factor 1 - v2 /c2 . This
“But look at the situation from the astronaut’s point of was just the factor needed to keep the interferometer
view. He is sitting there at rest, and a brick wall is from detecting the earth’s motion in the Michaelson-
approaching him at a speed v = .994c. He has been told Morley experiments. It was a short while later that
that there is a 10 foot hole in the wall, but he has also H.A. Lorentz showed that if the atoms in the arm of the
been told that the width of things contracts by a factor Michaelson interferometer were held together by elec-
1 - v2 /c2 . That means that the diameter of the hole tric forces, then such a contraction would follow from
should contract from 10 feet to 13 inches. He is sitting Maxwell’s theory of electricity. The big step, however,
there in a 10 foot diameter capsule, a brick wall with a was Einstein’s assumption that the principle of relativ-
13 inch hole is approaching him, and he is supposed to ity is correct. Then, if one object happens to contract
fit through. No way! He bails out and looks for another when moving, all objects must contract in exactly the
job.” same way so that the contraction could not be used to
detect one’s own motion. This contraction is called the
“That’s a good way to do thought experiments, Joan,”
replied the instructor. “Assume that what you want to Lorentz-Fitzgerald contraction, or Lorentz contrac-
test is correct, and then see if you can come up with an tion, for short.
inconsistency. In this case, by assuming that widths
contract, you predicted that the astronaut should easily Relativistic Calculations
make it through the hole in the wall. But the astronaut Although we have not quite finished with our discus-
faced disaster. The crash, from the astronaut’s point of sion of Einstein’s special theory of relativity, we have
view would have been an unfortunate violation of the covered two of the important consequences, time dila-
principle of relativity, which he could use as evidence tion and the Lorentz contraction, which will play
of his own uniform motion.” important roles throughout the text. At this point we
will take a short break to discuss easy ways to handle
“To sum it up,” the instructor added, “we now have calculations involving these relativistic effects. Then
time dilation where moving clocks run slow by a factor we will take another look at Einstein’s theory to see if
1 - v2 /c2 , and we see that moving lengths contract by there are any more new effects to be discovered.
the same factor. Only lengths in the direction of motion

;;;;;;;;;;;;;
contract, widths are unchanged.” After our discussion of time dilation, we pointed out the
importance of the quantity 1 - v2 /c2 which is a

;;;;;;;;;;;;;
;;;;;;;;;;;;; on
10 foot dia
number always less than 1. If we wanted to know how
much longer a moving observer’s time interval was, we
divided by 1 - v2 /c2 to get a bigger number. If we

;;;;;;;;;;;;;
me
ti
Cau

ter

wanted to know how much less was the frequency of a

;;;;;;;;;;;;;
moving clock, we multiplied by 1 - v2 /c2 to get a

;;;;;;;;;;;;;
smaller number.

;;;;;;;;;;;;;
With the Lorentz contraction we have another effect
that depends upon 1 - v2 /c2 . If we see an object go
by us, the object will contract in length. To predict its
Figure 24
contracted length, we multiply the uncontracted length
Do diameters contract?
by 1 - v2 /c2 to get a smaller number. If, on the other
 Lorentz Contraction 1-29

hand, an object moving by us had a contracted length Traveling by at nearly the speed of light, the 667 foot
l, and we stop the object, the contraction is undone and high Mt. Washington should take about 667 nanosec-
the length increases. We get the bigger uncontracted onds or .667 microseconds to go by. Since this is
length by dividing by 1 - v2 /c2 . considerably less than the 2.2 microsecond half life of
the muons, most of them should survive until sea level
As we mentioned earlier, first determine intuitively comes by.
whether the number gets bigger or smaller, then either
multiply by or divide by the 1 - v2 /c2 as appropriate. Example 2 Slow Speeds
This always works for time dilation, the Lorentz con- Joan walks by us slowly, carrying a meter stick point-
traction, and, as we shall see later, relativistic mass. ing in the direction of her motion. If her speed is v = 1
We will now work some examples involving the foot/second, what is the contracted length of her meter
Lorentz contraction to become familiar with how to stick as we see it?
handle this effect. This is an easy problem to set up. Since her meter stick
is contracted, we multiply 1 meter times the 1 - v2 /c2
Example 1 Muons and Mt Washington
with v = 1 foot/second. The problem comes in evalu-
In the Mt. Washington experiment, muons travel 6000 ating the numbers. Noting that 1 nanosecond = 10-9
feet from the top of Mt. Washington to sea level at a seconds, we can use the conversion factor 10-9 sec-
speed v = .994c. Most of the muons survive despite the onds/nanosecond to write
fact that the trip should take about 6 microseconds
(6000 nanoseconds), and the muon half life is τ = 2.2 ft sec
v = 1 × 10 -9
microseconds for muons at rest. sec nanosecond

We say that the muons survive the trip because their ft

= 10 -9 = 10 -9 c
internal timing device runs slow and their half life nanosecond
expands by a factor 1/ 1 - v2 /c2 = 9. The half life of Thus we have
the moving muons should be
half life of τ v = 10-9 , v2 = 10-18
moving muons = 1 - v2 /c2
c c2
= 2.2 microseconds × 9 and for Joan’s slow walk we have

= 19.8 microseconds 1 - v2 /c2 = 1 - 10-18 (19)

This is plenty of time for the muons to make the trip. If we try to use a calculator to evaluate the square root
in Equation (19), we get the answer 1. For the calcula-
From the muon’s point of view, they are sitting at rest tor, the number 10-18 is so small compared to 1, that it
and it is Mt. Washington that is going by at a speed v = is ignored. It is as if the calculator is telling us that when
.994c. The muon’s clocks aren’t running slow, instead Joan’s meter stick is moving by at only 1 foot/second,
the height of Mt. Washington is contracted. there is no noticeable contraction.
To calculate the contracted length of the mountain, start But there is some contraction, and we may want to
with the 6000 foot uncontracted length, multiply by know the contraction no matter how small it is. Since
1 - v2 /c2 = 1/9 to get calculators cannot handle numbers like 1 – 10-18 , we
contracted height 1 need some other way to deal with such expressions. For
= 6000 feet × this, there is a convenient set of approximation formu-
of Mt. Washington 9
las which we will now derive.
= 667 feet
1-30 Principle of Relativity

Approximation Formulas Some useful approximation formulas are the following

The approximation formulas deal with numbers close
to 1, numbers that can be written in the form (1 + a) or 1+α 2 ≈ 1 + 2α (20)
(1 – a) where a is a number much less than 1. For
example the square root in Equation (19) can be written 1-α 2 ≈ 1 - 2α (21)
as
1 ≈ 1-α (22)
1 - 10-18 = 1 - α 1+α
where α = 10-18 is truly a number much less than 1. 1 ≈ 1+α (23)
1-α
The idea behind the approximation formulas is that if
a is much less than 1, a2 is very much less than 1 and can 1-α ≈ 1-α (24)
be neglected. To see how this works, let us calculate (1 2

≈ 1+α
+ a)2 and see how we can neglect a2 terms even when 1 (25)
a is as large as .01. An exact calculation is 1-α 2
1+α 2 = 1 + 2α + α2
We have already derived Equation (20). Equation (21)
which for α = .01, α 2 = .0001 is follows from (20) if we replace α by –α.
2
1+α = 1 + .02 + .0001 Equation (22) can be derived as follows. Multiply the
quantity 1 – a by (1+a)/(1+a) which is 1 to get
= 1.0201
If we want to know how much 1 + α 2 differs from 1, 1+α 1 - α2 1
1-α = 1-α × = ≈
but do not need too much precision, we could round off 1+α 1+α 1+α
1.0201 to 1.02 to get

1+α 2 ≈ 1.02 In the last step we dropped the α2 terms.

(The symbol ≈ means “approximately equal to”). But To derive the approximate formula for a square root,
in replacing 1.0201 by 1.02, we are simply dropping the start with
α 2 term in Equation (19). We can write α α α α2
2 1- × 1- = 1- 2 + ≈ 1- α (26)
1 +α ≈ 1 + 2α = 1 +.02 = 1.02 (20) 2 2 2 4

Equation (20) is our first example of an approximation taking the square root of Equation (26) gives
formula.

In Equation (20) the smaller a is the better the approxi- 1-α ≈ 1-α
2
mation. If a = .0001 we have
which is the desired result. Again we only neglected α2
1.0001 2 = 1.00020001 exact terms.
Equation (20) gives To derive Equation (25), first use Equation (24) to get
2 1
1 + .0001 ≈ 1 + .0002 = 1.0002 ≈ 1
1-α 1-α
2
and we see that the neglected α2 terms become less and
less important.
 Approximation Formulas 1-31

Then use Equation (23), with a replaced by a/2 to get Exercise 6

1 α We saw that time dilation in a commercial jet was not a
≈ 1+
α 2 big effect either—clocks losing only one nanosecond
1- per hour in a jet traveling at 500 miles per hour. This was
2
not an unnoticed effect, however, because modern
which is the desired result. atomic clocks can detect this loss.
For those who are interested, the approximation formu- In our derivation of the one nanosecond loss, we stated
las we have written are the first term of the so called in Equation (17) that
binomial expansion:
1 for a speed of
n n-1 2 ≈ 1 + 2.7 × 10-13
1 + α n = 1 + nα + α + (27) 2
1 - v /c 2 500 miles/hour
2 (17)
where the coefficients of α, α2, etc. are known as the Starting with
binomial coefficients. If you need more accurate miles feet 1
approximations, you can use Equation (27) and keep v = 500 × 5280 ×
hour mile 3600 sec/hour
terms in α2 , α3 , etc. For all the work in this text, the
use the approximation formulas to derive the result
first term is adequate.
stated in Equation (17).

Exercise 5 Exercise 7
Show that Equations (20) through (25) are all examples Here is an exercise where you do not need the approxi-
of the first order binomial expansion mation formulas, but which should get you thinking
n about the Lorentz contraction. Suppose you observe
1+α ≈ 1 + nα (27a) that the Mars-17 spacecraft, traveling by you at a speed
of v = .995c, passes you in 20 nanoseconds. Back on
earth, the Mars-17 spacecraft is stored horizontally in
We are now ready to apply our approximation for-
a hanger that is the same length as the spacecraft. How
mulas to evaluate 1 - 10-18 that appeared in Equa-
long is the hanger?
tion (17). Since α = 10-18 is very small compared to
1, we have

1 - 10-18 = 1 - α ≈ 1 - α = 1 - 10
-18
2 2
Thus the length of Joan’s meter stick is

length of Joan's
= 1 meter × 1 - v 2 /c 2
contracted meter stick

-18
= 1 meter 1 - 10
2
= 1 meter - 5 × 10- 19 meters
1-32 Principle of Relativity

A CONSISTENT THEORY LACK OF SIMULTANEITY

As we gain experience with Einstein’s special theory of One of the foundations of our intuitive sense of time is
relativity, we begin to see a consistent pattern emerge. the concept of simultaneity. “Where were you when
We are beginning to see that there is general agreement the murder was committed,” the prosecutor asks. “At
on what happens, even if different observers have the time of the murder,” the defendant replies, “I was
different opinions as to how it happens. A good eating dinner across town at Harvey’s Restaurant.” If
example is the Mt. Washington experiment observing the defendant can prove that the murder and eating
muons traveling from the top of Mt. Washington to sea dinner at Harvey’s were simultaneous events, the jury
level. Everyone agrees that the muons made it. The will set him free. Everyone knows what simultaneous
muons are actually seen down at sea level. How they events are, or do they?
made it is where we get the differing points of view. We
say that they made it because their clocks ran slow. One of the most unsettling consequences of Einstein’s
They say they made it because the mountain was short. theory is that the simultaneity of two events depends
Time dilation is used from one point of view, the upon the point of view of the observer. Two events that
Lorentz contraction from another. from our point of view occurred simultaneously, may
not be simultaneous to an observer moving by. Worse
Do we have a complete, consistent theory now? In any yet, two events that occurred one after the other to us,
new situation will we always agree on the predicted may have occurred in the reverse order to a moving
outcome of an experiment, even if the explanations of observer.
the outcome differ? Or are there some new effects, in
addition to time dilation and the Lorentz contraction, To see what happens to the concept of simultaneous
that we will have to take into account? events, we will return to our thought experiment in-
volving the instructor and the class. The action takes
The answer is that there is one more effect, called the place on the earth before the trip to the star Zeta, and
lack of simultaneity which is a consequence of Joan has just brought in a paperback book on relativity.
Einstein’s theory. When we take into account this lack
of simultaneity as well as time dilation and the Lorentz “I couldn’t understand that book either,” the instructor
contraction, we get a completely consistent theory. says to Joan, “he starts with Einstein’s analogy of trains
Everyone will agree on the predicted outcome of any and lightening bolts, but then switches to wind and
experiment involving uniform motion. No other new sound waves, which completely confused me. There
effects are needed to explain inconsistencies. are many popular attempts to explain Einstein’s theory,
but most do not do very well when it comes to the lack
The lack of simultaneity turns out to be the biggest of simultaneity.”
effect of special relativity, it involves two factors of
1 - v2 /c2 . But in this case the formulas are not as “One of the problems with these popular accounts,” the
important as becoming familiar with some of the instructor continues, “is that we have to imagine too
striking consequences. We will find ourselves dealing much. In today’s lecture I will try to avoid that. In class
with problems such as whether we can get answers to we are going to carry out a real experiment involving
questions that have not yet been asked, or whether two simultaneous events. We are going to discuss that
gravity can crush matter out of existence. Strangely experiment until everyone in class is completely clear
enough, these problems are related. about what happened. No imagining yet, just observe
what actually occurred. When there are no questions
left, then we will look at our real experiment from the
point of view of someone moving by. At that point the
main features of Einstein’s theory are easy to see.”
 Lack of Simultaneity 1-33

“The apparatus for our experiment is set up here on the "OK, let’s do the experiment.”
lecture bench (Figure 25). On the left side of the bench
I have a red flash bulb and on the right side a green flash While the instructor is adjusting the apparatus, one of
bulb. These flash bulbs are attached to batteries and the flashbulbs goes off accidentally which amuses the
photocells so that when a light beam strikes their base, class, but finally the apparatus is ready, the laser beam
they go off.” turned on, and both bulbs fire.

“In the center of the desk is a laser and in front of it a “Well, were they simultaneous flashes?” the instructor
beam splitter that uses half silvered mirrors. When I turn asks the class.
the laser on, the laser beam comes out, strikes the beam
“I guess so,” Bill responds, a bit hesitantly.
splitter, and divides into two beams. One beam travels
to the left and sets off the red flash bulb, while the other “How do you know,” the instructor asks.
beam goes to the right and triggers the green flash bulb.
I will call the beams emerging from the beam splitter “Because you set it up that way,” answers Bill.
‘trigger beams’ or ‘trigger pulses’.”
Turning and pointing a finger at Joan who is sitting on
“Let us analyze the experiment before we carry it out,” the right side of the room nearer the green flash bulb (as
the instructor continues. “We will use the Einstein in Figure 26), the instructor says, “Joan, for you which
postulate that the speed of light is c to all observers. Thus flash was first?” Joan thought for a second and replied,
the left trigger pulse travels at a speed c and so does the “The green bulb is closer, I should have seen the green
right one as I showed on the sketch. Since the beam light first.”
splitter is in the center of the desk, the trigger pulses
which start out together, travel the same distance at the “But which occurred first?” the instructor interrupts.
same speeds to reach the flash bulbs. As a result the flash “What are you trying to get at?” Joan asks.
bulbs must go off simultaneously.”

“The flashing of the flashbulbs are an example of what

I mean by simultaneous events,” the instructor adds
with emphasis. “I know that they will be simultaneous
events because of the way I set up the experiment".

top view of lecture bench

laser laser

red green red green

c c trigger pulse trigger pulse
flash flash flash flash
bulb c c bulb
bulb bulb
beam splitter

Figure 25 Figure 26
Lecture demonstration experiment in which two Although Joan sees the light from the
flashbulbs are fired simultaneously by trigger green flash first, she knows that the two
signals from a laser. The laser and beam splitter flashes were simultaneous because of
are at the center of the lecture bench, so that the the way the experiment was set up.
laser light travels equal distances to reach the
red and green bulbs. A photocell, battery and
Joan
relay are mounted in each flashbulb base.
1-34 Principle of Relativity

“Let me put it this way,” the instructor responds. “Around “It is much easier than that.” the instructor exclaimed,
1000 BC, the city of Troy fell to the invading Greek “Don’t worry about when the light reaches you, just
army. About the same time, a star at the center of the look at the way I set up the experiment – two trigger
Crab Nebula exploded in what is known as a supernova pulses, starting at the same time, traveling the same
explosion. Since the star is 2000 light years away, the distance at the same speed. The flashes must have
light from the supernova explosion took 2000 years to occurred simultaneously. I chose this experiment
get here. The light arrived on July 4, 1057, about the because it is so easy to analyze when you look at the
time of the Battle of Hastings. Now which are simulta- trigger pulses.”
neous events? The supernova explosion and the Battle
of Hastings, or the supernova explosion and the fall of “Any other questions?” the instructor asks. But by this
Troy” time the class is ready to go on. “Now let us look at the
experiment from the point of view of a Martian moving
“I get the point,” replied Joan. “Just because they saw to the right a high speed v (Figure 27a). The Martian
the light from the supernova explosion at the time of the sees the lecture bench, laser, beam splitter and two flash
Battle of Hastings, does not mean that the supernova bulbs all moving to the left as shown (Figure 27b). The
explosion and that battle occurred at the same time. We lecture bench appears shortened by the Lorentz con-
have to calculate back and figure out that the supernova traction, but the beam splitter is still in the middle of the
explosion occurred about the time the Greeks were bench. What is important is that the trigger pulses,
attacking Troy, 2000 years before the light reached us.” being light, both travel outward from the beam splitter
at a speed c .
“As I sit here looking at your experiment,” Joan
continues, “I see the light from the green flash before As the bench passes by, the Martian sees that the green
the light from the red flash, but I am closer to the green flash bulb quickly runs into the trigger pulse like this
bulb than the red bulb. If I measure how much sooner ( c v ). But on the other side there is a race
the green light arrives, then measure the distances to the between the trigger pulse and the red flash bulb,
two bulbs, and do some calculations, I’ll probably find (v c ), and the race continues for a
that the two flashes occurred at the same time.” long time after the green bulb has fired. For the
Martian, the green bulb actually fired first, and the two
flashes were not simultaneous.”
v
Martian

laser laser
red green red green
flash c c flash v c c v
bulb bulb

v What the Martian sees

Figure 27a Figure 27b

In our thought experiment, a Martian astronaut The Martian astronaut sees the green flashbulb
passes by our lecture bench at a high speed v. running into its trigger signal and firing quickly.
The red flashbulb is running away from it's
trigger signal, and therefore will not fire for a
long time. Clearly the green flash occurs first.
 Causality 1-35

“How much later can the red flash occur?” asks Bill. “Let’s draw a sketch,” the instructor replies. The
result is in Figure (28b). “The Venetian astronaut
The instructor replied, “The faster the bench goes by, the sees the lecture bench moving to the left. Now the red
closer the race, and the longer it takes the trigger pulse to flash bulb runs into the trigger signal, and the race is
catch the red flash bulb. It isn’t too hard to calculate the with the green flash bulb. If the Venetian were going
time difference. In the notes I handed out before class, I by at the same speed as the Martian (Equation 30)
calculated that if the Martian sees our 12 foot long lecture then the green flash would occur one year after the
bench go by at a speed red one.”
v = .99999999999999999999999999999992c (28) “With Einstein’s theory, not only does the simulta-
then the Martian will determine that the red flash occurred neity of two events depend upon the observer’s point
one complete earth year after the green flash. Not only of view, even the order of the two events—which
are the two flashes not simultaneous, there is no funda- one occurred first—depends upon how the observer
mental limit as to how far apart in time that the two flashes is moving!”
can occur.”

The reader will find the instructor’s class notes in Appen-

dix A of this chapter.

At this point Joan asks a question. “Suppose an astronaut

from the planet Venus passed our experiment traveling
the other way. Wouldn’t she see the red flash first?”

v
Venusian

laser laser
red green red green
c c flash v c c v
flash
bulb bulb

What the Venetian sees v

Figure 28a Figure 28b

Now a Venusian astronaut passes by our lecture The Venetian astronaut sees the red flashbulb
bench at a high speed v in the other direction. running into its trigger signal and firing quickly.
The green flashbulb is running away from its
trigger signal, and therefore will not fire for a
long time. Clearly the red flash occurs first.
1-36 Principle of Relativity

Figure 29 red laser green

To test the speed of the computer, c c
Bill thinks of a question, and types it
in, when he sees the red flash. Joan
checks to see if the answer arrives at 4' 8'
the same time as the green flash.
computer

Bill Joan

CAUSALITY main difference in the setup is that the laser and beam
“You can reverse the order of two events that are years splitter have been moved from the center, over closer
apart!” Bill exclaimed. “Couldn’t something weird to the end where we type in questions. We have set it
happen in that time?” up so that the trigger pulse travels 4 feet to the red bulb
and 8 feet to the green bulb as shown in the sketch
“What about cause and effect,” asked Joan. “If you can (Figure 29). Since the trigger pulse takes 4 nanosec-
reverse the order of events, can't you reverse cause and onds to get to the red bulb, and 8 nanoseconds to reach
effect? Can’t the effect come before the cause?” the green bulb, the red flash will go off 4 nanoseconds
before the green one. We will use these 4 nanoseconds
“In physics,” the instructor responds, “there is a prin- to time the speed of the computer.”
ciple called causality which says that you cannot
reverse cause and effect. Causality is not equivalent to “Bill,” the instructor says, motioning to him, “you
the principle of relativity, but it is closely related, as we come over here, and when you see the red flash think
can see from the following thought experiment.” of a question. Then type it into the machine. Do not
think of the question until after you see the red flash, but
“Suppose,” she said, “we read an ad for a brand new then think of it and type it in quickly. We will assume
IBM computer that is really fast. The machine is so fast that you can do that in much less than a nanosecond.
that when you type a question in at one end, the answer You can always do that kind of thing in a thought
is printed out at the other end, 4 nanoseconds later. We experiment.”
look at the ad, see that the machine is 12 feet long, and
order one to replace our lecture bench. After the “OK, Joan,” the instructor says, motioning to Joan,
machine is installed, we decide to test the accuracy of “you come over here and look for the answer to Bill’s
the ad. Do we really get answers in 4 nanoseconds? To question. If the ad is correct, if the machine is so fast
find out, we set up the laser, beam splitter and flash that the answer comes out in 4 nanoseconds, then the
bulbs on the computer instead of the lecture bench. The answer should arrive when the green flash goes off.”

Martian
Figure 30
To a Martian passing by, our computer is moving to
the left at a speed near the speed of light. The race red laser green
between the red bulb and its trigger signal takes so c c
long that the green bulb fires first. As a result, v v
v
Joan sees the answer to a question that Bill has not
yet thought of. (This is what could happen if
What the Martian sees
information travels faster than the speed of light.)

Bill Joan
 Causality 1-37

The instructor positions Bill and Joan and the equip- answers to questions that have not yet been asked. That
ment as shown in Figure (29), turns on the laser and machine can be used to violate the principle of causal-
fires the flash bulbs. ity. But there was something peculiar about that
machine. When the answer went through the machine,
“Did you type in the question,” the instructor asks Bill, information went through the machine at three times
“when the red flash occurred?” the speed of light. Light takes 12 nanoseconds to cross
“Of course,” responds Bill, humoring the instructor. the machine, while the answer went through in 4
nanoseconds.”
“And did the answer arrive at the same time as the green
flash,” the instructor asks Joan. “Suppose,” asks Bill, “that the answer did not travel
faster than light. Suppose it took 12 nanoseconds
“Sure,” replies Joan, “why not?” instead of 4 nanoseconds for the answer to come out.”
The instructor replied, “To measure a 12 nanosecond
“Suppose it did,” replied the instructor. “Suppose the delay with our flash bulb apparatus, we would have to
ad is right, and the answer is printed when the green set the beam splitter right up next to the red bulb like this
flash goes off. Let us now look at this situation from the (Figure 31) in order for the trigger signal to reach the
point of view of a Martian who is traveling to the right green bulb 12 nanoseconds later. But with this setup,
at a very high speed. The situation to the Martian looks the red bulb flashes as soon as the laser is turned on. No
like this (Figure 30). Although the red bulb is closer to one, no matter how they are moving by, sees a race
the beam splitter, it is racing away from the trigger between the red bulb and the trigger signal. Everybody
pulse (v c ). If the computer is going agrees that the green flash occurs after the red flash.”
by fast enough, the race between the red bulb and its
trigger pulse will take much longer than the head-on “You mean,” interrupts Joan, “that you cannot violate
collision between the green bulb and its trigger pulse. causality if information does not travel faster than the
The green flash will occur before the red flash.” speed of light?”

“And I,” interrupts Joan, “will see the answer to a “That’s right,” the instructor replies, “that’s one of the
question that Bill has not even thought of yet!” important and basic consequences of Einstein’s theory.”

“I thought you would be in real trouble if you could “That’s interesting,” adds Bill. “It would violate the
reverse the order of events,” Joan added. principle of relativity if we observed the astronaut’s
capsule, or probably any other object, traveling faster
“It is not really so bad,” the instructor continued. “If the than the speed of light. The speed of light is beginning
ad is right, if the 12 foot long computer can produce to play an important role.”
answers that travel across the machine in 4 nanosec-
onds, we are in deep trouble. In that case we could see

Figure 31 laser
red green
If the answer to Bill's question takes
c
12 nanoseconds to travel through
the 12 foot long computer, then this
is the setup required to check the 12'
timing. The red bulb fires
instantaneously, and everyone computer
agrees that the red flash occurs first,
and the answer appears later. Bill Joan
1-38 Principle of Relativity

“That’s pretty far out,” replied Joan. “I didn’t know “Then,” Jim asked, “doesn’t that put a limit on how
that physics could say anything about how information incompressible, how rigid matter can be?”
– ideas – moved.”
“That looks like one of the consequences of Einstein’s
Jim, who had been sitting in the back of the classroom theory,” the instructor replies.
and not saying much, raised his hand. “At the begin-
ning of the course when we were talking about sound “Then that explains what they were trying to say in the
pulses, you said that the more rigid the material, the next chapter on black holes. They said that if you got
faster the speed of sound in the material. You used too much matter concentrated in a small region, the
Slinky pulses in your demonstrations because a Slinky gravitational force would become so great that it would
is so compressible that a Slinky pulse travels slowly. crush the matter out of existence.”
You can’t compress air as easily as a Slinky, and sound
“I didn’t believe it, because I thought that the matter
pulses travel faster in air. Since steel is very rigid,
would be squeezed down into a new form that is a lot
sound goes very fast in steel.”
more incompressible than nuclear matter, and the
“During these discussions about the speed of light, I collapse of the star would stop. But now I am beginning
have been wondering. Is there any kind of material that to see that there may not be anything much more rigid
is so rigid that sound waves travel at the speed of light?” than nuclear matter. Maybe black holes exist after all.”

“What made you ask that?” the instructor asked. “Will you tell us about neutron stars and black holes?”
Joan asks eagerly.
“I’ve been reading a book about the life and death of
stars,” Jim replied. I just finished the chapter on “Later in the course,” the instructor responds.
neutron stars, and they said that the nuclear matter in a
neutron star was very incompressible. It had to be to
resist the strong gravitational forces. I was wondering,
how fast is the speed of sound in this nuclear matter?”

“Up close to the speed of light,” the instructor replied.

“If the nuclear matter were even more rigid, more

incompressible, would the speed of sound exceed the
speed of light?” Jim asked.

“It can’t,” the instructor replied.

 1-39

APPENDIX A If we plug in the numbers t = 3 × 107 seconds (one

earth year), l = 6 feet, c = 109 feet/second, we get
Class Handout
To predict how long it takes for the trigger pulse to l 2 6 ft 2
catch the red bulb in Figure (27b), let l be the 1 - v/c = = 9 7
uncontracted half length of the lecture bench (6 feet for
ct 10 ft/sec × 3 × 10 sec
our discussion). To the Martian, that half of the lecture
bench has contracted to a length l 1 - v2 /c2 . = 2.8 × 10 -16

In the race, the red bulb traveling at a speed v, starts out Squaring this gives
a distance l 1 - v2 /c2 ahead of the trigger pulse,
which is traveling at a speed c. Let us assume that the 1 - v/c = 8 × 10 -32
race lasts a time t and that the trigger pulse catches the
red bulb a distance x from where the trigger pulse Thus if
started. Then we have
v = (1 - 8 × 10-32 )c
x = ct (30)
= .99999999999999999999999999999992c
In the same time t, the green bulb only travels a distance
x - l 1 - v2 /c2 , but this must equal vt; then the race will last a whole year. On the other side,
the trigger signal runs into the green flash bulb in far
vt = x - l 1 - v2 /c2 (31) less than a nanosecond because the lecture bench is
Using Equation (30) in (31) gives highly Lorentz contracted.

vt = ct - l 1 - v2 /c2
t c – v = l 1 – v 2 /c 2
Solving for t gives

2 2 l 1 + v/c 1 - v/c
t = l 1c--vv /c =
c 1 - v/c

l 1 + v/c
=
c 1 - v/c
If v is very close to c, then 1 + v/c ≈ 2 and we get

t ≈ l c2 1
1 - v/c
Chapter 2
Vectors

CHAPTER 2 VECTORS

In the first chapter on Einstein’s special theory of feature of projectile motion is that projectiles move
relativity, we saw how much we could learn from the with constant or uniform acceleration. We can think of
simple concept of uniform motion. Everything in the this as one step up in complexity from the uniform
special theory can be derived from (1) the idea that you motion discussed in the previous chapter.
cannot detect your own uniform motion, and (2) the
existence of a real clock that runs slow by a factor To study motion today, we have many tools that were
1 – v2 c2 .
not available to Galileo. In the laboratory we can slow
down the action, or stop it, using strobe photographs or
We are now about to study more complicated kinds of television cameras. To describe and analyze motion
motion where either the speed, the direction of motion, we have a number of mathematical tools, particularly
or both, are changing. Our work with non-uniform the concept of vectors and the subject of calculus. And
motion will be based to a large extent on a concept to predict motion, to predict not only where cannon
discovered by Galileo about 300 years before Einstein balls land but also the trajectory of a spacecraft on a
developed the special theory of relativity. It is interest- mission to photograph the solar system, we now have
ing that after studying complex forms of motion for over digital computers.
300 years, we still had so much to learn about simple
uniform motion. But the history of science is like that. As we enter the study of more complex forms of motion,
you will notice a shift in the way ideas are presented.
Major discoveries often occur when we see the simple
underlying features after a long struggle with complex Throughout the text, our goal is to construct a modern
situations. If our goal is to present scientific ideas in the view of nature starting as much as possible from the
basic underlying ideas. In our study of special relativ-
orderly progression from the simple to the complex, we
must expect that the historical order of their discovery ity, the underlying idea, the principle of relativity, is
more accurately expressed in terms of your experience
will not necessarily follow the same route.
flying in a jet than it is by any formal set of equations.
Galileo was studying the motion of projectiles, trying As a result we were able to extract the content of the
to predict where cannon balls would land. He devised theory in a series of discussions that drew upon your
a set of experiments involving cannon balls rolling experience.
along slightly inclined planes. These experiments
effectively slowed down the action, allowing Galileo to In most other topics in physics, common experience is
either not very helpful or downright misleading. If you
see the way the speed of a falling object changed as the
object fell. To explain his results Galileo invented the have driven a car, you know where the accelerator
concept of acceleration and pointed out that the simple pedal is located and have some idea about what
2-2 Vectors

acceleration is. But unless you have already learned it VECTORS

in a physics course, your view of acceleration will bear In this chapter we will study the vector as a mathemati-
little relationship to the concept of acceleration devel- cal object. The idea is to have the concept of vectors in
oped by Galileo and now used by physicists. It is our bag of mathematical tools ready for use in our study
perhaps unfortunate that we use the word acceleration of more complex motion, ready to be applied to the
in physics, for we often have to spend more time ideas of velocity, acceleration and later, force and
dismantling the students’ previous notions of accelera- momentum.
tion than we do building the concept as used in physics.
And sometimes we fail. In a sense, we will develop a new math for vectors. We
will begin with a definition of displacement vectors,
The physical ideas that we will study are often simply and will then explain how two vectors are added. From
expressed in terms of mathematical concepts like a this, we will develop a set of rules for the arithmetic of
vector, a derivative, or an integral. This does not mean vectors. In some ways, the rules are the same as those
that we will drop physical intuition and rely on math- for numbers, but in other ways they are different. We
ematics. Instead we will use them both to our best will see that most of the rules of arithmetic apply to
advantage. In some examples, the physical situation is vectors and that learning the vector convention is
obvious, and can be used to provide insight into the relatively simple.
related mathematics. The best way, for example, to
obtain a solid grip on calculus is to see it applied to
physics problems. On the other hand, the concept of a Displacement Vectors
vector, whose mathematical properties are easily de- A displacement vector is a mathematical way of ex-
veloped, is an extremely powerful tool for explaining pressing the separation or displacement between two
many phenomena in physics. objects. To see what is involved in describing the
separation between objects, consider a map such as the
one in Figure (1), which shows the position of the two
cities, New York and Boston. If we are driving on well-
marked roads, it is sufficient, when planning a trip, to
know that these two cities are separated by a distance
of 190 miles. However, the pilot of a small plane flying
from New York to Boston in a fog must know in what
direction to fly; he must also know that Boston is
located at an angle of 54 degrees east of north from New
York.

Corning, NY Boston

Pittsburgh New York

Figure 1
Displacement vectors. Boston and Corning, N. Y., have
equal displacements from New York and Pittsburgh,
respectively. These displacements are located at different
parts of the map, but they are the same displacement.
 Arithmetic of Vectors 2-3

The statement that Boston is located a distance of 190 Arithmetic of Vectors

miles and at an angle of 54 degrees east of north from Suppose that a pilot flies from New York to Boston and
New York provides sufficient information to allow a then to Buffalo. To his original displacement from New
pilot leaving New York to reach Boston in the thickest York to Boston he adds a displacement from Boston to
fog. The separation or displacement between the two Buffalo. What is the sum of these two displacements?
cities is completely described by giving both the dis- After these displacements he will be 300 miles from
tance and the direction. New York at an angle 57 degrees west of north, as
shown in Figure (2). This is the net displacement from
Looking again at Figure (1), we see that Corning, N.Y., New York, which is what we mean by the sum of the
is located 190 miles, at an angle of 54 degrees east of first two displacements.
north, from Pittsburgh. The very same instructions,
travel 190 miles at an angle of 54 degrees, will take a If the pilot flies to five different cities, he is adding
pilot from either Pittsburgh to Corning or New York to together five displacements, which we can represent by
Boston. If we say that these instructions define what we the vectors a, b, c, d, and e shown in Figure (3). (An
mean by the word displacement, then we see that arrow placed over a symbol is used to indicate that the
Corning has the same displacement from Pittsburgh as symbol represents a vector.) Since the pilot’s net
Boston does from New York. (For our discussion we displacement from his point of origin, represented by
will ignore the effects of the curvature of the earth.) The the bold vector, is simply the sum of his previous five
displacement itself is completely described when we displacements, we will say that the bold vector is the
give both the distance and direction, and does not sum of the other five vectors. We will write this sum as
depend upon the point of origin. ( a + b + c + d + e ), but remember that the addition of
vectors is defined graphically as illustrated in
The displacement we have been discussing can be Figure (3).
represented graphically by an arrow pointing in the
direction of the displacement (54 degrees east of north), If the numbers 405 and 190 are added, the answer is
and whose length represents the distance (190 mi). An 595. But, as seen in Figure (2), if you add the vector
arrow that represents a displacement is called a dis- representing the 405-mile displacement from Boston
placement vector, or simply a vector. One thing you to Buffalo to the vector representing the l90-mile
should note is that a vector that defines a distance and displacement from New York to Boston, the result is a
a direction does not depend on its point of origin. In vector representing a 300-mile displacement. Clearly,
Figure (1) we have drawn two arrows; but they both there is a difference between adding numbers and
represent the same displacement, and thus are the same vectors. The plus sign between two numbers has a
vector. different meaning from that of the plus sign between
two vectors.
d

Buffalo e c
405 mi
Boston b
30
0
m 57° 54° mi
i 0 a+
19
New York b+
c+
a

d +
Figure 3 e
Figure 2
The sum of five displacements
Addition of vectors. The vector sum of the
a ,b, c, d, and e equals the
displacement from New York to Boston plus vector a+b+c+d+e.
the displacement from Boston to Buffalo is
the displacement from New York to Buffalo.
2-4 Vectors

Although vectors differ from numbers, some similari- Rules for Vector Arithmetic
ties between the two can be noted, particularly with 1. The commutative law implies that
regard to the rules of arithmetic. First, we will review
the rules of arithmetic for numbers, and then see which a +b =b +a
of these rules also apply to vectors. Figure (4 ) verifies this rule graphically. The reader
should be able to see that a + b and b + a are the same
vectors.
Rules for Number Arithmetic b
1. Commutative law. In adding two numbers, a and b, a a
the order of addition makes no difference.
a+ b
a+b=b+a
b b+ a
2. Associative law. In adding three or more numbers,
a, b, and c, we have a
b
Figure 4
(a + b) + c = a + (b + c)
That is, if we first add a to b, and then add c, we get the
same result as if we had added a to the sum (b + c). 2. The associative law applied to vectors would imply

( a + b) + c = a + (b + c)
3. The negative of a number is defined by
From Figure (5 ) you should convince yourself that this
a + (– a) = 0 law works.
where (– a) is the negative of a. )
( a+ b
b
a + c = c
4. Subtraction is defined as the addition of the negative (a+
b )+ c
number. ( a+ b)
a – b = a + (– b)
These rules are so obvious when applied to numbers b
that it is hard to realize that they are rules. Let us apply a
a ( b+c )
the foregoing rules to vectors, using the method of + ( b+c ) =
c
addition of displacements. a+ (
b+c
)
Figure 5
 2-5

3. The negative of a vector is defined by Multiplication of a

Vector by a Number
a + –a = 0 Suppose we multiply a vector a by the number 5. What
The only way to get a zero displacement is to return to do we mean by the result 5a ? Let us again try to follow
the point of origin. Thus, the negative of a vector is a the rules of arithmetic to answer this question. In
vector of the same length but pointing in the opposite arithmetic we were taught that
direction (Figure 6). 5a=a+a+a+a+a
a
Let us try the same rule for vectors.
–a
Figure 6 5a = a + a + a + a + a
With this definition we see that 5a is a vector in the
4. The subtraction of vectors is now easy. If we want
same direction as a but five times as long (see Figure 8).
a - b , we just find a + –b. That is

a – b = a + –b
5a
To subtract, we just add the negative vector as shown
in Figure (7). a
–b
a –b Figure 8
b (a – b)
We may also multiply a vector by a negative number
a (see Figure 9); the minus sign just turns the vector
Figure 7 around. For example,

– 3a = 3 – a = – a + – a + – a

–a
a –a (–3)a
–a

Figure 9

When we multiply a vector by a positive number, we

merely change the length of the vector; multiplication
by a negative number changes the length and reverses
the direction.
2-6 Vectors

Magnitude of a Vector Example 1

Often we will want to discuss only the length or The vector s starts from point a and we would like to
magnitude of a vector, regardless of the direction in redraw it starting from point b, as shown in Figure (11).
which it is pointing. For example, if we represent the b
displacement of Boston from New York by the vector
s , then the magnitude of s (the length of this displace- s
ment) is 190 miles. We use a vertical bar on each side a
of the vector to represent the magnitude ; thus, we write
s = 190 mi (see Figure 10).
Solution: We want to draw a line through b that is
parallel to s . This can be done with a straightedge and
Boston
triangle as shown in Figure (12).
s = 190 mi s

Figure 10

Ruler or straight edge

New York

Vector Equations
s
Just as we can solve algebraic equations involving
numbers, we can do the same for vectors. Suppose, for a
Figure 12
example, we would like to find the vector x in the
vector equation

2a + 3b + 2x = c
Solving this equation the same way we would any Place the straight edge and triangle so that one side of
other, we get the triangle lies along the straight edge and the other
along the vector s. Then slide the triangle along the
x = 1/2c – a – 3/2b
straight edge until the side of the triangle that was
Graphically, we find (1/2)c , –a , and (–3/2)b ; we then originally along s now passes through b. Draw this line
vectorially add these quantities together to get the through b. If nothing has slipped, the line will be
vector x. parallel to s as shown in Figure (13).

Graphical Work
In the early sections of this text, we shall do a fair
b
amount of graphical work with vectors. As we can see
from the previous examples, the main problem in
graphical work is to move a vector accurately from one s
part of the page to another. This is easily done with a Figure 13 a
plastic triangle and ruler as described in the following
example.
 2-7

We now have the direction of s starting from b. Thus, Exercise 2 Associative law
we have only to put in the length. This is most easily Use the tear out page 2-20 for the vectors of Figure (16),
done by marking the length of s on the edge of a piece find
of paper and reproducing this length, starting from b as
shown in Figure (14). (a) a + b (in black);

(b) (a + b) + c (in red);

s (c) (b + c) (in black);

b (d) a + (b + c) (in blue).

s a
a b
Figure 16

Figure 14 c

Exercise 3 Subtraction
By being careful, using a sharp pencil, and practicing, Use the tear out page 2-21 for the three vectors a , b ,
you should have no difficulty in performing accurate and c shown in Figure (17), find the following vectors
graphically, labeling your results.
and rapid graphical work. The practice can be gained
by doing Problems 1 through 5. (Note that it is essential (a) a + b
to distinguish a vector from a number. Therefore, when
you are solving problems or working on a laboratory (b) a – b b
experiment, it is recommended that you always place (c) b – a
an arrow over the symbol representing a vector.) a c
(d) (a – b) + (b – a)
Figure 17
Exercise 1 Commutative law (e) b + c – a
The vectors a , b ,and c of Figure 15 are shown enlarged
on the tear out page 2-19. Using that page for your work, Exercise 4 Equations
find Suppose that a physical law is given by the vector
b equation
(a) a + b + c (in black);
a Pi = Pf
(b) b + c + a (in red);
Suppose that Pf is the sum of two vectors; that is,
(c) c + a + b (in blue). Pf = Pf 1 + Pf 2
c
(Label all your work.) Figure 15 Given the two vectors Pi and Pf 1 (Figure 18), find Pf 2 .
(These vectors are found on the tear out page 2-22.)
Does the commutative law work?
pi

Figure 18 pf1
2-8 Vectors

Exercise 5 COMPONENTS
Assume that the vectors Pf , Pf 1 , and Pf 2 are related by Another way to work with vectors, one that is espe-
the vector law: cially convenient for solving numerical problems, is
Pf = Pf 1 + Pf 2 through the use of a coordinate system and compo-
nents. To illustrate this method, suppose we were
In addition, the magnitudes of the vectors are related by giving instructions to a pilot on how to fly from New
2 2 2 York to Boston. One way, which we have mentioned,
Pf = Pf 1 + Pf 2
would be to tell the pilot both the direction and the
If you are given Pf and only the direction of Pf 1 (Figure distance she must fly, as “fly at an angle of 54 degrees
19), find Pf 1 and Pf 2 graphically. (These vectors are east of north for a distance of 190 miles.” But we could
found on the tear out page 2-22.) also tell her “fly 132 miles due east and then fly 112
miles due north.” This second routing, which describes
pf the displacement in terms of its easterly and northerly
components, as illustrated in Figure (20), is less direct,
direction but will also lead the pilot to Boston.
of p
f1
Figure 19 We can use the same alternate technique to describe a
vector drawn on a piece of paper. In Figure (20), we
drew two lines to indicate easterly and northerly direc-
tions. We have drawn the same lines in Figure (21), but
now we will say that these lines represent the x and y
directions. The lines themselves are called the x and y
axes, respectively, and form what is called a coordinate
system.

Just as the displacement from New York to Boston had

both an easterly and northerly component, the vector a
in Figure (21) has both an x and a y component. In fact,
the vector a is just the sum of its component vectors ax
and ay:
a = ax + ay (1)

north y

Boston

i
0 m a
112 mi

19 ay

54°
θ
east x
New York 154 mi ax
Figure 20 Figure 21
Two ways to reach Boston from New York. Component vectors. The sum of the component
vectors a x and a y is equal to the vector a .
 2-9

Trigonometry can be used to find the length or magni- Equation 4 gives us a new way to add vectors, as
tude of the component vectors; we get illustrated in Figure (23). Previously we would have
added the vectors a, b, and c directly, as shown in
ax ≡ ax = a cos θ (2) Figure (24). The new rule shows how we can first add
the x components (ax + bx + cx) as shown in Figure
ay ≡ ay = a sin θ (3) (23a), then separately add the y components
(ay + by + cy) as shown in Figure (23b), and then add
Often we will represent the magnitude of a component these vector sums vectorially, as shown in Figure (23c),
vector by not using the arrow, as was done in the to get the vector (a + b + c).
foregoing equations. (The equal sign with three bars,
ax ≡ ax , simply means that ax is defined to be the same ax bx
symbol as ax .) It is common terminology to call the (a)
(ax + bx + cx ) cx
magnitude of a component vector simply the compo-
nent; for example, ax ≡ ax is called the x component
of the vector a.
ay by
Addition of Vectors by Adding Components
An important use of components is as a means for (b) (a y + by + cy ) cy
handling vectors numerically rather than graphically.
We will show how this works by using an example of
the addition of vectors by adding components.
(ax + bx + cx )
Consider the three vectors shown in Figure (22). Since
each vector is the vector sum of its individual compo- (c) (a y + by + cy )
nents vectors, we have (a + b + c )
Figure 23

a = ax + ay b = bx + by c = cx + cy
By adding all three vectors a, b, and c together, we get
b
a
a + b + c = (ax + ay) + (bx + by ) + (cx + cy)
The right-hand side of this equation may be rearranged c
to give (a + b + c )

Figure 24
a + b + c = (ax + bx + cx) + (ay + by + cy) (4)

bx
a by
ay b
ax
cx

Figure 22 cy
c
2-10 Vectors

The advantage of using components is that we can Exercise 6

numerically add or subtract the lengths of vectors that Imagine you are given the vectors a, b, and c and the
point in the same direction. Thus, to add 500 vectors, two sets of coordinate axes (x1, y1) and (x2, y2) shown
we would compute the lengths of all the x components in Figure (27). Using the vectors found on the tear out
and add (or subtract) these together. We would then page 2-23
add the lengths of the y components, and finally, we y1
would vectorially add the resulting x and y compo-
nents. Since the x and y components are at right a
angles, we may find the total length and final direc- x1
tion by using the Pythagorean theorem and trigo- c y2
nometry, as shown in Figure (25). b

y
x2
2
a = a x2 + a y2 Figure 27
a
ay ay
tan θ = a) Find (a + b + c) by direct addition of vectors.
θ ax
x
ax b) Choose x 1 and y1 as your coordinate axes. Find (in
red) the x 1 and y1 components of a, b, c. Then
Figure 25
(i) Find (ax1 + bx1 + cx1)
It is not necessary to always choose the x components (ii) Find (ay1 + by1 + cy1)
horizontally and the y components vertically. We may (iii) Find (ax1 + bx1 + cx1) + (ay1 + by1 + cy1) .
choose a coordinate system (x', y') tilted at an angle, as
How does this compare with (a + b + c) ?
shown in Figure (26). To use the language of the
mathematician, ax' is the component of (or projection c) Repeat part (b) for the coordinate axis ( x 2, y2 ) .
of) a in the direction x'. We see that the vector sum of
all the component vectors still adds up to the vector
itself. Vector Equations in Component Form
Often we will run into a situation where we have a
y' y' vector equation of the form
a c = a+b
coordinate
system a = a x' + a y' but you have to solve the equation using components.
a y'
(x', y') This is easy to do, because to go from a vector equation
a x' to component equations, just rewrite the equation three
x'
x' (or two) times, once for each component. The above
Figure 26 equation becomes
cx = a x + b x
cy = a y + b y
cz = a z + b z
 2-11

VECTOR MULTIPLICATION One of the remarkable discoveries of the twentieth

We have seen how the rules work for vector addition, century is that there is a close relationship between the
subtraction, and the multiplication of a vector by a mathematical properties of scalars, vectors, and ten-
number. Does it make any sense to multiply two sors, and the physical properties of the various elemen-
vectors together? In considering the multiplication of tary particles. Later on we will discuss particles such
the two vectors, the first question to answer is: what is as the π meson now used in cancer research, the photon
the result? What kind of a thing do we get if we multiply which is the particle of light (a beam of light is a beam
a vector pointing east by a vector pointing north? Do of photons), and the graviton, the particle hypothesized
we get a vector pointing in some third direction? Do to be responsible for the gravitational force. It turns out
we get a number that does not point? Or do we get some that the physical properties of the π meson resemble the
quantity more complex than a vector? And perhaps a mathematical properties of a scalar, the properties of
more important question – why would one want to the photon are described by a vector (we will see this
multiply two vectors together? later in the text), and it requires a tensor to describe the
graviton (that is why people working with gravitational
We will see in the study of physics that there are various theories have to work with tensors).
reasons why we will want to multiply vectors, and we
can get various answers. One kind of multiplication One of the surprises of physics and mathematics is that
produces a number; this is called scalar multiplication there are particles like the electron, proton and neutron,
or the dot product. We will see examples of scalar the basic constituents of atoms, that are not described
multiplication shortly. A few chapters later we will by scalars, vectors, or tensors. To describe these
encounter the vector cross product where the result of particles, a new kind of a mathematical object had to be
the multiplication of two vectors is itself a vector, one invented—an object called the spinor. The spinor
that points in a direction perpendicular to the two describing the electron has properties half way be-
vectors being multiplied together. Finally there is a tween a scalar and a vector. No one knew about the
form of multiplication that leads to a quantity more existence of spinors until the discovery was forced by
complex than a vector, an object called a tensor or a the need to explain the behavior of electrons. In this text
matrix. A tensor is an object that maintains the we will not go into the mathematics of spinors, but we
directional nature of both vectors involved in the will encounter some of the unusual properties that
product. Tensors are useful in the formal mathematical spinors have when we study the behavior of electrons
description of the basic laws of physics, but are not in atoms. In a very real sense the spinor nature of
needed and will not be used in this text. electrons is responsible for the periodic table of ele-
ments and the entire field of chemistry.
The names scalar, vector, and tensor describe a hierar-
chy of mathematical quantities. Scalars are numbers In this text we can discuss a great many physical
like, 1, 3, and -7, that have a magnitude but do not point concepts using only scalars or vectors, and the two
anywhere. Vectors have both a magnitude and a kinds of vector products that give a scalar or vector as
direction. Tensors have the basic properties of both a result. We will first discuss the scalar or dot product
vectors used to construct them. In fact there are higher which is some ways is already a familiar concept, and
rank tensors that have the properties of 3, 4, or more then the vector or cross product which plays a signifi-
vectors. People working with Einstein’s generalized cant role later in the text.
gravitational theory have to work all the time with
tensors.
2-12 Vectors

The Scalar or Dot Product To formalize this concept, we will define the scalar
In a scalar product, we start with two vectors, multiply product of the vector a with itself as being the square
them together, and get a number as a result. What kind of the length of a . We will denote the scalar product
of a mathematical process does that involve? The by using the dot symbol to denote scalar multiplica-
Pythagorean theorem provides part of the answer. tion:
Scalar product 2
Suppose that we have a vector a whose x and y ≡ a⋅a ≡ a (6)
of a with itself
components are ax and ay as shown in Figure (28).
Then the magnitude or length a of the vector is given From Equations (4) and (6) we have in the (x, y)
by the Pythagorean theorem as coordinate system
a 2 = ax 2 + ay 2
(4) a ⋅ a = a x 2 + ay 2 (7)

y
In the ( x' , y' ) coordinate system we get
a ay 2 2
a⋅a = a x' + a y' (8)
x
ax
Figure 28 The fact that the length of the vector a is the same in
both coordinate systems means that this scalar or “dot”
product of a with itself has the same value even though
2 2
In some sense a 2 is the product of the vector a with the components or pieces ax2, ay2 or a x' , a y' are
itself, and the answer is a number that is equal to the different. In a more formal language, we can say that
square of the length of the vector a . the scalar product a ⋅ a is unchanged by, or invariant
under changes in the coordinate system. Basically we
Now suppose that we use a different coordinate system can say that there is physical meaning to the quantity
x' , y' shown in Figure (29) but have the same vector a . a ⋅ a (i.e. the length of the vector) that does not depend
In this new coordinate system the length of the vector upon the coordinate system used to measure the vector.
a is given by the formula
2 2 2 Exercise 7
a = a x' + a y' (5)
Find the dot product a⋅a for a vector with components
y' ax, ay, az in three dimensional space. How does the
Pythagorean theorem enter in this case?
a
x' ay'
Figure 29
ax'

The components a' x and a' y are different from ax and

ay, but we know that the length of a has not changed,
thus a 2 must be the same in Equations (4) and (5). We
have found a quantity a 2 which has the same value in
all coordinate systems even though the pieces ax2 and
ay2 change from one coordinate system to another.
This is the key property of what we will call the scalar
product.
 2-13

The example of calculating a ⋅ a above gives us a clue a ′x = a sin θ

to guessing a more general definition of dot or scalar b
a ′y = a cos θ

y'
products when we have to deal with the product of two θ
different vectors a and b . As a guess let us try as a a bx′ = 0
definition by′ = b

x'
a ⋅ b ≡ axbx + ayby (9) Figure 31

or in three dimensions Next choose a coordinate system x ′,y ′ rotated from

x,y by an angle 90 – θ as shown in Figure (31).
Here b lies along the y ′ axis and the dot product is given
a ⋅ b ≡ axbx + ayby + azbz (10)
by

This definition of a dot product does not represent the a⋅b = a x′ b ′x + a y′ b ′y

length of either a or b but perhaps a ⋅ b has the special = ab cos θ + 0
property that its value is independent of the choice of
coordinate system, just as a ⋅ a had the same value in Again we get the result
any coordinate system. To find out we need to calculate
the quantity a′xbx′ + a′yby′ + a′zbz′ in another coordi- a⋅b = ab cos θ (11)
nate system and see if we get the same answer. We will
do a simple case to show that this is true, and leave the Equation (11) holds no matter what coordinate system
more general case to the reader. we use, as you can see by working the following
exercise.

y ax = a Exercise 8
b bx = b cos θ Choose a coordinate system x″, y″ where the x axis
x θ by = b sin θ is an angle φ below the horizontal as shown in Figure
a (32). First calculate the components a″x, ay″, bx″, by″ and
Figure 30
then show that you still get
Suppose we have two vectors a and b separated by an a⋅ b ≡ a″x b″x + a″y b″y = abcos θ
angle q as shown in Figure (30). Let the lengths a and
b be denoted by a and b respectively. Choosing a y''
coordinate system (x, y) where the x axis lines up with b
a , we have θ
Figure 32 φ a
ax = a , ay = 0 x''
bx = b cos θ , by = b sin θ To do this problem, you need the following relation-
ships.
and the dot product, Equation 9, gives
sin θ + φ = sin θ cos φ + cos θ sin φ
a⋅b = a xb x + a yb y cos θ + φ = cos θ cos φ – sin θ sin φ )
= ab cos θ + 0 sin2 φ + cos2 φ = 1 for any angle φ

(This problem is much messier than the example we

did.)
2-14 Vectors

Interpretation of the Dot Product Physical Use of the Dot Product

When a and b are the same vector, then we had We have seen that the dot product a ⋅ b is given by the
a⋅ a = a
2
which is just the square of the length of simple formula a ⋅ b = a b cos θ and it has the special
the vector. If a and b are different vectors but parallel property that
to each other, then θ = 0°, cos θ = 1, and we get
b a ⋅ b ≡ axbx + ayby + azbz
a ⋅ b = ab
a has the same value in any coordinate system even
In other words the dot product of parallel vectors is just though the components ax, bx etc., are different in
the product of the lengths of the vectors. different coordinate systems. The fact that a ⋅ b is the
same number in different coordinate systems means
Another extreme is when
that it is truly a number with no dependence on direc-
the vectors are perpendicu-
tion. That is what we mean by a scalar quantity. This
lar to each other. In this
is a special property because a ⋅ b is made up of the
case θ = 90° , cos θ = 0 b
vectors a and b that do depend upon direction and
and a ⋅ b = 0. The dot θ = 90° whose values do change when we go to different
product of perpendicular
a coordinate systems.
vectors is zero. In a sense
the dot product of two vectors measures the parallelism In physics there are quantities like displacements x,
of the vectors. If the two vectors are parallel, the dot velocities v, forces F that all behave like vectors. All
product is equal to the full product ab. If they are point somewhere and have components that depend
perpendicular, we get nothing. If they are at some upon our choice of direction. Yet we will deal with
intermediate angle, we get a number between ab and other quantities like energy which does not point
zero. anywhere. Energy has a magnitude but no direction.
Yet our formulas for energy involve the vectors x, v,
Increasing θ
and F. How can we construct numbers or scalars from
more, we see
vectors? The answer is - take dot or scalar products of
that if the vectors b
are separated by an θ the vectors. This is the mathematical reason why most
of our formulas for energy will involve dot products.
angle between 90° and a
180° as in Figure (33), then Figure 33
the cos θ and the dot prod- Here cosθ is
uct are negative. A nega- negative.
tive dot product indicates
an anti-parallelism. The
extreme case is θ = 180°
where a ⋅ b = –ab.
 2-15

Vector Cross Product Exercise 9

The other kind of vector product we will use in this What direction would the vector c point if you used your
course is the vector cross product where we multiply left hand rather than your right hand in the above rule?
two vectors a and b together to get a third vector c .
The notation is We said that the vector cross product was not a particu-
larly intuitive concept when you first encounter it. In
a×b = c (12) the above exercise, you see that if by accident you use
where the name cross product comes from the cross we your left hand rather than your right hand, c = a × b
place between the vectors we are multiplying together. will point the other way. One can reasonably wonder
When you first encounter the cross product, it does not how a cross product could appear in any law of physics,
seem particularly intuitive. But we use it so much in for why would nature prefer right hand rules over left
later chapters that you will get quite used to it. Perhaps handed rules. It seems unbelievable that any basic
the best procedure is to skim over this material now, concept should involve anything as arbitrary as the
and refer back to it later when we start using it in various right hand rule.
physics applications. There are two answers to this problem. One is that in
most cases, nature has no preference for right handed-
To define the cross product b ness over left handedness. In these cases it turns out that
a × b = c, we have to define
not only the magnitude but any law of physics that involves right hand rules turns
also the direction of the result- out to involve an even number of them so that any
ing vector c. Starting with a physical prediction does not depend upon whether you
two vectors a and b pointing
Figure 34 used a right hand rule or a left hand rule, as long as you
in different directions as in Figure 34, what unique use the same rule throughout. Since there are more
direction is there for c to point? Should c point half way right handed people than left handed people, the right
between a and b, or should it be closer to a because a is hand rule has been chosen as the standard convention.
longer than b ? No, there is c=a b
nothing particularly unique or c = a b
obvious about any of the direc- b
tions in the plane defined by a b
and b. The only truly unique θ a
direction is perpendicular to a
this plane. We will say that c c=a b
points in this unique direction
as shown in Figure 35. Figure 35

The direction perpendicular to the plane of a and b is b

not quite unique. The vector c could point either up or
down as indicated by the solid or dotted vector in Figure
35. To select between these two choices, we use what
a
is called the right hand rule which can be stated as
follows: Point the fingers of your right hand in the
direction of the first of the two vectors in the cross Figure 36
Right hand rule for the vector cross product.
product a × b (in this case the vector a). Then curl
your fingers until they point in the direction of the
second vector (in this case b ), as shown in Figure 36.
If you orient your right hand so that this curling is
physically possible, then your thumb will point in the
direction of the cross product vector c .
2-16 Vectors

Exercise 10 Until 1956 it was believed that the basic laws of physics
There is left and right handedness in the direction of the
did not distinguish between left and right handedness.
threads on a screw or bolt. In Figure (37a) we show a The fact that there are more right handed than left
screw with a right handed thread. By this, we mean that handed people, or that the DNA used by living organ-
if we turn the screw in the direction that we can curl the isms had a right handed spiral structure (like a right
fingers of our right hand, the screw will move through handed thread) was simply an historical accident. But
wood in the direction that the thumb of our right hand then in 1956 it was discovered that the elementary
points. particle called the neutrino was fundamentally left
handed. Neutrinos spin like a top. If a neutrino is
passing by you and you point the thumb of your left
hand in the direction the neutrino is moving, the fingers
of your left hand curl in the direction that the neutrino
is spinning. Or we may say that the neutrino turns in the
direction of a left handed thread, as shown in Figure 38.

Figure 37a neutrino

Right handed thread. left-handed
screw
In Figure (37b), we have a left hand thread. If we turn direction
the screw in the direction we can curl the fingers of our of motion
left hand, the screw will move through the direction direction
of rotation
pointed by our left thumb.
Figure 38
The neutrino is inherently a left handed object.
When one passes by you, it spins in the direction
that the threads on a left handed screw turn.

Another particle, called the anti-neutrino, is right handed.

If you point the thumb of your right hand in the
direction of motion of an anti-neutrino, the fingers of
your right hand can curl in the direction that the anti-
Figure 37b neutrino rotates. T.D. Lee and N.C. Yang received the
Left handed thread 1957 Nobel prize in physics for their discovery that
some basic phenomena of physics can be used to
For this exercise find some screws and bolts, and distinguish between left and right handedness.
determine whether the threads are right handed or left
handed. Manufacturers use one kind of thread pre- The idea of right or left handedness in the laws of
dominately over the other. Which is the predominant physics will appear in several of our later discussions of
thread? Can you locate examples of the other kind of the basic laws of physics. The point for now is that
thread? (The best place to look for the other kind of having a quantity like the vector cross product that uses
thread is in the mechanism of some water faucets. Can
the right hand convention may be a useful tool to
you find a water faucet where one side uses a right hand
thread and the other a left hand thread? If you find one,
distinguish between left and right handedness.
determine which is the right and which the left hand
thread.)
 2-17

Exercise 11
c = a × b = ab sinθ (13)
Go back to Figure 34 where we show the vectors a and
b , and draw the vector c' = b × a. Use the right hand
rule as we stated it to determine the direction of c' . From where a ≡ a and b ≡ b are the lengths of a and b
your result, decide what happens when you reverse the respectively, and θ is the angle between them. Equa-
order in which you write the vectors in a cross product. tion 13 is the definition we will use for the magnitude
Which of the arithmetic rules does this violate? of the vector cross product.

In Equation 13, we see that not only is the cross product

Magnitude of the Cross Product zero when the vectors are parallel, but is a maximum
Now that we have the right hand rule to determine the when the vectors are perpendicular. In the sense that
direction of c = a × b, we now need to specify the the dot product a ⋅ b was a measure of the parallelism
magnitude of c. of the vectors a and b, the cross product is a measure of
their perpendicularity. If a and b are perpendicular,
plane of vectors
then the length of c is just the product ab. As the vectors
perpendicular to become parallel the length of c reduces to zero.
a and b
Component Formula for the Cross
b
Product
a Sometimes one needs the formula for the components
of c = a × b expressed in terms of the components of
a and b. The result is a mess, and is remembered only
Figure 39 by those who frequently use cross products. The
answer is
A clue as to a consistent definition of the magnitude of
c is the fact that when a and b are parallel, they do not cx = ay bz - az by
define a plane. In this special case there is an entire cy = az bx - ax bz
plane perpendicular to both a and b, as shown in Figure
39. Thus there is an infinite number of directions that cz = ax by - ay bx (14)
c could point and still be perpendicular to both a and b.
These formulas are not so bad if you are doing a
We can avoid this mathematical ambiguity only if c has
computer calculation and you are letting the computer
zero magnitude when a and b are parallel. We do not
evaluate the individual components.
care where c points if it has no length.
Exercise 12
c
Assume that a points in the x direction and b is in the
xy plane as shown in Figure 41. By the right hand rule,
b c will point along the z axis as shown. Use Equation
θ 14 to calculate the magnitude of cz and compare your
a result with Equation 13.
Figure 40

z
c
The simplest formula for the magnitude c = a × b ,
that is related to the product of a and b , yet has zero y y b
length when a and b are parallel is x θ a
Figure 41
2-18 Vectors

RIGHT HANDED COORDINATE

SYSTEM
Notice in Figure 41, we have drawn an (x, y, z)
coordinate system where z rises up from the xy plane.
We could have drawn z down and still have three
perpendicular directions. Why did we select the up-
ward direction for z?

The answer is that the coordinate system shown in

Figure 41 is a right hand coordinate system, defined
as follows. Point the fingers of your right hand in the
direction of the first coordinate axis (x). Then curl your
fingers toward the second coordinate axis (y). If you
have oriented your right hand so that you can curl your
fingers this way, then your thumb points in the direction
of the third coordinate axis (z).

The importance of using a right handed coordinate

system is that Equation 14 for the cross product ex-
pressed as components works only for a right handed
coordinate system. If by accident you used a left
handed coordinate system, the signs in the equation
would be reversed.

Exercise 13
Decide which of the (x, y, z) coordinate systems are right
handed and which are left handed.
z z z

y
y x x
x y
(a) (b) (c)

y y

z
x x
z
(d) (e)
Figure 42
 Tear out page 2-19

Figure 15
Vectors for Exercise 1, page 7. Find
(a) a + b + c (in black);

(b) b + c + a (in red);

(c) c + a + b (in blue).

b
a

c
2-20 Vectors Tear out page

Figure 16
Vectors for Exercise 2, page 7. Find
(a) a + b (in black);

(b) (a + b) + c (in red);

(c) (b + c) (in black);

(d) a + (b + c) (in blue).

a
b

c
 Tear out page 2-21

Figure 17
Vectors for Exercise 3, page 7. Find
(a) a + b

(b) a - b

(c) b - a

(d) (a - b) + (b - a)

(e) b + c - a

a c
2-22 Vectors Tear out page

Figure 18
Vectors for Exercise 4, page 7.
Pi = Pf
Suppose that P f is the sum of two vectors; that is,

Pf = Pf1 + Pf2
Given the two vectors P i and P f1 (Figure 18), find P f2 .

pi

pf1

Figure 19
Vectors for Exercise 5, page 8.
Pf = Pf1 + Pf2
In addition, the magnitudes of the vectors are related by
2 2 2
Pf = Pf1 + Pf2
If you are given Pf and only the direction of P f1 , find P f1
and P f2 graphically.

pf
direction
of p
f1
 Tear out page 2-23

Figure 27
Vectors for Exercise 6, page 10.
(a) Find a + b + c by direct addition of vectors. y1
(b) Choose x1 and y1 as your coordinate axes. Find (in
red) the x1 and y1 components of a, b, c. Then

(i) Find ax1 + bx1 + cx1

x
(ii) Find ay1 + by1 + cy1 1
(iii) Find (ax1 + bx1 + cx1 ) + (ay1 + by1 + cy1 ). y2
How does this compare with (a + b + c)?

(c) Repeat part B for the coordinate axis (x2 , y2). (you
can use the back side of this page.)

x2
a

c
b
2-24 Vectors Tear out page

Figure 27
Vectors for Exercise 6, page 10, repeated.
(a) Find a + b + c by direct addition of vectors. y1
(b) Choose x1 and y1 as your coordinate axes. Find (in
red) the x1 and y1 components of a, b, c. Then

(i) Find ax1 + bx1 + cx1

x1
(ii) Find ay1 + by1 + cy1

(iii) Find (ax1 + bx1 + cx1 ) + (ay1 + by1 + cy1 ). y2

How does this compare with (a + b + c)?

(c) Repeat part B for the coordinate axis (x2, y2).

x2
a

c
b
Figure 1
Marcel Duchamp, Nude Decending a Staircase
Philadelphia Museum of Art: Louise and Walter
Arensberg Collection
Chapter 3
Description of Motion

CHAPTER 3 DESCRIPTION OF MOTION

On the facing page is a reproduction (Figure 1) of In the next few chapters, a similar technique will be
Marcel Duchamp’s painting, Nude Descending a used to describe motion. We now have devices
Staircase, which was first displayed in New York at available, such as the stroboscope (called the strobe),
The International Exhibition of Modern Art, gener- that produce short bursts of light at regular inter-
ally known as the Armory Show, in 1913. The objec- vals; with the strobe, we can photograph the succes-
tive of the painting, to convey a sense of motion, is sive positions of an object, such as a ball moving on
achieved by repeating the stylized human form five the end of a string (see Figure 2). Although we do not
times as it descends the steps. At the risk of obscur- have the artist’s freedom of expression to convey the
ing the artistic qualities of the painting, we may concept of motion by using a strobe photograph, we
imagine this work as a series of five flash photo- do obtain a more accurate measure of the motion.
graphs taken in sequence as the model walked
downstairs.

Figure 2
Strobe photograph showing the motion of a ball on the end of a string.
3-4 Description of Motion

Figure 3
Strobe photograph of a moving object. In this photograph, the time
between flashes is so long that the motion is difficult to understand.

The photograph in Figure (2) was taken with the out the picture; that is, in Figure (2) at intervals of
strobe flashing five times per second while the ball 1/5 sec and in Figure (4) at intervals of 1/15 sec. For
was moving slowly. As a result, we see a smooth this text, we will use a special symbol, ∆t, to repre-
curve and have a fairly complete idea of the ball’s sent the time interval between flashes of the strobe.
entire motion. When we run the strobe at a rate of The t stands for time, while the ∆ (Greek letter delta)
five flashes per second but move the ball more indicates that these are short time intervals between
rapidly in a complicated pattern, the result is as flashes. Thus, ∆t = 1/5 sec in Figures (2) and (3), and
shown in Figure (3). From this picture it is difficult ∆t = 1/15 sec in Figure (4).
to guess the ball’s path; thus Figure (3) provides us
with a poor representation of the motion of the ball. For objects that are moving slowly along fairly
But if we turn the strobe up from 5 to 15 flashes per smooth paths, we can use fairly long time intervals
second (as in Figure 4), the rapid and complicated ∆t between strobe flashes and their motion will be
motion of the ball is easily understood. adequately described. As the motion becomes faster
and more complicated, we turn the strobe up to a
The motion of any object can be described by locat- higher flashing rate to follow the object, as in Figure
ing its position at successive intervals of time. A (4). To study complicated motion in more detail, we
strobe photograph is particularly useful because it locate the position of the object after shorter and
shows the position at equal time intervals through- shorter time intervals ∆t.
 3-5

Figure 4
DISPLACEMENT VECTORS Strobe photograph of a similar motion. In this
When we represent the motion of an object by photograph, the time between flashes was reduced
a strobe photograph, we are in fact representing and the motion is more easily understood.
this motion by a series of displacements, the
successive displacements of the object in equal
intervals of time. Mathematically, we can de-
scribe these displacements by a series of dis-
placement vectors, as shown in Figure (5). This
Figure 5
illustration is a reproduction of Figure (2) with Displacement vectors. The displacement between
the successive displacement vectors drawn from flash number 1 and flash number 2 is represented
the center of the images. by the displacement vector s 1 and so on. The
entire path taken by the ball is represented by the

s1 s2 series of eight displacement vectors.

s3
2 3 s4
1 4
s5 s8
5 s7 9
s6
8
6
7
3-6 Description of Motion

s1

(a) (b)

s2

s1 s1

(c) (d)

s1
s1

(e) (f)
Figure 6
Representation of the path of a ball for various ∆t. As the shorter and shorter ∆t is
used, the path of the ball is more accurately represented, as in figures (b) through (d).
 3-7

In a sense we are approximating the path of the ball seven displacement vectors in (e) represent the path
by a series of straight lines along the path. This is fairly accurately. When we reduce ∆t to 1/15 sec (f),
reasonably accurate provided that ∆t is short enough, we get more pictures of the same kink and the curve
as shown in Figure (6). becomes smoother. It now appears that in most
places the 14 displacement vectors form a fairly
In Figure (6), (a) is the strobe photograph shown in accurate picture of the true path. We notice, how-
Figure (4), taken at a strobe interval of ∆t = 1/15 sec; ever, that the very bottom of the kink is cut off
(b) shows how this photograph would have looked if abruptly; here, shorter time intervals are needed to
we had set the strobe for ∆t = 10/15 sec, or 2/3 sec. get an accurate picture of the motion.
Only one out of ten exposures would have been
produced. If we had represented the path of the ball A Coordinate System
by the vector s1 it would have been a gross misrep-
In the strobe photographs discussed so far, we have
resentation. In (c), which would be the strobe
a precise idea of the time scale, 1/5 second between
picture at ∆t = 6/15 sec, we see that the ball is no flashes in Figure (2), 1/15 second in Figure (4), but
longer moving in a straight line, but still s1 and s2 no idea about the distance scale. As a result we know
provide a poor representation of the true motion. the direction of the succeeding displacement vec-
Cutting ∆t in half to get (d), ∆t = 3/15 sec, we would tors, but do not know their magnitude.
discover that there is a kink in the path of the ball.
While taking the picture, we would have had to be One way to introduce a distance scale is to photo-
careful in noticing the sequence of positions in order graph the motion in front of a grid as shown in Figure
to draw the correct displacement vectors. (7). With this setup we obtain photographs like that
shown in Figure (8), where we see the strobe motion
Reducing ∆t to 2/15 sec (e), would give us a more of a steel ball projectile superimposed on the grid.
detailed picture of the kink. This is not too different The grid is illuminated by room lights which are
from (d); moreover, we begin to suspect that the dimmed to balance the exposure of the grid and the
strobe flashes.

Figure 7
Experimental setup for taking strobe photographs. A
Polaroid camera is used record the motion of a ball Figure 8
moving in front of a grid. The grid, made of stretched Strobe photograph of a steel ball projectile.
fish line, is mounted in front of a black painted wall. The strobe flashes were 1/10 second apart.
3-8 Description of Motion

Using techniques like that illustrated in Figure (9) to (10). The grid, which has now become the graph
locate the centers of the images, we can transfer the paper in Figure (10), serves as our coordinate system
information from the strobe photograph to graph for locating the images.
paper and obtain the results shown in Figure (10).
Figure (10) is the end result of a fair amount of
tedious lab work, and the starting point for our
analysis. For those who do not have strobe facilities, Manipulation of Vectors
or the time to extract the information from a strobe Figure (10) represents the kind of experimental data
photograph, we will include in the text a number of upon which we will base our description of motion.
examples already transferred to graph paper in the We have, up to now, described the motion of the
form of Figure (10). projectile in terms of a series of displacement vec-
tors labeled s-1 , s0 , s3 as shown. To go further, to
Using a television camera attached to an Apple II introduce concepts like velocity and acceleration,
computer, we can, in under 2 minutes, obtain results we need to perform certain routine operations on
that look like Figure (10). We will include a few of these displacement vectors, like adding and sub-
these computer strobe photographs in our examples tracting them. A number of vector operations were
of motion. However the computer strobe is not yet discussed in Chapter 2, let us briefly review here
commercially available because we plan to use a those that we need for the analysis of strobe photo-
computer with more modern graphics capabilities. It graphs. We will also introduce the concept of a
is likely that within a few years, one will be able to coordinate vector which will be useful in much of
easily and quickly obtain results like those in Figure our work.

0 s0 1
s1
2

s2

0 10 20

Figure 9
Using a pin and cylinder to locate the center of the
ball. Move the cylinder until it just covers the image of
the ball and then gently press down on the pin. The Figure 11
pin prick will give an accurate location for the center. Measuring the length of the vector S1 .
 3-9

Measuring the Length of a Vector the individual vectors from a graph like Figure (10).
One of the first pieces of information we need from Take a piece of scrap paper and mark off the length
a strobe photograph is the magnitude or length of the of the vector as shown in the upper part of the figure.
displacement vectors we have drawn. Figure (11) Then rotate the paper until it is parallel to the grid
illustrates the practical way to obtain the lengths of lines, and note the distance between the marks.

0 10 20 30 40 50 60 70 80 90 100

0 s0 1
90
s-1 s1 90

-1 2
80 80

s2
70 70

3
60 60

50 s3 50

40 40

4
30 30
Ball coordinates
-1) ( 8.4, 79.3)
20 0) (25.9, 89.9) 20
1) (43.2, 90.2)
2) (60.8, 80.5)
3) (78.2, 60.2)
10 10
4) (95.9, 30.2)

0 10 20 30 40 50 60 70 80 90 100

Figure 10
Strobe photograph transferred to graph paper. Using the pin and cylinder of Figure (9), we located the
coordinates of the center of each image in Figure (8), and then reconstructed the strobe photograph as
shown. We can now perform our analysis on the large graph paper rather than the small photograph.
3-10 Description of Motion

In Figure (11), we see that the marks are 20 small For those who are mathematically inclined, this
grid spacings apart. In Figure (10), we see that each simple graphical work with vectors may seem el-
grid spacing represents a distance of 1 centimeter. ementary, especially compared to the exercises en-
Thus in Figure (11), the vector s1 has a magnitude of countered in an introductory calculus course. But,
20 centimeters. We can write this formally as as we shall see, this graphical work emphasizes the
basic concepts. We will have many opportunities
s 1 = 20 cm later to extract sophisticated formulas from these
This technique may seem rather simple, but it works basic graphical operations.
well and you will use it often.
For these exercises, you may use the practice
Graphical Addition and Subtraction graph on page 3-28, and the tear out sheet on page
3-29.
Since we are working with experimental data in
graphical form, we need to use graphical techniques Exercise 1
to add and subtract vectors. These techniques,
Find the magnitudes of the vectors s 0, s 1, s 2, and s 3
originally introduced in Chapter 2, are reviewed
here in Figures (12) and (13). Figure (12a) and (12b) in Figure (10).
show the addition of two vectors by placing them
head to tail. Think of the vectors A and B as separate Exercise 2
trips; the sum A + B is our net displacement as we Explain why the vector s 0 4 , given by
take the trips A and B in succession. To subtract B s0 4 = s0 + s1 + s2 + s3
from A, we simply add (–B) to A as shown in Figure
(12c). has a magnitude of 91.3 cm which is quite a bit less
than the sum of the lengths s 0 + s 1 + s 2 + s 3 .
To perform vector addition and subtraction, we need
to move the vectors from one place to another. This Exercise 3
is easily done with a triangle and a straight edge as Use graphical methods to find the vector s 3 - s 2 .
indicated in Figure (13). The triangle and straight (The result should point vertically downward and
edge allows you to draw a parallel line; then mark a have a length of about 10 cm.)
piece of paper as in Figure (11), to make the new
vector have the same length as the old one.

B
A
(b)
A+B
A
B
–B

(a) (A – B)
(c)
B
A
Figure 13
Figure 12 Moving vectors around. (This
Addition and subtraction of vectors. was discussed in Figure 2-12.)
 3-11

Coordinate System ANALYSIS OF

and Coordinate Vectors STROBE PHOTOGRAPHS
A coordinate system allows us to convert graphical In our analysis of the strobe photograph of projectile
work into a numerical calculation that can, for ex- motion, Figure (10), we are representing the path of
ample, be carried out on a computer. Figure (14) the ball by a series of displacement vectors S 0 ... S 3
illustrates two convenient ways of describing the (We will think of the photograph as starting at point
location of a point. One is to give the x and y (0). The point labeled (-1) will be used later in our
coordinates of the point (x,y), and the other is to use calculation of the instantaneous velocity at point (0).
a coordinate vector R which we define as a vector In a sense, we “know” that the ball actually went
that is drawn from the origin of the coordinate along a smooth continuous curve, and we could have
system to the point of interest. represented the curve more accurately by reducing
Figure (15) illustrates the way an arbitrary vector S can ∆ t as we did in Figure (6). But with many images to
mark the trajectory, each displacement vector S i
be expressed in terms of coordinate vectors. From the
becomes too short for accurate graphical work. In
diagram we see that R2is the vector sum of R1 + S, thus
taking a strobe photograph, one must reach a com-
we can solve for the vector S to get the result
promise where the displacement vectors S i are long
S = R2 – R1. enough to work with, but short enough to give a
reasonable picture of the motion.

Velocity
Y The series of displacement vectors in Figure (10)
show not only the trajectory of the projectile, but
(X,Y)
because the images are located at equal time inter-
vals, we also have an idea of the speed of the
R projectile along its path. A long displacement vector
R ≡ (X,Y)
indicates a higher speed than a short one. For each
of the displacement vectors we can calculate what
X one would call the average speed of the projectile
Figure 14 during that interval.
The coordinate vector R , which starts
at the origin, locates the point (x,y). The idea of an average speed for a trip should be
fairly familiar. If, for example, you went on a trip for
a total distance of 90 miles, and you took 2 hours,
Y you divide 90 miles by 2 hours to get an average
speed of 45 miles per hour. For more detailed
R1 + S = R 2 information about your speed, you break the trip up
R1 S
into small segments. For example, if you wanted to
S = R 2 – R1 know how fast you were moving down the interstate
R2
highway, you measure how long it takes to pass two
X consecutive mile markers. If it took one minute,
Figure 15 then your average speed during this short time
Expressing the vector S in terms of coordinate vectors. interval is one mile divided by 1/60 hour which is 60
miles per hour. If you broke the whole trip down into
1 minute intervals, measured how far you went
during each interval, and calculated your average
speed for each interval, you would have a fairly
complete record of your speed during your trip. It is
3-12 Description of Motion

this kind of record that we get from a strobe photo- describes both the direction of motion and the speed.
graph of the motion of an object. We will also use this convention throughout the
text.)
In physics, we use a concept that contains more
information than simply the speed of the object. We In our discussion of strobe photographs, we noted
want to know not only how many miles per hour or that if we used too long a time interval ∆t, we got a
centimeters per second an object is moving, but also poor description of the motion as in Figures (6b) and
what direction the object is moving. This informa- (6c). As we used shorter time intervals as in Figures
tion is all contained in the concept of a velocity (6d, e, and f), we got a better and better picture of the
vector. path.

To construct a velocity vector for the projectile We have the same problem in dealing with the
shown in Figure (10), when, for example, the ball is velocity of an object. If we use a very long ∆ t, we
at position 1, we take the displacement vector S 1, get a crude, average, description of the object’s
divide it by the strobe time interval ∆ t, to get what velocity. As we use a shorter and shorter ∆ t, our
we will call the velocity vector v1 : description of the velocity, Equation (1), becomes
more and more precise.
v1 ≡ S 1 (1) Since, in this chapter, we will be working with
∆t
experimental data obtained from strobe photographs,
In Equation (1), what we have done is multiply the there is a practical limit on how short a time interval
vector S 1 by the number (1/∆ t) to get v1 . From our ∆ t we can use and have vectors big enough to work
earlier discussion of vectors we know that multiply- with. We will see that, for the kinds of motion that
ing a vector by a number gives us a vector that points we encounter in the introductory physics lab, a
in the same direction, but has a new length. Thus v1 reasonably short ∆ t like .1 sec gives reasonably
is a vector that points in the same direction as S 1, but accurate results.
it now has a length given by
If you make more precise measurements of the
position of an object you generally find that as you
S1
v1 = = 20 cm = 200 cm use shorter and shorter ∆ t to measure velocity, you
∆t .1 sec sec
reach a point where the velocity vector no longer
where we used S 1 = 20 cm from Figure (11) and changes. What happens is that you reach a point
we knew that ∆t = .1 sec for this strobe photograph. where, if you cut ∆ t in half, the particle goes in the
same direction but only half as far. Thus both the
Not only have we changed the length of S1 by displacement S 1 and the time interval ∆ t are both cut
multiplying by (1/∆t), we have also changed the in half, and the ratio v1 = S 1 / ∆t is unchanged. This
dimensions from that of a distance (cm) to that of a limiting process, where we see that the velocity
speed (cm/sec). Thus the velocity vector v1 contains vector changes less and less as ∆ t is reduced, is
two important pieces of information. It points in the demonstrated graphically in our discussion of in-
direction of the motion of the ball, and has a length stantaneous velocity at the end of the chapter.
or magnitude equal to the speed of the ball.

(Physics texts get rather picky over the use of the Exercise (4)
words speed and velocity. The word speed is re- What is the magnitude of the velocity vector v3 , for the
served for the magnitude of the velocity, like 200 ball in Figure (10). Give your answer in cm/sec.
cm/sec. The word velocity is reserved for the veloc-
ity vector as defined above; the velocity vector
 3-13

Equation (1) is well suited for graphical work but for Acceleration
numerical calculations it is convenient to express S i in In Chapter 1 on Einstein’s special theory of relativ-
terms of the coordinate vectors R i. This is done in ity, we limited our discussion to uniform motion,
Figure (16), where we see that the vector sum motion in a straight line at constant speed. If we took
R i + S i = R i+1 thus S i = R i+1 – R i and Equation (1) a strobe photograph of an object undergoing uni-
becomes form motion, we would get a result like that shown
R i+1 – R i in Figure (17). All the velocity vectors would point
vi = (2) in the same direction and have the same length. We
∆t
will, from now on, call this motion with constant
If we call R i+1 – R i the “change in the position R velocity , meaning that the velocity vector is con-
stant, unchanging.
during the time ∆t ”, and denote this change by ∆R,
Equation (2) becomes
V1 V2 V3 V4 V5 V6 V7

= ∆R
R i+1 – R i
vi = (2a) Figure 17
∆t ∆t
Motion with constant velocity.
which is perhaps a more familiar notation for those
who have already studied calculus. In a calculus From the principle of relativity we learned that there
course, one would define the velocity vi by taking is something very special about motion with con-
the limit as ∆ t → 0 (i.e., by turning the strobe stant velocity—we cannot feel it. Recall that one
flashing rate “all the way up”). In our experimental statement of the principle of relativity was that there
work with strobe photographs, we reduce ∆ t only is no experiment that you can perform to detect your
to the point where we have a reasonable representa- own uniform motion relative to empty space. You
tion of the path; using too short a time interval makes cannot tell, for example, whether the room you are
the experimental analysis impossible. sitting in is at rest or hurdling through space at a
speed of 100,000 miles per hour.

i Although we cannot feel or detect our own uniform

motion, we can easily detect non uniform motion.
S i = R i +1 – R i We know what happens if we slam on the brakes and
come to a sudden stop—everything in the car falls
Ri forward. A strobe photograph of a car using the
brakes might look like that shown in Figure (18a).
Each successive velocity vector gets shorter and
R i +1 shorter until the car comes to rest.

V1 V2 V3 V4 V5
Si (R i + 1 – R i )
Vi = = Figure 18a
t t
Put on the brakes, and your velocity changes.
Figure 16
Expressing the velocity vector vi , in terms
of the coordinate vectors R i and R i+ 1.
3-14 Description of Motion

Another way the velocity of a car can change is by moving due east at 60 miles per hour, as indicated in
going around a corner as illustrated in Figure (18b). Figure (20a). Now we have a non zero change in
In that figure the speed does not change, each veloc- velocity ∆v as indicated in Figure (20b).
ity vector has the same length, but the directions are
changing. It is also easy to detect this kind of change In our two examples, we find that if we have uniform
in velocity—all the packages in the back seat of your motion which we cannot feel, the change in velocity
car slide to one side of the seat. ∆v is zero. If we have non uniform motion, ∆v is not
zero and we can feel that. Is it ∆v, the change in
velocity, that we feel?
V1
V2 Almost, but not quite. Let us look at our second
example, Figure (20), more carefully. There are two
V3 distinct ways that our velocity can change from
pointing south to pointing east. In one case there
could have been a gradual curve in the road. It may
V4 have taken several minutes to go around the curve
and we would be hardly aware of the turn.
Figure 18b
When you drive around a corner, your speed may not
In the other extreme, we may have been driving
change, but your velocity vector changes in direction.
south, bounced off a stalled truck, and within a
fraction of a second finding ourselves traveling due
The point we want to get at is, what do we feel when our
east. In both cases our change in velocity ∆v = vf – vi
velocity changes? Consider two examples. In the first, is the same, as shown in Figure (20b). But the effect
we are moving at constant velocity, due east at 60 miles on us is terribly different. The difference in the two
per hour. A strobe photograph showing our initial and cases is that the change in velocity ∆v occurred
final velocity vectors vi and vf would look like that in much more rapidly when we struck the truck than
Figure (19a). If we define the change in velocity ∆v by when we went around the curve. What we feel is not
the equation ∆v alone, but how fast ∆v happens.
∆v ≡ vf - vi If we take the change in velocity ∆v and divide it by
then from Figure (19b) we see that ∆v = 0 for the time ∆t over which the change takes place, then
uniform motion. the smaller ∆t, the more rapidly the change takes
place, the bigger the result. This ratio ∆v/∆t which
For the second example, suppose we are traveling more closely represents what we feel than ∆v alone,
due south at 60 miles per hour, and a while later are is given the special name acceleration.
Vi
Vi Vf
(a) (a)
Vf

Vf
(b) ∆V ≡ Vf – Vi = 0
V
i
f –

– Vi
V

(b)
=

–Vi
∆V

Figure 19
We see that ∆v = 0 for motion with constant velocity.
Vf
Figure 20
∆v ≠ 0 when we change our direction of motion.
 3-15

The physicists’ use of the word acceleration for the The Acceleration Vector
quantity ∆v/∆t presents a problem for students. The The quantity ∆v/∆t, which we call acceleration, is
difficulty is that we have grown up using the word usually denoted by the vector a
acceleration, and already have some intuitive feel-
ing for what that word means. Unfortunately this
a = ∆v (3)
intuition usually does not match what physicists ∆t
mean by acceleration. Perhaps physicists should
have used a different name for ∆v/∆t, but this did not where ∆v is the change in the velocity vector during
happen. The problem for the student is therefore not the time ∆t. To see how to apply Equation (3) to a
only to develop a new intuition for the quantity strobe photograph, suppose that Figure (21) repre-
∆v/∆t, but also to discard previous intuitive ideas of sents a photograph of a particle moving with some
what acceleration might be. This can be uncomfort- kind of non uniform velocity. Labeling the image
able. positions 1, 2, 3, etc. and the corresponding velocity
vectors v1 , v2 , v3 , let us consider what the
The purpose of the remainder of this chapter is to particle’s acceleration was during the time it went
develop a new intuition for the physics definition of from position 2 to position 3. At position 2 the
acceleration. To do this we will consider three particle’s velocity was v2 . When it got to position 3
examples of motion; projectile motion, uniform its velocity was v3 . The time it took for the velocity
circular motion, and projectile motion with air resis- to change from v2 to v3 , a change ∆v = v3 - v2 , was
tance. In each of these cases, which can be carefully the strobe time ∆t. Thus according to Equation (3),
studied in the introductory lab or simulated, we will the particle’s acceleration during the interval 2 to 3,
use strobe photographs to determine how the accel- which we will call a3, is given by
eration vector ∆v/∆t behaves. In each case we will
a 3 ≡ ∆v2 3 = 3 2
see that there is a simple relationship between the v –v
(4a)
behavior of the acceleration vector and the forces ∆t ∆t
pulling or pushing on the object. This relationship
between force and acceleration, which is the corner- (One could object to using the label a3 for the
stone of mechanics, will be discussed in a later acceleration during the interval 2 to 3. But a closer
chapter. Here our goal is to develop a clear picture inspection shows that a3 is an accurate name. Actu-
of acceleration itself. ally the velocity v3 is the average velocity in the
interval 3 to 4, and v2 is the average velocity in the
interval 2 to 3. Thus ∆v = v3 - v2 is a change in
Determining Acceleration velocity centered on position 3. As a result Equation
from a Strobe Photograph (4a) gives surprisingly accurate results when work-
We will use strobe photographs to provide an ex- ing with experimental strobe photographs. In any
plicit experimental definition of acceleration. In the case such errors become vanishingly small when we
next chapter we will see how the strobe definitions use sufficiently short ∆t 's.)
go over to the calculus definition that you may have
already studied. We prefer to start with the strobe
definition, not only because it provides a more 1
intuitive approach to the concept, but also because of V1
its experimental origin. With an experimental defi-
nition we avoid some conceptual problems inherent 2
in calculus. It turns out, surprisingly, that some of V2
the concepts involved in the calculus definition of V3
acceleration are inconsistent with physics. We can V4
3
more clearly understand these inconsistencies when Figure 21 4
we use an experimental definition of acceleration as Determining a for non uniform motion.
the foundation for our discussion.
3-16 Description of Motion

If we have a strobe photograph with many images, PROJECTILE MOTION

then by extending Equation (4a), the acceleration at As our first example in the use of our strobe defini-
position i is tion of acceleration, let us calculate the acceleration
of the ball at position 2 in our strobe photograph,
Figure (10), of projectile motion.
∆v i v –v strobe definition
ai ≡ = i i–1 (4)
∆t ∆t of acceleration The first problem we face is that Equation (4)
expresses the acceleration vector a2 in terms of the
velocity vectors v1 and v2 , while the strobe photo-
We will call Equation (4) our strobe definition of graph shows only the displacement vectors S 1 and
acceleration. Implicit in this definition is that we S 2, as seen in Figure (10a), a segment of Figure (10)
use a short enough ∆t so that all the kinks in the reproduced here.
motion are visible, but a long enough ∆t so that we
The easiest way to handle this problem is to use the
have vectors long enough to work with.
formulas

v1 = S 1 ; v2 = S 2
∆t ∆t
1
in Equation (4a) to express a2 directly in terms of the
s1 known vectors S1 and S2 . The result is
2
v2 – v1 S /∆t – S2/∆t
a2 = = 1
∆t ∆t
s2
S2 – S1 experimental
a2 = measurment (5)
3 ∆t 2 of acceleration

Equation (5) tells us that we can calculate the accelera-

Figure 10a tion vector a2 by first constructing the vector S2 – S 1,
A section of the projectile motion photograph, Figure and then dividing by ∆t 2 . That means that a2 points in
(10), showing the displacement vectors S1 and S2 . the direction of the vector S2 – S1 , and has a length
equal to the length S2 – S1 (in cm) divided by ∆t 2.
1 As a result the magnitude of the acceleration vector has
s1 the dimensions of cm /sec2.
2 Let us apply Equation (5) to our projectile motion
photograph, Figure (10), to see how all this works.
The first step is to use vector subtraction to construct
s2 – s1 s2 the vector S2 – S 1. This is done in Figure (22). First
we draw the vectors S1 and S2 , and then construct
the vector –S 1 as shown. (The vector –S 1 is the
– s1 3 same as S1 except that it points in the opposite
Figure 22 direction.) Then we add the vectors S2 and –S 1 to
The vector S2 – S1 points straight down get the vector S2 – S 1 by the usual technique of
and has a length of about 10 cm. vector addition as shown

S2 – S1 = S 2 + –S 1 (6)
 3-17

Note that even if S 2 and S 1 had the same length, the UNIFORM CIRCULAR MOTION
difference S 2 – S 1 would not necessarily be zero be- To give the reader some time to think about the
cause this is vector subtraction, NOT NUMERICAL above exercise on projectile motion, we will change
SUBTRACTION. the topic for a while and analyze what is called
Once we have constructed the vector S 2 – S 1 , we know uniform circular motion. In uniform circular mo-
the direction of the acceleration vector a2 because it tion, the particle travels like a speck of dust sitting on
points in the same direction as S 2 – S 1 . In Figure (22), a revolving turntable.
we see that S 2 – S 1 points straight down, thus a2 points The explicit example we would like to consider is a
straight down also. golf ball with a string attached, being swung in a
Now that we have the direction of a2, all that is left circle over the instructor’s head, as indicated in
is to calculate its magnitude or length. This magni- Figure (23a). We could photograph this motion, but
tude is given by the formula it is very easy to simulate a strobe photograph of
uniform circular motion by drawing a circle with a
compass, and marking off equal intervals as shown
S2 – S1 length of vector S 2 – S 1 in Figure (23b). In that figure we have also sketched
a2 = = (7)
∆ t2 ∆ t2 in the displacement vectors as we did in our analysis
of the projectile motion photograph.
To get the length of the S 2 – S 1 , we can use the
technique shown in Figure (11). Mark off the length
of the vector S 2 – S 1 on a piece of scrap paper, and Golf ball
then use the grid to see how many centimeters apart
r g
the marks are. In this case, where S 2 – S 1 points rin V
St
straight down, we immediately see that S 2 – S 1 is
about 10 cm long. Thus the magnitude of a2 is given
by

S2 – S1
a2 = = 10 cm 2 = 1000 cm (8) Figure 23a
∆t 2 .1 sec sec2 Swinging a golf ball around at constant speed in a
circle.
where we knew ∆ t = .1 sec for the strobe photograph
in Figure (10).

Our conclusion is that, at position 2 in the projectile S3

S2 S4
motion photograph, the ball had an acceleration a2
that pointed straight down, and had a magnitude of S1
about 1000 cm/sec2.

For this exercise, you may use the tear out sheet
on page 3-30.
Figure 23b
Exercise 5
Simulating a strobe photograph of a golf
(Do this now before reading on.) Find the ball swinging at constant speed in a circle.
acceleration vectors a0,a1, and a3 for the projectile We marked off equal distances using a
motion in Figure (10). From your results, what can you compass.
say about the acceleration of a projectile?
3-18 Description of Motion

Figure (23b) shows the kind of errors we have to deal Exercise (6)
with in using a strobe to study motion. Clearly the
(Do this now.) Find the direction of at least 4 more
golf ball travels along the smooth circular path
acceleration vectors around the circle. In each case
rather than the straight line segments marked by the show that ai points toward the center of the circle.
vectors . As we use shorter and shorter ∆ t our
approximation of the path gets better and better, but We said earlier that the physicists’ definition of
soon the vectors get too short for accurate graphical acceleration, which becomes a i = Si – S i–1 /∆ t2,
work. Choosing images spaced as in Figure (23b) does not necessarily agree with your own intuitive
gives vectors a reasonable length, and a reasonable idea of acceleration. We have just discovered that,
approximation of the circular path. using the physicists’ definition, a particle moving at
(It will turn out that when we use our strobe defini- constant speed along a circular path accelerates
tion of acceleration, most errors caused by using a toward the center of the circle. Unless you had a
finite ∆ t cancel, and we get a very accurate answer. previous physics course, you would be unlikely to
Thus we do not have to worry much about how far guess this result. It may seem counter intuitive. But,
apart we draw the images.) as we said, we are using these examples to develop
an intuition for the physics definition of accelera-
Now that we have the displacement vectors we can tion. Whether you like it or not, according to the
construct the acceleration vectors a1,a2, using physics definition, a particle moving at constant
Equation (5). The construction for a2 is shown in speed around a circle, is accelerating toward the
Figure (24). To the vector S 2 we add the vector –S 1 center. In a little while, the reason for this will
to get the vector S2 – S 1 as shown. become clear.

The first thing we note is that the vector S2 – S1

points toward the center of the circle! Thus the Magnitude of the Acceleration
acceleration vector a2 given by for Circular Motion
Although perhaps not intuitive, we have gotten a
fairly simple result for the direction of the accelera-
S2 – S 1
a2 = (9) tion vector for uniform circular motion. The center
∆t2 is the only unique point for a circle, and that is where
also points toward the center of the circle. the acceleration vector points. The next thing we
need to know is how long the acceleration vectors
S3 are; what is the magnitude of this center pointing
S2 S4 acceleration. From the strobe definition, the magni-
– S1 tude a2 is a2 = S2 – S 1 /∆ t2, a rather awkward
S1 result that appears to depend upon the size of ∆ t that
we choose. However with a bit of geometrical
( S 2 – S 1) construction we can re-express this result in terms of
the particle’s speed v and the circle’s radius r. The
derivation is messy, but the result is simple. This is
one case, where, when we finish the derivation, we
Figure 24 recommend that the student memorize the answer
We find that the vector S2 – S1 , and therefore the rather than try to remember the derivation. Uniform
acceleration, points toward the center of the circle.
circular motion appears in a number of important
physics problems, thus the formula for the magni-
tude of the acceleration is important to know.
 3-19

In Figure (25a) we have constructed two triangles,

which are shown separately in Figures (25b) and v∆t = S2 – S1 (10)
r v∆t
(25c). As seen in Figure (25b), the big triangle
which goes from the center of the circle to positions Multiplying Equation (10) through by v and divid-
(1) and (2) has two equal sides of length r, the radius ing both sides by ∆ t gives
of the circle, and one side whose length is equal to
the particle’s speed v times the strobe time ∆t.
v 2 = S2 – S1 = a (11)
r ∆t 2
1
The second triangle, shown in Figure (25c), has
sides of length S2 and –S1 , but both of these are
of length v∆t as shown. The third side is of length we got S2 – S1 ∆t 2 on the right side, but this is just
S2 – S1 , the length we need for our calculation of the magnitude of the acceleration vector a1. Since
the magnitude of the acceleration vector. the same derivation applies to any position around
the circle, we get the simple and general result that,
The trick of this calculation is to note that the angles for a particle moving with uniform circular motion,
labeled θ in Figures (29b, c) are the same angle, so the particle’s acceleration a points toward the center
that these two triangles are similar isosceles tri- of the circle, and has a magnitude
angles. The proof that these angles are equal is given
in Figure (26) and its caption.
2
a = vr
acceleration of a particle
(12)
With similar triangles we can use the fact that the in uniform circular motion
ratios of corresponding sides are equal. Equating the
ratio of the short side to the long side of the triangle where v is the speed of the particle and r is the radius
of Figure (25b), to the ratio of the short side to the of the circle. As we said, this simple result should be
long side of the triangle in Figure (25c), we get memorized.
S2 =
V∆t
2
S1 S2 θ C
S2 – S1
φ
V∆t
1 3
( S 2 – S 1) α 2α
S1 =
D
B
–S1 90 – θ
r (c)
θ
V∆t
S1 = θ

r A
(a) Figure 26
r
θ That the two angles labeled θ in Figure (25) are the
(b) same, may be seen in the following geometrical
construction. Since the sum of the angles in any
Figure 25 triangle is 180°, we get α + ϕ + 90° = 180° (from
Derivation of the formula for the magnitude of the triangle BCD). Because BAC is an isosceles triangle,
acceleration of a particle with uniform circular motion. 90° – θ + α = ϕ. Eliminating ϕ we get 2α = θ , which
is the result we expected.
3-20 Description of Motion

AN INTUITIVE DISCUSSION We said that the string was pulling on the ball. To see
OF ACCELERATION that this is true, try swinging a ball on the end of a
string (or a shoe on the end of a shoelace) in a circle.
We have now studied two examples of non uniform
To keep the ball (or shoe) moving in a circle, you
motion, the projectile motion seen in the strobe
have to pull in on the string. In turn, the string pulls
photograph of Figure (10), and the circular motion
of a golf ball on the end of a string, a motion we in on the ball (or shoe). If you no longer pull in on
the string, i.e., let go, the ball or shoe flies away and
illustrated in Figure (23). In each case we calculated
no longer undergoes circular motion. The string
the acceleration vector of the particle at different
points along the trajectory. Let us now review our pulling on the ball is necessary in order to have
circular motion.
results to see if we can gain some understanding of
why the acceleration vector behaves the way it does. What is the common feature of projectile and circu-
lar motion? In both cases the object accelerates in
If you worked Exercise (5) correctly, you discov-
ered that all the acceleration vectors are the same, at the direction of the force acting on the object. When
you throw a steel ball in the air, the ball does not
least to within experimental accuracy. As shown in
escape earth’s gravity. As the ball moves through
Figure (27), as the steel ball moves along its trajec-
tory, its acceleration vector points downward to- the air, gravity is constantly pulling down on the
ball. The result of this gravitational pull or force is
ward the earth, and has a constant magnitude of
to accelerate the ball in the direction of the gravita-
about 1000 cm/sec2.
tional force. That is why the projectile motion
As shown in Figure (28), the golf ball being swung acceleration vectors point down toward the earth.
at constant speed around in a circle on the end of a
When we throw a ball a few feet up in the air, it does
string, accelerates toward the center of the circle, in
not get very far away from the surface of the earth.
the direction of the string pulling on the ball. The
In other words we expect the gravitational pull to be
magnitude of the acceleration has the constant value.
equally strong throughout the trajectory. If the ball’s
acceleration is related to the gravitational pull, then
0 1 we expect the acceleration to also be constant through-
2 out the trajectory. Thus it is not surprising that all the
vectors have the same length in Figure (27).
a0 a1
a2 3
golf ball

a3
r ing
str
Figure 27 a
All the acceleration vectors for
projectile motion point down toward
the earth.

Figure 28
The golf ball accelerates in the
direction of the string that is pulling
on it.
 3-21

In the case of circular motion, the string has to pull From the relationship we have seen between force
in on the golf ball to keep the ball moving in a circle. and acceleration we can understand why a projectile
As a result of this pull of the string toward the center that goes only a few feet above the surface of the
of the circle, the ball accelerates toward the center of earth should have a constant acceleration. The
the circle. Again the acceleration is in the direction gravitational force does not change much in those
of the force on the object. few feet, and therefore we would not expect the
acceleration caused by gravity to change much ei-
This relationship between force and acceleration, ther.
which we are just beginning to see in these two
examples, forms the cornerstone of what is called On the other hand there is no obvious reason, at this
classical or Newtonian mechanics. We have more point, why in the absence of air, a steel ball and a
details to work out, but we have just glimpsed the feather should have the same acceleration. Galileo
basic idea of much of the first half of this course. To believed that all projectiles, in the absence of air
give historical credit for these ideas, it was Galileo resistance, have the same acceleration. But it was
who first saw the importance of the concept of not until Newton discovered both the laws of me-
acceleration that we have been discussing, and Isaac chanics (the relationship between acceleration and
Newton who pinned down the relationship between force) and the law of gravity, that it became a
force and acceleration. physical prediction that all projectiles have the same
gravitational acceleration.
Acceleration Due to Gravity In the early part of the 20th century, Einstein went a
Two more topics, both related to projectile motion, step farther than Newton, and used the fact that all
will finish our discussion in this section. The first is objects have the same gravitational acceleration to
the fact that, if we can neglect air resistance, all develop a geometrical interpretation of the theory of
projectiles near the surface of the earth have the gravity. The gravitational force was reinterpreted as
same downward acceleration a. If a steel ball and a a curvature of space, with the natural consequence
feather are dropped in a vacuum, they fall together that a curvature of space affects all objects in the
with the same acceleration. This acceleration, which same way. This theory of gravity, known as
is caused by gravity, is called the acceleration due Einstein’s general theory of relativity, was a result
to gravity and is denoted by the symbol g. The vector of Einstein’s effort to make the theory of gravity
g points down toward the earth, and, at the surface of consistent with the principle of relativity.
the earth, has a magnitude.
It is interesting how the simplest ideas, the principle
acceleration due to of relativity, and the observation that the gravita-
g ≡ g = 980 cm/sec2 gravity at the sur- (13) tional acceleration is the same for all objects, are the
face of the earth
cornerstones of one of the most sophisticated theo-
This is quite consistent with our experimental result ries in physics, in this case Einstein’s general theory
of about 1000 cm/sec2 that we got from the analysis of relativity. Even today, over three quarters of a
of the strobe photograph in Figure (10). century since Einstein developed the theory, we still
do not understand what many of the predictions or
If we go up away from the earth, the acceleration due consequences of Einstein’s theory will be. It is
to gravity decreases. At an altitude of 1,600 miles, exciting, for these predictions may help us under-
the acceleration is down to half its value, about stand the behavior of the universe from its very
500 cm/sec2. On other planets g has different val- beginning.
ues. For example, on the moon, g is only about 1/6
as strong as it is here on the surface of the earth, i.e.

gmoon = 167 cm/sec2 (14)

3-22 Description of Motion

Exercise (7) Projectile Motion

The first earth satellite, Sputnik 1, traveled in a low, with Air Resistance
nearly perfect, circular orbit around the earth as Back to a more mundane subject, we wish to end this
illustrated in Figure (29). discussion of acceleration with the example of pro-
jectile motion with air resistance. Most introductory
(a) What was the direction of Sputnik 1’s acceleration physics texts avoid this topic because they cannot
vector as it went around the earth?
deal with it effectively. Using calculus, one can
(b) What was the direction of the force of gravity on handle only the simplest, most idealized examples,
Sputnik 1 as the satellite went around the earth? and even then the analysis is beyond the scope of
most texts. But using strobe photographs it is easy
(c) How is this problem related to the problem of the to analyze projectile motion with air resistance, and
motion of the golf ball on the end of a string? Give an we learn quite a bit from the results.
answer that your roommate, who has not had a
physics course, would understand. What turned out to be difficult, was to find an
example where air resistance affected the motion of
Sputnik 1 orbit a projectile enough to produce a noticeable effect.
We found that a golf ball and a ping pong ball have
almost the same acceleration when thrown in the air,
despite the considerable difference in weight or
EARTH
mass. Only when we used the rough surfaced
Styrofoam balls used for Christmas tree ornaments
did we finally get enough air resistance to give a
significant effect.
Figure 29
Sputnik 1's circular orbit.

Figure 30a
Motion of a Styrofoam ball. This is the lightest ball we
could find.
 3-23

A strobe photograph of the projectile motion of the figure riding on the ball in Figure (31). You will feel
Styrofoam ball is seen in Figure (30a), and an a wind in your face, a wind directed oppositely to v3 .
analysis showing the resulting acceleration vectors This wind will push on the ball in the direction
in Figure (30b). In Figure (30b) we have also drawn opposite to v3 , i.e., in the direction of aair. Thus we
the acceleration vectors g that the ball would have conclude that the acceleration aair is created by the
had if there had not been any air resistance. We see force of the wind on the ball.
that the effect of air resistance is to bend back and
shorten the acceleration vectors. What we learn from this example is that if we have
two forces simultaneously acting on an object,
Figure (31) is a detailed analysis of the Styrofoam’s each force independently produces an accelera-
acceleration at point (3). (We used an enlargement tion, and the net acceleration is the vector sum of
of the strobe photograph to improve the accuracy of the independent accelerations. In this case the
our work, such detailed analysis is difficult using independent accelerations are caused by gravity and
small Polaroid photographs.) In Figure (31) v 3 is the the wind. The net acceleration a3 of the ball is given
velocity of the ball, g is the acceleration due to by the vector Equation (15), a3 = g + aair. As we
gravity, and a3 the ball’s actual acceleration. The will see in later chapters, this vector addition of
vector aair, which represents the change in a caused accelerations plays a fundamental role in mechan-
by air resistance is given by the vector equation ics.

a3 = g + aair (15) 3
The important feature of Figure (31) is that aair is
oppositely directed to the ball’s velocity v3 . To
"
understand why, imagine that you are the stick a3 ind
"w
g
0
1 a air
a0 a1 2
a2
v3

3
a3 a air = –Kv
-1) ( 5.2, 94.9) g
0) (24.0, 101.4)
1) (40.8, 97.8) 4
2) (56.5, 85.3) a4 Figure 31
3) (70.8, 64.7) g When we do a detailed comparison of a and
4) (83.4, 37.1) g at point 3, we see that the air resistance
5) (95.2, 3.9) produces an acceleration a air that points in the
direction of the wind felt by the ball.
5

Figure 30b
Acceleration of the Styrofoam ball.
3-24 Description of Motion

INSTANTANEOUS VELOCITY tory of the ball. In Figure (32a) where ∆ t = 0.4 sec
In calculus, instantaneous velocity is defined by and there are only two images the only possible
starting with the equation vi = Ri+1 - Ri /∆ t and definition of v0 is the displacement between these
then taking the limiting value of vi as we use shorter images, divided by ∆ t as shown. Clearly ∆ t is too
and shorter time steps ∆ t. This corresponds in a large here for an accurate representation of the ball’s
strobe photograph to using a higher and higher motion.
flashing rate which would give increasingly short A better description of motion is obtained in Figure
displacement vectors Si . In the end result one (32b) where∆ t = 0.1 sec as in the original photo-
pictures the instantaneous velocity being defined at graph. We used this value of in our analysis of the
each point along the continuous trajectory of the projectile motion, Figure (10). Reducing ∆ t by
object. another factor 1/4 gives the results shown in Figure
The effect of using shorter and shorter ∆ t is illus- (32c). At this point the images provide a detailed
trated in Figure (32). In each of these sketches the picture of the path and v0 = S0 /∆ t is now tangent to
dotted line represents the smooth continuous trajec- the path at (0). A further decrease in ∆ t would
produce a negligible change in v0 .

0 0 S0 1
v0
v0
S0
v0 =
∆t
S0
v0 = ∆t = 0.1 Sec
∆t b)

a) ∆t = 0.4 Sec

S0
1

v0 v instantaneous
01 ~i velocity
S0
Si
vi =
S0 ~ ∆t
v0 =
∆t d) ∆t 0 Sec
c) ∆t = 0.025 Sec

Figure 32
We approach the instantaneous velocity
as we make ∆t smaller and smaller.
 3-25

The instantaneous velocity at point (0) is the final

value of v0 , the value illustrated in Figure (32d)
which no longer changes as ∆ t is reduced. This is
an abstract concept in that we are assuming such a
final value exists. We are assuming that we always
reach a point where using a stroboscope with a still
higher flashing rate produces no observable change
in the value of v0 . This assumption, which has
worked quite well in the analysis of large objects
such as ping pong balls and planets, has proven to be
false when investigated on an atomic scale. Accord-
ing to the quantum theory which replaces classical
mechanics on an atomic scale when one uses a
sufficiently short ∆ t in an attempt to measure veloc-
ity, the measurement destroys the experiment rather
than giving a better value of v0 .
3-26 Description of Motion

Instantaneous Velocity v–1 + v0

from a Strobe Photograph vi = (16)
2
In the case of projectile motion (i.e., motion with
as illustrated in Figure (33). (Note that the vector
constant acceleration) there is a simple yet precise
sum v–1 + v0 ∆t is the same as the line 2v i ∆t
method for determining an object’s instantaneous
which connects the preceding and following image.)
velocity vi from a strobe photograph. (Vectors
This is a reasonable estimate of the ball’s instanta-
representing instantaneous velocity will be under-
neous velocity because v–1 is the average velocity
lined in order to distinguish them from the vectors
during the time ∆ t before the ball got to (0), and v0
representing the strobe definition of velocity.) This
the average velocity during the interval after leaving
method, which also gives quite good approximate
(0). The ball’s velocity at (0) should have a value
values for other kinds of motion, will be used in our
intermediate between v–1 and v0 , which is what
computer calculations for determining the initial
Equation (16) says.
velocity of the object.
The constant acceleration formula
To see what the method is, consider Figure (33)
where we have drawn the vector obtained from S = vit + 1 at 2 (17)
2
Figure (32d). We have also drawn a line from the
center of image (–1) to the center of image (+1) and which may be familiar from a high school physics
notice that vi ∆t is parallel to and precisely half as course, provides a direct application of the concept
long as this line. Thus we can construct vi ∆t by of instantaneous velocity. (Remember that this is not
connecting the preceding and following images and a general formula; it applies only to motion with
taking half of that line. constant acceleration where the vector a changes
neither in magnitude or direction.) As illustrated in
The vector constructed by the above rule is actually Figure (34) the total displacement of the projectile
the average of the preceding velocity vector v–1 and
the following vector v0 .

V~ i t
Vi ∆t
~
_ at 2
1
2
2 Vi ∆t
~
S
V0 ∆t
∆t
V-1
∆t
V–1 + V0 )
( 1 2
_
S =V
~ i t + 2 at

~ i ∆t = ( –1
2V V + V0 )∆t

V
~ i = ( V–1 + V0 )
2 t = 3∆t

Figure 33 Figure 34
Illustration of the constant acceleration
formula as a vector equation.
 3-27

during a time t (here t = 3∆t) is the vector sum of vi t For these exercises, use the tear out sheet on
and 1/2at2. To draw this figure, we used pages 3-31,32.

v i t = v i 3∆t = 3 v i ∆t Exercise 8
Use Equation 17 to predict the displacements of the
and obtained v i ∆t from our method of determining ball
instantaneous velocity. We also used
(a) Starting at position (0) for a total time t = 4∆t.
1 a t2 = 1 a 3∆t 2 = 9 a∆t 2
2 2 2
(b) Starting at position (1) for a total time t = 3∆t.
where we obtained a ∆t2 from the relation
Do the work graphically as we did in Figures 33-35.
S2 - S1
a =
∆t 2 Exercise 9
The other constant acceleration formula is
a∆t 2 = S2 - S1
v f = v i + at
as illustrated in Figure (35).
where v i is the initial velocity, and v f the object’s
velocity a time t later. Apply this equation to Figure 10
to predict the ball’s instantaneous velocity v f at point
(3 ∆t) (3) for a ball starting at point (0). Check your predic-
V~i t = V~i tion by graphically determining the instantaneous
velocity at point (3).

Show your results on graph paper.

V~i ∆t Exercise 10
Show that the constant acceleration formulas would
correctly predict projectile motion even if time ran
a∆t 2 backward. (For example, assume that the ball went
backward as shown in Figure (36), and repeat Exer-
cise 8b, going from position 3 to position 0.)
4.5 a ∆t 2 S2
– S1
0

_ 2
1
2 at
= _ a(3∆t)2
1
2
= 4.5 a ∆t 2 3

appearence of ball
Figure 35 moving backward
How to construct the vectors v i t and in time
1 2 at 2 from a strobe photograph.

Figure 36
Run the motion of the ball backward in time, and
it looks like it was launched from the lower right.
 3-28 Description of Motion

0 10 20 30 40 50 60 70 80 90 100

0 s0 1
90
s-1 s1 90

-1 2
80 80

s2
70 70

3
60 60

50 s3 50

40 40

4
30 30

20 20

10 10

0 10 20 30 40 50 60 70 80 90 100
Extra copy of Figure 10
Use this graph for practice with vectors.
 Tear out page 3-29

0 10 20 30 40 50 60 70 80 90 100

0 s0 1
90
s-1 s1 90

-1 2
80 80

s2
70 70

3
60 60

50 s3 50

40 40

4
30 30

20 20

10 10

0 10 20 30 40 50 60 70 80 90 100
Extra copy of Figure 10
Use this graph for homework that you
pass in. This page may be torn out.
 3-30 Description of Motion Tear out page

0 10 20 30 40 50 60 70 80 90 100

0 s0 1
90
s-1 s1 90

-1 2
80 80

s2
70 70

3
60 60

50 s3 50

40 40

4
30 30

20 20

10 10

0 10 20 30 40 50 60 70 80 90 100
Extra copy of Figure 10
Use this graph for homework that you
pass in. This page may be torn out.
 Tear out page 3-31

0 10 20 30 40 50 60 70 80 90 100

0 s0 1
90
s-1 s1 90

-1 2
80 80

s2
70 70

3
60 60

50 s3 50

40 40

4
30 30

20 20

10 10

0 10 20 30 40 50 60 70 80 90 100
Extra copy of Figure 10
Use this graph for homework that you
pass in. This page may be torn out.
 3-32 Description of Motion Tear out page

0 10 20 30 40 50 60 70 80 90 100

0 s0 1
90
s-1 s1 90

-1 2
80 80

s2
70 70

3
60 60

50 s3 50

40 40

4
30 30

20 20

10 10

0 10 20 30 40 50 60 70 80 90 100
Extra copy of Figure 10
Use this graph for homework that you
pass in. This page may be torn out.
 3-33

0 10 20 30 40 50 60 70 80 90 100

90 90

80 80

70 70

60 60

50 50

40 40

30 30

20 20

10 10

0 10 20 30 40 50 60 70 80 90 100
Spare graph paper
Use this graph paper if you
want practice with vectors.
This is not a tear out page.
 4-0
3-34 Calculus in Physics
Description of Motion

0 0 v0 1

v0

1
∆t = 0.4 sec ∆t = 0.1 sec

(a) (c)

v0 v~i
01

∆t = 0.025 sec Instantaneous Velocity

(b) (d)

Figure 1
Transition to instantaneous velocity.
 4-1

Chapter 4
Calculus in Physics
This chapter, which discusses the use of calculus in physics, is for those who have had a calculus
course which they remember fairly well. For those whose calculus is weak or poorly remembered,
or for those who have not studied calculus, you should replace this chapter with Chapter 1 of
Calculus 2000.

CHAPTER 4 CALCULUS IN PHYS-

ICS

In the previous chapter we used strobe photographs to LIMITING PROCESS

define velocity and acceleration vectors. The basic In our discussion of instantaneous velocity we concep-
approach was to turn up the strobe flashing rate as we tually turned the strobe all the way up as illustrated in
did in going from Figure (3-3) to (3-4) until all the kinks Figures (2-32a) through (2-32d), redrawn here in Fig-
are clearly visible and the successive displacement ure (1). In these figures, we initially see a fairly large
vectors give a reasonable description of the motion. change in v0 as the strobe rate is increased and ∆t
We did not turn the flashing rate too high, for the reduced. But the change becomes smaller and it looks
practical reason that the displacement vectors became as if we are approaching some final value of v0 that does
too short for accurate work. Calculus corresponds to not depend on the size of ∆t, provided ∆t is small
conceptually turning the strobe all the way up. enough. It looks as if we have come close to the final
value in Figure (1c).

The progression seen in Figure (1) is called a limiting

process. The idea is that there really is some true value
of v0 which we have called the instantaneous velocity,
and that we approach this true value for sufficiently
small values of ∆t . This is a calculus concept, and in
the language of calculus, we are taking the limit as ∆ t
goes to zero.

The Uncertainty Principle

For over 200 years, from the invention of calculus by
Newton and Leibnitz until 1924, the limiting process
and the resulting concept of instantaneous velocity was
one of the cornerstones of physics. Then in 1924
Werner Heisenberg discovered what he called the
uncertainty principle which places a limit on the
accuracy of experimental measurements.
4-2 Calculus in Physics

Heisenberg discovered something very new and unex- motion, we would expect get results like those shown
pected. He found that the act of making an experimen- in Figure (2). This figure represents projectile motion
tal measurement unavoidably affects the results of an with an acceleration g = 980 cm/sec2 and ∆t = .01sec,
experiment. This had not been known previously as the reader can easily check.
because the effect on large objects like golf balls is
undetectable. But on an atomic scale where we study When we study the uncertainty principle in Chapter 30,
small systems like electrons moving inside an atom, the we will see that a measurement that is accurate enough
effect is not only observable, it can dominate our study to show that Position (2) is below Position (1), could
of the system. disturb the electron enough to reverse its direction of
motion. The next position measurement could find the
One particular consequence of the uncertainly prin- electron over where we drew Position (3), or back
ciple is that the more accurately we measure the where we drew Position (0), or anywhere in the region
position of an object, the more we disturb the motion of in between. As a result we could not even determine
the object. This has an immediate impact on the what direction the electron is moving. This uncertainty
concept of instantaneous velocity. If we turn the strobe would not be the result of a sloppy experiment, it is the
all the way up, reduce ∆t to zero, we are in effect trying best we can do with the most accurate and delicate
to measure the position of the object with infinite measurements possible.
precision. The consequence would be an infinitely big
disturbance of the motion of the object we are studying. The uncertainty principle has had a significant impact
If we actually could turn the strobe all the way up, we on the way physicists think about motion. Because we
would destroy the object we were trying to study. now know that the measuring process affects the results
of the measurement, we see that it is essential to provide
It turns out that the uncertainty principle can have a experimental definitions to any physical quantity we
significant impact on a larger scale of distance than the wish to study. A conceptual definition, like turning the
atomic scale. Suppose, for example, that we con- strobe all the way up to define instantaneous velocity,
structed a chamber 1 cm on a side, and wished to study can lead to fundamental inconsistencies.
the projectile motion of an electron inside. Using
Galileo’s idea that objects of different mass fall at the Even an experimental definition like our strobe defini-
same rate, we would expect that the motion of the tion of velocity can lead to inconsistent results when
electron projectile should be the same as more massive applied to something like the electron in Figure (2). But
objects. If we took a strobe photograph of the electron’s these inconsistencies are real. Their existence is telling
us that the very concept of velocity is beginning to lose
meaning for these small objects.
0 1
On the other hand the idea of the limiting process and
-1 2
v1 instantaneous velocity is very convenient when ap-
plied to larger objects where the effects of the uncer-
1 centimeter

3 tainty principle are not detectable. In this case we can

apply all the mathematical tools of calculus developed
over the past 250 years. The status of instantaneous
4 velocity has changed from a basic concept to a useful
mathematical tool. Those problems for which this
mathematical tool works are called problems in classi-
cal physics; and those problems for which the uncer-
1 centimeter tainty principle is important, are in the realm of what we
call quantum physics.
Figure 2
Hypothetical electron projectile motion experiment.
 4-3

CALCULUS DEFINITION The velocity vector vi is now given by

OF VELOCITY
∆R i
With the above perspective on the physical limita- vi = (2)
∆t
tions on the limiting process, we can now return to
This is just our old strobe definition vi = Si /∆ t, but
the main topic of this chapter—the use of calculus in
using a notation which emphasizes that the displace-
defining and working with velocity and accelera-
ment Si = ∆Ri is the change in position that occurs
tion.
during the time ∆ t. The Greek letter ∆ (delta) is used
In discussing the limiting process in calculus, one both to represent the idea that the quantity ∆R i or ∆t
traditionally uses a special set of symbols which we can is small, and to emphasize that both of these quantities
understand if we adopt the notation shown in Figure change as we change the strobe rate.
(3). In that figure we have drawn the coordinate vectors
The limiting process in Figure (1) can be written in the
R i and R i+1 for the i th and (i + 1) th positions of the
form
object. We are now using the symbol ∆R i to represent
the displacement of the ball during the i to i+1 interval. Limit ∆R i
vi ≡ ∆t (3)
The vector equation for ∆R i is → 0 ∆t
where the word “Limit” with ∆ t→0 underneath, is to
∆R i = R i+1 – R i (1) be read as “limit as ∆ t goes to zero”. For example we
In words, Equation (1) tells us that ∆R i is the change, would read Equation (3) as “the instantaneous veloc-
during the time ∆ t, of the position vector R describing ity vi at position i is the limit, as ∆t goes to zero, of
the location of the ball. the ratio ∆R i /∆t . ”

For two reasons, Equation (3) is not quite yet in

i standard calculus notation. One is that in calculus, only
∆R i the limiting value, in this case, the instantaneous veloc-
i +1
ity, is considered to be important. Our strobe definition
vi = ∆R i /∆t is only a step in the limiting process.
Therefore when we see the vector vi , we should
assume that it is the limiting value, and no special
Ri symbol like the underline is used. For this reason we
will drop the underline and write
R i +1
Limit ∆R i
vi = ∆t (3a)
→ 0 ∆t

∆R i = R i +1 – R i

V i = ∆R i /∆t

Figure 3
Definitions of ∆ Ri and Vi .
4-4 Calculus in Physics

The second change deals with the fact that when ∆ t ing. But to a physicist, there is a different, more
goes to zero we need an infinite number to time steps practical meaning. Think of dt as a short ∆ t, short
to get through our strobe photograph, and thus it is not enough so that the limiting process has essentially
possible to locate a position by counting time steps. occurred, but not too short to see what is going on. In
Instead we measure the time t that has elapsed since the Figure (1), a value of dt less than .025 seconds is
beginning of the photograph, and use that time to tell us probably good enough.
where we are, as illustrated in Figure (4). Thus instead
of using vi to represent the velocity at position i, we If dt is small but finite, then we know exactly what
write v t to represent the velocity at time t. Equation the dR t is. It is the small but finite displacement
(3) now becomes vector at the time t. It is our old strobe definition of
velocity, with the added condition that dt is such a
∆R(t) short time interval that the limiting process has
v(t) = Limit
∆t → 0 ∆t
(3b) occurred. From this point of view, which we will use
where we also replaced ∆R i by its value ∆R(t) at throughout this text, dt is a real time interval, and
time t. dR t a real vector which we can work with in a
normal way. The only thing special about these
Although Equation (3b) is in more or less standard quantities is that when we see the letter d instead of
calculus notation, the notation is clumsy. It is a pain to ∆ , we must remember that a limiting process is
keep writing the word Limit with a ∆ t→0 underneath. involved. In this notation, the calculus definition of
To streamline the notation, we replace the Greek letter velocity is
∆ with the English letter d as follows
dR t
vt = (5)
Limit ∆R(t) ≡ dR(t) dt
∆t → 0 ∆t dt (4)
where R t and v t are the particle’s coordinate vector
(The symbol ≡ means defined equal to.) To a math- and velocity vector respectively as shown in Figure (5).
ematician, the symbol dR t /dt is just shorthand Remember that this is just fancy shorthand notation for
notation for the limiting process we have been describ- the limiting process we have been describing.

t = .1sec t = .2sec

t = 0sec t = .3sec
V(t)

t = .4sec
R(t) at t = .3 sec R(t)

t = .5sec

Figure 4 Figure 5
Rather than counting individual images, we can Instantaneous position and velocity at time t.
locate a position by measuring the elapsed time t.
In this figure, we have drawn the displacement
vector R(t) at time t = .3 sec.
 4-5

ACCELERATION The strobe definition of a i can now be written

In the analysis of strobe photographs, we defined both strobe v(t + ∆t) – v(t) ∆v(t)
a(t) definition = ≡ (8)
a velocity vector v and an acceleration vector a. The ∆t ∆t
definition of a, shown in Figure (2-12) reproduced here
in Figure (6) was Now go through the limiting process, turning the
strobe up, reducing ∆ t until the value of a t settles
vi+1 – vi
ai ≡ (6) down to its limiting value. We have
∆t
In our graphical work we replaced vi by S i /∆t so that
calculus v t + ∆t – v t
we could work directly with the displacement vectors a(t) definition = Limit
S i and experimentally determine the behavior of the ∆t→0 ∆t
acceleration vector for several kinds of motion. (9)
∆v(t)
= Limit
∆t→0 ∆t
Let us now change this graphical definition of accelera-
tion over to a calculus definition, using the ideas just
applied to the velocity vector. First, assume that the ball Finally use the shorthand notation d/dt for the limiting
reached position i at time t as shown in Figure (6). process:
Then we can write
vi = v(t) dv t
a(t) = (10)
vi+1 = v(t+∆t) dt
to change the time dependence from a count of strobe
flashes to the continuous variable t. Next, define the Equation (10) does not make sense unless you remem-
vector ∆ v(t) by ber that it is notation for all the ideas expressed above.
Again, physicists think of dt as a short but finite time
∆v(t) ≡ v(t+∆t) – v(t) = vi+1 – vi (7) interval, and dv t as the small but finite change in the
velocity vector during the time interval dt. It’s our
We see that ∆v(t) is the change in the velocity vector as strobe definition of acceleration with the added re-
the time advances from t to t+∆t . quirement that ∆ t is short enough that the limiting
process has already occurred.

position at
time t position at
time t + ∆t
Vi

( Vi+1–Vi ) Vi+1

–Vi
a i = ( Vi+1–Vi )
∆t

Figure 6
Experimental definition of the acceleration vector.
4-6 Calculus in Physics

Components The limiting process in calculus does not affect the

Even if you have studied calculus, you may not recall decomposition of a vector into components, thus Equa-
encountering formulas for the derivatives of vectors, tion (5) for v(t) and Equation (10) for a(t) become
like dR(t)/dt and dv(t)/dt which appear in Equations
(5) and (10). To bring these equations into a more v(t) = dR(t)/dt (5)
familiar form where you can apply standard calculus
formulas, we will break the vector Equations (5) and vx (t) = dRx (t)/dt (5a)
(10) down into component equations. vy (t) = dRy (t)/dt (5b)
vz(t) = dRz(t)/dt (5c)
In the chapter on vectors, we saw that any vector
equation like and
a(t) = dv(t)/dt (10)
A = B+C (11)
ax(t) = dvx (t)/dt (10a)
is equivalent to the three component equations
ay(t) = dvy (t)/dt (10b)
Ax = Bx + Cx az(t) = dvz(t)/dt (10c)
Ay = By + Cy
Az = Bz + Cz (12) Often we use the letter x for the x coordinate of the
vector R and we use y for Ry and z for Rz. With this
The advantage of the component equations was that notation, Equation (5) assumes the shorter and perhaps
they are simply numerical equations and no graphical more familiar form
work or trigonometry is required.
vx (t) = dx(t)/dt y (5a’)
vy (t) = dy(t)/dt R (5b’)
vz(t) = dz(t)/dt x
(5c’)
At this point the notation has become deceptively short.
You now have to remember that x(t) stands for the x
coordinate of the particle at a time t.

We have finally boiled the notation down to the point

where it would be familiar from any calculus course. If
we restrict our attention to one dimensional motion
along the x axis. Then all we have to concern ourselves
with are the x component equations

dx(t)
vx(t) =
dt
(10a)
dvx(t)
a x(t) =
dt
 4-7

Distance, Velocity and If the particle is moving at constant speed

Acceleration versus Time Graphs vx (t) = v0 (11)
One of the ways to build an intuition for Equations (5a)
then the graph of velocity versus time is a straight
and (10a) is through the use of graphs of position,
horizontal line of height v0 as shown in Figure (7b).
velocity and acceleration versus time. Suppose, for
example, we had a particle moving at constant speed in If you travel away from home at constant speed, then
the x direction, the uniform motion that the principle of your distance from home is proportional to the time you
relativity tells us that we cannot detect. Graphs of have traveled. If you start at t = 0, then at time t your
distance x(t), velocity v(t) and acceleration a(t) for this distance from home is
motion are shown in Figure (7).
x t = v0 t (12)
x(t) This is graphed as the straight line as shown in Figure
(7a). The slope of this line, the tangent of the angle θ
tanθ = x/t = V0 is x/t, which from Equation (12) is v0.
x(t) = V0 t
When a particle moves at constant velocity, there is no
change in the succeeding velocity vectors, thus the
x acceleration a(t) is zero for all time
a) θ a(t) = 0 (13)
t t as shown in Figure (7c).
v(t) In summary, we have seen that for this example of
uniform motion in the x direction
x(t) = v0 t (12)
v(t) = v0 (11)
V0 a(t) = 0 (13)
v(t) = V0
Now let us see if these results agree with our calculus
definitions (5a) and (10a). From Equation (5a) we get
b) dx(t)
t v(t) = = d v0 t (14)
dt dt
The v0 being constant comes outside and we have
a(t)
v(t) = v0 dt = v0 (15)
dt
where we used dt/dt = 1. Our calculus result agrees with
Equation (12).

From Equation (10), we get

dv(t)
c) a(t) = 0 a(t) = = d v0 = 0 (16)
t dt dt
because the derivative of a constant is zero.
Figure 7
Motion with constant velocity.
4-8 Calculus in Physics

x(t) What we should begin to see from this example, is that

x(t) = at + bt 2 if we have the formula for x(t) then it is easy to use
calculus to figure out the particle’s velocity and accel-
eration. Let us consider one more example. Suppose
x(t) is given by the formula
x(t) = at + bt2 (17)
where a and b are constants. Then the calculus formu-
a) t las (5a) and (10a) give

v(t) dx(t)
v(t) = = a + b d t2
dt dt
v(t) = a + 2bt = a + 2bt (18)
b
=2
2 bt/t where we used d t2 /dt = 2t. Equation (10a) gives
θ=
tan 2bt
dv(t)
a(t) = = 2b (19)
θ dt
a
t The results in Equations (17), (18) and (19) are graphed
b) in Figures (8a, b and c) The position vs time a straight
t
line with a slope 2b, and the acceleration is a constant
a(t) 2b. Figure (8) therefore represents an example of
motion with constant acceleration.

2b a(t) = 2b

c)
t
Figure 8
Motion with constant acceleration.
 4-9

THE CONSTANT ACCELERATION The integral on the left is simply v(t) evaluated between
FORMULAS 0 and T.
Unfortunately life is not as simple as one might think T T
dv(t) = v(t) = v(t) – v(0) (22)
from the preceding example. If you have the formula 0 0
for x(t), then you can calculate v(t) and a(t) very easily On the right side of Equation (21), we set a(t) = a0 (for
by differentiation. But usually you have to go the other constant acceleration) to get
way. From the physics you figure out what the accel-
T T T
eration is, then you have to work back to get v(t) and a(t)dt = a 0 dt = a 0 dt
finally x(t). At best, this reverse process involves 0 0 0
integration which is typically quite a bit harder than
T
differentiation. = a0 t = a 0T – a 0 × 0 = a 0T (23)
0
Let us work out an example where we know the
Using Equations (22) and (23) in (21) we get
acceleration and have to integrate to get the velocity
and position. We will take the easiest non trivial case v(T) - v(0) = a0T (24)
where the acceleration is constant. The result will be
The next step is to recognize that Equation (24) applies
the constant acceleration formulas.
to any time T, so that we can replace T by t to get
If we know a(t), the first step is to solve equation (10a) v(t) = v(0) + a0t (25)
by turning it into an integral equation as follows
To emphasize that v(0), the particle’s speed at time t =
a(t) =
dv(t)
(10a) 0, is not a variable, we will use the notation v 0 ≡ v0
dt and Equation (25) becomes
First multiply both sides by dt. (Remember that
physicists keep dt very small but finite, so that we can v(t) = v0 + a 0t
(26)
move it around.) We get
(If the steps we have used to derive Equation (26) were
dv(t) = a(t) dt (20) familiar and comfortable, then your calculus back-
Now integrate both sides of Equation (20) from time ground is in good shape and you should not have much
t = 0 up to time t = T. (This is called a definite integral.) of a problem with calculus in reading this text. If, on the
We get other hand what we did was strange, if the notation was
unfamiliar and the steps unpredictable, a review of
T T
dv(t) = a(t)dt (21) calculus is indicated. What we have done in the
0 0 derivation of Equation (26) is use the concept of a
definite integral. We will use definite integrals through-
out the course and now is the time to learn how to use
them. You should also be sure that you can do simple
differentiations like d/dt at2 = 2at.)
4-10 Calculus in Physics

To get the other constant acceleration formula, start

with Equation (5a) T T
v(t)dt = (v0 + a 0t)dt
dx(t) 0 0
v(t) = (5a)
dt T T
= v0 dt + a 0t dt (30)
and multiply through by dt to get 0 0

dx(t) = v(t)dt (27) T T

= v0 dt + a 0 t dt
Again integrate both sides from t = 0 to t = T to get 0 0

T T
dx(t) = v(t)dt (28) Knowing that
0 0 T 2
We can immediately do the integral on the left hand t dt = t (31)
0 2
side
we get
T T
dx(t) = x(t) = x(T) – x(0) (29) T T 2 T
+ a0 t
0 0
v(t)dt = v0 t
At this point we cannot do the integral on the right side 0 0 2 0
of Equation (28) until we know explicitly how v(t) (32)
2
depends on the variable t. If, however, the acceleration = v0T + a 0 T
2
is constant, we can use Equation (26) for v(t) to get
Using Equations (29) and (32) in (28) gives

x T - x0 = v0 T + 1 a0T2 (33)
2
Since Equation (33) applies for any arbitrary time T, we
can replace T by t to get

x(t) = x0 + v0 t + 1 a0t2 (34)

2
where we have written x0 for x(0), the position of the
particle at time t = 0.
 4-11

Three Dimensions Using this notation, we define the following vectors by

Equations (26) and (34) are the constant acceleration their components
formulas for motion in one dimension, along the x axis.
(We can, of course, choose the x axis to point any way coordinate
Rt ≡ (x(t),y(t),z(t)) (38)
we want.) vector

If we want to describe motion in three dimensions with velocity

constant acceleration, we repeat the steps leading to vt ≡ (vx (t),vy (t),vz (t)) (39)
vector
Equations (26) and (34), but starting with (5b) and
(10b) for motion along the y axis, and (5c) and (10c) constant
for motion along the z axis. The steps are essentially a ≡ ax,ay,az (40)
acceleration
identical, and we end up with the six equations
1 With this vector notation, the six constant acceleration
x(t) = x 0 + vx (0)t + a xt2 (35a) formulas (35a, b, c) and (36 a, b, c) reduce to the two
2
vector equations
1 2
y(t) = y 0 + vy (0)t + at (35b)
2 y x(t) = x(0) + v(0) t + 1 at2 (35)
2
1 2
z(t) = z 0 + vz (0)t + at (35c) v(t) = v0 (0) + at (36)
2 z
or using the notation R0 = R(0), v0 = v(0), we have
vx (t) = vx (0) t + a x t (36a)
R(t) = R0 + v0 t + 1 at2 (35’)
vy (t) = vy (0) t + a y t (36b) 2
vz (t) = vz (0) t + a z t (36c)
v(t) = v0 + at (36’)
where we have temporarily gone back to the notation
x(0) for x0, vx (0) for vx0, etc., and ax, ay, and az are the These are the set of vector equations that we tested in
x, y, z components of the assumed constant accelera- our studies in Chapter 3 of instantaneous velocity with
tion. constant acceleration.
In Chapter 3 we introduced a notation that allowed us We have gone through all the details of the derivation
to conveniently express a vector S in terms of its of Equations (35) and (36), because they represent one
components Sx, Sy and Sz, by writing the components, of the major successes of the use of calculus in the
separated by commas, inside a parenthesis as follows prediction of motion. Whenever a particle’s accelera-
tion a is constant, and we know a, R0, and v0, we can use
S ≡ Sx ,Sy ,Sz (37) these equations to predict the particle’s position R(t)
and velocity v(t) at any time t in the future.
4-12 Calculus in Physics

PROJECTILE MOTION Our strobe photograph does not eliminate the possibil-
WITH AIR RESISTANCE ity that a air is more complicated, something like
In our experimental study of projectile motion, we saw aair = K2 v 2 (42)
that when we used a styrofoam projectile, air resistance
affected the acceleration of the projectile. From the or perhaps some combination like
point of view that we are riding on the ball, we would 2
a air = K1 v + K2 v (43)
feel a wind in our face, blowing in a direction -v,
opposite to the velocity v of the projectile. The effect It turns out that the motion of a sphere through a liquid
of this wind was to blow the acceleration vector back (in our case a Styrofoam ball through air) has been
as shown in Figure (3-28), reproduced here as Figure studied extensively by both physicists and engineers.
(9). For slow speeds the motion is like Equation (41) but as
the speed increases it looks more like Equation (43) and
We saw that the experimental vector a 3 was the accel- soon becomes even more complicated. The only
eration g we would have in the absence of air resistance, simple fact is that a air always points in the direction -v,
plus a correction a air which pointed in the direction of in the direction of the wind in our face (until vortex
the wind, in the -v direction as shown. shedding occurs).

The magnitude a air cannot accurately be determined As an exercise to test the ability of calculus to predict
from the strobe photograph. About all we can tell is that motion, let us assume that our simple guess aair = -K v
aair is zero if the ball is at rest, and increases as the is good enough. We would then like to solve the
speed v of the ball increases. The simplest guess is that calculus Equations (5) and (10) for the case where the
a air is proportional to v and we have the formula acceleration is not constant, but is given by the formula

simple a = g - Kv (44)
a air = –Kv (41)
guess
where g is the constant acceleration due to gravity, v is
the instantaneous velocity of the particle, and K is what
3 we will call the air resistance constant. Equation (44)
is pictured in Figure (9).

d" y
a3 "win
g
a air a
g
v
v3

–Kv

a air = –Kv x
Figure 9 Figure 10
The acceleration produced by air resistance.
 4-13

Our first step is to introduce a coordinate system as Integrating from t = 0 to t = T gives

shown in Figure (10), and break the motion up into x T T
and y components. Since the acceleration g due to
gravity points down, we have gx = 0 and the vector dvx (t) = - K vx (t)dt (48)
0 0
Equation (44) can be written as the two component
equations We can do the integral on the left, and remove the K
from the integral on the right giving
ax = -K vx (gx = 0) (44a)
T
ay = g - K vy (g = -980 cm/sec2) (44b) vx (t)| T
0
= vx (T) - vx (0) = -K vx(t)dt (49)
0
The calculus Equations (10a, b) that we have to solve
become Now we are in trouble, because we have to integrate
vx t in order to find vx t . We can’t do the integral until
ax = dvx = - K vx (45) we know the answer, and we have to do the integral to
dt
get the answer. It boils down to the fact that the
dvy techniques we used to solve the calculus equations for
ay = = g - K vy (46)
dt constant acceleration do not work now. As soon as the
Let us focus on the simpler of the two equations, acceleration is not constant, we have a much more
Equation (45) for the horizontal velocity of the projec- difficult problem.
tile. We want to solve the equation

dvx(t)
+ Kvx(t) = 0 (45’)
dt

Suppose we try to solve Equation (45) using the same

steps we used to predict vx for constant acceleration
(Equations 20 through 26). Multiplying through by dt
gives
dvx (t) = -K vx (t)dt (47)
4-14 Calculus in Physics

DIFFERENTIAL EQUATIONS kind of problems that can be solved. Simple physical

Equation (45) is an example of what is called a differ- modifications of a problem can turn an easy problem
ential equation. (An equation with derivatives in it.) into an unsolvable one.
Only in very special cases, as in our example of Before inventing calculus, Isaac Newton invented a
constant acceleration, can these equations be solved in simple step-by-step method that we will discuss in the
a straightforward manner by integration. In slightly next chapter. Newton’s step-by-step method has the
more complicated cases, these equations can be solved great advantage that slight complications in the physi-
by certain standard tricks that one learns in an advanced cal setup lead to only slightly more work in obtaining
calculus course on differential equations. We will use a solution. We will see that it is almost no harder to
one of these tricks to solve Equation (45). predict projectile motion with air resistance, even with
In general, however, differential equations cannot be v2 terms, than it is to predict projectile motion without
solved without numerical methods that are now handled air resistance. The step-by-step method will allow us
by digital computers. If, for example we assumed that to handle problems in this course, realistic problems,
the air resistance was proportional to v2 as in Equation that do not have a calculus solution.
(42), then Equation (45) for the x component of veloc- There are two disadvantages to the step-by-step method,
ity would be replaced by however. One, is that you get a numerical answer, like
dvx (t) an explicit orbit, rather than a general result. In
+ K2 vx (t) 2 = 0 (45a) contrast, the constant acceleration formulas describe
dt
all possible trajectories for motion with constant accel-
Equation (45a) is what is called a non linear differen-
eration.
tial equation, the word non linear coming from the
appearance of the square of the unknown variable vx (t). The second problem is that in the step-by-step method,
a simple calculation is repeated many times, perhaps
At the current time, there is no general way to solve non
thousands or millions of times to obtain an accurate
linear differential equations except by computer. Non
answer. Before digital computers, lifetimes were spent
linear differential equations have marvelously compli-
doing this kind of calculation by hand to predict the
cated features like chaotic behavior that have been
motion of the moon. But modern digital computers
discussed extensively in the popular press in the last
have changed all that. In minutes, the digital computer
few years. It is currently a hot research topic.
running your word processor can do what used to be
The point of this discussion is that when we use months of work.
calculus to predict motion, a very slight increase in the
complexity of the problem can lead to enormous Solving the Differential Equation
increases in the difficulty in solving the problem. We have essentially finished what we wanted to say
When the projectile’s acceleration was constant, we about applying calculus to the problem of projectile
could easily solve the calculus equations to get the motion with air resistance. The gist is that adding air
constant acceleration formulas. If the air resistance has resistance turns a simple problem into a hard one. Even
the simple form aair = -K v, then we have to solve a for the simplest form of air resistance, aair = -K v, we
differential equation, but we can still get an answer, a end up with the differential equation
formula that predicts the motion of the particle. If we
go up one step in complexity, if aair is proportional to dvx (t)
+ K vx (t) = 0 (45)
the square of the speed, then we have a non linear dt
differential equation that we cannot solve without which cannot be solved directly with integration.
numerical or approximation techniques.
Later in the course we will encounter several other
Calculus gives marvelous results when we can solve differential equations, one having the same form as
the problem. We get formulas describing the motion at Equation (45). When we meet these equations, we will
all future times. But we are extremely limited in the show you how to solve them.
 4-15

At this time, we do not really need the solution to Differentiating Equation (50) gives
Equation (45). This equation does not represent a basic dvx (t)
physics problem because our formula for air resistance = -vx0 α e-α t (51)
dt
is an approximation of limited validity. We include a where we used the fact that
solution for those who are interested, who want to see
de-α t = - α e-α t (52)
the problem completed now. Those for whom calculus dt
is new or rusty may wish to skip to the next chapter.
Substituting Equations (50) and (51) into Equation (45)
The reason that differential equations are hard to solve gives
is that the solutions are curves or functions rather than = 0 (53)
numbers. For example, the solution to the ordinary
First note that the exponential function e-α t cancelled
Equation x2 = 4 is the pair of numbers x = + 2 and x =
out. This indicates that we have guessed the correct
- 2. But Equation (45) has the decaying exponential
function.
curve shown in Figure (11) for a solution. What this
curve tells us is that the vx or the horizontal motion, dies Next note that vx0 cancels. This means that any value
out in time and the projectile will eventually have only of vx0 in Equation (50) is a possible solution. The
y motion. After enough time the ball will be falling particular value we want will be determined by the
straight down. experimental situation.
One of the standard techniques for solving differential What we have left is
equations is to guess the answer and then plug your
α = K (54)
guess into the equation to see if you are right. When you
take a course in solving differential equations, you Thus the differential Equation (45) has the exponen-
learn how to make educated guesses. If you had been tially decaying solution
through such a course, you would guess that Equation
(45) should have an exponentially decaying solution, vx (t) = vx0 e-Kt (55)
and try a solution of the form
where the decay rate is the air resistance constant K .
vx (t) = vx0 e-α t (guess ) (50)
For those of you who have actually had a course in
where α and vx0 are constants whose values we wish solving differential equations, see if you can solve for
to find. the vertical motion of the projectile. The differential
vx equation you have to solve is
dvy (t)
= g - Kvy (t) (46)
dt
The answer for long times turns out to be simple – the
projectile ends up coasting at a constant terminal
velocity. See if you can get that result.

The answer is vy = g/K 1 - e-Kt if vy = 0 at

t = 0.

t
Figure 11
Air resistance causes the horizontal component
of the velocity to decay exponentially.
4-16 Calculus in Physics

Appendix A Example A1
A boy throws a ball straight up into the air and catches
it (at the same height from which he threw it) 2 sec later.
How high did the ball go?
SOLVING PROJECTILE MOTION
Solution: To solve all projectile problems, we use the
PROBLEMS equations
In high school physics texts and most college level
1
introductory physics texts, there is considerable em- S = vi t + at2
phasis on solving projectile motion problems. A good 2
reason for this is that these problems provide practice vf = v i + at
in problem solving techniques such as drawing clear
However these are vector equations. Using a coordi-
sketches, developing an orderly approach, and check-
nate system in which the y axis is in the vertical
ing units. Not such a good reason is that, in texts that
direction and the x axis is in the horizontal direction, we
rely solely on algebra and calculus, the only thing they
get the following equations.
can solve in the early stages are projectile motion or
circular motion problems. Vertical motion:
The disadvantage of over emphasizing projectile mo-
1
tion problems is that students begin to use the projectile Sy = viy t + a yt2
motion formulas as a general way of predicting motion, 2
using the formulas in circumstances where they do not
apply. The important point to remember is that the vfy = v iy + a y t
formulas v = vi + at and x = vit + 1/2at2 are very Horizontal motion:
limited in scope. They apply only when the accelera-
1
tion a is constant, a not very likely circumstance in the Sx = vix t + a xt2
real world. The acceleration a is not constant for 2
circular motion, projectile motion with air resistance, vfx = v ix + a x t
satellite motion, the motion of electrons in a magnetic
field, and most interesting physics problems. Now projectiles near the surface of the earth accelerate
downward at a rate of nearly 980 cm/sec2. This value
From the point of view that solving projectile motion varies slightly at different points on the surface of the
problems is basically for practice in problem solving earth, but is always quite close to 980 cm/sec2. This
techniques, we will show you an orderly way of acceleration due to gravity is usually designated g;
handling these problems. The approach which we will since it is directed downward in the minus y direction,
illustrate using several examples should allow you with we have
practice to handle any constant acceleration problems
a = –g = – 980 cm/sec2 ( – 32 ft/sec 2)
y
test makers throw at you. In these examples, we are
demonstrating not only how the problem is solved but ax = 0
also how you should go about doing it. As a result, we get the equations
(Note -- in this appendix, all velocity vectors are
instantaneous velocities, thus we will not bother under-
lining them.)
 4-17

Vertical motion: Example A2

1
(a) Sy = viy t – gt2 (A1a) A ball is thrown directly upward at a speed of 48 ft/sec.
2 How high does it go?
(b) vfy = viy – gt (A1b)
Solution: First, find the time it takes to reach to top of
Horizontal motion: its trajectory. We have
(c) S x = vixt (A1c)
viy = 48 ft/sec
(d) vfx = vix (A1d)
vfy = 0 at the top of the trajectory
Horizontal motion and vertical motion are entirely
independent of each other. We see, for example, from From Equation (4-A1b) we have
Equation (A1d), that the horizontal speed of a projectile vfy = viy – gt
does not change; but this has already been obvious
or
from the strobe photographs.
viy 48 ft/sec
Now let us apply Equation (A1) to the situation where t = = = 1.5 sec
g 32 ft/sec2
the boy throws the ball straight up and catches it 2 sec
later. Since there is no horizontal motion, we only need Now we can use Equation A1a to calculate how high
equations (A1a, b). the ball goes.
1 2
One good technique for solving projectile problems is Sy = viyt – gt
2
to work up to and back from the top of the trajectory.
The reason is that at the top of the trajectory, Equations We have viy = 48 ft/sec, t = 1.5 sec to reach the top;
(A1a, b) are very easily applied. thus Sy, the distance to the top, is
ft 1 ft 2
In our problem, the ball spent half its time going up and Sy = 48 × 1.5 sec – × 32 2 × (1.5 sec)
sec 2 sec
half its time falling; thus, the fall took 1 sec. The
distance that it fell is Sy = 72 ft – 36 ft = 36 ft
1
Sy = viy t – gt2
2
where t is 1 sec, and since we are starting at the top of
the trajectory. We get
1 1
Sy = – gt2 = – × 32 ft/sec2 × 1 sec2
2 2
Sy = – 16 ft
The minus sign indicated the ball fell 16 ft below the
top of the trajectory.
4-18 Calculus in Physics

Vertical motion. A ball is thrown straight up at a speed

Example A3 viy = 48 ft/sec ; how long a time t does it take to come
An outfielder throws a ball at a speed of 96 ft/sec at an back to the ground?
angle of 30° above the horizontal. How far away from
the outfielder does the ball strike the ground? Horizontal motion. A ball travels horizontally at a
speed vix = 83 ft/sec . If it travels for a time t (result of
Solution: When solving problems, the first step is to vertical motion problem) how far does it travel?
draw a neat diagram of the situation, as in Figure (A1).
The first calculation is to find the x and y components We see that the vertical motion problem is exactly the
of vi . From our diagram we see that one we solved in Example A-2, viy = 48 ft/sec in both
cases. Thus, using the same solution, we find that the
vix = vicos θ ball takes 1.5 sec to go up and another 1.5 sec to come
down, for a total time of
= 96 ft/sec × 0.864
t = 3 sec
= 83 ft/sec
Now solve the horizontal motion from
viy = vi sin θ Sx = vix t
= 96 ft/sec × 0.50 We get
= 48 ft/sec Sx = 83 ft/sec × 3 sec = 249 ft
which is the answer.
where cos 30° = 0.864 and sin 30° = 0.50.

Now we are in a position to separate the problem into

two parts – vertical motion and horizontal motion.
These may be treated as two independent problems.

Figure A1
Sketch of the problem. On the sketch, label the
symbols used, show what is given, and state what you
are to find. It is generally better to work the problem
in terms of letters, substituting numbers only at the
end, or at convenient breaks in the problem.
 4-19

Checking Units Exercise A1

It is easy to make a mistake when working a problem. A 22-caliber rifle with a muzzle velocity of 600 ft/sec is
One of the best ways to avoid mistakes is to write out fired straight up. How high does the bullet go? How long
the dimensions of each number used in the calculation; before it hits the ground?
if the answer has the wrong dimension, you will know
there is a mistake somewhere. For example, in the Exercise A2
preliminary edition of this text the following formula (The rifle of Exercise A1 is fired at an angle of 45°. How
accidentally appeared. far does the bullet travel? (Give answer in ft and in mi.)
1
S = vi + at2 Exercise A3
2
A right fielder is 200 ft from home plate. Just at the time
Putting in the dimensions, we find he throws the ball into home plate, a runner leaves third
ft 1 ft 2 base and takes 3.5 sec to reach home plate. If the
S ft = vi + a 2 × t sec maximum height reached by the ball is 64 ft, did the
sec 2 sec
runner make it to home plate in time? (Problem from J.
or Orear, Fundamental Physics, Wiley, New York, 1961.)
ft 1 2
S ft = vi + at ft Exercise A4
sec 2
A steel ball is bouncing up and down on a steel plate
Clearly the (v i ) ft /sec has the wrong dimensions, with a period of oscillation of 1 sec. How high does it
since we cannot add ft/sec to ft. Thus, through a check bounce? (Problem from J. Orear, Fundamental Phys-
of the dimensions we would immediately spot an error ics, Wiley, New York, 1961.)
in this formula, even if we had no idea what the formula
is about. To correct this formula, the v i must be Exercise A5
multiplied by t sec so that the result is (v i )ft/sec× t sec A small rocket motor is capable of providing an accel-
equals (v i t) ft . eration of 0.01 g to a space capsule. If the capsule starts
from a far-out space station and the rocket motor runs
As another instance, in the solution of Example (A3) continuously, how far away is the capsule at the end of
we had 1 year? What is the capsule’s speed relative to the
viy space station at the end of the year?
t =
g Exercise A6
At this point you might begin to worry that you have A car traveling at 60 mi/hr strikes a tree. Inside the car
made a mistake; your doubts will be dispelled, how- the driver travels 1 ft from the time the car struck the tree
ever, once dimensions are inserted until he is at rest. What is the deceleration of the driver
if his deceleration is constant? Give the answer in ft/sec2
viy ft/sec viy and in g’s.
t sec = = sec
g ft/sec2 g
4-20 Calculus in Physics

Exercise A7
During volcanic eruptions, chunks of solid rock can be
blasted out of the volcano. These projectiles are called
volcanic blocks. Figure (2) shows a cross-section of Mt.
Fuji, in Japan.

At what initial speed v0 would a block have to be

ejected, at 45°, in order to fall at the foot of the volcano
as shown.

What is the time of flight?

(Problem from Halliday and Resnick.)

V~ i t
Hint Use the vector equation
1/ at 2
2
S = vit + 1 at 2
2
S
which is illustrated in Figure 3-34 reproduced to the
right. In this problem S is the total displacement of the Figure 3-34 (reproduced)
rock, from the time it left the volcano until it hit the
ground. Separate the vector equation into x and y
components.
0
V

45º

33km
Mt Fuji

9.4km
Figure A2
The farthest out blocks are the ones ejected at the greatest speed v0 at an
angle of 45°. By noting that the most distant blocks are 9.4 km away, you
can thus determine the maximum speed at which the blocks were ejected.
Chapter 5
Computer
Prediction
of Motion

CHAPTER 5 COMPUTER The clue for developing such a method is to note that
STEP-BY-STEP
PREDICTION OF MOTION
CALCULATIONS
In the last chapter we saw that for the special case of in our analysis of strobe photographs, we have been
breaking the motion into short time intervals of
constant acceleration, calculus allowed us to obtain a
rather remarkable set of formulas that predicted the length ∆ t. During each of these time intervals, not
much happens. In particular, the Styrofoam
object’s motion for all future times (as long as the
projectile’s acceleration vector did not change much.
acceleration remained constant). We ran into trouble,
Only over the span of several intervals was there a
however, when the situation got a bit more compli-
significant change in the acceleration vector. This
cated. Add a little air resistance and the analysis using
suggests that we could predict the motion by assum-
calculus became considerably more difficult. Only for
ing that the Styrofoam ball’s acceleration vector
the very simplest form of air resistance are we able to
was essentially constant during each time interval,
use calculus at all.
and at the end of each time step correct the accelera-
On the other hand, adding a little air resistance had tion vector in order to predict the motion for the next
only a little effect on the actual projectile motion. time step. In this way, by a series of short calcula-
Without air resistance the projectile’s acceleration tions, we can predict the motion over a long time
vectors pointed straight down and were all the same period. This is a rough outline of the step-by-step
length, as seen in Figure (3-27). Include some air method of predicting motion that was originally
resistance using the Styrofoam projectile, and the developed by Isaac Newton and that we will discuss
acceleration vectors tilted slightly as if blown back by in this chapter.
the wind one would feel riding along with the ball, as
The problem with the step-by-step prediction of motion
seen in Figure (3-31). Since projectile motion with air
is that it quickly gets boring. You are continually
resistance is almost the same as that without, one
repeating the same calculation with only a small
would like a method of predicting motion that is almost
change in the acceleration vector. Worse yet, to get
the same for the two cases, a method that becomes only
very accurate results you should take very many, very
a little harder if the physical problem becomes only a
small, time steps. Each calculation is almost identical
little more complex.
to the previous one, and the process becomes tedious.
If these calculations are done by hand, one needs an
enormous incentive in order to obtain meaningful
results.
5-2 Computer Prediction of Motion

COMPUTER CALCULATIONS Calculating and Plotting a Circle

Because of the tedium involved, step-by-step calcula- Figure (1) shows 100 points on the circumference of
tions were used only in desperate circumstances until a circle of radius r. To make this example somewhat
the invention of the digital computer in the middle of similar to the analysis of strobe photographs, we will
the twentieth century. The digital computer is most choose a circle of radius r = 35 cm, centered at
effective and easiest to use when we have a repetitive x = 50, y = 50, so that the entire circle will fit in the
calculation involving many, very similar steps. It is the region x = 0 to 100, y = 0 to 100, as shown. The i th
ideal device for handling the step-by-step calculations point around the circle has x and y coordinates given
described above. With a digital computer we can use by
very small time steps to get very accurate results, doing x i = r cos θ i
thousands or millions of steps to predict far into the
future. We can cover the same range of prediction as y i = r sin θ i (1)
the calculus-derived formulas, but not encounter sig- where θ i , the angle to the i th point, is given by
nificant difficulties when there is a slight change in the
360 2π
problem, such as the addition of air resistance. θi = * i degrees = i radians
100 100 *
To illustrate how to use the computer to handle a
repetitive problem, we will begin with the calculation (We know that it is easier to draw a circle using a
and plotting of the points on a circle. We will then go compass than it is to calculate and plot all these
back to our graphical analysis of strobe photographs individual points. But if we want something more
and see how that analysis can be turned into a series of complicated than a circle, like an ellipse or Lissajous
steps for a computer prediction of motion. figure, we cannot use a compass. Then we have to
calculate and plot individual points as we are doing.)
y
If we wrote out the individual steps required to
calculate and plot these 100 points, the result might
look like the following:
yi i i=0
r sin(θ)

r
θ0 = (2π/100)*0 = 0 radians
θ i=0
50 x0 = 50 + r cos(θ0) = 50 + 35 cos(0)
r cos(θ)
i = 99
= 50 + 35*1 = 85

y0 = 50 + r sin(θ0) = 50 + 35 sin(0)

= 50 + 35*0 = 50
x
xi
Plot a point at (x = 85, y = 50)
50
Figure 1
Points on a circle.
 5-3

i=1 1) Let r = 35
2) Start with i = 0
θ1 = (2π/100)*1 = .0628 radians
3) Let θ i = (2 π /100) * i
x1 = 50 + r cos(θ1) = 50 + 35 cos(.0628)
4) Let x i = 50 + r cos θ i
= 50 + 35*.9980 = 84.93
5) Let y i = 50 + r sin θ i
y1 = 50 + r sin(θ1) = 50 + 35 sin(.0628)
6) Plot a point at x i,y i
= 50 + 35*.0628 = 52.20
7) Increase i by 1
Plot a point at (x = 84.93, y = 52.20) 8) If i is less than 100, then go back to step 3
and continue in sequence
...
9) If you got here, i = 100 and you are done
i = 50
Figure 2
θ50 = (2π/100)*50 = π radians A program for calculating the points around a circle.

x50 = 50 + r cos(θ50) = 50 + 35 cos(π)

Exercise 1
= 50 + 35*(–1) = 15
Follow through the instructions in Figure (2) and see that
y50 = 50 + r sin(θ50) = 50 + 35 sin(π) you are actually creating the individual steps shown
earlier.
= 50 + 35*0 = 50

Plot a point at (x = 15, y = 50)

...

In the above, not only will it be tedious doing the

calculations, it is even tedious writing down the steps.
That is why we only showed three of the required 100
steps.
The first improvement is to find a more efficient way
of writing down the steps for calculating and plotting
these points. Instead of spelling out all of the details of
each step, we would like to write out a short set of
instructions, which, if followed carefully, will give us
all the steps indicated above. Such instructions might
look as follows:
5-4 Computer Prediction of Motion

PROGRAM FOR CALCULATION The DO LOOP

The set of instructions shown in Figure (2) could be In a sense, the set of instruction in Figure (2) is already
called a plan or program for doing the calculation. A in the form of sloppy BASIC, or you might say pidgin
similar set of instructions typed into a computer is BASIC. We only have to clean up a few grammatical
called a computer program. Our instructions in Figure rules and it will work well. The first problem we will
(2) would not be of much use to a person who spoke address is the statement in instruction #8.
only German. But if we translated the instructions into 8) If i is less than 100, then go back to step 3
German, then the German speaking person could fol- and continue in sequence
low them. Similarly, this particular set of instructions
is not of much use to a computer, but if we translate There are two problems with this instruction. One is
them into a language the computer “understands”, the that it is long and wordy. Computer languages are
computer can follow the instructions. usually designed with shorter, crisper instructions. The
second problem is that the instruction relies on number-
The computer language we will use in this course is ing instructions, as when we say “go back to step 3”.
called BASIC, a language developed at Dartmouth There is no problem with numbering instructions in
College for use in instruction. The philosophy in the very short programs, but clarity suffers in long pro-
design of BASIC is that it be as much as possible like grams. The name “step 3” is not a particularly descrip-
an ordinary spoken language so that students can tive name; it does not tell us why we should go back
concentrate on their calculations rather than worry there and not somewhere else. It is much better to state
about details of operating the computer. Like human that we have a cyclic calculation, and that we should go
languages, the computer language has evolved over back to the beginning of this particular cycle.
time, becoming easier to use and clearer in meaning.
The version of BASIC we will use is called True The grammatical construction we will use, one of the
BASIC, a modern version of BASIC written by the variations of the so-called “DO LOOP”, has the follow-
original developers of the language. ing structure. We mark the beginning of the cyclic
process with the word “DO”, and end it with the
The way we will begin teaching you the language command “LOOP UNTIL...”. Applied to our instruc-
BASIC is to translate the set of instructions in Figure tions in Figure (2), the DO LOOP would look as
(2) into BASIC. We will do this in several steps, follows:
introducing a few new ideas at a time, just as you learn
a few rules of grammar at a time when you are learning LET r = 35
a foreign language. We will know that we have arrived LET i = 0
at the actual language BASIC when the computer can DO
successfully run the program. It is not unlike testing
your knowledge of a foreign language by going out in LET θ i = (2π /100) * i
the street and seeing if the people in that country LET x i = 50 + r cos θ i
understand you.
LET y i = 50 + r sin θ i
Plot a point at x i,y i
Increase i by 1
LOOP UNTIL i = 100
All done
Figure 3
Introducing the DO LOOP.
 5-5

In the instructions in Figure (3), we begin by establish- is to first evaluate the right hand side and store the
ing that r = 35 and that i will start with the value 0. Then results in the memory cell mentioned on the left side. In
we mark the beginning of the cyclic calculation with this example the computer evaluates i + 1 by first
the command DO, and end it with the command looking in cell “i” to see what number is stored there.
“LOOP UNTIL i = 100”. The idea is that we keep It then adds 1 to that stored value to get the value (i + 1).
repeating all the stuff between the “DO” line and the To finish the command, it looks for a cell labeled “i”,
“LOOP...” line until our value of i has been incrimi- removes the number stored there and replaces it with
nated up to the value i = 100. When i reaches 100, then the value just calculated. The net result of all this is that
the loop command is ignored and we have finished both the numerical value stored in cell i is increased by 1.
the loop and the calculation.
There is a good mnemonic that helps you remember
how a LET statement works. In the command LET
The LET Statement i = i + 1, the computer takes the old value of i, adds
Another major grammatical rule is needed before one to get the new value, and stores that in cell i. If
Figure (3) becomes a BASIC program that can be read we write the LET statement as
by the computer. That involves a deeper understanding
of the LET statement that appears in many of the LET i new = i old + 1
instructions.
then it is clear what the computer is doing, and we are
One example of a LET statement is the following not tempted to cancel the i’s. In this text we will often
use the subscripts “old” and “new” to remind us what
LET i = i + 1 (2) the computer is to do. When we actually type in the
commands, we will omit the subscripts “old” and
At first sight, statement (2) looks a bit peculiar. If we
“new”, because the computer does that automatically
think of it as an equation, then we would cancel the i’s
when performing a LET command.
and be left with
With this understanding of the LET statement, our
LET = 1
program for calculating the points on a circle becomes
which is clearly nonsense. Thus the LET statement is
not really an equation, and we have to find out what it LET r = 35
is. LET i = 0
DO
The LET statement combines the computer’s ability to
do calculations and to store numbers in memory. To LET θ i = (2π /100) * i
understand the memory, think of the mail boxes at the
post office. Above each box there is a name like LET x i = 50 + r cos θ i
“Jones”, and Jones’ mail goes inside the box. In the LET y i = 50 + r sin θ i
computer, each memory cell has a name like “i”, and a
number goes inside the cell. Unlike a mail box, which Plot a point at x i,y i
can hold several letters, a computer memory cell can
store only one number at a time. LET i new = i old + 1
LOOP UNTIL i = 100
The rule for carrying out a LET statement like
All done
LET i = i + 1
Figure 4
Handeling the LET statement.
5-6 Computer Prediction of Motion

Variable Names
In Figure (4), we begin our repetitive DO LOOP by Our command
calculating a new value of the angle θ . This new
value is stored in the memory cell labeled θ , and LET θ = (2*π/100)*i
later used to calculate new values of x = r cos θ and
has been rewritten in the form
y = r sin θ . Since we are using the updated values of
θ , we can drop the subscripts i on the variables θ i, LET Theta = (2*Pi/100)*i
x i , yi . After we plot the point at the new coordinate Unfortunately, only a few special symbols are avail-
(x, y), we calculate the next value of i with the able in the font chosen by True BASIC. When we want
command LET i = i + 1, and then go back for the next a symbol like θ and it is not available, we can spell it out
calculation. as we have done.

To get a working BASIC program, there are a few other We have spelled out the name “Pi” for π, because
small changes that are easily seen if we compare our BASIC understands that the letters “Pi” stand for the
program in Figure (4) with the working BASIC pro- numerical value of π. (“Pi” is what is called a reserved
gram in Figure (5). Let us look at each of the changes. word in True BASIC.)

Multiplication
We are used to writing an expression like

r cos(θ)
and assuming that the variable r multiplies the function
cos(q). In BASIC you must always use an “ *” for
multiplication, thus the correct way to write r cos( θ ) is

r*cos(θ)
Similarly we had to write 2*Pi rather than 2Pi in the line
defining Theta.

Plotting a Point
Our command

Plot a point at (x, y)

becomes in BASIC
Figure 5
Listing of the BASIC program. PLOT x,y
It is not as descriptive as our command, but it works the
same way.
 5-7

Comment Lines This setting gives us plenty of room to plot anything in

In a number of places in the BASIC program we have the range 0 to 100 as shown by the dotted square in
added lines that begin with an exclamation point "!". Figure (6).
These are called “comment lines” and are included to
make the program more readable. A comment line has When we are plotting a circle, we would like to have it
no effect on the operation of the program. The com- look like a circle and not get stretched out into an
puter ignores anything on a line following an exclama- ellipse. In other words we would like a horizontal line
tion point. Thus the two lines 10 units long to have the same length as a vertical line
10 units long. True BASIC for the Macintosh com-
LET i = i + 1 puter could have easily have done this because Macin-
LET i = i + 1 ! Increment i tosh pixels are square, so that equal horizontal and
vertical distances should simply contain equal num-
are completely equivalent. (If you write a command bers of pixels. (A pixel is the smallest dot that can be
that does something peculiar, you can explain it by drawn on the screen. A standard Macintosh pixel is 1/
adding a comment as we did above.) 72 of an inch on a side, a dimension consistent with
typography standards.)
Plotting Window However True BASIC also works with IBM comput-
The only really new thing in the BASIC program of ers where there is no standard pixel size or shape. To
Figure (5) is the SET WINDOW command. We are handle this lack of standardization, True BASIC left it
going to plot a number of points whose x and y values up to the user to guess what choice in the SET WIN-
all fall within the range between 0 and 100. We have DOW command will give equal x and y dimensions.
to tell the computer what kind of scale to use when This is an unfortunate compromise.
plotting these points.
If you are using a Macintosh MacPlus, Classic or SE,
In the command one of the computers with the 9" screens, set the
horizontal dimension 1.5 times bigger than the vertical
SET WINDOW -40, 140, -10, 110
one, use the full screen as an output window, and the
the computer adjusts the plotting scales so that the dimensions will match (circles will be circular and
computer screen starts at -40 and goes to +140 along the squares square.) If you have any other screen or com-
horizontal axis, and ranges from -10 to + 110 along the puter, you will have to keep adjusting the SET WIN-
vertical axis, as shown in Figure (6). DOW command until you get the desired results.
(Leave the y axis range from –10 to +110, and adjust
the x axis range. For the 15" screen of the iMac, we got
110
100 a round circle plot for x values from –33 to +133.)

0 100
–10

–40 140

Figure 6
Using the SET WINDOW command.
5-8 Computer Prediction of Motion

Practice Exercise 3 Plotting a Circular Line

The best way to learn how to handle BASIC programs It’s pretty hard to see the dots in Figure (7). The output
is to start with a working program like the one in Figure can be made more visible if lines are drawn connecting
(5), and make small modifications and see what hap- the dots to give us a circular line. In BASIC it is very easy
pens. Below are a series of exercises designed to give to connect the dots you are plotting. You simply add a
you this practice, while at the same time introducing semicolon after the PLOT command. I.e., change the
some techniques that will be useful in the analysis of command
strobe photographs. When you finish these exercises, PLOT x, y ! Plots dots
you will be ready to use BASIC as a tool for predicting
the motion of projectiles, both without and with air to the command
resistance, which is the subject of the remainder of the
PLOT x, y; ! Plots lines
chapter.
The result is shown in Figure (8).
Exercise 2 A Running Program
Get a copy of True BASIC (preferably version 2.0 or Modify your program by changing the PLOT command
later), launch it, and type in the program shown in Figure as shown, and see that your output looks like Figure (8).
(5). Type it in just as we have printed it, with the same (Optional—There is a short gap in the circle on the right
indentations at the beginning of the lines, and the same hand side. Can you modify your program to eliminate
comments. Then run the program. You should get an this gap?)
output window that has the circle of dots shown in Figure
(7).

If something has gone wrong, and you do not get this

output, first check that you have typed exactly what we
printed in Figure 5. If that doesn’t work, get help from a
friend, advisor, computer center, whatever. Sometimes
the hardest part of programming is turning on the
equipment and getting things started properly.

Once you get your circle of dots, save a copy of the

program.
Figure 8
The circle of lines plotted by adding a
semicolon to the end of the PLOT command.

Figure 7
The circle of dots plotted by the
program shown in Figure (5).
 5-9

Exercise 4 Labels and Axes Exercise 5a Numerical Output

Although we have succeeded in drawing a circle, the Sometimes it is more useful to see the numerical results
output is fairly bare. It is impossible to tell, for example, of a calculation than a plot. This can easily be done by
that we have a circle of radius 35, centered at x = y = 50. replacing the PLOT command by a PRINT command.
We can get this information into the output by drawing
axis and labeling them. This can be done by adding the To do this, go back to your original circle plotting
following lines near the beginning of the program, just program (the one shown in Figure (5) which we asked
after the SET WINDOW command you to save), and change the line
PLOT x, y

to the two lines

PRINT "x = "; x, "y = ";y

!PLOT x, y

What we have done is added the PRINT line, and then

The results of adding these lines are shown in Figure (9). put an exclamation point at the beginning of the PLOT
The BOX LINES command drew a box around the line so that the computer would ignore the PLOT com-
region of interest, and the three PLOT TEXT lines gave mand. (We left the PLOT line in so that we could use it
us the labels seen in the output. later.) If we ran the program we get a whole bunch of
printing, part of which is shown in Figure (10). Do this
Add the 5 lines shown above to your program and see and see that you get the same results.
that you get the results shown in Figure (9). Save a copy
of that version of the program using a new name. Then
find out how the BOX LINES and PLOT TEXT com-
mands work by making some changes and seeing what
happens.

Figure 9
A box, drawn by the BOX LINES command Figure 10
makes a good set of axes. You can then plot If we print the coordinates of every
text where you want it. point, we get too much output.
5-10 Computer Prediction of Motion

Selected Printing (MOD Command) The command MOD(i, 10) means evaluate the number
The problem with the output in Figure (10) is that we i counting modulus 10. Thus when i gets to 10, MOD(i,
print out the coordinates of every point, and we may 10) goes back to zero. When i gets to 20, MOD(i, 10)
not want that much information. It may be more goes back to zero again. Thus as i increases, MOD(i,
convenient, for example, if we print the coordinates 10) goes back to zero every time i hits a power of 10.
for every tenth point. To do this, we use the following
trick. We replace the PRINT command In the command
IF MOD(i,10) THEN PRINT "x = ";x, "y = ";y
PRINT "x = ";x, "y = ";y
no printing occurs until i increases to a power of ten.
by the command Then we do get a print. The result is that with this
command the coordinates of every tenth point are
IF MOD(i,10) THEN PRINT "x = ";x, "y = ";y printed, and there is no printing for the other points, as
we see in Figure (11).
To understand what we did, remember that each time
we go around the loop, the variable i is incriminated by
Exercise 5b
1. The first time i = 0, then it equals 1, then 2, etc.
Take your program from Exercise (5a), modify the
The function MOD( ), stands for the mathematical print command with the MOD statement, and see that
term “modulus”. If we count modulus 3, for ex- you get the results shown in Figure (11). Then figure
ample, we count: 0, 1, 2, and then go back to zero out how to print every 5th point or every 20th point.
when we hit 3. Comparing regular counting with See if it works.
counting MOD 3, we get:

regular counting: 0123456789

counting MOD 3: 0120120120

Counting MOD 10, we go:

0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 0, 1, 2, 3, ... etc. Every time
we get up to a power of ten, we go back to zero.

Figure 11
The coordinates of every tenth point is
printed when we use the MOD command.
 5-11

Exercise 6 Plotting Crosses Next add in the subroutine lines as shown in Figure (13)
(your program should look just like Figure 13) and see
Our last exercise will be to have the computer plot both
if you get the results shown in Figure (12). When you
a circle, and a set of crosses located at every tenth point
have a running program, figure out how to make the
along the circle as shown in Figure (12). This is about
crosses bigger or smaller. How can you plot twice as
as fancy a plot as we will need in the course, so that you
many crosses?
are almost through practicing the needed fundamen-
tals.

To plot the crosses seen in Figure (12) we added what

is called a “subroutine” shown at the bottom of Figure
(13).

To get the program shown in Figure (13), go back to the

program of Exercise (3) (we asked you to save it), and
add the command
IF MOD(i, 10) = 0 THEN CALL CROSS
where “CROSS” is the name of the subroutine at the
bottom of Figure (13). You can see that the
IF MOD(i,10) = 0 part of the command has the
subroutine called at every tenth point.

Figure 13
Figure 12 The complete BASIC program for drawing
Here we use the MOD command and a the picture shown in Figure (12).
subroutine to plot a cross at every tenth dot.
5-12 Computer Prediction of Motion

PREDICTION OF MOTION Equation (5) can be interpreted as an equation that

Now that we have the techniques to handle a repetitive predicts the projectile’s new position R i+1 in terms
calculation we can return to the problem of using the of the old position R i, the old velocity vector vi , and
step-by-step method to predict the motion of a projec- the time step ∆t. To emphasize this predictive nature
tile. The idea is that we will convert our graphical of Equation (5), let us rename R i+1 the new vector
analysis of strobe photographs, discussed in Chapter 3, R new , and the old vectors R i and vi , as R old and vold .
into a pair of equations that predict the motion of the With this renaming, the equation becomes
projectile one step at a time. We will then see how these
R new = R old + vold * ∆t (6)
equations can be applied repeatedly to predict motion
over a long period of time. which is illustrated in Figure (14b).
Figure (14a) is essentially our old Figure (3-16) Equation (6) predicts the new position of the ball
where we used a strobe photograph to define the using the old position and velocity vectors. To use
velocity of the projectile in terms of the projectile’s Equation (6) over again to predict the next new
coordinate vectors R i and R i+1 . The result was position of the ball, we need updated values for R
and v. We already have Ri + 1 or Rnew for the updated
Si R – Ri coordinate vector; what we still need is an updated
vi = = i+1 (4)
∆t ∆t velocity vector vi+1 or vnew.
If we multiply Equation (4) through by ∆t and
rearrange terms, we get

R i+1 = R i + vi ∆t (5)
which is the vector equation pictured in Figure (14a).

Si = Vi ∆t Vold ∆t

Ri R old
R i+1 R new

R i+1= R i + Si
R i+1= R i + Vi ∆t R new = R old + Vold ∆t

Figure 14a Figure 14b

To predict the next position R i + 1 of the ball, we add the So that we do not have to number every point
ball's displacement S i = vi ∆t to the present position R i . in our calculation, we label the current
position "old", and the next position "new".
 5-13

To obtain the updated velocity, we use Figure There are two important criteria for using this step-
(3-17), drawn again as Figure (15a), where the by-step method of predicting motion described above.
acceleration vector ai was defined by the equation One is that we must have an efficient method to
vi + 1 – v i handle the repetitive calculations involved. That is
ai = (7) where the computer comes in. The other is that we
∆t
Multiplying through by ∆t and rearranging terms, must know the acceleration at each step. In the case
Equation (7) becomes of projectile motion, where a is constant, there is no
problem. We can also handle projectile motion with
vi + 1 = v i + a i ∆ t (8) air resistance if we can use formulas like
which expresses the new velocity vector in terms of
the old velocity vi and the old acceleration ai, as aair = – K v
illustrated in Figure (15a). Changing the subscripts a = g + aair
from i + 1 and i to “new” and “old” as before, we get shown in Figure (3-31). To handle more general
vnew = vold + a old* ∆ t (9) problems, we need a new method for determining the
as our basic equation for the projectile’s new velocity. acceleration vector. That new method was devised by
Isaac Newton and will be discussed in the chapter on
We have now completed one step in our prediction of Newtonian Mechanics. In this chapter we will focus on
the motion of the projectile. We start with the old projectile motion with or without air resistance so that
position and velocity vectors R old and vold, and used we know the acceleration vectors throughout the mo-
Equations (6) and (9) to get the new vectors Rnew and tion.
vnew. To predict the next step in the motion, we change
the names of Rnew, vnew to R old and vold and repeat
Equations (6) and (9). As long as we know the
acceleration vector ai at each step, we can predict the
motion as far into the future as we want.

Vold Vold
Vold
Vnew
Vnew
A ∆t
R old
R new

( V new – Vold) Vnew = Vold + A *∆t

A=
∆t

Figure 15a Figure 15b

Once we get to the "new" position, we will Yhe value of vnew is obtained from the
need the new velocity vector vnew in order definition of acceleration A = ( vnew – vold ) /∆ t .
to predict the next new position.
5-14 Computer Prediction of Motion

TIME STEP AND When we are doing numerical calculations, however,

INITIAL CONDITIONS we are not limited by graphical techniques and can get
more accurate results by using shorter time steps. We
Equation (6) and (9) are the basic components of our
will see that for the analysis of our strobe photographs,
step-by-step process, but there are several details to be
time steps in the range of .01 second to .001 second
worked out before we have a practical program for
work well. Much shorter time steps, like a millionth of
predicting motion. Two of the important ones are the
a second, greatly increase the computing time required
choice of a time step ∆t, and the initial conditions that
while not giving more accurate results. If we use
get the calculations started.
ridiculously short time steps like a nanosecond, the
In our strobe photographs we generally used a time step computer must do so many calculations that the round-
∆t = .1 second so that we could do effective graphical off error in the computer calculations begins to accu-
work. If we turn the strobe up and use a shorter time mulate and the answers get worse, not better. Just as
step, then the images are so close together, the arrows with graphical work there is an optimal time step.
representing individual displacement vectors are so (Later we will have some exercises where you try
short, that we cannot accurately add or subtract them. various time steps to see which give the best results.)
Yet if we turn the strobe down and use a longer ∆t, our
analysis becomes too coarse to be accurate. The choice
∆t = .1 sec is a good compromise.

V 0∆
t
V–0 0 – 1
0
V0 –1 )
R –1
( R1 –
R –1
R0 R1

V
( R 1 – R –1 )
– 0 = (2 ∗∆t)

Figure 16 Figure 17
By using a very short time Step dt in our computer The displacement v0 ∆ t is just half the displacement
calculation, we will closely follow the continuous ( R 1 – R –1 ) . This is an exact result for projectile
path shown by the dotted lines. Thus we should use motion, and quite accurate for most strobe
the instantaneous velocity vector v0 , rather than the photographs.
strobe velocity v0 as our initial velocity.
 5-15

When we use a short time step of .01 seconds or less for To avoid confusing the longer strobe time step and the
analyzing our projectile motion photographs, we are shorter computer time step that we will be using in the
close to what we have called the instantaneous velocity same calculation, we will give them two different
illustrated in Figure (3-32). But, as shown in Figure names as follows. We will use ∆t for the longer strobe
(16), the instantaneous velocity v0 and the strobe time step, which is needed for calculating the initial
velocity v0 are quite different if the strobe velocity was instantaneous velocity, and the name dt for the short
obtained from a strobe photograph using ∆t = .1 computer time step.
second. To use the computer to predict the motion we
see in our strobe photographs, we need the initial ∆t = time between strobe flashes
position R0 and the initial velocity v0 as the start for our dt = computer time step (12)
step-by-step calculation. If we are going to use a very
short time step in our computer calculation, then our This choice of names is more or less consistent with
first velocity vector should be the instantaneous veloc- calculus, where ∆t is a small but finite time interval and
ity v0, not the strobe velocity . dt is infinitesimal.

This does not present a serious problem, because

back in Chapter 3, Figure (3-33) reproduced here as
Figure (17), we showed a simple method for obtain-
ing the ball’s instantaneous velocity from a strobe
photograph. We saw that the instantaneous velocity
v0 was the average of the previous and following
strobe velocities v–1 and v1 :
v–1 + v1
v0 = (10)
2
where v–1 = S –1 /∆t and v0 = S 0 /∆t .
However, the sum of the two displacement vectors
( S –1 + S 0 ) is just the difference between the coor-
dinate vectors R 1 and R –1 as shown in Figure (17).
Thus the instantaneous velocity of the ball at Posi-
tion (0) in Figure (17) is given by the equation
R 1 – R –1
v0 = (11)
∆t
If we use Equation (11) as the formula for the initial
velocity in our step-by-step calculation, we are starting
with the instantaneous velocity at Position (0) and can
use very short time intervals in the following steps.
5-16 Computer Prediction of Motion

AN ENGLISH PROGRAM The first version of the English projectile motion

FOR PROJECTILE MOTION program is shown in Figure (18). This program is
We are now ready to write out a program for predicting designed to predict the motion of the steel ball projec-
the motion of a projectile. The first version will be what tile shown in Figure (3-8) and used for the drawings
we call an “English” program -- one that we can easily seen in Figures (15) and (16).
read and understand. Once we have checked that the
program does what we want it to do, we will see what In the program we begin with a statement of the initial
modifications are necessary to translate the program conditions – the starting point for the analysis of the
into BASIC. motion. In this photograph, the strobe time step is ∆t =
.1 seconds, and we are beginning the calculations at the
position labeled R0 in Figure (16). The instantaneous
velocity at that point is given by the formula
English Program v0 = (R1 – R-1)/2∆t as shown in Figure (17). These
results appear in the program in the lines
! --------- Initial conditions
LET ∆t = .1 LET ∆t = .1
LET Rold = R0 LET Rold = R0
LET Vold = ( R 1 – R -1) 2*∆t LET Vold = ( R 1 – R –1) (2*∆t )
LET Told = 0 Our new thing we are going to do in this program is
keep track of the time by including the variable T in
our calculations. We begin by setting T = 0 in the
! --------- Computer Time Step
initial conditions, and then increment the clock by a
LET dt = .01 computer time step dt every time we go around the
calculation loop. This way T will keep track of the
! --------- Calculational loop elapsed time throughout the calculations. The clock
is initialized by the command
DO
LET Rnew = Rold + Vold*dt LET Told = 0
LET A = g The computer time step dt plays a significant role in the
LET Vnew = Vold + A*dt program because we will want to adjust dt so that each
LET Tnew = Told + dt calculational step is short enough to give accurate
results, but not so short to waste large amounts of
PLOT R computer time. We will start with the value dt = .01
LOOP UNTIL T > 1 seconds, as shown by the command
Figure 18 LET dt = .01
Later we will try different time steps to see if the results
change or are stable.
 5-17

The important part of the program is the calculational The calculational loop itself is bounded by the DO and
loop which is repeated again and again to give us the LOOP UNTIL commands:
step-by-step calculations. The calculations begin with
the command DO
...
LET Rnew = Rold + Vold*dt ...
which is the calculation pictured in Figure (14b). Here ...
we are using the short computer time step dt so that LET Tnew = Told + dt
Rnew will be the position of the ball dt seconds after it
...
was at Rold.
LOOP UNTIL T > 1
The next line
Remember that with a DO – UNTIL loop there is a
LET A = g test to see if the condition, here T > 1, is met. If T has
simply tells us that for this projectile motion the ball’s not reached 1, we go back to the beginning of the
acceleration has the constant value g. (Later, when we loop and repeat the calculations. Because of the
predict projectile motion with air resistance, we change command LET Tnew = Told + dt , T increments by dt
this line to include the acceleration produced by the air each time around. At some point T will get up to one,
resistance.) the condition will be met, and we leave the loop. At
that point the program is finished. (We chose the
To calculate the new velocity vector, we use the condition T > 1 to stop the calculation because the
command projectile spends less than one second in the strobe
photograph. Later we may use some other criterion
LET Vnew = Vold + A*dt to stop the calculation.)
which is pictured in Figure (15b). Again we are using
the short computer time step dt rather than the longer
strobe rate ∆t.

The last two lines inside the calculational loop are

LET Tnew = Told + dt

PLOT R
The first of these increments the clock so that T will
keep track of the elapsed time. Then we plot a point at
the position R so that we can get a graph of the motion
of the ball.
5-18 Computer Prediction of Motion

A BASIC PROGRAM FOR To translate the initial conditions, we used the

PROJECTILE MOTION experimental values of the ball's coordinates given in
Figure (3-10), the steel ball projectile motion strobe
The program in Figure (18) is quite close to a BASIC
photograph we have been using for all of our drawings.
program. We have the LET statements and the
These coordinates are reproduced below in Figure
Do – LOOP commands that appeared in our working
(19).
BASIC program back in Figure (5). The only problem
is that BASIC unfortunately does not understand vec-
Ball coordinates
tor equations. In order to translate Figure (18) into a
-1) ( 8.3, 79.3)
workable BASIC program, we have to convert all the
0) (25.9, 89.9)
vector equations into numerical equations.
1) (43.2, 90.2)
To do this conversion, we write the vector equation out 2) (60.8, 80.5)
as three component equations as shown below. 3) (78.2, 60.2)
4) (95.9, 30.2)
A=B+C (13)
becomes
Figure 19
Experimental coordinates of the steel
Ax = Bx + Cx (14a)
ball projectile, from Figure (3-10).
Ay = By + Cy (14b)
Az = Bz + Cz (14c) Using the fact that R0 = (25.9, 89.9), we can write the
We saw this decomposition of a vector equation into equation
numerical or scalar equations in Chapter 2 on vectors
and Chapter 4 on calculus. (It should have been in LET Rold = R0
Chapter 2 but was accidently left out. It will be put in.) as the two equations
If the motion is in two dimensions, say in the x–y plane,
then we only need the x and y component Equations LET Rx = 25.9
(14a) and (14b).
LET Ry = 89.9
Let us apply this rule to translate the vector LET
In a similar way we use the experimental values for
statement
R1 and R –1 to evaluate the initial value of Vold.
LET Rnew = Rold + Vold*dt (15) In Figure (20) we have converted the vector LET
into two numerical LET statements. If we use the statements into scalar ones to obtain a workable
notation BASIC program. We have also included the vector
statements to the right so that you can see that the
R = (Rx, Ry) ; V = (Vx, Vy) English and BASIC programs are essentially the
same. We also added the SET WINDOW command
we get, dropping the subscripts “new” and “old”, so that the output could be plotted.
Rx = Rx + Vx*dt (16a) In Figure (21), we show the output from the Basic
Ry = Ry + Vy*dt (16b) program of Figure (20). It looks about as bad as
Figure (7), the output from our first circle plotting
We can drop the subscripts “new” and “old” because in program. In the following exercises we will add
carrying out the LET statement the computer must use axes, plot points closer together, and plot crosses
the old values of Rx and Vx to evaluate the sum every tenth of a second. In addition, we will get
Rx + Vx*dt, and this result which is the new value of numerical output that can be compared directly with
Rx is stored in the memory cell labeled “Rx”. the experimental values shown in Figure( 19).
 5-19

BASIC Program English Program

! --------- Initial conditions

LET ∆t = .1
LET Rold = R0
LET Vold = ( R 1 – R -1) 2*∆t
LET Told = 0

! --------- Computer Time Step

LET dt = .01

! --------- Calculational loop

DO
LET Rnew = Rold + Vold*dt
LET A = g
LET Vnew = Vold + A*dt
LET Tnew = Told + dt
PLOT R
LOOP UNTIL T > 1

END
Figure 20
Projectile Motion program in both BASIC and English.

Figure 21
Output from the BASIC
program in Figure (20).
(Look closely for the dots.)
5-20 Computer Prediction of Motion

Exercise 7 Exercise 11 Reducing Numerical Output

Start BASIC, type the BASIC projectile motion program Because the MOD function works reliably only with
shown in Figure 20, and run it. Keep fixing it up until it integers, we will introduce a counter variable i like we
gives output that looks like that shown in Figure 21. had in our circle plotting program.

Exercise 8 Changing the Time Step First we must initialize i . We can do that at the same time
we initialize dt as shown.
Reduce the time step to dt = .001 seconds. The plot
should become essentially a continuous line. ! --------- Computer Time Step and Counter
LET dt = .01
Exercise 9 Numerical Output LET i = 0
Change the plot command to a print statement to see
numerical output. You can do this by turning the PLOT Then we will increment i by 1 each time we go around
command into a comment, and adding a PRINT com- the calculational loop, using the now familiar command
mand as shown below. LET i = i+1. If we are using a time step dt = .001 then
we have to go around the calculational loop 100 times
!PLOT Rx,Ry to reach a time interval of .1 seconds. To do this, our
print command should start with IF MOD(i,100) = 0...
PRINT "Rx = ";Rx, "Ry = ";Ry
Thus, inside the calculational loop, the Print command
Just as in Exercise 5, you will get too much output when of Exercise 9 should be replaced by
you run the program. If you have done Exercise 8, the LET i = i+1
coordinates of the ball will be printed every thousandth IF MOD(i,100) = 0 THEN PRINT "RX = ";RX, "RY = ";RY
of a second. Yet from the strobe photograph, you have
data for tenth second intervals. The next two exercises Make the changes shown above, run your program,
are designed to reduce the output. and see that you get the output shown below in Figure
22. Compare these results with the experimental values
Exercise 10 Attempt to reduce output shown in Figure 19.
Replace the PRINT command of exercise 9 by the
command
IF MOD(T,.1) = 0 THEN PRINT "Rx = ";Rx, "Ry = ";Ry

The idea is to pull the same trick we used in reducing the

output in Exercise 5, going from Figure 10 to Figure 11.
In The above MOD statement, we would hope that we
would get output every time T gets up to a multiple of 0.1.

Try the modification of the PRINT command using

MOD(T,.1) as shown above. When you do you will not
get any output. The MOD(T,.1) command does not
work, because the MOD function generally works only
with integers. We will fix the problem in the next Figure 22
exercise. Numerical output from the projectile motion
program, printed at time intervals of .1 seconds.
These predicted results should be compared with
the experimental results seen in Figure 3-10.
 5-21

Exercise 12 Plotting Crosses Projectile Motion Program

Now we have the MOD statement to reduce the printing
output, we can use the same trick to plot crosses in the
output at .1 second intervals. All we have to do is restore
the PLOT command, change the MOD statement to
IF MOD(i,100) = 0 THEN CALL CROSS
and add a cross plotting subroutine which should now
look like
! --------- Subroutine "CROSS" draws a cross at Rx,Ry.
SUB CROSS
PLOT LINES: Rx-2,Ry; Rx+2,Ry
PLOT LINES: Rx,Ry-2; Rx,Ry+2
END SUB
The only change from the CROSS subroutine in the
circle plotting program is that the cross is now centered
at coordinates (Rx,Ry) rather than (x,y) as before.

The complete cross plotting is shown in Figure (23), and

the results are plotted in Figure (24). Modify your
projectile motion program to match Figure (23), and
see that you get the same results. (How did we stop the
plotting outside the square box?)

Figure 24 Figure 23
Output from our BASIC projectile Projectile motion program that plots
motion program of Figure 23. crosses every tenth of a second.
5-22 Computer Prediction of Motion

PROJECTILE MOTION intuition for the role played by the acceleration vector.
WITH AIR RESISTANCE We will see that if we know a particle’s acceleration,
have a formula for it, and know how the particle started
Projectile motion is an example of a very special kind
moving, we can predict where the particle will be at any
of motion where the acceleration vector is constant –
time in the future.
does not change in either magnitude or direction. In this
special case we can easily use calculus to predict Once we have gained experience with this kind of
motion far into the future. prediction, we can then focus our attention on the core
problem in mechanics, namely finding a general method
But let the acceleration vector change even by a small
for determining the acceleration vector. As we men-
amount, as in the case of projectile motion with air
tioned, the general method was discovered by Newton
resistance, and a calculus solution becomes difficult or
and will be discussed shortly in the chapter on Newto-
impossible to obtain. This illustrates the important role
nian Mechanics.
the acceleration vector plays in the prediction of mo-
tion, but overemphasizes the importance of motion In our study of the effects of air resistance, we will use
with constant acceleration. as our main example the styrofoam ball projectile
shown in Figures (3-30a, b) and reproduced here as
With a computer solution, very little additional effort is
Figures (25a, b). To obtain the coordinates listed in
required to include the effects of air resistance. We will
be able to adjust the acceleration for different amounts Figure (25b), each image was enlarged and studied
separately. As a result, these coordinates should be
or kinds of air resistance. The point is to develop an
accurate to within half a millimeter (except for possible
errors due to parallax in taking the photograph).
0
-1 1
A g
A g 2

A g

3
-1) ( 5.2, 94.9) A
g
0) (24.0, 101.4)
1)
-1) (40.8,
( 5.35,97.8)
94.84)
2)0) (56.5,
(24.03,101.29)
85.3) 4
3)1) (70.8,
(40.90,64.7)
97.68)
A
2) (56.52, 85.15)
4)3) (83.4,
(70.77,37.1)
64.56)
g
5)4) (95.2,
(83.48, 3.9)
36.98)
5) (95.18, 3.86)
5

Figure 25b
Figure 25a To obtain as accurate a value as we could
The styrofoam projectile of Figure (3-30a). for each ball coordinate, each image was
We have printed a negative of the photograph enlarged and studied separately.
to show the grid lines more distinctly.
 5-23

Figure (26), a reproduction of Figure (3-31), is a The simplest formula we can write which has Aair
detailed analysis of the ball’s acceleration at Position pointing in the –V direction is Equation (18),
(3). As shown in Figure (26) we can write the formula Aair = –KV, where K is a constant that we have to find
for the ball’s acceleration vector A in the form from the experiment. If some choice of the constant K
allows us to accurately predict all the experimental
A = g + Aair (17) points in Figure (25), then we will have verified that
Equation (18) is a reasonably accurate description of
where one possible formula for Aair is the effects of air resistance.
Aair = –KV (18) It may happen, however, that one choice of K will lead
to an accurate prediction of one position of the ball,
V being the instantaneous velocity of the ball.
while another choice leads to an accurate prediction of
In Equation (17), Aair is defined as the change from the another point, but no value of K gives an accurate
normal acceleration g the projectile would have with- prediction of all the points. If this happens, equation
out air resistance. As we see, Aair points opposite to (18) may be inadequate, and we may need a more
V, which is the direction of the wind we would feel if complex formula.
we were riding on the ball. Figure (26) suggests the
The next level of complexity is that K itself depends on
physical interpretation that this wind is in effect blow-
the speed of the ball. Then Aair would have a
ing the acceleration vector back. It suggests that
magnitude related to V2, V3, or something worse. In
acceleration vectors can be pushed or pulled around,
this case the air resistance is “nonlinear” and exact
which is the underlying idea of Newtonian mechanics.
calculus solutions are not possible. But, as we see in
In Figure (26) the earth is pulling down on the ball
Exercise 15, we can still try out different computer
which gives rise to the component g of the ball’s
solutions.
acceleration, and the wind is pushing back to give rise
to the component Aair. In reality, when a sphere moves through a fluid like air
or water, the resistance of the fluid can become very
complex. At high enough speeds, the sphere can start
3
shedding vortices, the fluid can become turbulent, and
the acceleration produced by the fluid may no longer be
directed opposite to the instantaneous velocity of the
"
a3 ind sphere. In Exercise 13 we take a close look at Aair for
"w
all interior positions for the projectile motion shown in
g Figure (25). We find that to within experimental
a air accuracy, for our styrofoam projectile Aair does point
in the –V direction. Thus a formula like Equation (18)
is a good starting point. We can also tell from the
a air = –K v3 experimental data whether K is constant and what a
v3

good average value for K should be.

Figure 26
The air resistance is caused by
the wind you would feel if you
were riding on the ball.
5-24 Computer Prediction of Motion

Air Resistance Program resistance constant K should be much larger. With the
Figure (20) was our BASIC program for projectile computer, you can simply use larger and larger values
motion. We would now like to modify that program so of K to see the effects of increasing the air or fluid
that we can predict the motion of the Styrofoam ball viscosity. We ask you to do this in Exercise 15. This
shown in Figure (25). To do this, we must change the is a very worthwhile exercise, for as the fluid viscosity
command increases, as you increase K, you get an entirely new
kind of motion. There is a change in the qualitative
LET A = g character of the motion which you can observe by
rerunning the program with different values of K.
to the new command

LET A = g – KV (19)
and try different values for K until we get the best
agreement between prediction and experiment.

A complete program with this modification is shown in

Figure (27). In this program we see that Equation (19)
Use initial values
has been translated into the two component equations
from Figure (25).
LET Ax = 0 – K*Vx
Try different
LET Ay = –980 – K*Vy values of K
In addition, we are printing numerical output at .1 sec
intervals so that we can accurately compare the pre-
dicted results with the experimental ones. In the line

LET K = ...
which appears in the Initial Conditions, we are to plug New formula for A
in various values of K until we get the best agreement
that we can between theory and experiment.

Finding K does not have to be complete guesswork. In

Exercise 13 we ask you to do a graphical analysis of the
Styrofoam ball’s acceleration at several positions us-
ing the enlargements provided. From these results you
should choose some best average value for K and use
Figure 27
that as your initial guess for K in your computer
BASIC program for projectile motion with air
program. Then fine tune K until you get the best resistance. It is left to the reader to insert
agreement you can. We ask you to do this in Exercise appropriate initial conditions, and choose
14. values of the air resistance constant K.

Once you have a working program that predicts the

motion of the Styrofoam ball in Figure (25), you can
easily do simulations of different strengths of air resis-
tance. What if you had a steel ball being projected
through a viscous liquid like honey? The viscous liquid
might have the same effect as air, except that the
 5-25

In Exercise 16, we show you one way to modify the air

resistance formulas to include nonlinear effects, i.e., to 0
allow Aair to depend on V2 as well as V. What we do 1 direction of V
–
is first use the Pythagorean theorem to calculate the
magnitude V of the ball’s speed and then use that in a
more general formula for Aair. The English lines for A∆t 2
this are
g∆t 2 2
LET V = V2x + V2y ∆t = .1 sec

LET A = g – K(1 + K2*V)V (20) 0 10 20 30 40 50 cm

where we now try to find values of K and K2 that
Figure 28b
improve the agreement between prediction and experi- Blowup of position 1 in Figure 25b.
ment. The translation of these lines into BASIC is
shown in Exercise 16.

Exercise 13 Graphical Analysis

Figures (28 a,b,c,d,e) are accurate enlargements of 1
sections of Figure (25b). In each case we show three
positions of the Styrofoam projectile so that you can
determine the ball's instantaneous velocity V at the
center position. Using the section of grid you can
2
determine the magnitude of both V ∆t and A air ∆t2 . direction of V
–
From that, and the fact that ∆t = .1 sec, you can then A∆t 2
determine the size of the air resistance constant K using
the equation A air = – K V .
g∆t 2
Do this for each of the diagrams, positions 0 through 4
and then find a reasonable average value of K. How
3
constant is K? Do you have any explanation for changes
in K? ∆t = .1 sec

direction of V 0 10 20 30 40 50 cm
–
Figure 28c
0 Blowup of position 2 in Figure 25b.

-1 1
A∆t 2 g∆t 2

0 10 20 30 40 50
Figure 28a
Blowup of position 0 in Figure 25b.
5-26 Computer Prediction of Motion

Exercise 14 Computer Prediction Exercise 15 Viscous Fluid

Starting with the Basic program shown in Figure (27) After you get your program of Exercise 14 working,
use the experimental values shown in Figure (25b), allow the program to print out numerical values for up to
reproduced below, to determine the initial conditions for T = 15 seconds. After about 10 seconds, the nature of
the motion of the ball. Then use your best value of K from the motion is very different than it was at the beginning.
Exercise 13 as your initial value of K in the program. By Explain the difference. (You may be able to see the
trial and error, find what you consider the best value of difference better by printing Vx and Vy rather than Rx
K to bring the predicted coordinates into reasonable and Ry.)
agreement with experiment.
You will see the same phenomenon much faster if you
-1) ( 5.35, 94.84) greatly increase the air resistance constant K. Redo
0) (24.03,101.29) your program to plot the output, drawing crosses every
1) (40.90, 97.68) .1 seconds. Then rerun the program for ever increasing
2) (56.52, 85.15) values of K. Explain what you see.
3) (70.77, 64.56)
4) (83.48, 36.98)
5) (95.18, 3.86) 3

4
3 A∆t 2

direction of V
–
A∆t 2 direction of V g∆t 2
–

g∆t 2

4 5
∆t = .1 sec
∆t = .1 sec

0 10 20 30 40 50 cm
0 10 20 30 40 50 cm
Figure 28e
Figure 28d Blowup of position 4 in Figure 25b.
Blowup of position 3 in Figure 25b.
 5-27

Exercise 16 Nonlinear Air Resistance (optional) Exercise 17 Fan Added

In Exercise 14, you probably found that you were not In Figure (30), on the next page, we show the results of
able to precisely predict all the ball positions using one placing a rack of small fans to the right of the styrofoam
value of K. In this exercise, you allow K to depend on the ball's trajectory in order to increase the effect of air
ball's speed v in order to try to get a more accurate resistance. Now, someone riding with the ball should
prediction. One possibility is to use the following feel not only the wind due to the motion of the ball, but
formulas for Aair, which we mentioned earlier: also the wind of the fans, as shown in Figure (29).

LET V = V2x + V2y Our old air resistance formula

LET A = g – K (1 + K2*V) V (20) LET A = g + K(–Vball)

With Equations (20), you can now adjust both K and K2 should probably be replaced by a command like
to get a better prediction. These equations are trans-
lated into BASIC as follows. LET A = g + K(–Vball + Vfan)
LET V = SQR(Vx*Vx + Vy*Vy) Translated into BASIC, this would become
LET Ax = 0 – K*(1 + K2*V)*Vx LET Ax = 0 + K*(–Vx –Vfan)
LET Ay = –980 – K*(1 + K2*V)*Vy LET Ay = –980 + K*(–Vy + 0 ) (21)
Make these modifications in the program of Exercise 14, where Vball = (Vx,Vy) is the current velocity of the ball,
and see if you can detect evidence for some V2 and Vfan = (–Vfan,0) is the wind caused by the fan. We
dependence in the air resistance. assume that this wind is aimed in the –x direction and
has a magnitude Vfan. We now have two unknown
parameters K and Vfan which we can adjust to match
the experimental results shown in Figure (30).
(Vfan )

Do this, starting with the value of K that you got from

the analysis of the styrofoam projectile in Figure 25b
(Exercise 13 or 14). Does your resulting value for Vfan
wind of fan

seem reasonable? Can you detect any systematic

error in your analysis? For example, should Vfan be
"
ind stronger near the fans, and get weaker as you move
"w
left?
Vbal
l

Figure 29
Additional wind created by fan.
5-28 Computer Prediction of Motion

Figure 31
Figure 30
In this diagram, the Apple II computer has calculated
Styrofoam projectile with a bank of fans. In order
and plotted the centers of each of the images seen in
to get more air resistance, we added a bank of
the composite strobe photograph on the left.
small fans as shown. This Strobe "photograph"
was taken with the Apple II Strobe system.

Figure 32
The Apple II also prints out the Coordinates of each image. The time ∆t between crosses
is 1/10 sec. Between the dots there is a 1/30 sec time interval. The coordinates of the
initial 7 dots are printed to help determine the initial instantaneous velocity of the ball.
Chapter 6
Mass

CHAPTER 6 MASS

By now we have learned how to use either calculus or stration lectures, the results are straightforward and
the computer to predict the motion of an object whose are what one expects. But when we consider what
Chapter 6 is known. But in most problems we do not
acceleration would happen if similar experiments were carried out
Massthe acceleration, at least initially. Instead we
know with one of the objects moving at speeds near the speed
may know the forces acting on the object, or something of light, we predict a very different behavior for mass.
about the object’s energy, and use this information to This new behavior is summarized by the Einstein mass
predict motion. This approach, which is the heart of formula, a strikingly simple result that one might guess,
the subject of mechanics, involves mass, a concept but which we cannot quite derive from the definition of
which we introduce in this chapter. mass, and the principle of relativity alone. What is
needed in addition is the law of conservation of linear
In the metric system, mass is measured in grams or momentum which we will discuss in the next chapter.
kilograms, quantities that should be quite familiar to
the reader. It may be surprising that we devote an One of the striking features of Einstein’s special theory
entire chapter to something that is measured daily by of relativity is the fact that nothing, not even informa-
grocery store clerks in every country in the world. But tion, can travel faster than the speed of light. We can
the concept of mass plays a key role in the subject of think of nature as having a speed limit c. In our world,
mechanics. Here we focus on developing an experi- speed limits are hard to enforce. We will see that the
mental definition of mass, a definition that we can use Einstein mass formula provides nature with an auto-
without modification throughout our discussion of matic way of enforcing its speed limit.
physics.
Einstein’s mass formula appears to predict that no
After introducing the experimental definition, we will particle can quite reach the speed of light. We end the
go through several experiments to determine how chapter with a discussion of how to handle particles,
mass, as we defined it, behaves. In low speed experi- like photons and possibly neutrinos, that do travel at
ments, the kind we can do using air tracks in demon- precisely the speed of light.
6-2 Mass

DEFINITION OF MASS an explicit prescription for measuring mass. Then,

In everyday conversation the words mass and weight using this prescription, we will perform several experi-
are used interchangeably. Physicists use the words ments to see how mass behaves.
mass and weight for two different concepts. Briefly,
we can say that the weight of an object is the force that Recoil Experiments
the object exerts against the ground, and we can mea- As a crude experiment suppose that the two skaters
sure weight with a device such as a bathroom scale. The shown in Figure (6-1), a father and a child, stand in front
weight of an object can change in different circum- of each other at rest and then push each other apart. The
stances. For example, an astronaut who weighs 180 father hardly moves, while the child goes flying off.
pounds while standing on the ground, floats freely in an The father is more massive, harder to budge. No matter
orbiting space capsule. If he stood on a bathroom scale how hard or gently the skaters push apart, the big one
in an orbiting space craft, the reading would be zero, always recoils more slowly than the smaller one. We
and we would say he is weightless. On the other hand will use this observation to define mass.
the mass of the astronaut is the same whether he is in
orbit or standing on the ground. An astronaut in orbit In a similar but more controlled experiment, we replace
does not become massless. Mass is not what you the skaters by two carts on what is called an air track.
measure when you stand on the bathroom scales. An air track consists of a long square metal tube with
a series of small holes drilled on two sides as shown in
What then is mass? One definition, found in the dictio- Figure (6-2). A vacuum cleaner run backwards blows
nary, describes mass as the property of a body that is a air into the tube, and the air escapes out through the
measure of the amount of material it contains. Another small holes. The air carts have V-shaped bottoms
definition, which is closer to the one we will use, says which ride on a thin film of air, allowing the carts to
that the more massive an object, the harder it is to move almost without friction along the track.
budge.
To represent the two skaters pushing apart on nearly
Both of these definitions are too vague to tell us how to frictionless ice, we set up two carts with a spring
actually measure mass. In this section we will describe between them as shown in Figure (6-3a). A thread is
an experimental definition of mass, one that provides tied between the carts to keep the spring compressed.
When we burn the thread, the carts fly apart as shown

small cart
holes

pressurized film of air

air

end view
Figure 2
End view of an air track. Pressurized air from the
back side of a vacuum cleaner is fed into a square
hollow metal tube, and flows out through a series
Figure 1
of small holes. A cart, riding on a film of air, can
Two skaters, a father and a son, standing at rest
move essentially without friction along the track.
on frictionless ice, push away from each other.
The smaller, less massive child recoils faster
than the more massive father.
 6-3

in Figure (6-3b). If the two carts are made of similar Properties of Mass
material, but one is bigger than the other, the big one Since we now have an explicit prescription for measur-
will recoil at lesser speed than the small one. We say ing mass, we should carry out some experiments to see
that the big cart, the one that comes out more slowly, if this definition makes sense. Our first test is to see if
has more mass than the small one. the mass ratio m A / m B changes if we use different
strength springs in the recoil experiment. If the ratio of
Because we can precisely measure the speeds vA and vB recoil speeds vB / vA, and therefore the mass ratio,
of the recoiling air carts, we can use the experiment depends upon what kind of spring we use, then our
pictured in Figures (6-3a,b) to define the mass of the definition of mass may not be particularly useful.
carts. Let us call m A and m B the masses of carts A and
B respectively. The simplest formula relating the In the appendix to this chapter, we describe apparatus
masses of the carts to the recoil speeds, a formula that that allows us to measure the recoil speeds of the carts
has the more massive cart recoiling at less speed is with fair precision. To within an experimental accu-
racy of 5% to 10% we find that the ratio vB / vA of the
mA v recoil speeds does not depend upon how hard the spring
= B recoil definition of mass (1) pushes the carts apart. When we use a stronger spring,
mB vA
both carts come out faster, in such a way that the speed
In words, Equation 1 says that the ratio of the masses is ratio is unchanged. Thus to the accuracy of this
inversely proportional to the recoil speeds. I.e., if m A experiment we conclude that the mass ratio does not
is the small mass, the vB is the small speed. depend upon the strength of the spring used.

thread spring
Standard Mass
A B So far we have talked about the ratio of the masses of
the two carts. What can we say about the individual
frictionless Air Track
masses m A or m B alone? There is a simple way to
(a) discuss the masses individually. What we do is select
one of the masses, for instance m B, as the standard
VA VB mass, and measure all other masses in terms of m B. To
A B express m A in terms of the standard mass m B, we
multiply both sides of Equation (1) through by m B to
(b) get

V
B formula for m A
vB
V mA = mB in terms of the (2)
A vA standard mass mB

For a standard mass, the world accepts that the plati-

num cylinder kept by the International Bureau of
Weights and Measures near Paris, France, is precisely
(c)
one kilogram. If we reshaped this cylinder into an air
Figure 3 cart and used it for our standard mass, then we would
Recoil experiment. To simulate the two skaters
pushing apart, we place two carts on an air track
with a compressed spring between them. The
carts are held together by a string. When the
string is burned, the carts fly apart
as did the skaters. The more massive cart recoils
at a smaller speed vB < vA .
6-4 Mass

have the following explicit formula for the mass of cart carts related to the individual masses m C and m D? If we
A recoiled from the standard mass. perform the experiment shown in Figure (6-4), we find
using the that
v m C + D = m C + m D mass adds (4)
m A = 1 kilogram × vstd
one kilogram
A cylinder for our (3)
standard mass The experimental result, shown in Equation (4), is that
where vstd is the recoil speed of the standard mass. mass adds. The mass of the two carts recoiled together
Once we have determined the mass of one of our own is the sum of the masses of the individual carts. This
carts, using the standard mass and Equation (3), we can is the reason we can associate the concept of mass
then use that cart as our standard and return the plati- with the quantity of matter. If, for example, we have
num cylinder to the French. two identical carts, then together the two carts have
twice as much matter and twice as much mass.
Of course the French will not let just anybody use their
standard kilogram mass. What they did was to make Exercise 1
accurate copies of the standard mass, and these copies In physics labs, one often finds a set of brass cylinders
are kept in individual countries, one of them by the of various sizes, each cylinder with a number stamped
National Institute of Standards and Technology in on it, representing its mass in grams. The set usually
Washington, DC which then makes copies for others in includes a 50-gm, 100-gm, 200-gm, 500-gm, and
1000-gm cylinder. Suppose that you were given a rod
the United States to use.
of brass and a hacksaw; describe in detail how you
would construct a set of these standard masses. At your
Addition of Mass disposal you have a frictionless air track, two carts of
Consider another experiment that can be performed unknown mass that ride on the track, the standard
using air carts. Suppose we have our standard cart of 1000-gm mass from France (which can be placed on
one of the carts), and various things like springs, thread,
mass m B, and two other carts which we will call C and
and matches.
D. Let us first recoil carts C and D from our standard
mass m B, and determine that C and D have masses m C A Simpler Way to Measure Mass
and m D given by
The preceding problem illustrates two things. One is
v v
m C = m B vB ; m D = m B vB that with an air track, carts, and a standard mass, we can
C D
use our recoil definition to measure the mass of an
Now what happens if, as shown in Figure (6-4), we tie object. The second is that the procedure is clumsy and
carts C and D together and recoil them from cart B. rather involved. What we need is a simpler way to
How is the mass m C + D of the combination of the two measure mass.
vC + D vB The simpler way involves the use of a balance, which
mC mD mB is a device with a rod on a pivot and two pans suspended
from the rod, as shown in Figure (6-5). If the balance
Figure 4 is properly adjusted, we find from experiment that if
Addition of mass. If we tie two carts C and D together equal masses are placed in each pan, the rod remains
and recoil the pair from our standard mass mA , and balanced and level. This means that if we place an
use the formula unknown mass in one pan, and add brass cylinders of
v known mass to the other pan until the rod becomes
mC + D = mB v B
C+D
balanced, the object and the group of cylinders have the
for the combined mass mC + D , we find from same mass. To determine the mass of the object, all we
experiment that mC + D = mC + mD . In other words the have to do is add up the masses of the individual
mass of the pair of carts is the sum of the masses of the
individual carts, or we can say that mass adds. cylinders.
 6-5

Inertial and Gravitational Mass It is common terminology to call what we measure

The pan balance of Figure (6-5) is actually compar- in a recoil experiment the inertial mass of the object,
ing the downward gravitational force on the con- and what we measure using a pan balance the gravi-
tents of the two pans. If the gravitational forces are tational mass. The experiments of Etvös and Dicke
equal, then the rod remains balanced. What we are demonstrate that inertial mass and gravitational mass
noting is that there are equal gravitational forces on are equivalent to each other to one part in 10 11. Is
equal masses. This is an experimental result, not an this a coincidence, or is there some fundamental
obvious conclusion. For example, we could con- reason why these two definitions of mass turn out to
struct two air carts, one from wood and one from be equivalent? Einstein addressed this question in
platinum. Keep adjusting the size of the carts until his formulation of a relativistic theory of gravity
their recoil speeds are equal, i.e., until they have known as Einstein’s General Theory of Relativity.
equal recoil masses. Then put these carts on the pan We will have more to say about that later.
balance of Figure (6-5). Although the wood cart has
a much bigger volume than the platinum one, we Mass of a Moving Object
will find that the two carts still balance. The gravi- One reason we chose the recoil experiment of Figure
tational force on the two carts will be the same (3) as our experimental definition of mass is that it
despite their large difference in size. allows us to study the mass of moving objects,
In 1922, the Swedish physicist Etvös did some very something that is not possible with a pan balance.
careful experiments, checking whether two objects, From the air track experiments we have discussed so
which had the same mass from a recoil type of far, we have found two results. One is that the ratio
experiment would experience the same gravitational of the recoil speeds, and therefore the ratio of the
force as measured by a pan balance type of experi- masses of the two objects, does not depend upon the
ment. He demonstrated that we would get the same strength of the spring or the individual speeds vA and
result to one part in a billion. In 1960, R. H. Dicke
vB. If we use a stronger spring so that m A emerges
improved Etvös’ experiments to an accuracy of 1 twice as fast, m B also emerges twice as fast so that
part in 10 11. the ratio m A /m B is unchanged.
rod
In addition, we found that mass adds. If carts C and
D have masses m C and m D when recoiled individu-
pivot ally from cart B, then they have a combined mass
m C,D = m C + m D when they are tied together and
object standard both recoiled from cart B.
masses

Figure 5
Schematic drawing of a pan balance. If the balance
is correctly adjusted and if equal masses are placed
in the pans, the rod will remain level. This allows us
to determine an unknown mass simply by comparing
it to a known one.
6-6 Mass

RELATIVISTIC MASS In the next section we will discuss experiments in

In our air track experiments, we found that the ratio of which, instead of a bullet being fired by a gun, an
the recoil speeds did not depend upon the strength of the electron is ejected by an atomic nucleus. The electron
spring we used. However, when the recoil speeds is such a small particle that it is often ejected at speeds
approach the speed of light, this simple result can no approaching the speed of light. The nuclei we will
longer apply. Because of nature’s speed limit c, the consider are so much more massive that they recoil at
ratio of the recoil speeds must in general change with low speeds familiar to us, speeds like that of a jet plane
speed. or earth satellite. At these low speeds the mass of an
object does not change noticeably with speed. Thus in
To see why the recoil speed ratio must change, imagine these electron recoil experiments, the mass of the
an experiment involving the recoil of two objects of nuclei is not changing due to its motion. Any change
very different size, for example a bullet being fired in the ratio of recoil speeds is due to a change in the mass
from a gun as shown in Figure (6). Suppose, in an initial of the electron as the speed of the electron approaches
experiment not much gunpowder is used and the bullet the speed of light.
comes out at a speed of 100 meters per second and the
gun recoils at a speed of 10 cm/sec = .1 m/sec. For this We will see that as we push harder and harder on the
case the speed ratio is 1000 to 1 and we say that the gun electron, trying to make it go faster than the speed of
is 1000 times as massive as the bullet. light, the mass of the electron increases instead. It is
precisely this increase in mass that prevents the elec-
In a second experiment we use more gun powder and tron emerging at a speed greater than the speed of light
the bullet emerges 10 times faster, at a speed of 1000 and this is how nature enforces the speed limit c.
meters per second. If the ratio of 1000 to 1 is main-
tained, then we predict that the gun should recoil at a
speed of 1 meter per second. If we did the experiment, Beta (β) Decay
the prediction would be true. The electron recoils we just mentioned occur in a
process called β (beta) decay. In a β decay, a radioac-
But, as a thought experiment, imagine we used such tive or unstable nucleus transforms into the nucleus of
powerful gun powder that the gun recoiled at 1% the another element by ejecting an electron at high speeds
speed of light. If the speed ratio remained at 1000 to 1, as illustrated in Figure (7). In the process the nucleus
we would predict that the bullet would emerge at a itself recoils as shown.
speed 10 times the speed of light, an impossible result.
The bullet cannot travel faster than the speed of light,
the speed ratio cannot be greater than 100 to 1, and thus
the ratio of the masses of the two objects must have
changed. mn
vb mb mg vg ve me vn
nucleus
bullet electron
gun
Figure 6 Figure 7
To discuss higher speed recoils, consider a bullet Radioactive decay of a nucleus by β decay. In
being fired from a gun. We are all aware that the this process the unstable nucleus ejects an electron,
bullet emerges at a high speed, but the gun itself also often at speeds ve near the speed of light.
recoils. (The recoil of the gun becomes obvious the
first time you fire a shotgun.) In this setup, the
gunpowder is analogous to the spring, and the gun
and bullet are analogous to the two carts.
 6-7

The name β decay is historical in origin. When Ernest Electron Mass in β Decay
Rutherford (who later discovered the atomic nucleus) Applying our definition of mass to the β decay process
was studying radioactivity in the late 1890s, he noticed of Figure (7) we have
that radioactive materials emitted three different kinds
of radiation or rays, which he arbitrarily called me vn ve
mn vn (5)
α (alpha) rays, β (beta) rays and γ (gamma) rays, after m n = ve me
the first three letters of the Greek alphabet. Further
investigation over the years revealed that α rays were where m e and ve are the mass and recoil speed of the
beams of helium nuclei, which are also known as electron and m n and vn of the nucleus. We are assuming
α particles. The β rays turned out to be beams of that the nucleus was originally at rest before the
electrons, and for this reason a nuclear decay in which β decay.
an electron is emitted is known as a β decay. The γ rays
To develop a feeling for the speeds and masses in-
turned out to be particles of light which we now call
volved in the β decay process, we will analyze two
photons. (The particle nature of light will be discussed
examples of the β decay of a radioactive nucleus. In the
in a later section of this chapter.)
first example, which we introduce as an exercise to give
In the 1920s, studies of the β decay process raised you some practice calculating with Equation (5), we
serious questions about some fundamental laws of can assume that the electron’s mass is unchanged and
physics. It appeared that in the β decay, energy was still predict a reasonable speed for the ejected electron.
sometimes lost. (We will discuss energy and the basic In the second example, the assumption that the electron’s
law of conservation of energy in Chapter 9.) In the mass is unchanged leads to nonsense.
early 1930s, Wolfgang Pauli proposed that in β decay,
two particles were emitted—an electron and an unde-
tectable one which later became known as the neutrino.
(We will discuss neutrinos at the end of this chapter.)
Pauli’s hypothesis was that the missing energy was
carried out by the unobservable neutrino. Thirty years
later the neutrino was finally detected and Pauli’s
hypothesis verified.

Some of the time the neutrino created in a β decay

carries essentially no energy and has no effect on the
behavior of the electron and the nucleus. When this is
the case, we have the genuine 2-particle recoil experi-
ment illustrated in Figure (7). This is a recoil experi-
ment in which one of the particles emerges at speeds
near the speed of light.
6-8 Mass

Plutonium 246 Exercise 2 β Decay of Plutonium 246

We will begin with the decay of a radioactive nucleus A Plutonium 246 nucleus has an average lifetime of just
called Plutonium 246. This is not a very important over 11 days, upon which it decays by emitting an
nucleus. We have selected it because of the way in electron. If the nucleus is initially at rest, and the decay
which it β decays. is one in which the neutrino plays no role, then the
nucleus will recoil at the speed
The number 246 appearing in the name tells us the
number of protons and neutrons in the nucleus. Protons recoil speed of
vn = 572 meters Plutonium246 (9)
and neutrons have approximately the same mass m p second in a β decay
which has the value
This recoil speed is not observed directly, but enough
m p = 1.67 × 10 –27
kg mass of proton (6) is known about the Plutonium 246 β decay that this
number can be accurately calculated. Note that a
The Plutonium 246 nucleus has a mass 246 times as speed of 572 meters/second is a bit over 1000 miles per
great, thus hour, the speed of a supersonic jet.

m Plutonium 246 = 246 × m p Your exercise is to predict the recoil speed ve of the
(7) electron assuming that the mass of the electron me is
= 4.10 × 10 –25
kg
the same as the mass (me)0 of an electron at rest.
An electron at rest or moving at slow speeds has a mass
Your answer should be
m e 0 given by
ve = .86 c (10)
me 0
= 9.11 × 10 –31
kg (8) where

This is called the rest mass of an electron. We have c = 3 × 108 meters (11)
sec ond
added the subscript zero to remind us that this is the is the speed of light.
mass of a slowly moving electron, one traveling at
speeds much less than the speed of light. The above exercise, which you should have done by
now, shows that we do not get into serious trouble if we
assume that the mass of the electron did not change due
to the electron’s motion. The predicted recoil speed
ve = .86c is a bit too close to the speed of light for
comfort, but the calculation does not exhibit any obvi-
ous problems. This is not true for the following
example.
 6-9

Protactinium 236 Exercise 4 Increase in Electron Mass.

An even more obscure nucleus is Protactinium 236 Reconsider the Protactinium 236 decay, but this time
which has a lifetime of about 12 minutes before it β assume that the electron emerges at essentially the
decays. The Protactinium β decay is, however, much speed of light ( ve = c). (This is not a bad approximation,
more violent than the Plutonium 246 decay we just it actually emerges at a speed v = .99 c). Use the
discussed. If the Protactinium 236 nucleus is initially definition of mass, Equation 5, to calculate the mass of
at rest, and the neutrino plays no significant role in the the recoiling electron. Your answer should be
decay, then the recoil velocity of the nucleus is
me = 6.8 × 10– 30 kg = 7.47× (me )0 (13)
recoil speed of
vn = 5170 meters Protactinium236 (12)
second nucleus
In Exercise 4, you found that by assuming the electron
This is nine times faster than the recoil speed of the could not travel faster than the speed of light, the
Plutonium 246 nucleus. electron mass had increased by a factor of 7.47. The
emerging electron is over 7 times as massive as an
Exercise 3 Protactinium 236 β decay. electron at rest! Instead of emerging at 7 times the
speed of light, the electron comes out with 7 times as
Calculate the recoil speed of the electron assuming that
much mass.
the mass of the recoiling electron is the same as the
mass of an electron at rest. What is wrong with the
answer? Exercise 5 A Thought Experiment.
To illustrate that there is almost no limit to how much the
You do not have to work Exercise 3 in detail to see that mass of an object can increase, imagine that we per-
we get a into trouble if we assume that the mass of the form an experiment where the earth ejects an electron
recoiling electron is the same as the mass of an electron and the earth recoils at a speed of 10 cm/sec. (A β
at rest. We made this assumption in Exercise 2, and decay of the earth.) Calculate the mass of the emitted
predicted that the electron in the Plutonium 246 β electron. By what factor has the electron’s mass in-
decay emerged at a speed of .86 c. Now a nucleus of creased?
about the same mass recoils 9 times faster. If the
electron mass is unchanged, it must also recoil 9 times
faster, or over seven times the speed of light. This
simply does not happen.
6-10 Mass

THE EINSTEIN MASS FORMULA Essentially the same formula applies to the mass of a
A combination of the recoil definition of mass with the moving object. If an object has a mass m o when at rest
observation that nothing can travel faster than the speed or moving slowly as in air cart experiments (we call m o
of light, leads to the conclusion that the mass of an the rest mass of the object), then when the object is
object must increase as the speed of the object ap- moving at a speed v, its mass m is given by the formula
proaches the speed of light. Determining the formula
for how mass increases is a more difficult job. It turns mo Einstein
out that we do not have enough information at this point m = mass (14)
in our discussion to derive the mass formula. What we 1–v2 /c2 formula
have to add is a new basic law of physics called the law
of conservation of linear momentum. a result first deduced by Einstein.

We will discuss the conservation of linear momentum Equation (14) has just the properties we want. When
in the next chapter, and in the appendix to that chapter, the particle is moving slowly as in our air cart recoil
derive the formula for the increase in mass with veloc- experiments, v << c, 1–v 2 /c2 ≈ 1 and the mass of
ity. We put the derivation in an appendix because it is the object does not change with speed. But as the speed
somewhat involved. But the answer is very simple, of the object approaches the speed of light, the 1–v 2 /c 2
almost what you might guess. approaches zero, and m = m o / 1–v 2 /c 2 increases
without bounds. If we could accelerate an object up to
In our discussion of moving clocks in Chapter 1, we the speed of light, it would acquire an infinite mass.
saw that the length T ′ of the astronaut’s second in-
creased according to the formula Exercise 6
At what speed does the mass of an object double (i.e.,
T′ = T
(1-11) at what speed does m = 2 m 0?) (Answer: v = .866 c.)
1–v 2 /c 2
where T was the length of one of our seconds. For Exercise 7
slowly moving astronauts where v << c, we have Electrons emerging from the Stanford Linear Accelera-
T ′ ≈ T and the length of the astronaut’s seconds is tor have a mass 200,000 times greater than their rest
nearly the same as ours. But as the astronaut ap- mass. What is the speed of these electrons? (The
proaches the speed of light, the number 1–v 2 /c 2 answer is v = .9999999999875 c. Use the approxima-
becomes smaller and smaller, and the astronaut’s sec- tion formulas discussed in Chapter 1 to work this
onds become longer and longer. If the astronaut goes problem.)
at the speed of light, 1/ 1–v 2 /c2 becomes infinitely
Exercise 8
large, the astronaut’s seconds become infinitely long,
and time stops for the astronaut. A car is traveling at a speed of v = 68 miles per hour. (68
miles/hr = 100 ft/second = 10 –7 ft/nanosecond = 10 –7
c.) By what factor has its mass increased due to its
motion. (Answer: m/m o = 1.000000000000005.)
 6-11

Nature’s Speed Limit ZERO REST MASS PARTICLES

When the police try to enforce a 65 mile/hr speed limit, If you think about it for a while, you may worry that
they have a hard job. They have to send out patrol cars nature’s enforcement of its speed limit c is too effec-
to observe the traffic, and chase after speeders. Even tive. With the formula m = m o / 1–v 2 /c 2 , we ex-
with the most careful surveillance, many drivers get pect that nothing can reach the speed of light, because
away with speeding. it would have an infinite mass, which is impossible.
Nature is more clever in enforcing its speed limit c. By What is light? It travels at the speed of light. If light
having the mass of an object increase as the speed of the consists of a beam of particles, and these particles travel
object approaches c, it becomes harder and harder to at the speed c, then the formula m = m o / 1–v 2 /c 2
change the speed of the object. If you accelerated an suggests that these particles have an infinite mass,
object up to the speed of light, its mass would become which is impossible.
infinite, and it would be impossible to increase the
particle’s speed. Then perhaps light does not consist of particles, and is
therefore exempt from Einstein’s formula. Back in
Historically it was noted that massive objects were hard Newton’s time there was considerable debate over the
to get moving, but when you got them moving, they nature of light. Isaac Newton supported the idea that
were hard to stop. This tendency of a massive object to light consisted of beams of particles. Red light was
keep moving at constant velocity was given the name made up of red particles, green light of green particles,
inertia. That is why our recoil definition of mass, blue light of blue particles, etc. Christian Huygens, a
which directly measures how hard it is to get an object well known Dutch physicist of the time, proposed that
moving, measures what is called inertial mass. Nature light was made up of waves, and that the different
enforces its speed limit c by increasing a particle’s colors of light were simply waves with different wave-
inertia to infinity at c, making it impossible to acceler- lengths. Huygens developed the theory of wave mo-
ate the particle to higher speeds. Because of this tion in order to support his point of view. We will
scheme, no one speeds and no police are necessary. discuss Huygen’s theory later in the text.

In 1801, about 100 years after the time of Newton and

Huygens, Thomas Young performed an experiment
that settled the debate they started. With his so called
two slit experiment Young conclusively demonstrated
that light was a wave phenomena.
6-12 Mass

Another century later in 1905, the same year that he The number 0/0 is not a disaster, it is simply undefined.
published the special theory of relativity, Einstein also It can be 1 or 2.7, or 6 × 10 – 23 . It can be any number
published a paper that conclusively demonstrated that you want. (How many nothings fit into nothing? As
light consisted of beams of particles, particles that we many as you want.) In other words, if the rest mass m 0
now call photons. (Einstein received the Nobel Prize of a photon is zero, the Einstein mass formula says
in 1921 for his paper on the nature of light. At that time nothing about the photon’s mass m photon. Photons do
his special theory of relativity was still too controver- have mass, but the Einstein mass formula does not tell
sial to be awarded the prize.) us what it is. (Einstein presented a new formula for the
photon’s mass in his 1905 paper. He found that the
Thus by 1905 it was known that light was both a particle
photon’s mass was proportional to the frequency of the
and a wave. How this could happen, how to picture
light wave.)
something as both a particle and a wave was not
understood until the development of quantum mechan- We will study Einstein’s theory of photons in detail
ics in the period 1923 through 1925. later in the text. All we need to know now is that light
consists of particles called photons, these particles
Despite the fact that light has a wave nature, it is still
travel at the speed of light, and these particles have no
made up of beams of particles called photons, and these
rest mass. If you stop photons, which you do all the time
particles travel at precisely the speed c. If we apply
when light strikes your skin, no particles are left. There
Einstein’s mass formula to photons, we get for the
is no residue of stopped photons on your skin. All that
photon mass m photon
is left is the heat energy brought in by the light.
m0 m0 m A photon is an amazing particle in that it exists only
m photon = = = 0
1–v 2 /c 2 v = c 1–1 0 when moving at the speed of light. There is no lapse of
(15) time for photons; they cannot become old. (They
where m 0 is the rest mass of the photon. cannot spontaneously decay like muons, because their
half life would be infinite.) There are two different
At first sight it looks like we are in deep trouble with worlds for particles. Particles with rest mass cannot get
Equation (15). Division by zero usually leads to a up to the speed of light, while particles without rest
disaster called infinity. There is one exception to this mass travel only at the speed of light.
disaster. If the rest mass m 0 of the photon is zero, then
we get
m0
m photon = = 0 (16)
0 0
 6-13

NEUTRINOS Neutrinos are now detected by what one might call a

Another particle that may have no rest mass is the brute force technique. Aim enough neutrinos at a big
neutrino. According to current theory there should be enough detector and a few will be stopped and ob-
three different kinds of neutrinos, but for now we will served. The first time neutrinos were detected was in
not distinguish among them. an experiment by Clyde Cowan and Fred Reines,
performed in 1956, almost 30 years after Pauli had
In our discussion of the β decay process, we mentioned proposed the existence of the particle. Noting that
that when a radioactive nucleus decays by emitting an nuclear reactors are a prodigious source of neutrinos,
electron, a neutrino is also emitted. Most of the time the Cowan and Reines succeeded in detecting neutrinos by
energy given up by the nucleus is shared between the building a detector the size of a railroad tank car and
electron and the neutrino, thus the electrons carried out placing it next to the reactor at Savannah River,
only part of the energy. The very existence of the Georgia.
neutrino was predicted from the fact that some energy
appeared to be missing in β decay reactions and it was The largest neutrino detectors now in use were origi-
Pauli who suggested that this energy was carried out by nally built to detect the spontaneous decay of the proton
an undetected particle. (a process that has not yet been observed). They consist
of a swimming pool sized tank of water surrounded by
Neutrinos are difficult to detect. They can pass through arrays of photocells, all located in deep mines to shield
immense amounts of matter without being stopped or them from cosmic rays. If a proton decays, either
deflected. In comparison photons are readily absorbed spontaneously or because it was struck by a neutrino,
by matter. As any scuba diver knows, even in the a tiny flash of light is emitted in the subsequent particle
clearest ocean, a good fraction of the sunlight is ab- reaction. The flash of light is then detected by one of
sorbed by the time you get down to a depth of 50 or the photocells.
more feet. At that depth most of the red light has been
absorbed and objects have a grayish blue cast. In Solar Neutrinos
muddy water photons are absorbed much more rapidly,
Aside from nuclear reactors, another powerful source
and opaque objects like your skin stop photons in the
of neutrinos is the sun. Beta decay processes and
distance of a few atomic diameters.
neutrino emission are intimately associated with the
On the other hand, neutrinos can pass through the earth nuclear reactions that power the sun. As a result
with almost no chance of being stopped. As a writer neutrinos emerge from the small hot core at the center
discussing the 1987 supernova explosion phrased it, of the sun where the nuclear reactions are taking place.
the neutrinos from the supernova explosion swept The sun produces so many neutrinos that we can detect
through the earth, the earth being far more transparent them here on earth.
to the neutrinos than a thin sheet of the clearest glass to
light.
6-14 Mass

There is a good reason to look for these solar neutrinos. Neutrino Astronomy
The neutrinos created in the core of the sun pass directly An event on the night of February 23, 1987 changed the
through the outer layers of the sun and reach us eight role of neutrinos in modern science. On that night
minutes after they were created in a nuclear reaction. In neutrinos were detected from the supernova explosion
contrast, light from the hot bright core of the sun takes in the Magellanic cloud, a small neighboring galaxy.
the order of 14,000 years to diffuse its way out to the This was the first time neutrinos were detected from an
surface of the sun. If for some reason the nuclear astronomical source other than our sun. The informa-
reactions in the sun slowed down and the core cooled, tion we obtained from this observation represented
it would be about 14,000 years before the surface of the what one could call the birth of neutrino astronomy.
sun cooled. But the decrease in neutrinos could be
detected here on earth within 8 minutes. Looking at the A supernova is an exploding star, an event so powerful
solar neutrinos provides a way of looking at the future that, for a short period of time of about 10 seconds, the
of the sun 14,000 years from now. star radiates more power than all the rest of the visible
universe. And this energy is radiated in the form of
Solar neutrinos have been studied and counted since neutrinos.
the 1960s. Computer models of the nuclear reactions
taking place in the sun make explicit predictions about The supernova explosion occurs when the core of a
how many neutrinos should be emitted. The neutrino large star runs out of nuclear fuel and collapses. (This
detectors observe only about 1/3 to 1/2 that number. only happens to stars several times larger than our sun.)
There have been a number of experiments using vari- The gravitational energy released in the collapse is
ous kinds of detectors, and all the experiments show what provides the energy for the explosion. We know
this deficiency. that sometimes a neutron star is formed at the center of
the collapsed core, and computer simulations predict
If the deficiency is really an indication that the nuclear
reactions in the sun’s core have slowed, then we can
expect a cooling of the sun within 14,000 years, a
cooling that might have a significant impact on the
earth’s climate. On the other hand there may be some
part of the nuclear reactions in the sun that we do not
fully understand, with the result that the computer
predictions are in error. We are not sure yet which is
correct; the solar neutrino deficiency is one of the
current areas of active research.

Figure 8
1987 Supernova at age 3 1 2 years, photographed
by the Hubble telescope. The ring is gas blown
off by the explosion.
 6-15

that much of the energy released in the collapse is The exact time of the arrival of the light from the
carried out in a burst of neutrinos. The core material is supernova explosion is harder to pin down, but some
so dense that even the neutrinos have some difficulty fortunate coincidences occurred there too. The super-
getting out. They take about 10 seconds to diffuse out nova was first observed by a graduate student Ian
of the core, and as a result the neutrino pulse is about 10 Sheldon working at the Las Campanas Observatory in
seconds long. Chile. Ian was photographing the large Magellanic
cloud on the night of February 23, 1987, and noted that
The collapsing core also creates a shock wave that a plate that he had exposed that night had a bright stellar
spreads out through the outer layers of the star, reaching object that was not on the plate exposed the night
the surface in about three hours. When the shock wave before. The object was so bright it should be visible to
reaches the surface, the star suddenly brightens and we the naked eye. Ian went outside, looked up, and there
can see from the light that the star has exploded. it was.
The details about the core collapse, the neutrino burst Once the supernova had been spotted, there was an
and the shock wave are all from computer models of immediate search for more precise evidence of when
supernova explosions, models developed over the past the explosion had occurred. A study of the records of
25 years. Whenever you model a physical process, you the neutrino detectors turned up the ten second neutrino
like to test your model with the real process. Computer pulse at 7:36 AM on February 23. Three hours after that
models of supernova explosions are difficult to test Robert McNaught, an observer in Siding Spring, Aus-
because there are so few supernova explosions. The tralia, had exposed two plates of the large Magellanic
last explosion in our galaxy, close enough to study in clouds. When the plates were developed later, the
detail, occurred in 1604, shortly before the invention of supernova was visible. One hour before McNaught
the telescope. exposed his plates, Albert Jones, an amateur astrono-
mer in New Zealand happened to be observing at the
The supernova explosion on February 23, 1987 was not
precise spot where the supernova occurred and saw
only close enough to be studied, several fortunate
nothing unusual. Thus the light from the supernova
coincidences provided much detailed information. The
explosion arrived at some time between two and three
first coincidence was the fact that theoretical physicists
hours after the neutrino pulse.
had predicted in the 1960s that the proton might
spontaneously decay (with a half life of about 10 32 The fact that the photons from the supernova explosion
years.) To detect this weak spontaneous decay, several arrived two to three hours later than the neutrinos, is not
large detectors were constructed. As we mentioned, only a good test of the computer models of the super-
these large detectors were also capable of detecting nova explosion, it also provides an excellent check on
neutrinos. On February 23, at 7:36 AM universal time, the rest mass of the neutrino. The 1987 supernova
the detectors in the Kamokande lead mine in Japan, the occurred 160,000 light years away from the earth.
Morton Thekol salt mine near Cleveland, Ohio and at After the explosion, neutrinos and photons raced to-
Baksam in the Soviet Union all detected a 10 second ward the earth. The neutrinos had a 3 hour head start,
wide pulse of neutrinos. Since the Magellanic cloud and after traveling for 160,000 years, the neutrinos
and the supernova are visible only from the southern were still 2 hours, and perhaps 3 hours ahead. That is
hemisphere and all the neutrino detectors are in the as close a race as you can expect to find. From this we
northern hemisphere, all the detected neutrinos had to can conclude that neutrinos travel at, or very, very close
pass through the earth. The 10 second width of the to the speed of light. And therefore their mass must be
pulse verified earlier computer models about the diffu- precisely zero, or very close to it.
sion of neutrinos out of the collapsing core.
Chapter 7
Conservation of
Linear and Angular
Momentum

CHAPTER 7 CONSERVATION OF LINEAR & ANGULAR MOMENTUM

The truly basic laws of physics, like the principle of momentum, which does not change when the spring
relativity, not only have broad applications, but are is released and the carts recoil. We will then look at
often easy to describe. The principle of relativity a wider class of experiments, in which objects not
says that there is a quantity, namely your own only recoil, but collide at different angles. Again we
uniform motion, that you cannot detect. The hard will see that linear momentum does not change.
part is working out the implications of the simple
idea. We will also see that linear momentum is conserved
not just for familiar objects like billiard balls, but
In this chapter we discuss two more basic laws of also for objects as small as protons colliding in a
physics, laws that apply with no known exceptions to hydrogen bubble chamber.
objects as large as galaxies and as small as sub-
atomic particles. These are the laws of the conser- In the appendix to this chapter we will show how the
vation of linear momentum and the law of the recoil definition of mass, when combined with the
conservation of angular momentum. These are the law of conservation of linear momentum and the
first of several so-called conservation laws that we principle of relativity, leads to Einstein's relativistic
will encounter in our study of physics. A conserva- mass formula m = m0 / 1 – v2 /c2 .
tion law states that there is some quantity which does The second conservation law deals with angular
not change in a given set of experiments. momentum. The concept of angular momentum is a
We will introduce our first example of a conserva- bit more subtle than that of linear momentum. As a
tion law by going back to the results of the aircart result, in this chapter we will focus on developing an
recoil experiments that we used in the last chapter to intuitive feeling for the concept. A more formal
define mass. In analyzing these results, we will see mathematical treatment will be put off until a later
that there is a quantity, which we will call linear chapter where the formalism is needed.
7-2 Conservation of Linear and Angular Momentum

The law of conservation of angular momentum has CONSERVATION OF

many applications that range from the astronomical LINEAR MOMENTUM
scale to the subatomic scale of distance. The very In our discussion of the recoil definition of mass in
existence of planets in the solar system is a conse-
the last chapter, we looked at a number of experi-
quence of the conservation of angular momentum.
ments that were performed to determine how mass
The law also allows us to understand the behavior of behaves. One of the crucial observations was that,
atomic nuclei in a magnetic field, a behavior that is
at least with carts on an air track, the ratio of the
involved in the creation of the marvelous images
recoil speeds did not change as we changed the
seen in magnetic resonance imaging apparatus. On strength of the spring pushing the carts apart. If the
the very smallest scale of distance, angular momen-
big cart came out moving half as fast as the small one
tum turns out to be one of the basic intrinsic proper-
when we used a weak spring, then it still came out
ties of all elementary particles. half as fast when we used a strong spring. With the
strong spring both speeds were greater, but the ratio
was still the same.

We used this unchanging ratio in our definition of

mass. If, as shown in Figure (6-3) reproduced here,
cart A recoils at a speed vA and cart B at a speed vB,
then the mass ratio m A /m Bwas defined by Equation
(6-1) as
mA vB
m B = vA (6-1)

What we will do now is manipulate Equation (6-1)

until we end up with a quantity that does not change
when the spring is released.

thread spring

A B

frictionless Air Track

(a)

VA VB
A B

(b)

Figure 6-3
In Chapter 6 we defined the ratio of the mass of the
carts m A/mB to be equal to the inverse ratio of the recoil
speeds vB /vA .
 7-3

Multiplying Equation (6-1) through by m B and vA Comparing Equations (3) and (6) we see that the sum
gives of the linear momenta of the two carts was not
m A vA = m B vB (1) changed by the recoil. This sum was zero before the
Next, note that vA and vB are the magnitudes of the carts were released, and it is zero afterward.
recoil velocity vectors vA and vB. Since vA and vB We will call the sum of the linear momentum of the
point in opposite directions, we can write Equation two carts the total linear momentum ptot of the
(1) as a vector equation in the form system of two carts.
m A vA = –m B vB (2) p tot ≡ p A + p B = m AvA + m BvB
where the minus sign handles the fact that vA and vB (7)
definition of total linear momentum
are oppositely directed.
Then we can restate our observation that Equation
Now move the –m B vB to the left hand side to give (3) and (6) look the same by saying that the total
the result linear momentum ptot of the system was unchanged
by the recoil. Another way of phrasing it is to say
m AvA + m BvB = 0 after recoil (3) that in the recoil experiment, the total linear momen-
tum of the carts is conserved.
where vA and vB are the cart's velocity vectors after
the recoil. At this point, we do not have a new law of physics,
instead, we have merely reformulated our definition
We will now introduce a new interpretation. Let us of mass. But the result turns out to be far more
define the linear momentum of a particle as the general than we have seen so far. The general law
product of the particle's mass times its velocity. may be stated as follows. If we have a system of
Using the letter p to denote linear momentum, we particles, and there is no net external force acting
have on them, then the total linear momentum of the
system of particles is conserved.
definitionof
p ≡ mv linear momentum (4)
So far, we have not said much about forces and how
to recognize them. Thus we will, for now, limit our
Note that linear momentum p is a vector because it discussion of the conservation of linear momentum
is the product of a number, the mass m, times a to examples where it is fairly clear that there is no net
vector, the velocity v . external force or influence. In our recoil experi-
ment, gravity is pulling down on the carts, the air is
Looking back at Equation (3), we see that m A vA is
pushing up, and the two effects cancel. The air track
the linear momentum of cart A after the recoil, and
was explicitly designed so that there would be no net
m B vB is the linear momentum of cart B after the
force on the cart.
recoil. Equation (3) tells us that the sum of these two
linear momenta is zero. In our recoil experiment, our system consists of the
two carts and the spring. When the thread is burned,
Before the string was cut and the carts released, both
the spring exerts a force on both carts, but the spring
carts were sitting at rest. Before the release, we have
is part of the system. The spring forces are internal,
vA = 0 , vB = 0 before release (5) not external forces, and therefore cannot change the
Thus the sum of the linear momenta before the linear momentum of the system. If the linear mo-
release was mentum was zero before the thread was burned, it
must still be zero afterward.
m AvA + m BvB = 0 before recoil (6)
7-4 Conservation of Linear and Angular Momentum

COLLISION EXPERIMENTS
A more common example of where external forces
can be ignored and linear momentum is conserved is
during the collision of two objects like billiard balls.
While two objects are colliding, the forces between
the objects, the internal forces, are usually much
greater than any outside external forces. As a result,
just before, during, and just after the collision, exter-
ceiling
nal forces can be neglected and linear momentum is
conserved.

string In an experiment, that is easily carried out in the

introductory physics lab, two steel balls are sus-
strobe 1 pended by strings from the ceiling as shown in
P Figure (1). One of the two balls is pulled back and
2 1 1
grid
released. It strikes the ball at rest and the two balls
bounce off as seen in the strobe photograph of the
m motion, Figure (2). The strobe photograph is ana-
irr
or camera lyzed in Figure (3) and the resulting momentum
vectors are plotted in Figure (4).
floor
What we see in the strobe photograph is ball 2 at rest
Figure 1 and ball 1 coming in, attaining a velocity v1i just
Experimental setup to study the conservation of before the collision. After the collision, balls 1 and
linear momentum during the collision of two balls. 2 bounce off in different directions, with velocity
vectors v1f and v2f respectively.

Figure 2
Strobe photograph taken using the setup of Figure (1). The data for the experiment are
m1 = 70.3 gm (the ball initially released), m2 = 240 gm (ball initially at rest), ∆ t = 1/ 10 sec (period
between flashes). Spacing between grid lines = 1 cm. (Photograph from a student lab notebook.)
 7-5

In Figure (4) we have plotted the momentum p1i of momentum being

pf ≡ p1f + p2f = carried out of (9)
ball 1 just before the collision the collision
momentumof From Figure (4) we see that pf is equal to p1i the
p 1i = mv1i = ball 1 before (8) momentum brought into the collision by ball 1.
the collision
Since the same amount of momentum came out of
and also plotted the momenta p1f and p2f of balls the collision as was carried in, the total linear mo-
1 and 2 after the collision. The vector sum of these mentum did not change. The total linear momentum
two momenta is the total momentum pf being of the system of the two balls was conserved during
carried out by the two balls the collision.

P2f

P1i
m1 = 70.3 gm p = 6.43 x 10 3 gm cm
1i sec
m2 = 24.0 gm
P1f v1i = 91.5 cm/sec p = 6.05 x 10 3 gm cm
1f sec
v1f = 86 cm/sec
v2f = 37 cm/sec p = 0.89 x 10 3 gm cm
2f sec
Figure 3
Analysis of Figure (2). Ball 1 enters with a momentum p1 i and collides with Ball 2 which
is initially at rest. After the collision, Balls 1 and 2 emerge with momenta p1 f and p2 f
respectively. (Each large square on this graph paper represents a distance of 10 cm.)

P1i

P1f

P2f

0 2 x10 3 4 x10 3 6 x10 3

momentum scale gm cm/sec

Figure 4
Here we see that the momentum p1 i brought in by Ball 1 is
equal to the momentum p1 f + p2 f carried out after the collision.
7-6 Conservation of Linear and Angular Momentum

Exercise 1
Figures (5 a and b) show the collision between two balls of equal mass m 1 = m 2 = 73 grams . Again
∆t = 1/10 sec . Using one of these figures, construct a graph similar to Figure (4), and compare the momentum
brought in by Ball 1 with the momentum carried out by the two balls after the collision.

Figures 5 a, b
Strobe photographs of the collision of two equal mass balls.
 7-7

Subatomic Collisions The following examples and exercises are chosen to

In the study of subatomic particles, you cannot show some of the more practical applications of the
photograph or image the particles themselves, the conservation of linear momentum.
best you can do is study the tracks left behind in a
particle detector. A common particle detector, de- Example 1 Rifle and Bullet
veloped by Don Glaser in the early 1950s, is the
A 2-kilogram rifle fires a 10-gram bullet at a speed
bubble chamber. We will discuss the bubble cham- of 400 meter/sec. What is the recoil velocity of the
ber in more detail in later chapters. However the
gun? In this case, the rifle and bullet are initially at
basic idea is that the bubble chamber is filled with
rest and have zero total linear momentum. Just after
liquid hydrogen, and that a charged particle moving the bullet leaves the gun, before any external forces
through the liquid hydrogen leaves a track that can
have had time to act on the system, the total momen-
be made visible as a string of bubbles.
tum of the system (gun plus bullet) is still zero. We
Bubble chambers are used primarily to study the get
collisions between subatomic particles. In Figure pgun + pbullet = 0
(6) we have a bubble chamber photograph in which
an incoming proton from a particle accelerator moves m gvg = – m bvb
through the liquid hydrogen until it strikes a hydro- where the minus sign indicates that vg is in the
gen nucleus, namely another proton. The two pro- opposite direction to the motion of the bullet. Solv-
tons emerge from the collision, coming out at right ing for the magnitude vg of the recoil velocity, we get
angles as shown. m 10 gm
vg = m b vb = × 400 meters
sec
After we discuss the law of conservation of energy, g 2000 gm
we will show that if two identical particles collide,
vg = 2 meters
sec
one of them being initially at rest, then if both energy
and linear momentum are conserved during the Thus, we see that the initial recoil velocity of the gun
collision the particles must emerge at right angles. is 2 meters/sec.
Thus we can use the right angle between the emerg-
ing proton tracks in Figure (6) as experimental
evidence that linear momentum is conserved even
among the interactions of subatomic particles.

proton or hydrogen nucleus

initially at rest

Figure 6
Collision between two protons. When a charged elementary particle passes through the liquid
hydrogen in a bubble chamber, it leaves a trail of bubbles that can be photographed. Here we see a
proton coming into the picture from the upper left, and striking the nucleus of one of the hydrogen
atoms in the liquid hydrogen. The hydrogen nucleus is itself a proton, and after the collision the
two protons emerge as shown in the sketch.
7-8 Conservation of Linear and Angular Momentum

In this example we applied the law of conservation Exercise 3 Frictionless Ice

of linear momentum over such a short time that Suppose you are sitting in the middle of a completely
outside forces did not have time to act on the system. frictionless surface, such as an idealized pond of ice.
The conservation of linear momentum applies over Propose a method of getting out of such a predica-
longer times, but we must enlarge our concept of the ment. (Problem from J. Orear, Fundamental Physics,
system, as seen in Example (2), Wiley, New York, 1961.)

Exercise 4 Bullet and Block

Example 2
A 10-gram bullet traveling 300 meters/sec strikes and
A 78 kilogram hunter standing on nearly frictionless
lodges in a 3-kilogram block of wood initially at rest on
ice fires the gun of the preceding example. What is a pond of ice. What is the final velocity of the block
the recoil velocity of the hunter? and bullet after the collision?
Our system now consists of the bullet, gun and
Exercise 5 Two Skaters Throwing Ball
hunter. Initially the total linear momentum of the
Two skaters, each of mass 60 kilogram, are standing
system is zero. After the bullet is fired, and after the
a slight distance apart on nearly frictionless ice.
gun is firmly lodged against the shoulder of the Initially at rest, they throw a 1-kilogram ball back and
hunter, the gun and hunter together recoil at a veloc- forth between them; each time the ball travels at a
ity vh . Applying the law of conservation of linear speed of 10 meters/sec over the ice.
momentum, we have (remembering that vg now
equals vh ) (a) What is the recoil velocity of the first skater
immediately after he throws the ball for the
phunter + pgun + pbullet = 0 first time?

m hvh + m gvh = – m bvb (b) After the second skater catches the ball
for the first time, what is his recoil velocity?
m 10 gm
vh = m +bm vb = × 400 meters
sec
(c) After the second skater has thrown the ball
h g 78 + 2 kg back for the first time, what is his recoil
velocity?
10 gm
vh = × 400 meters 1 meter
sec = 20 sec (d) After the ball has made 10 complete round
80,000 gm
trips and the first skater is holding the ball,
cm
vh = 5 sec what is the velocity of each skater? What
is the total momentum of the system of the
two skaters and the ball?
Exercise 2
a) Starting from the preceding two examples, further Exercise 6 Rocket
enlarge the system. Assume that the hunter is stand-
An 11-ton rocket consists of 10 tons of fuel. If the fuel
ing firmly on the earth when the gun is fired. Taking
is discharged as exhaust gasses that travel at an
the point of view that the earth is initially at rest,
average speed of 1 mile/sec (relative to the earth),
calculate the recoil velocity of the gun, hunter, and
how fast will the rocket be traveling when the fuel is
earth. (mearth = 6 × 1027 gm)
used up? Neglect gravity and air resistance. (Hint:
b) After the bullet strikes the ground, what is the Consider the total momentum of all the exhaust gas.)
velocity of the earth, assuming it was at rest before
the gun was fired?
 7-9

CONSERVATION OF Even those of us who are not skilled figure skaters

ANGULAR MOMENTUM can repeat the skaters experience of a spin using a
rotating platform and two iron dumbbells. When
Anyone who has watched figure skating in the
done as a classroom demonstration, this is some-
winter Olympics has seen an example of the conser-
times known as the “three dumbbell experiment”.
vation of angular momentum. When a figure skater
The instructor stands on the rotating platform and
like the one shown in Figure (7) starts her spin, her
holds the dumbbells out as shown in Figure (8a). A
arms are outstretched and she is turning slowly. As
student helps in the demonstration by starting the
she brings her arms in, she turns faster and faster
instructor rotating slowly. The instructor then pulls
until the maneuver is completed. She starts the spin
in his arms and rotates even faster than a figure
with a certain amount of angular momentum, and
skater because of the mass in the dumbbells (Figure
that amount does not change, is conserved, through-
8b). However unless the instructor is skilled at this
out the spin. To understand the concept of angular
demonstration, he is likely to make a far less grace-
momentum, we have to see why the skater had the
ful exit from the spin than do the Olympic figure
same angular momentum when rotating slowly with
skaters .
her arms outstretched and rotating rapidly with her
arms pulled close to her body.

R1 R2

(a) (b)

Figure 8
The "three dumbbell" experiment. The instructor,
standing on a platform that is free to rotate, holds two
dumbbells out at arm length as shown in (a). With a
slight push from a student, the instructor starts to
rotate slowly. The instructor then pulls his arms in,
and the rotation increases significantly (b).

Figure 7
Figure skater doing a spin. As the skater
pulls her arms in, she turns faster and
faster. This is an example of the
conservation of angular momentum.
7-10 Conservation of Linear and Angular Momentum

In a more controlled and idealized experiment, we Since the mass m of the ball did not change, and
can set a ball swinging in a circle at the end of a string because v1 r 1 = v2 r 2 , we see that
as shown in Figure (9). Let the other end of the string
pass down through the small end of a plastic funnel 1 = 2 (14)
mounted on a board as shown in Figure (9a). If we and in this example angular momentum did not
pull down on the string to reduce the radius r of the change. The ball's angular momentum was con-
circle around which the ball is traveling, the speed v served while we pulled on the string and the ball sped
of the ball increases. An analysis of this motion up. It was also conserved when the figure skater
shows a simple result—the product of the radius of pulled in her arms during the spin, and the instructor
the circle times the speed of the ball remains con- pulled in on the dumbbells. The only difference in
stant. If the ball is initially moving in a circle of the three examples is that we have a more complex
radius r1 and a speed v1 as shown in Figure (9b), and formula for angular momentum for the figure skater
we reduce the radius to a length r2 as shown in Figure and instructor.
(9c), the new speed v2 is given by the equation
v1 r 1 = v2 r 2 (10) Exercise 7
What are the dimensions of angular momentum when
Since r2 is smaller than r1, v2 must be bigger than v1 mass is measured in grams, length in centimeters,
to keep the product constant. and time in seconds?

The angular momentum of a ball of mass m traveling Exercise 8

at a speed v in a circle of radius r is defined to be the Figure (9d) is a strobe photograph from a student lab
product mvr. Using the letter to represent angular notebook. The string was suddenly pulled down,
momentum, we have shortening the radius of the ball's circular orbit. Using
this photograph, show that the angular momentum of
angular momentumof a the ball did not change.
≡ mvr mass m traveling at a speed (11)
v in a circle of radius r
Exercise 9
In Figure (9a), the ball has an angular momentum Neglecting the mass of the spokes, what is the
angular momentum of a bicycle wheel of mass m and
1 = mv1 r1 (12) radius r, spinning with a period T? (T is the length of
after the string is shortened, the angular momentum time the wheel takes to go around once.)

2 is
2 = mv2 r2 (13)
 7-11

a) Side view b) Top view c) String pulled in

r m v
1
funnel m v2
string
r1 r2
board string

pull

d)

Figure 9
A more controlled demonstration of the conservation of angular momentum. One end of a string
is tied to a ball of mass m, and the other is fed down through a plastic funnel mounted on the end
of a board shown in (a). The ball is then swung in a circle of radius r1 , and speed v1 as shown in
(b). Then pull down on the free end of the string, to reduce the radius of the circle to r2. It takes a
fairly strong tug, but the speed of the ball increases to v2 as shown in (c). From the experimental
results shown in (d), you can check that r1 v1 = r2 v2 . Since the angular momentum of the ball is
proportional to rv, angular momentum was conserved in this experiment. (Photo from lab of
G. Sheldon.)
7-12 Conservation of Linear and Angular Momentum

A MORE GENERAL DEFINITION Applying Equation (15) to the situation shown in

OF ANGULAR MOMENTUM Figure (10a), before the ball is hooked, we see that
as the ball heads toward the hook, its initial angular
The concept of angular momentum applies to more
momentum i is
general situations than mass traveling in a circle.
For a more general definition of angular momentum, (16)
i = pi r⊥ = mv r⊥
consider the situation shown in Figure (10). In
Figure (10a) a ball with linear momentum p = mv is Since this is the same as the angular momentum after
traveling along a path that will take it a distance r⊥ the ball is hooked and traveling in a circle, the
from some point labeled O. To make the situation angular momentum is unchanged, is conserved,
more realistic, imagine that there is a light rod during the process of being captured by the hook.
pivoted at point O with a hook at the other end. The
p = mv
length of the rod is r⊥ , so that the hook will just catch
the ball as the ball passes by (Figure 10b). path of b
all
a)
Once the ball has been hooked, it will travel in a
ball heading
circle as shown in Figure (10c). If the rod is for hook r = pe
rpendicu
perpendicular to the path of the ball when the ball is distance lar
from path
hooked (as shown in Figure 10a) then there will be of ball to
point O
O
no disruption in the speed of the ball and the ball will
move around the circle at the same speed v.
p = mv
Once the ball is traveling in a circle we know that its
angular momentum about the pivot O is given by b)
ball catches
Equation (11) as = mvr⊥. In our generalization of on hook r
the definition of angular momentum, the ball has this
same amount of angular momentum before it was
hooked as it did after. O

The more general definition is as follows. Consider

the path of the ball shown in Figure (10a) and let us
use the name lever arm for the distance of closest
p = mv

approach from the path to the axis at point O. This

lever arm is the perpendicular distance to the path,
the distance we have labeled r⊥. Then as our new c) r
ball swinging in circle,
definition of angular momentum, we say that the with angular momentum
O
magnitude of the ball's angular momentum is = mvr
equal to the product of the magnitude of the linear
momentum p = mv times the length of the lever arm
r⊥ Figure 10
We say that the ball initially has an angular
= pr⊥ (15) momentum mvr⊥ , in (a), that remains unchanged
when the ball is caught by the hook and travels in
c circle, in (c).
 7-13

When you have a conserved quantity like angular Exercise 10

momentum, it takes on a reality that goes beyond the If you are standing directly over the axis of the
formulas that define it. With our generalized defini- platform and a ball is thrown directly toward you, as
tion of angular momentum, we find that angular shown in Figure (12), do you start to rotate after
momentum can be passed from one object to an- catching the ball? Try the experiment yourself and
other. For example suppose a student is standing see if the prediction is correct.
motionless on the rotating platform as in Figure
(11a), and the instructor tosses a softball off to the student
side of the student as shown. The softball has a lever
arm r⊥ and therefore an angular momentum mv r⊥
p
about the axis of the rotating platform. If the student
reaches out and grabs the ball, she acquires the mass
angular momentum of the ball and starts rotating as
shown in Figure (11b). If she brings the ball closer
in toward her body, she will rotate faster because the rotating platform
Figure 12
ball has a shorter lever arm.
How much angular momentum
does the student catch in this case?
p

r student

rotating platform

(a) Ball thrown to student

(b) Student, gaining angular momentum

|p| r = mv r from ball, starts to rotate
Figure 11
Student catching angular momentum
from a ball thrown off to the side.
7-14 Conservation of Linear and Angular Momentum

ANGULAR MOMENTUM Exercise 11

AS A VECTOR What would be the direction of positive rotation if we
Our definition of angular momentum is clearly not used a left hand convention. Draw a sketch and
yet complete. Even when we are standing on a explain.
rotating platform so that we can freely rotate only
about the axis of the platform, there are still two The next generalization of our definition of angular
different directions we can rotate—clockwise and momentum is best illustrated by the use of the
counter clockwise. The definition of angular mo- rotating bicycle wheel mounted on a handle as
mentum must somehow account for these two direc- shown in Figure (14). To make an effective demon-
tions of rotation. stration, the tire of the bicycle wheel has been
replaced by wire wrapped along the rim of the wheel
The study of rotations can be complex, particularly to give the wheel added mass. If we spin the wheel
if you allow rotations in three dimensions, about any all the mass m on the rim is moving at the same speed
of the three coordinate axes x, y or z. The rotating v and has the same lever arm r about the axis of the
platform used in Figure (8) and (11) greatly simpli- wheel. Thus the angular momentum of the rotating
fies the situation by restricting our motion to rotation wheel is = mv r .
about one axis, the axis of the platform which is
conventionally called the z axis as shown in Figure The next step will at first seem arbitrary, and perhaps
(13). downright silly. What we are going to do now is to
define the angular momentum of the wheel as a
One way to distinguish positive and negative rota- vector, of length = mv r , pointing along the axis
tion is by using a right-hand rule. The rule is to point of the wheel. Which way it points is defined by a
the thumb of your right hand along the positive z right-hand rule. As shown in Figure (14), curl the
axis, and then say that the direction of positive fingers of your right hand in the direction that the
rotation is the direction that the fingers of your right wheel is rotating and the angular momentum vector
hand curl, as seen in Figure (13). Looking down points along the axis in the direction of your
(with the z axis pointing up toward us), we find that thumb.
positive rotation is counter clockwise and negative
rotation is clockwise. z This method of turning angular momentum into a
vector seems doubly arbitrary. First of all, the
axis of rotation

angular momentum vector points perpendicular

positive
y
rotation = mv r
rotating
x platform

Figure 13
Definitions of the z axis and of positive and negative Figure 14
rotation. For convenience we will call the axis about A further generalization of the concept of angular
which our platform rotates the "z axis". To distinguish momentum is to say that it is a vector . The magnitude
the two kinds of rotation, we will use the right hand is just our old definition mvr⊥. We now say that the
rule. Curl the finger of your right hand in the direction angular momentum vector points in the direction of the
the platform is rotating. If your thumb points up, we axis of rotation, in the direction given by the right hand
say the rotation is positive. If your thumb points down, rule. (Curl the fingers of your right hand in the direction
the rotation is negative. the wheel is rotating, and your thumb points in the
direction of .)
 7-15

to the plane in which the motion of the wheel is The special thing that happens when you stand on a
occurring, and then one arbitrarily selects a right- freely rotating platform is that you cannot change
handed instead of a left-hand convention to decide your own z component of angular momentum. If no
which way along the axis the vector should point. one off the platform passes or tosses in some angular
It is hard to believe that such arbitrary choices could momentum, your z component of angular momen-
have any relationship to physical reality. tum is conserved and there is no way you can change
But it works, as we can easily demonstrate using the it.
bicycle wheel and the rotating platform. In the first Have the student turn the bicycle wheel upside down
demonstration, have a student stand at rest on the as shown in Figure (15b). Now of the bicycle
rotating platform and let the instructor spin the wheel is pointing down so that the bicycle wheel
bicycle wheel and orient it so that the bicycle wheel's now has a negative z component of angular momen-
angular momentum vector is pointing up as tum
shown in Figure (15a). The instructor then hands the
bicycle wheel, and its angular momentum, to the z bicycle wheel = – mvr (16)
student as shown. Since the total z of the student and bicycle wheel
On the rotating platform motion is restricted to must be conserved, the student must gain a positive
rotation about the z axis (axis of the platform), thus z component of angular momentum
the only component of angular momentum that is of = + 2mvr (17)
z student
interest is the z component. In Figure (15a) the
bicycle has a positive z component of angular mo- so that the sum remains + mvr.
mentum z = mvr , and the student has none. Thus
What happens when the student turns the bicycle
the total angular momentum of the wheel and stu-
wheel over is that she starts rotating counter clock-
dent is
wise; she gains a positive z component of angular
z total = z bicycle wheel + z student momentum. If at any time she turns the wheel back
= +mvr + 0 (15) up, she will stop rotating.
= +mvr

Figure 15a Figure 15b Figure 15 Movie

A student, at rest on the rotatable The student turns the bicycle wheel over. The student turns the
platform, is handed a rotating bicycle If angular momentum is conserved, the bicycle wheel over.
wheel whose angular momentum student must gain an angular
vector is up as shown. momentum 2 directed up as shown.
7-16 Conservation of Linear and Angular Momentum

z If the instructor hands the student the rotating wheel

oriented horizontally as shown in Figure (16a), and
the student is initially at rest, then neither the student
or the bicycle wheel have a z component of angular
momentum. The total z component is zero and must
remain that way no matter what the student does. In
the following exercise, you are to predict what will
happen if she turns the wheel up or turns it down.

Exercise 12
Explain what will happen if the student orients the
Figure 16a
bicycle wheel up as shown in Figure (16b). What
The student at rest is handed a bicycle wheel pointed happens when she turns it down as shown in Figure
sideways. In this orientation the wheel has no z (16c).
component of angular momentum. Once the student
has the wheel, the z component of the angular Exercise 13
momentum of the plus wheel is conserved. The student at rest on the rotating platform is handed
a bicycle wheel at rest. She spins the bicycle wheel
and orients it so that the bicycle wheel's angular
momentum vector points up. Explain carefully what
happens to the student.

If you watch, or better yet try these angular momen-

tum demonstrations yourself, you begin to believe
that there is really a quantity called angular momen-
tum that you can pass around and manipulate. Now
our emphasis is on gaining an intuitive feeling for
the concept, later we will come back to the topic
with more mathematical machinery. But it is impor-
Figure 16b
tant to already have an intuitive grasp of the concept
What happens to the student
if she turns the wheel up? or the mathematical machinery will not make sense.

Figure 16c
What happens to the student
if she turns the wheel down?

Figure 17
Detail of Hubble photograph of the Eagle
nebula. Each nub is a star surrounded by
its own gas cloud. (See page 18 for a more
complete photograph of the nebula.)
 7-17

Formation of Planets and the hotter the resulting ball. If the ball is of the
Applications of the law of conservation of angular order of one tenth the mass of the sun or larger, the
momentum are not confined to classroom demon- ball will be hot enough to ignite nuclear reactions
strations. We will end this chapter with a discussion and a star is born.
of two astronomical applications. One deals with
the formation of planets, and the other their motion The lumpiness in the gas clouds is probably related
about the sun. to a large scale turbulence in the flow of the gas in the
cloud. Such lumpines is usually associated with
Stars are formed in the large clouds of gas that rotational motion (vorticies), thus as a lump of gas
stretch throughout the galaxy. An example of such breaks away from the rest of the cloud, it is likely to
a gas cloud is the Eagle Nebula shown in Figure (17). have some rotation. While the rotational velocity
A particularly active area of star formation is in the may be small for a two light year diameter lump, as
nebula in the constellation of Orion shown in Figure the lump contracts, the speed increasas due to the
(18). The Orion constellation rises after sunset in conservation of angular momentum. If there is enough
early winter, and the nebula is located in the middle angular momentum in the gas cloud, a point is
of the sword dangling from the three bright stars of reached where the rotation inhibits further collapse
Orion's belt. Using binoculars, one can see the of the cloud.
nebula as a bright patch of gas. The gas is illumi-
nated by the newly formed stars inside. The only way the cloud can continue to collapse is to
leave some mass and most of the angular momentum
A star forms when a lump of gas in the cloud begins outside. Computer models indicate that a rotating
to collapse due to the gravitational attraction be- disk of gas forms outside the newly born star, a disk
tween the gas particles. Judging from the spacing that contains most of the original angular momen-
between stars in the neighborhood of the sun, the sun tum. After a while the gas in the disk condenses into
was formed from the collapse of a region of gas planets that orbit the star. As a result of being formed
about two light years in radius. As the cloud col- from a rotating disk of gas, we expect most planets
lapses the gravitational attraction between the par- to lie in a plane, and go around the star in the same
ticles becomes stronger and stronger. The particles direction, which the sun and planets do. If the
rush toward each other at a faster and faster rate, and collapsing gas had no angular momentum, the disk
finally collapse into a hot ball. The more mass in the would not form and there would be no planets.
collapse, the more gravitational energy released,

Figure 18 Figure 18a

The nebula in the constellation of Orion. This is a Hubble photograph of forming stars in the Orion
particularly active area for the formation of new stars. nebula. They are still surrounded by clouds of gas.
7-18 Conservation of Linear and Angular Momentum

from a ring of dust 2 light-years in radius, what must

Exercise 14
have been the initial rotational speed of these par-
Most of the angular momentum of the solar system is ticles? (The distance from the sun to Jupiter is 43
light-minutes or 2.6 × 10 light-seconds. Jupiter
taken up by the distant massive planets Jupiter, 3

travels at an orbital speed of 1.3 × 10 cm/ sec for

Saturn and Uranus. If Jupiter were originally formed 6

a period of nearly 12 years. 1 year ≈ π × 10 sec )

7

Figure 17b
Hubble photograph of the Eagle nebula. The nubs at the top of the tallest column are
young stars with their own gas clouds. Extremely bright light from a star in the
background is pushing away all gas that is not gravitationally attached to a star.
Chapter 8
Newtonian Mechanics

CHAPTER 8 NEWTONIAN MECHANICS

In Chapters 4 and 5 we saw how to use calculus and the

computer in order to predict the motion of a projectile.
We saw that if we knew the initial position and velocity
of an object, and had a formula for its acceleration
vector, then we could predict its position far into the
future.

To go beyond a discussion of projectile motion, to

develop a general scheme for predicting motion, two
new concepts are needed. One is mass, discussed in
chapter 6, and the other is force, to be introduced now.
We will see that once we know the forces acting on an
object, we can obtain a formula for the object’s accel-
eration and then use the techniques of Chapters 4 and
5 to predict motion. This scheme was developed in the
late 1600s by Isaac Newton and is known as Newtonian
Mechanics.
8-2 Newtonian Mechanics

FORCE Only when Galileo focused on a problem without

The concept of a force—a push or a pull—is not as much friction – projectile motion – did the important
strange or unfamiliar as the acceleration vector we have role of the acceleration vector become apparent. Later,
been discussing. When you push on an object you are Newton compared the motion of a projectile (the apple
exerting a force on that object. The harder you push, the that supposedly fell on his head) with the motion of the
stronger the force. And the direction you push is the planets and the moon, giving him more examples of
direction of the force. From this we see that force is a motion without friction. These examples led Newton
quantity that has a magnitude and a direction. As a to the discovery that force is directly related to accelera-
result, it is reasonable to assume that a force is de- tion, not velocity.
scribed mathematically by a vector, which we will In our discussion of projectile motion, and projectile
usually designate by the letter F. motion with air resistance, we have begun to see the
It is often easy to see when forces are acting on an relation between force and acceleration. While a
object. What is more subtle is the relationship between projectile is in flight, and we can neglect air resistance,
force and the resulting acceleration it produces. If I the projectile’s acceleration is straight down, in the
push on a big tree, nothing happens. I can push as hard direction of the earth as shown in Figure (1). As we
as I want and the tree does not move. (No bulldozers stand on the earth, we are being pulled down by gravity.
allowed.) But if I push on a chair, the chair may move. While the projectile is in flight, it is also being pulled
The chair moves if I push sideways but not if I push down by gravity. It is a reasonable guess that the
straight down. projectile’s downward acceleration vector g is caused
by the gravitational force of the earth.
The ancient Greeks, in particular, Aristotle, thought
that there was a direct relationship between force and When we considered the motion of a particle at con-
velocity. He thought that the harder you pushed on an stant speed in a circle as shown in Figure (2), we saw
object, the faster it went. There is some truth in this if that the particle’s acceleration vector pointed toward
you are talking about pushing a stone along the ground the center of the circle. A simple physical example of
or pulling a boat through water. But these examples, this circular motion was demonstrated when we tied a
which were familiar problems in ancient time, turn out golf ball to a string and swing it over our head.
to be complex situations, involving friction and viscous
forces. golf ball
0 1
2 r ing
str
a0 a1 a

a2 3

a3

Figure 1
The earth's gravitational force produces a uniform
downward gravitational acceleration. (Figure 3-27) Figure 2
The acceleration of the ball is in the same direction as
the force exerted by the string. (Figure 3-28)
 8-3

While swinging the golf ball, it was the string pulling THE ROLE OF MASS
on the ball that kept the ball moving in a circle. (Let go Our three examples, projectile motion, motion in a
of the string and the ball goes flying off.) The string is circle, and projectile motion with air resistance, all
capable of pulling only along the length of the string, demonstrate that a force produces an acceleration in the
which in this case is toward the center of the circle. direction of the force. The next question is – how much
Thus the force exerted by the string is in the direction acceleration? Clearly not all forces have the same
of the golf ball’s acceleration vector. This makes our effect. If I shove a child’s toy wagon, the wagon might
second example in which the particle’s acceleration accelerate rapidly and go flying off. The same shove
vector points in the same direction as the force exerted applied to a Buick automobile will not do very much.
on it.
There is clearly a difference between a toy wagon and
The example of projectile motion with air resistance, a Buick. The Buick has much more mass than the
shown in Figure (3), presented a more complex situa- wagon, and is much less responsive to my shove.
tion. In our study of the motion of a Styrofoam
projectile, we had two forces acting on the ball. There In our recoil definition of mass discussed in Chapter 6
was the downward force of gravity, and also the force and illustrated in Figure (4), we defined the ratio of two
exerted by the wind we would feel if we were riding masses as the inverse ratio of their recoil speeds
along with the ball. We saw that gravity and the wind
each produced an acceleration vector, and that the m 1 = v2
ball’s actual acceleration was the vector sum of the two m2 v1
individual accelerations. This is an important clue as to The intuitive idea is that the more massive the object,
how we should handle situations where more than one the slower it recoils. The more mass, the less respon-
force is acting on an object. sive it is to the shove that pushed the carts apart.

Think about the spring that pushes the cart apart in our
3
recoil experiment. Once we burn the thread holding the
carts together, the spring pushes out on both carts,
causing them to accelerate outward. If the spring is
" pushing equally hard on both carts (later we will see
a3 ind
"w that it must), then we see that the resulting acceleration
g and final velocities are inversely proportional to the
mass of the cart. If m1 is twice as massive as m2, it gets
a air only half as much acceleration from the same spring
force. Our recoil definition and experiments on mass
suggests that the effectiveness of a force in producing
an acceleration is inversely proportional to the object’s
v3

mass. For a given force, if you double the mass, you get
a 3 = g + a air only half the acceleration. That is the simplest relation-
ship between force and mass that is consistent with our
general experience, and it turns out to be the correct
Figure 3 one.
Gravity and the wind each produce an VA VB
acceleration, g and aair respectively. A B
The net acceleration of the ball is the
vector sum of the two accelerations.
Figure 4
Definition of mass. When two carts recoil from
rest, the more massive cart recoils more slowly.
8-4 Newtonian Mechanics

NEWTON’S SECOND LAW At this point Equation (1) or (1a) serves more as a
We have seen that a force F acting on a mass m, definition of force than a basic scientific result. We can,
produces an acceleration a that 1) is in the direction for example, see from Equation (1a) that force has the
of F , and 2) has a magnitude inversely proportional dimensions of mass times acceleration. In the MKS
to m. The simplest equation consistent with these system of units this turns out to be kg(m/sec2), a
observations is collection of units called the newton. Thus we can say
that we push on an object with a force of so many
F newtons. In the CGS system, the dimensions of force
a = m (1)
are gm(cm/sec2), a set of units called a dyne. A dyne
Equation (1) turns out to be the correct relationship, and turns out to be a very small unit of force, of the order of
is known as Newton’s Second Law of Mechanics. the force exerted by a fly doing push-ups. The newton
(The First Law is a statement of the special case that, is a much more convenient unit. The real confusion is
if there are no forces, there is no acceleration. That was in the English system of units where force is measured
not obvious in the late 1600s, and was therefore stated in pounds, and the unit of mass is a slug. We will
as a separate law.) A more familiar form of Newton’s carefully avoid doing Newton’s law calculations in
second law, seen in all introductory physics texts is English units so that the student does not have to worry
about pounds and slugs.

F = ma (1a) At a more fundamental level, we can use Equation (1)

to detect the existence of a force by the acceleration it
produces. In projectile motion, how do we know that
If there is any equation that is essentially an icon for the
introductory physics course, Equation (1a) is it. there is a gravitational force Fg acting on the projectile?
Because of the gravitational acceleration. The accel-
eration a due to gravity is equal to g (9.8 m/ sec2
directed downward), thus we can say that the gravita-
tional force Fg that produces this acceleration is

gravitational force
Fg = mg (2)
on a mass m

where m is the mass of the projectile.

Fg Fg
m1 m2

Figure 5
The gravitational force between small masses
is proportional to the product of the masses,
and inversely proportional to the square of the
separation between them.
 8-5

NEWTON’S LAW OF GRAVITY Exercise 1

Newton went beyond using the second law to define Combine Newton’s second law F = ma with the law of
force; he also discovered a basic law for the gravita- gravity F g = Gm1m2 r2 and show that the dimen-
tional force between objects. With Newton’s law of sions for G in Equation (4) are correct.
gravity combined with Newton’s second law, we can
make detailed predictions about how projectiles, satel-
lites, planets, and solar systems behave. This combina- Big Objects
tion, where one has an explicit formula for gravitational In our statement of Newton’s law of gravity, we were
forces, and the second law to predict what accelerations careful to say that Equation (3) applied to two small
these forces produce, was one of the most revolution- objects. To be more explicit, we mean that the two
ary scientific discoveries ever made. objects m1 and m2 should be small in dimensions
compared to the separation r between them. We can
Newton’s so-called universal law of gravitation can think of Equation (3) as applying to two point particles
most simply be stated as follows. If we have two small or point masses.
masses of mass m1 and m2, separated by a distance r as
shown in Figure (5), then the force between them is What happens if one or both of the objects are large
proportional to the product m1m2 of their masses, and compared to their separation? Suppose, for example,
inversely proportional to the square of the distance r that you would like to calculate the gravitational force
between them. This can be written as an equation of the between you and the earth as you stand on the surface
form of the earth. The correct way to do this is to realize that
you are attracted, gravitationally, to every rock, tree,
m 1m 2 Newton's law every single piece of matter in the entire earth as
Fg = G (3)
r2 of gravity indicated in Figure (6). Each of these pieces of matter
is pulling on you, and together they produce a net
where the proportionality constant G is a number that gravitational force Fg which is the force mg that we
must be determined by experiment. saw in our discussion of projectile motion.
Equation (3) itself is not the whole story, we must make
several more points. First, and very important, is the
fact that gravitational forces are always attractive; m1
is pulled directly toward m2, and m2 directly toward
m1. Second, the strength of these forces are equal, even
if m2 is much bigger than m1, the force of m2 on m1 is
the same in strength as the force of m1 on m2. That is
why we used the same symbol Fg for the two attractive
forces in Figure (5).

Newton’s law of gravity is called the universal law of

gravitation because Equation (3) is supposed to apply
to all masses anywhere in the universe, with the same
numerical constant G everywhere. G is called the
universal gravitational constant, and has the numeri- Figure 6
cal value, in the MKS system of units You are attracted to every piece of matter in the earth.
3 universal
m
G = 6.67 × 10
–11
2
gravitational (4)
kg sec constant
We will discuss shortly how this number was first
measured.
8-6 Newtonian Mechanics

It appears difficult to add up all the individual forces Using the results of Figure (7), i.e., calculating Fg by
exerted by every chunk of matter in the entire earth, to replacing the earth by a point mass me located a
get the net force Fg. Newton also thought that this was distance re below us, we get
difficult, and according to some historical accounts,
invented calculus to solve the problem. Even with Gmm e
calculus, it is a fairly complicated problem to add up all Fg = (5)
2
of these forces, but the result turns out to be very simple. re
For any uniformly spherical object, you get the cor- for the strength of the gravitational force on a particle
rect answer in Newton’s law of gravity if you think of of mass m at the surface of the earth. Combining this
all the mass as being concentrated at a point at the with Newton’s second law
center of the sphere. (This result is an accidental
consequence of the fact that gravity is a 1/r2 force, i.e.,
Fg = mg or Fg = mg (6)
that it is inversely proportional to the square of the
distance. We will have much more to say about this we get
accident in later chapters.)
Gmm e
Since the earth is nearly a uniformly spherical object, mg = (7)
r2
you can calculate the gravitational force between you
and the earth by treating the earth as a point mass The important result is that the particle’s mass m
located at its center, 4000 miles below you, as indicated cancels out of Equation (7), and we are left with the
in Figure (7). formula

Gm e
Galileo’s Observation g = (8)
As we mentioned earlier, Galileo observed that, in the re2
absence of air resistance, all projectiles should have the
same acceleration no matter what their mass. This for the acceleration due to gravity. We note that g
leads to the striking result that, in a vacuum, a steel ball depends on the earth mass m e , the earth radius re, and
and a feather fall at the same rate. Now we can see that the universal constant G, but not on the particle’s mass
this is a consequence of Newton’s second law com- m. Thus objects of different mass should have the same
bined with Newton’s law of gravity. acceleration.

Fg Fg
rearth

earth =
me

Figure 7
The gravitational force of the entire earth acting
on you is the same as the force of a point particle
with a mass equal to the earth mass, located at the
earth's center, one earth radius below you.
 8-7

THE CAVENDISH EXPERIMENT

A key feature of Newton’s law of gravitation is that all
objects attract each other via gravity. Yet in practice, glass fiber
the only gravitational force we ever notice is the force
of attraction to the earth. What about the gravitational
force between two students sitting beside each other, or
between your two fists when you hold them close to
each other? The reason that you do not notice these
forces is that the gravitational force is incredibly weak, small lead balls
weak compared to other forces that hold you, trees, and
rocks together. Gravity is so weak that you would a) Side view of the small balls.
never notice it except for the fact that you are on top of
a huge hunk of matter called the earth. The earth mass
is so great that, even with the weakness of gravity, the
resulting force between you and the earth is big enough
to hold you down to the surface.

The gravitational force between two reasonably sized

objects is not so small that it cannot be detected, it just Fg
requires a very careful experiment that was first per-
formed by Henry Cavendish in 1798. In the Cavendish Fg
experiment, two small lead balls are mounted on the
end of a light rod. This rod is then suspended on a fine
glass fiber as shown in Figure (8a).

As seen in the top view in Figure (8b), two large lead

balls are placed near the small ones in such a way that
the gravitational force between each pair of large and b) Top view showing two large lead balls.
small balls will cause the rod to rotate in one direction.
Once the rod has settled down, the large lead balls are
moved to the position shown in Figure (8c). Now the
gravitational force causes the rod to rotate the other
way. By measuring the angle that the rod rotates, and
by measuring what force is required to rotate the rod by Fg
this angle, one can experimentally determine the strength
of the gravitational force Fg between the balls. Then by Fg
using Newton’s law of gravity
m1
r
Fg = G m1m2 m2
Figure 9
r2 c) Top view with large balls rotated to
new position.
applied to Figure (9), one can solve for G in terms of the Figure 8
known quantities Fg , m1, m2 and r2. This was the way The Cavendish experiment. By moving the large lead
that Newton’s universal constant G, given in Equation balls, the small lead balls are first pulled one way, then
(4) was first measured. the other. By measuring the angle the stick holding the
small balls is rotated, one can determine the
gravitational force Fg .
8-8 Newtonian Mechanics

"Weighing” the Earth SATELLITE MOTION

Once you know G, you can go back to the formula The key idea that led Newton to his universal law of
(8) for the acceleration g due to gravity, and solve for gravitation was that the moon, while traveling in its
the earth mass me to get orbit about the earth, was subject to the same kind of
2 force as an apple falling from a tree. We have seen that
gre2 2 6
9.8 m/sec × 6.37 × 10 m a projectile in flight, such as an apple, accelerates down
me = = –11 3 2 toward the center of the earth. The moon, in its nearly
G 6.67 × 10 m /kg sec circular orbit around the earth, also accelerates toward
= 6.0 x 1024kg (9) the center of the earth, as illustrated in Figure (10).
Newton proposed that the accelerations of the falling
As a result, Cavendish was able to use his value for G apple and of the orbiting moon were both caused by the
to determine the mass of the earth. This was the first gravitational pull of the earth.
determination of the earth’s mass, and as a result the
Cavendish experiment became known as the experi- Golf ball
ment that “weighed the earth”. V
F
Exercise 2
The density of water is 1 gram/ cm3 . The average density String
of the earth’s outer crust is about 3 times as great. Use
Cavendish’s result for the mass of the earth to decide if
the entire earth is like the crust. (Hint —the volume of a
sphere of radius r is 4 3π3 r3 ). Relate your result to what
you have read about the interior of the earth.

Inertial and Gravitational Mass

The fact that, in the absence of air resistance, all
projectiles have the same acceleration— the fact that
the m’s canceled in Equation (7), has a deeper conse- Moon
V
quence than mere coincidence. In Newton’s second
law, the m in the formula F = ma is the mass defined F
by the recoil definition of mass discussed in Chapter 6.
Called inertial mass, it is the concept of mass that we
get from the law of conservation of linear momentum.

In Newton’s law of gravity, the projectile’s mass m in Earth

2
the formula Fg = Gmm e /r e is what we should call
the gravitational mass for it is defined by the gravita-
tional interaction. It is the experimental observation
that the m’s cancel, the observation that all projectiles
Figure 10
have the same acceleration due to gravity, that tells us When we swing a golf ball in a circle, the ball
that the inertial mass is the same as gravitational mass. accelerates toward the center of the circle, in the
This equivalence of inertial and gravitational mass has direction it is pulled by the string. Similarly, the
been tested with extreme precision to one part in a moon, in its circular orbit about the earth,
billion by Etvös in 1922 and to even greater accuracy accelerates toward the center of the earth, in the
by R. H. Dicke in the 1960s. direction it is pulled by the earth's gravity.
 8-9

The moon, being farther away from the center of the I.e., the moon’s acceleration is 27 thousand times
earth should be expected to feel a weaker gravitational weaker than the apple’s.
force and therefore have a weaker acceleration. From
direct calculation Newton could determine how much To understand the meaning of this result, let us look
weaker the moon’s acceleration was, and thus deter- at the square of the ratio of the distances from the
mine how the gravitational acceleration and force apple to the center of the earth, and the moon to the
decreases with distance. center of the earth. We have
2
2
To repeat Newton’s calculation, we know that the rapple to center of earth 6.37 × 10 6 m
apple on the surface of the earth has an acceleration =
gapple = 9.8 m/sec2. To determine the magnitude rmoon orbit 3.82 × 10 8 m
of the moon’s orbital acceleration toward the earth,
gmoon orbit , we can use the formula derived in Chap- = 2.78 × 10 -4 (13)
ter 3 for uniform circular motion, namely
which, to the accuracy of our work, is the same as the
v2 uniform
a = gmoon orbit = r circular (3-12) ratio of accelerations.
motion
Equating the results in Equations (12) and (13), we
To calculate the speed v of the moon, we note that the
get
moon takes 27.32 days or 2.36 x 106 seconds for one
complete orbit. The radius of the moon orbit is 3.82 g moon orbit r2e
x 108 meters, so that = 2
g apple rmoon orbit
vmoon = orbital circumference = t2π r
time for one orbit orbit g apple × r 2e
g moon orbit = 2 ∝ 2 1 (14)
r moon orbit r moon orbit
2π × 3.82 × 10 meters
8
=
2.36 × 10 6 sec Where g apple × r 2e can be thought of as a constant.
m
= 1.02 × 10 3 sec (10)
From such calculations Newton saw that the gravita-
or very close to 1 kilometer per second. Substituting tional acceleration of the moon, and thus the gravita-
this value of v into the formula v2 /r, gives tional force, decreased as the square of the distance
2 from the moon to the center of the earth. This was how
1.02 × 10 3 m/sec Newton deduced that gravity was a 1/r2 force law.
g moon orbit =
3.82 × 10 8 m Exercise 3
(11) How far above the surface of the earth do you have to
–3 m
= 2.70 × 10 be so that, in free fall, your acceleration is half that of
sec2 objects near the surface of the earth?
The ratio of the moon’s orbital acceleration to the
apple’s acceleration
g moon orbit 2.70 × 10 -3 m/sec2
=
g apple 9.8 m/sec2
(12)
= 2.71 × 10 - 4
8-10 Newtonian Mechanics

Other Satellites is the same. By using the constant acceleration formu-

To explain to the world the similarity of projectile and las with g = 32 ft/sec2, we can easily calculate that at the
satellite motion, that both the apple and the moon were end of one second both projectiles will have fallen 16
simply falling toward the center of the earth, Newton ft, and at the end of two seconds a distance of 64 ft.
drew the sketch shown in Figure (11). In the sketch,
Newton shows a projectile being fired horizontally In Figure (13), we have sketched the curved surface of
from the top of a mountain, and shows what would the earth. Due to this curvature, the surface of the earth
happen if there were no air resistance. If the horizontal will be 16 ft below a horizontal line out at a distance of
velocity were not too great, the projectile would go a 4.9 miles, and 64 ft below at a distance of 9.8 miles.
short distance along the typical parabolic path we have This effect can be seen from a small boat as you leave
studied in the strobe labs. As the projectile is fired faster shore. When you are 10 miles off shore, you cannot see
it would travel farther before hitting the ground. Fi- lighthouses under 64 ft tall, unless you climb your own
nally we reach a point where the projectile keeps falling mast. (For landlubbers sunning on the beach, sailboats
toward the earth, but the earth keeps falling away and with 64 ft high masts disappear from sight at a distance
the projectile goes all the way around the earth without of 10 miles.)
hitting it. Comparing Figures (12) and (13), we see that in the
Another perspective of the same idea is illustrated in absence of air resistance, if a projectile were fired
Figures (12) and (13). Figure (12) is a strobe photo- horizontally at a speed of 4.9 miles per second, during
graph showing two steel balls launched simultaneously, the first second it would fall 16 ft, but the earth would
one being dropped straight down and the other being have also fallen 16 ft, and the projectile would be no
fired horizontally. The photograph clearly demon- closer to the surface. By the end of the 2nd second the
strates that the downward motion of the two projectiles projectile would have fallen 64 ft, but still not have
come any closer to the surface of the earth. Such a
projectile would keep traveling around the earth, never
hitting the surface. It would fall all the way around,
becoming an earth satellite.

Figure 11
Newton's sketch showing that the difference
between projectile and satellite motion is that Figure 12
satellites travel farther. Both are accelerating Two projectiles, released simultaneously. The
toward the center of the earth. horizontal motion has no effect on the vertical
motion: they both fall at the same rate.
 8-11

Exercise 4 If the astronaut in an orbiting space capsule is weight-

less, but still subject to the gravitational force of the
An earth satellite in a low orbit, for instance 100 miles up,
earth, we cannot directly associate the word weight
is so close to the surface of the earth (100 miles is so
small compared to the earth’s radius of 4000 miles) that
with the effects of gravity. In order to come up with a
the satellite’s acceleration is essentially the same as the definition of the word weight that has some scientific
acceleration of projectiles here on earth. Use this result value, and is reasonably consistent with the use of the
to predict the period T of the satellite’s orbit. (Hint – the word in the popular press, we can define the weight of
satellite travels one earth circumference 2 πre in one an object as the magnitude of the force the object exerts
period T. This allows you to calculate the satellite’s on the bathroom scales. Here on earth, if you have an
speed v. You then use the formula v2/r for the magnitude object of mass m and you set it on the bathroom scales,
of the satellite’s acceleration.) it will exert a downward gravitational force of magni-
tude
Weight Fg = mg
The popular press often talks about the astronauts in
spacecraft orbiting the earth as being weightless. This Thus we say that the object has a weight W given by
is verified by watching them on television floating W = mg (15)
around inside the space capsule. You might jump to the
conclusion that because the astronauts are floating For example, a 60 kg boy standing on the scales
around in the capsule, they do not feel the effects of exerts a gravitational force
gravity. This is true in the same sense that when you
m
jump off a high diving board, you do not feel the effects W 60 kg boy = 60 kg × 9.8
of gravity—until you hit the water. While you are sec2
falling, you are weightless just like the astronauts. = 588 newtons
The only significant difference between your fall from We see that weight has the dimensions of a force, which
the high diving board, and the astronaut’s weightless in the MKS system is newtons. If the same boy stood
experience in the space capsule, is that the astronaut’s on the same scales in an orbiting spacecraft, both the
experience lasts longer. As the space capsule orbits the boy and the scales would be in free fall toward the
earth, the capsule and the astronauts inside are in center of the earth, the boy would exert no force on the
continuous free fall. They have not escaped the earth’s scales, and he would therefore be weightless.
gravity, it is gravity that keeps them in orbit, accelerat-
ing toward the center of the earth. But because they are
in free fall, they do not feel the acceleration, and are
considered to be weightless.
m
bathroom scales
9.8 mi
4.9 mi line of
16 ft sight mg
surface of 64 ft
the earth

Figure 14
We will define the weight of an object as
the force it exerts on the bathroom scales.

Figure 13
The curvature of the earth causes the horizon
to fall away 64 feet at a distance of 9.8 miles.
8-12 Newtonian Mechanics

Although we try to make the definition of the word Earth Tides

weight consistent with the popular use of the word, we An aspect of Newton’s law of gravity that we have not
do not actually succeed. In almost any country except said much about is the fact that gravity is a mutual
the United States, when you buy a steak, the butcher attraction. As we mentioned, two objects of mass m1
will weigh it in grams. The grocer will tell you that a and m2 separated by a distance r, attract each other
banana weighs 200 grams. You are not likely find a with a gravitational force of magnitude
grocer who tells you the weight of an object in newtons. F g = Gm 1m 2/r2. The point we want to emphasize
It is a universal convention to tell you the mass in grams now is that the force on each particle has the same
or kilograms, and say that that is the weight. About the strength Fg.
only place will you will find the word weight to mean
a force, as measured in newtons, is in a physics course. Let us apply this idea to you, here on the surface of the
earth. Explicitly, let us assume that you have just
(In the English system of units, a pound is a force, so jumped off a high diving board as illustrated in Figure
that it is correct to say that our 60 kg mass boy weighs (15), and have not yet hit the water. While you are
132 lbs. That, of course, leaves us with the question of falling, the earth’s gravity exerts a downward force Fg
what mass is in the English units. From the formula which produces your downward acceleration g.
F = mg, we see that m = F/g, or an object that weighs
32 lbs has a mass 32 lbs/32ft/sec2 = 1. As we mentioned According to Newton’s law of gravity, you are exerting
earlier, this unit mass in the English units is called a an equal and opposite gravitational force Fg on the
slug. This is the last time we will mention slugs in this earth. Why does nobody talk about this upward force
text.) you are exerting on the earth? The answer, shown in the
following exercise, is that even though you are pulling
up on the earth just as hard as the earth is pulling down
on you, the earth is so much more massive that your
pull has no detectable effect.

Exercise 5
Assume that the person in Figure (15) has a mass of 60
kilograms. The gravitational force he exerts on the earth
causes an upward acceleration of the earth a earth.
Show that a earth = 10 — 22 m/sec2.

diving board

Fg

Fg

water
Figure 15
As you fall toward the water, the earth is
pulling down on you, and you are pulling up on
the earth. The two forces are of equal strength.
 8-13

More significant than the force of the diver on the earth strength as the force the earth exerts on the moon. Since
is the force of the moon on the earth. It is well known the earth is more massive, the effect on the earth is not
that the ocean tides are caused by the moon’s gravity as great, but it is noticeable. The reason for the second
acting on the earth. On the night of a full moon, high bulge of water on the far side of the earth is that the
tide is around midnight when the moon is directly center of the earth is closer to the moon than the water
overhead. The time of high tide changes by about an on the back side, and therefore accelerates more rapidly
hour a day in order to stay under the moon. toward the moon than the water on the back side. The
water on the back side gets left behind to form a bulge.
The high tide under the moon is easily explained by the
idea that the moon’s gravity sucks the ocean water up The result, the fact that there are two high tides a day,
into a bulge under the moon. As the earth rotates and the fact that there is a second bulge on the back side, is
we pass under the bulge, we see a high tide. This direct experimental evidence that the earth is acceler-
explains the high tide at midnight on a full moon. ating toward the moon. It is direct evidence that the
moon’s gravity is pulling on the earth, just as the earth’s
The problem is that there are 2 high tides a day about 12 gravity is pulling on the moon.
hours apart. The only way to understand two high tides
is to realize that there are two bulges of ocean water, one As a consequence of the earth’s acceleration, the moon
under the moon and one on the opposite side of the is not traveling in a circular orbit centered precisely on
earth, as shown in Figure (16). In one 24 hour period the center of the earth. Instead both the earth and the
we pass under both bulges. moon are traveling in circles about an axis point located
on a line joining the earth’s and moon’s centers. This
Why is there a bulge on the backside? Why isn’t the axis point is located much closer to the center of the
water all sucked up into one big bulge underneath the earth than that of the moon, in fact it is located inside the
moon? earth about 3/4 of the way toward the earth’s surface as
The answer is that the moon’s gravity not only pulls on shown in Figure (17).
the earth’s water, but on the earth itself. The force of
gravity that the moon exerts on the earth is just the same

Moon Moon

axis
point

Earth rotating
Earth under the two
bulges of water
Earth

Figure 16 Figure 17
The two ocean bulges cause two high tides per day. Both the earth and the moon travel in circular
orbits about an axis point located about 1/4 of
the way down below the earth's surface.
8-14 Newtonian Mechanics

Planetary Units In our work with projectiles in the lab the CGS system
In introductory physics texts, it has become almost an of units was excellent. The projectiles typically went
article of religion that all calculations shall be done distances from 10 to 100 cm, in times of the order of 1
using MKS units. This has some advantages – we do second, and had masses of the order of 100 gm. There
not have to talk about pounds and slugs, but practicing were no large exponents involved.
physicists seldom follow this rule. Physicists studying
the behavior of elementary particles, for example, Now that we are studying the motion of earth satellites,
routinely use a system of units that simplify their we are faced with large exponents in quantities like the
calculations, units in which the speed of light and other earth mass and the gravitational constant G which are
5.98 × 10 24kg and 6.67 × 10 –11m3/kg sec2 respec-
fundamental constants have the numerical value 1.
Using these special units they can quickly solve simple tively. The calculations we have done so far using these
problems and gain an intuitive feeling for which quan- numbers have required a calculator, and we have had to
tities are important and which quantities are not. work hard to gain insight from the results.

Table 1 Planetary Units

Constant Symbol Planetary units MKS units

Gravitational Constant G 20 6.67 x 10-11 m 3 2

kg sec
Acceleration due
to gravityat the ge 20 9.8 m/sec2
earth's surface
Earth mass me 1 5.98 x 1024 kg

Moon mass mmoon .0123 7.36 x 1022 kg

Sun mass msun 3.3 x 105 1.99 x 10 30kg

Metric ton ton 1.67 x 10- 22 1000 kg

Earth radius re 1 6.37 x 106 m

Moon radius rmoon .2725 1.74 x 106 m

Sun radius rsun 109 6.96 x 108 m

Earth orbit radius rearth orbit 23400 1.50 x 1011 m

Moon orbit radius rmoon orbit 60 3.82 x 108 m

Hour hr 1 hr 3600 sec

Moon period lunar month 656 hrs 2.36 x 106 sec

(= 27.32 days)
(siderial)

Year yr 8.78 x 103 hrs 3.16 x 107 sec

 8-15

We now wish to introduce a new set of units, which we Exercise 6

will call planetary units, that makes satellite calcula- We will have you convert Newton’s universal gravita-
tions much simpler and more intuitive. One way to tional constant G into planetary units. Start with
design a new set of units is to first decide what will be
our unit mass, our unit length, and our unit time, and meters3
G = 6.67 x 10 -11
then work out all the conversion factors so that we can kg sec2
convert a problem into our new units. For working
earth satellite problems, we have found that it is conve- Then multiply or divide by the conversion factors
nient to take the earth mass as the unit mass, the earth sec
3600
radius as the unit length, and the hour as the unit time. hr
kg
mearth = 1 earth mass 5.98 x 10 24
earth mass
Rearth = 1 earth radius meters
6.37 x 10 6
hour = 1 earth radii

With these choices, speed, for example, is measured until all the dimensions in the formula for G are con-
in (earth radii)/ hr, etc. verted to planetary units. (I.e., convert from seconds to
hours, kg to earth mass, and meters to earth radii.) If you
This system of units has a number of advantages. We do the conversion correctly, you should get the result
can set me and re equal to 1 in the gravitational force earth radii 3
formulas, greatly simplifying the results. We know G = 20
earth mass hr2
immediately that a satellite has crashed if its orbital
radius becomes less than 1. Typical satellite periods are
a few hours and typical satellite speeds are from 1 to 10 Exercise 7
earth radii per hour. What may be a bit surprising is that Explain why g and G have the same numerical value in
both the acceleration due to gravity at the surface of the planetary units.
earth, g, and Newton’s universal gravitational constant
G, have the same numerical value of 20. As an advertisement for how easy it is to use planetary
units in satellite calculation, let us repeat Exercise (4)
Table 1 shows the conversion from MKS to planetary using these units. In that exercise we wished to
units of common quantities encountered in the study of calculate the period of Sputnik 1, a satellite traveling in
satellites moving in the vicinity of the earth and the a low earth orbit. We were to assume that Sputnik’s
moon. orbital radius was essentially the earth’s radius re as
Sputnik 1 shown in Figure (18), and that Sputnik’s acceleration
toward the center of the earth was essentially the same
as the projectiles we studies in the introductory lab, i.e.,
re ge = 9.8 m /sec2.

Earth

Figure 18
A satellite in a low earth orbit.
8-16 Newtonian Mechanics

Using the formula COMPUTER PREDICTION

a = vr
2 OF SATELLITE ORBITS
In this chapter we have discussed two special kinds of
we get motion that a projectile or satellite can have. One is the
v2Sputnik v2Sputnik parabolic trajectory of a projectile thrown across the
ge = = room – motion that is easily described by calculus and
re 1
Therefore the constant acceleration formulas. The other is the
orbital motion of the moon and man-made satellites
vSputnik = ge = 20 earth radii that are in circular orbits. These orbits can be analyzed
hr
using the fact that their acceleration is directed toward
Now the satellite travels a total distance 2π re to go the center of the circle and has a magnitude v2/r.
one orbit, therefore the time it takes is
These two examples are deceptively simple. Newton’s
Sputnik period = v 2π re = 2π = 1.4 hrs diagram, Figure (11), shows that there is a continuous
Sputnik 20 range of orbital shapes starting from simple projectile
Compare the algebra that we just did with what you had motion out to circular orbital motion and beyond. For
to go through to get an answer in Exercise (4). (You all these orbital shapes, we know the projectile’s accel-
should have gotten the same answer, 1.4 hrs, or 84 eration is the gravitational acceleration toward the
minutes, or 5,040 seconds. This is in good agreement center of the earth. But to go from a knowledge of the
with the observed time for low orbit satellites.) If you acceleration to predicting the shape of the orbit is not
have watched satellite launches on television, you may necessarily an easy task.
recall waiting about an hour and a half before the
satellite returned. There are no simple formulas like the constant accel-
eration formulas that allow us to predict where the
satellite will be at any time in the future. Using
Exercise 8
advanced calculus techniques one can show that the
A satellite is placed in a circular orbit whose radius is
orbits should have the shape of conic sections, one
2re (it is one earth radii above the surface of the earth.)
example being the elliptical orbits discovered by Kepler.
(a) What is the acceleration due to gravity at this But if we go to more complicated problems like trying
altitude? to predict the motion of the Apollo 8 spacecraft from
the earth to the moon and back, then a calculus ap-
(b) What is the period of this satellite’s orbit?
proach is completely inadequate.
(c) What is the shortest possible period any earth
satellite can have? Explain your answer. On the other hand these problems are easily handled
using the step-by-step method of predicting motion,
Exercise 9 the method, discussed in Chapter 5, that we implement
Communication satellites are usually placed in circular
using the computer. With a slight modification of our
orbits over the equator, at an altitude so that they take old projectile motion program, we can predict what
precisely 24 hours to orbit the earth. In this way they will happen to an earth satellite no matter how it is
hover over the same point on the earth and can be in launched and what orbit it has. Adding a few more lines
continuous communication with the same transmitters to the program allows us to send the satellite to the
and receivers. This orbit is called the Clarke orbit, moon and back.
named after the science fiction writer Arthur Clark who
first emphasized the importance of such an orbit. Cal- Once we are familiar with a basic satellite motion
culate the radius of the Clark orbit. program, we can easily add new features. We can, for
example, change the exponent in the gravitational
force law from 1/r2 to 1/r2.1 to see what happens if the
gravitational force law is modified. Similar modifica-
 8-17

tions were in fact predicted by Einstein’s general This loop expresses the method of predicting motion
theory of relativity, thus we will be able to observe the that we developed from the analysis of strobe pho-
kind of effects that were used to verify Einstein’s tographs. The idea behind the command
theory.
LET Rnew = Rold + Vold * dt
New Calculational Loop is illustrated in Figure (5-15a) reproduced here. The
In Chapter 5, we set up the machinery to do computer new position of the particle is obtained from the old
calculations. This involved learning the LET state- position by adding the vector Vold * dt to the old
ment, constructing loops, plotting crosses, etc. Al- coordinate vector R old.
though this may have been a bit painful (but perhaps not
as painful as learning calculus), we do not have to do Once we get to the new position of the particle, we
much of that again. We can use essentially the same need the new velocity vector in order to calculate the
machinery to predict satellite orbits. The only signifi- next new position. The new velocity vector is
cant change is in the calculational loop where we obtained from the command
predict the particle’s new position and velocity. LET Vnew = Vold + A * dt

In the projectile motion program, the English version as illustrated in Figure (5-15b). The DO–LOOP part of
of the calculational loop was, from Figure (5-18) the program tells us to keep repeating this step-by-step
process until we get as much of the trajectory as we
! --------- Calculational Loop want (in this case until one second has elapsed).
DO
The calculational loop of Figure (19) works for
LET Rnew = Rold + Vold * dt projectile motion because we always know the
LET A = g projectile’s acceleration A which is given by the
LET Vnew = Vold + A * dt line
LET Tnew = Told + dt LET A = g projectile motion (16)
PLOT R
LOOP UNTIL T > 1 This is the line that characterizes projectile motion, the
line that tells the computer that the projectile has a
Figure 19
constant acceleration g.

Vold Vold
Vold
Vnew
Vnew
A ∆t
R old
R new

( V new – Vold) Vnew = Vold + A *∆t

A=
∆t

Figure 5-15a Figure 5-15b

Predicting the next new position. Predicting the next new velocity.
8-18 Newtonian Mechanics

The only fundamental change we need to make in In our satellite motion problem, the gravitational
going from projectile motion to satellite motion is to force Fg points toward the center of the earth. Thus
change our command for the particle’s acceleration to define the direction of Fg , we need a unit vector
A . Instead of assuming that the particle’s accelera- that points toward the center of the earth. In Figure
tion is constant, we use Newton’s law of gravity (20a) we show the coordinate vector R which de-
Fg = Gm1m 2 /r2 to calculate the force acting on fines the position of the satellite in a coordinate
the satellite, and then Newton’s second law A = Fg /m system whose origin is at the center of the earth. In
to obtain the resulting acceleration. Figure (20b) we see the vector –R, which points
from the satellite to the center of the earth, the same
There are of course some other details. We have to find direction as the gravitational force. Therefore we
a way to express the vector nature of the gravitational would like to turn –R into a unit vector, which we do
force – i.e., to tell the computer which way the gravita- by dividing by the length of R, namely the distance
tional force is pointing, and we are going to change our from the center of the earth to the satellite.
plotting scale since we are no longer working in front
of a 100 cm by 100 cm grid. But essentially we are Since we will often use unit vectors in this text, we will
replacing the command designate them by a special symbol. Instead of an
arrow over the letter, we will use what is called a caret
LET A = g by typographers, or more familiarly a hat by physicists.
Thus our unit vector in the –R direction will be denoted
by the new lines
by –R and is given by the formula
LET Fg = GM em/R 2 with instructions
for a direction
–R unit vector in
LET A = Fg /m –R = the –R direction (16)
R
and then using the same old program.
Satellite

Unit Vectors
R
We have no problem describing the direction of the
gravitational force on the satellite—the force is di-
a)
rected toward the center of the earth. But how do we tell
Earth
the computer that? What mathematical technique can
we use to express the direction of Fg?

The technique that we will use throughout the course is

the use of the unit vector. A unit vector is a dimension- –R
less vector of length 1 that points in the direction of
interest. If we want a vector of length 5 newtons that
b)
points in the same direction, then we multiply our unit
vector by the number 5 newtons to get the desired
result. (Recall that multiplying a vector by a number,
for example n, gives a vector n times as long, pointing
in the same direction.)
–R
There is an easy way to construct unit vectors. If we can
find some vector that points in the desired direction, we
c)
divide that vector by its own length, and we end up with
a vector of length 1, the required unit vector.
Figure 20
The unit vector –R
 8-19

In Equation (16), the length R is given by the The command

Pythagorean theorem
LET Rnew = Rold + Vold * dt
R = R2x + R2y (16a) for Rnew becomes
Rx and Ry being the x and y coordinates of the satellite. LET Rx = Rx + Vx * dt (18a)
With the unit vector –R , we can now write an LET Ry = Ry + Vy * dt (18b)
explicit formula for the gravitational force vector
Fg . We multiply the unit vector –R by the magni- where we drop the subscripts “new” and “old” because
the computer automatically takes the old values on the
tude GMm/R2 of the gravitational force to get
right side of the LET statement, calculates a new value,
and stores the new value in the memory cell named on
GMem
Fg = –R (17) the left side of the LET statement. (See our discussion
R2 of the LET statement on page 5-5).

In Equations (18a) and (18b) we obtain numerical

Calculational Loop values for the new coordinates Rx and Ry of the
for Satellite Motion satellite. However, we will also need to know the
We are now ready to go in an orderly way from the distance R from the satellite to the center of the earth (in
calculational loop for projectile motion to a calcula- order to construct the unit vector – R ). The value of R
tional loop for satellite motion. We can focus our is easily determined by adding the command
attention on the following three lines of the projec-
tile motion calculation loop (Figure 21) because the LET R = SQR (Rx*Rx + Ry*Ry) (18c)
other lines remain unchanged.
where SQR is BASIC’s way of saying square root.
LET Rnew = Rold + Vold * dt
The translation of the command for Fg only requires
LET A = g
the translation of the unit vector R into x and y
LET Vnew = Vold + A * dt coordinates. Remembering that R = R/R , we get
Figure 21 R x = Rx/R ; R y = Ry/R (19)

The first step is to replace LET A = g by Newton’s thus the translation of the LET statement for Fg can be
law of gravity and Newton’s second law as shown in written as
Figure (22). LET Fg = G * Me * M / (R*R)
LET R new = R old + Vold * dt LET Fgx = (–Rx / R) * Fg
LET Fg = – R GM em/R 2 LET Fgy = (–Ry / R) * Fg
LET A = Fg/m The computer can handle these lines because it already
knows the new values of Rx, Ry and R from Equations
LET Vnew = Vold + A * dt (18a, b, and c).
Figure 22

Because BASIC is limited to working with numerical

commands rather than vectors (an unfortunate limita-
tion), the next step is to make sure that we can translate
each of these vector commands into the separate x and
y components. We will do this separately for each of
the 4 lines.
8-20 Newtonian Mechanics

The translation of LET statements for A and Vnew are Working Orbit Program
straightforward. We get We are now ready to convert a working projectile
motion program, Figure (5-23) reproduced here, into a
LET Ax = Fgx / M working orbital motion program. In addition to con-
LET Ay = Fgy / M verting the calculational loop as we have just discussed,
we need to change some constants and plotting ranges,
LET Vx = Vx + Ax * dt but the general structure of the program will be un-
LET Vy = Vy + Ay * dt changed.
LET V = SQR(Vx*Vx + Vy*Vy)
Plotting Window
We included a calculation of the magnitude V of the We will initially consider satellite motion that stays
satellite’s speed for future use. We may, for example, reasonably close to the earth, within several earth radii.
want to construct a unit vector in the -V direction to Using planetary units, and placing the earth at the
represent the direction of air resistance on a reentering center of the plot, we can get a reasonable range of
satellite. We have found it convenient to routinely orbits if we let Rx vary for example from - 9 to +9 earth
calculate the magnitude of any vector whose x and y radii. If we have a standard 9" Macintosh screen, the x
coordinates we have just calculated. dimension should be 1.5 times the y dimension, thus Ry
should go only from -6 to +6. The following command
Summary sets up this plotting window
To summarize our translation, we started with the
vector commands SET WINDOW -9, 9, -6, 6
To show where the earth is located, we can use the
following lines to plot a cross at the center of the earth
LET R new = R old + Vold * dt
LET Rx = 0
LET Fg = –R GMe m/R 2 LET Ry = 0
LET A = Fg /m CALL CROSS
Constants and Initial Conditions
LET Vnew = Vold + A * dt
In going from the projectile motion to the satellite
motion program, we have to change the constants and
and ended up with the BASIC commands initial conditions. Using planetary units, our constants
G, Me, and m are
LET Rx = Rx + Vx * dt
LET Ry = Ry + Vy * dt LET G = 20
LET R = SQR (Rx*Rx + Ry*Ry) LET Me = 1
LET m = .001
LET Fg = G * Me * M / (R*R)
LET Fgx = (–Rx / R) * Fg (Our choice of the satellite mass m does not matter
LET Fgy = (–Ry / R) * Fg because it cancels out of the calculation.)

LET Ax = Fgx / M For initial conditions, we will start the satellite .1 earth
radii above the surface of the earth on the + x axis;
LET Ay = Fgy / M

LET Vx = Vx + Ax * dt LET Rx = 1.1

LET Vy = Vy + Ay * dt LET Ry = 0
LET V = SQR(Vx*Vx + Vy*Vy) LET R = SQR(Rx*Rx + Ry*Ry)
CALL CROSS
 8-21

Projectile Motion Program Orbit-1 Program

Figure 23
Projectile motion program that plots
crosses every tenth of a second.

Figure 24
Our new orbital motion program.
8-22 Newtonian Mechanics

We also calculated an initial value of R for use in the Exercise 10

gravitational force formula, and plotted a cross at this
Convert your projectile motion program to the Orbit 1
initial point.
program. Use the same initial conditions so that you get
We are going to fire the satellite in the +y direction, the same orbit as that shown in Figure (25). (It is
important to get your Orbit 1 program running correctly
parallel to the surface of the earth. Trial and error shows
now, for it will be used as the basis for studying several
us that a reasonable value for the speed of the satellite
phenomena during the rest of this chapter. If you are
is 5.5 earth radii per hour, thus we write for our initial having problems, simply type the program in precisely
velocity commands as shown in Figure (24).

LET Vx = 0
LET Vy = 5.5
LET V = SQR(Vx*Vx + Vy*Vy)

In our projectile motion program of Figure (5-23) we

wanted a cross plotted every 100 time steps dt. This
was done with the command

IF MOD(i,100) = 0 THEN CALL CROSS

For our orbit program, trial and error shows that we get
a good looking plot if we draw a cross every 40 time Figure 25
steps, each time step dt being .01 hours. Thus our new Output of the Orbit 1 program. The satellite is initially
MOD line will be out at a distance x = 1.1 earth radii, and is fired in the
+y direction at a speed of 5.5 earth radii per hour.
IF MOD(i,40) = 0 THEN CALL CROSS

and we will get a cross every .01 * 40 = .4 hours. Once your program is working, it is easy to make small
modifications to improve the results. To create Figure
The final change is to stop plotting after one orbit. From (25a) we added the command
running the program we find that one orbit takes about
9 hours, thus we can stop plotting just before one orbit BOX CIRCLE -1,1,-1,1
with the LOOP instruction
to draw a circle to represent the earth. We also changed
LOOP UNTIL T > 9 dt to .001 and changed the MOD command to
MOD(i,539) to get an even number of crosses around
Putting all these steps together gives us the complete the orbit. We then plotted until T = 9 hours. (With dt ten
BASIC program shown in Figure (24). times smaller, our i counter has to be ten times bigger
to get the old crosses.)
When we run the Orbit 1 program, we get the elliptical
orbit shown in Figure (25).

Figure 25a
 8-23

Satellite Motion Laboratory you can hear the statement "That's good there, Neil". A
In our study of projectile motion, we could go to the short while later you hear the clear echo "That's good
laboratory and take strobe photographs in order to see there, Neil". The time delay from the original statement
how projectiles behaved. Obtaining experimental data and the echo is the time it takes a radio wave, traveling
for the study of satellite motion is somewhat more at the speed of light, to go to the moon and back. Using
difficult. What we will do is to use the Orbit 1 program an inexpensive stop watch, one can easily measure the
or slight modification of it to stimulate satellite motion, time delay as being about 2 2/5 seconds. Thus the one-
using it as our laboratory for the study of the behavior way trip to the moon is 1 1/5 seconds. Since light
of satellites. travels 1 ft/nanosecond, or 1 billion feet per second,
from this one determines that the moon is about 1.2
But first we wish to check that the Orbit 1 program billion feet away. You can convert this distance to earth
makes predictions that are in agreement with experi- radii to check the astronomer's value of 60 earth radii as
ment. The program is based on Newton's laws of the average distance to the moon.)
2
gravity, Fg = GMm/r , Newton's law of motion
a = F/m, and the procedures we developed earlier Exercise 11
for predicting the motion of an object whose accel- Adjust the initial conditions in your Orbit 1 program so
eration is known. Thus a verification of the results that the satellite is in a low earth orbit, and see what the
of the Orbit 1 program can be considered a verifica- period of the orbit is.
tion of these laws and procedures.
(To adjust the initial conditions, start, for example, with
Some tests of the Orbit 1 program can be made using R x = 1.01, R y = 0, vx = 0 and adjust vy until you get a
the results of your own experience. Anyone who has circular orbit centered on the earth. As a check that the
listened to the launch of a low orbit satellite should be satellite did not go below the surface of the earth, you
aware that the satellite takes about 90 minutes to go could add the line
around the earth once. The Orbit 1 program should give IF R < 1 THEN PRINT "CRASHED"
the same result, which you can check in Exercise 11.
Another obvious test is the prediction of the period of Adding this line just after you have calculated R in the
the moon in its orbit around the earth. It is about 4 DO LOOP will immediately warn you if the satellite has
weeks from full moon to full moon, thus the period crashed. You can then adjust the initial vy so that you
should be approximately 4 weeks or 28 days. The fact just avoid a crash. Once you have a circular orbit, you
that the apparent diameter of the moon does not change can adjust the time in the "LOOP UNTIL T > ..." com-
mand so that just one orbit is printed. This tells you how
much during this time indicates that the moon is
long the orbit took. You can also see how long the orbit
traveling in a nearly circular orbit about the earth. If you took by adding the line in the DO LOOP
accept the astronomer's measurements that the moon
orbit radius is about 60 earth radii away, then you can IF MOD(I, 40) = 0 THEN PRINT T, RX, RY
check the Orbit 1 program to see if it predicts a 4 week
Looking at the values of R x and R y you can tell when
period for an earth satellite in a circular orbit of that
one orbit is completed, and the value of T tells you how
radius (Exercise 12). long it took.
(An easy way to measure the distance to the moon was
Exercise 12
provided by the first moon landing. Because of a
problem with Neil Armstrong's helmet, radio signals Put the satellite in a circular orbit whose radius is equal
sent to Neil from Houston were retransmitted by Neil’s to the radius of the moon's orbit. (See Table 1, Planetary
Units, for the value of the moon orbit radius.) See if you
microphone, giving an apparent echo. The echo was
predict that the moon will take about 4 weeks to go
particularly noticeable while Neil was setting up a TV
around this orbit.)
camera. On a tape of the mission supplied by NASA,
8-24 Newtonian Mechanics

KEPLER'S LAWS visible without a telescope, the situation is more com-

A more detailed test of Newton's laws and the Orbit 1 plicated. Mars, for example, moves in one direction
program is provided by Kepler's laws of planetary against the background of the fixed stars, then reverses
motion. and goes backward for a while, then forward again as
shown in Figure (26). None of the planets has the
To get a feeling for the problems involved in studying simple uniform motion seen in the case of the moon and
planetary motion, imagine that you were given the job the sun.
of going outside, looking at the sky, and figuring out
how celestial objects moved. The easiest to start with After a lot of observation and the construction of many
is the moon, which becomes full again every four plots, you might make a rather significant discovery.
weeks. On closer observation you would notice that the You might find what the early Greek astronomers
moon moved past the background of the apparently learned, namely that if you assume that the planets
fixed stars, returning to its original position in the sky Mercury, Venus, Mars, Jupiter, and Saturn travel in
every 27.3 days. Since, as we mentioned, the diameter circular orbits about the sun, while the sun is traveling
of the moon does not change much, you might then in a circular orbit about the earth, then you can explain
conclude that the moon is in a circular orbit about the all the peculiar motion of the planets. This is a
earth, with a period of 27.3 days. remarkable simplification and compelling evidence
that there is a simple order underlying the motion of
The time it takes the moon to return to the same point celestial objects.
in the sky is not precisely equal to the time between full
moons. A full moon occurs when the sun, earth, and One of the features of astronomical observations is that
moon are in alignment. If the sun itself appears to move they become more accurate as time passes. If you
relative to the fixed stars, the full moons will not occur observe the moon for 100 orbits, you can determine the
at precisely the same point, and the time between full average period of the moon nearly 100 times more
moons will not be exactly the time it takes the moon to accurately than from the observation of a single period.
go around once. You can also detect any gradual shift of the orbit 100
times more accurately.
To study the motion of the sun past the background of Background stars
Apparent retrogra
the fixed stars is more difficult because the stars are not de o
rbit
visible when the sun is up. One way to locate the
position of the sun is to observe what stars are overhead
at "true" midnight, half way between dusk and dawn.
The sun should then be located on the opposite side of
the sky. (You also have to correct for the north/south
position of the sun.) After a fair amount of observation
and calculations, you would find that the sun itself
moves past the background of the fixed stars, returning
to its starting point once a year.

From the fact that the sun takes one year to go around Mars orbit
the sky, and the fact that its apparent diameter remains
essentially constant, you might well conclude that the Earth orbit
sun, like the moon, is traveling in a circular orbit about
the earth. This was the accepted conclusion by most
astronomers up to the time of Nicolaus Copernicus in
the early 1500s AD.
Figure 26
If you start looking at the motion of the planets like Retrograde motion of the planet Mars.
Mercury, Venus, Mars, Jupiter, and Saturn, all easily Modern view of why Mars appears to
reverse its direction of motion for a while.
 8-25

Even by the time of the famous Greek astronomer Ptolemy planets moved in circular orbits about the sun. Such a
in the second century AD, observations of the positions theory not only conflicted with common sense about
of the planets had been made for a sufficiently long time feeling the motion of the earth, but also displaced the earth
that it had become clear that the planets did not travel in and mankind from the center of the universe, two results
precisely circular orbits about the sun. Some way was quite unacceptable to many scholars and theologians.
needed to explain the non circularity of the orbits.
Copernicus' theory was not quite as simple as it first
The simplicity of a circular orbit was such a compelling sounds. Because of the accuracy with which planetary
idea that it was not abandoned. Recall that the apparently motion was know by 1543, it was necessary to include
peculiar motion of Mars could be explained by assuming epicycles in the planetary orbits in Copernicus' model.
that Mars traveled in a circular orbit about the sun which
in turn traveled in a circular orbit about the earth. By Starting around 1576, the Dutch astronomer Tycho Brahe
having circular orbits centered on points that are them- made a series of observations of the planetary positions
selves in circular orbits, you can construct complex that were a significant improvement over previous mea-
orbits. By choosing enough circles with the correct radii surements. This work was done before the invention of
and periods, you can construct any kind of orbit you wish. the telescope, using apparatus like that shown in Figure
(27). Tycho Brahe did not happen to believe in the
Ptolemy explained the slight variations in the planetary Copernican sun-centered theory, but that had little
orbits by assuming that the planets traveled in circles
around points which traveled in circles about the sun,
which in turn traveled in a circle about the earth. The extra
cycle in this scheme was called an epicycle. With just a
few epicycles, Ptolemy was able to accurately explain all
observations of planetary motion made by the second
century AD.

With 1500 more years of planetary observations, Ptolemy's

scheme was no longer working well. With far more
accurate observations over this long span of time, it was
necessary to introduce many more epicycles into Ptolemy's
scheme in order to explain the positions of the planets.

Even before problems with Ptolemy's scheme became

apparent, there were those who argued that the scheme
would be simpler if the sun were at the center of the solar
system and all the planets, including the earth, moved in
circles about the sun. This view was not taken seriously
in ancient times, because such a scheme would predict
that the earth was moving at a tremendous speed, a
motion that surely would be felt. (The principle of
relativity was not understood at that time.)

For similar reasons, one did not use the rotation of the
earth to explain the daily motion of sun, moon, and stars.
That would imply that the surface of the earth at the
equator would be moving at a speed of around a thousand
miles per hour, an unimaginable speed!

In 1543, Nicolaus Copernicus put forth a detailed plan for

the motion of the planets from the point of view that the Figure 27
Tycho Brahe’s apparatus.
sun was the center of the solar system and that all the
8-26 Newtonian Mechanics

effect on the reason for making the more accurate Kepler's First Law
observations. Both the Ptolemaic and Copernican Kepler's first law states that the planets move in ellip-
systems relied on epicycles, and more accurate data tical orbits with the sun at one focus. By analogy we
was needed to improve the predictive power of these should find from our Orbit 1 program that earth satel-
theories. lites move in elliptical orbits with the center of the earth
at one focus. To check this prediction, we need to know
Johannes Kepler, a student of Tycho Brahe, started how to construct an ellipse and determine where the
from the simplicity inherent in the Copernican system, focus is located.
but went one step farther than Copernicus. Abandon-
ing the idea that planetary motion had to be described The arch above the entrance to many of the old New
in terms of circular orbits and epicycles, Kepler used England horse sheds was a section of an ellipse. The
Tycho Brahe's accurate data to look for a better way to carpenters drew the curve by placing two nails on a
describe the planet's motion. Kepler found that the wide board, attaching the ends of a string to each nail,
planetary orbits were accurately and simply described and moving a pencil around while keeping the string
by ellipses, where the sun was at one of the focuses of taut as shown in Figure (28). The result is half an ellipse
the ellipse. (We will soon discuss the properties of with a nail at each one of the focuses. (If you are in the
ellipses.) Kepler also found a simple rule relating the Mormon Tabernacle’s elliptical auditorium and drop a
speed of the planet to the area swept out by a line drawn pin at one focus, the pin drop can be heard at the other
from the planet to the sun. And thirdly, he discovered focus because the sound waves bouncing off the walls
that the ratio of the cube of the orbital radius to the all travel the same distance and add up constructively
square of the period was the same for all planets. These at the second focus point.)
three results are known as Kepler's three laws of
planetary motion. To see if the satellite orbit from the Orbit 1 program is
an ellipse, we first locate the second focus using the
Kepler's three simple rules for planetary motion, which output shown in Figure (25a) by locating the point
we will discuss in more detail shortly, replaced and symmetrically across from the center of the earth as
improved upon the complex system of epicycles needed shown in Figure (29). Then at several points along the
by all previous theories. After Kepler's discovery, it orbit we draw lines from that point to each focus as
was obvious that the sun-centered system and elliptical shown, and see if the total length of the lines (what
orbits provided by far the simplest description of the would be the length of the stretched string) remains
motion of the heavenly objects. For Isaac Newton, half constant as we go around the orbit).
a century later, Kepler's laws served as a fundamental
test of his theories of motion and gravitation. We will
now use Kepler's laws in a similar way, as a test of the c b
validity of the Orbit 1 program and our techniques for a
predicting motion. d

board focus focus

string

Figure 29
nail nail Checking that our satellite orbits are an ellipse. We
Figure 28 construct a second focus, and then see if the sum of
Ellipse constructed with two nails and a string. the distances from each focus to a point on the
ellipse in the same for any point around the ellipse.
For this diagram, we should show that a+b = c+d.
 8-27

Exercise 13 have in Figure (30) reproduced the output of Figure

Using the output from your Orbit 1 program, check that
(25a), shaded the areas swept out as R moves from
the orbit is an ellipse. positions A to B, from C to D, and from E to F. These
areas should look approximately equal; you will check
Exercise 14 that they are in fact equal in Exercise 15.
Slightly alter the initial conditions of your Orbit 1 program The most significant consequence of Kepler's second
to get a different shaped orbit. (Preferably, make the law is that in order to sweep out equal areas while the
orbit more stretched out.) Check that the resulting orbit
radius vector is changing length, the planet or satellite
is still an ellipse.
must move more rapidly when the radius vector is
short, and more slowly when the radius vector is long.
Kepler's Second Law The planet moves more rapidly when in close to the
Kepler's second law relates the speed of the planet to the sun, and more slowly when far away.
area swept out by a line connecting the sun to the planet. An extreme example of elliptical satellite orbits are the
If we think of the sun as being at the origin of the orbits of some of the comets that periodically visit the
coordinate system, then the line from the sun to the sun. Halley's comet, for example, visits the sun once
planet is what we have been calling the coordinate every 76 years. The comet spends about 1 year in the
vector R . It is also called the radius vector R . Kepler's close vicinity of the sun, where it is visible from the
second law explicitly states that the radius vector R earth, and the other 75 years on the rest of its orbit which
sweeps out equal areas in equal times. goes out beyond the edge of the planetary system. The
To apply Kepler's second law to the output of our Orbit comet moves rapidly past the sun, and spends the
1 program, we note that we had the computer plot a majority of the 76 year orbital period creeping around
cross at equal times along the orbit. Thus the area swept the back side of its orbit where its radius vector is very
out by the radius vector should be the same as R moves long.
from one cross to the next. To check this prediction, we

B
C
R
D

A
R

E
Figure 30 F
Kepler’s Second Law. The radius vector R should sweep out equal areas in equal time.
8-28 Newtonian Mechanics

Exercise 15 Exercise 16
For both of your plots from Exercises 13 and 14, check Consider the example of a planet of mass mp in a
that the satellite's radius vector sweeps out equal areas circular orbit about the sun whose mass is Ms . Using
in equal times. Explicitly compare the area swept out Newton's second law and Newton's law of gravity, and
during a time interval where the satellite is in close to the the fact that for circular motion the magnitude of the
earth to an equal time interval where the satellite is far acceleration is v 2/ R , solve for the radius R of the orbit.
from the earth. Then use the fact that the period T is the distance 2 π R
divided by the speed v, and construct the ratio R3/ T2 .
This exercise requires that you measure the areas of All the variables except Ms should cancel and you
lopsided pie-shaped sections. There are a number of should get the result shown in Equation 20.
ways of doing this. You can, for example, draw the
sections out on graph paper and count the squares, you Exercise 17 (optional)
can break the areas up into triangles and calculate the
A more general statement of Kepler's third law, that
areas of the triangles, or you can cut the areas out of
applies to elliptical orbits, is that R3/ T2 is the same for
cardboard and weigh them.
all the planets, where R is the semi major axis of the
ellipse (as shown in Figure (31)). Check this prediction
Kepler's Third Law for the two elliptical orbits used in Exercises (13) and
(14). In both of those examples the satellite was orbiting
Kepler's third law states that the ratio of the cube of the the same earth, thus the ratios should be the same.
orbital radius R to the square of the period T is the same
ratio for all the planets. We can easily use Newton's
laws of gravity and motion to check this result for the
case of circular orbits. The result, which you are to
calculate in Exercise 16, is Semi major axis

R3 GMs
2
= 2
(20)
T 4π
Figure 31
where M s is the mass of the sun. In this calculation, the The semi major axis of an ellipse.
mass m p of the planet, the orbital radius R, the speed
v all cancelled, leaving only the sun mass M s as a
variable. Since all the planets orbit the same sun, this
ratio should be the same for all the planets.

When the planet is in an elliptical orbit, the length of the

radius vector R changes as the planet goes around the
sun. What Kepler found was that the ratio of R 3 /T2
was constant if you used the "semi major axis" for R.
The semi major axis is the half the maximum diameter
of the ellipse, shown in Figure (31). As an optional
Exercise (17), you can compare the ratio of R 3 /T2 for
the two elliptical orbits of Exercises (13) and (14),
using the semi major axis for R.
 8-29

MODIFIED GRAVITY start with Coulomb’s electric force law, include the
AND GENERAL RELATIVITY effects of special relativity, and find that magnetism is
After we have verified that the Orbit 1 program calcu- one of the essential consequences of this combination.
lates orbits that are in agreement with Kepler’s laws of Einstein’s relativistic theory of gravity is more com-
motion, we should be reasonably confident that the plex than the theory of electricity and magnetism, and
program is ready to serve as a laboratory for the study the new predictions of the theory are much harder to
of new phenomena we have not necessarily encoun- test. It turns out that Newtonian gravity accurately
tered before. To illustrate what we can do, we will describes almost all planetary motion we can observe
begin with a question that cannot be answered in the in our solar system. Einstein calculated that his new
lab. What would happen if we modified the law of theory of gravity should predict new observable effects
gravity? What, for example, would happen if we only in the case of the orbit of Mercury and in the
changed the universal constant G, or altered the expo- deflection of starlight as it passed the rim of the sun. In
nent on the r dependence of the force? With the 1917 Sir Arthur Eddington led a famous eclipse expe-
computer program, these questions are easily answered. dition in which the deflection of starlight past the rim
We simply make the change and see what happens. of the eclipsed sun could be observed. The deflection
These changes should not be made completely without predicted by Einstein was observed, making this the
first clear correction to Newtonian gravity detected in
thought. I have seen a project where a student tried to
250 years. Einstein’s real fame began with the success
observe the effect of changing the mass of the satellite.
After many plots, he concluded that the effect was not of the Eddington expedition.
great. That is not a surprising result considering the fact While Einstein set out to construct a theory of gravity
that the mass ms of the satellite cancels out when you consistent with special relativity, he was also im-
equate the gravitational force to m s a . pressed by the connection between gravity and space.
Because all projectiles here on the surface of the earth
One can also see that, as far as its effect on a satellite’s
have the same downward acceleration, if you were in
orbit, changing the universal constant G will have an
a sealed room you could not be completely sure whether
effect equivalent to changing the earth mass Me . Since
your room was on the surface of the earth, and the
Kepler’s laws did not depend particularly on what mass
downward accelerations were caused by gravity, or
our sun had, one suspects that Kepler’s laws should
whether you were out in space, and your room was
also hold when G or Me are modified. This guess can
accelerating upward with an acceleration g. These
easily be checked using the Orbit 1 program.
equivalent situations are shown in Figure (32).
Changing the r dependence of the gravitational force is
another matter. After developing the special theory of g
relativity, Einstein took a look at Newton’s theory of
gravity and saw that it was not consistent with the stationary elevator accelerating elevator
principle of relativity. For one thing, because the
Newtonian gravitational force is supposed to point to
the current instantaneous position of a mass, it should falling ball floating ball
be possible using Newtonian gravity to send signals
faster than the speed of light. (Think about how you g
might do that.)

From the period of time between 1905 and 1915

Einstein worked out a new theory of gravity that was gravity no gravity
consistent with special relativity and, in the limit of
Figure 32
slowly moving, not too massive objects, gave the same Equivalent situations. Explain why you would
results as Newtonian gravity. We will get to see how feel the same forces if you were sitting on the
this process works when, in the latter half of this text we floor of each of the two rooms.
8-30 Newtonian Mechanics

2
The equivalence between a gravitational force and an A 1/r force law is unique in that only for this
acceleration turned out to be the cornerstone of exponent, – 2, does the perihelion, the axis of the
Einstein’s relativistic theory of gravity. It turned out elliptical orbit, remain steady. For any other value of
that Einstein’s new theory of gravity could be inter- the exponent, the perihelion rotates or precesses one
preted as a theory of space and time, where mass way or another.
caused a curvature of space, and what we call gravita-
tional forces were a consequence of this curvature of It turns out that a number of effects can cause the
space. This geometrical theory of gravity, Einstein’s perihelion of a planet’s orbit to precess. The biggest
relativistic theory, is commonly called the General effect we have not yet discussed is the fact that there are
Theory of Relativity. a number of planets all orbiting the sun at the same time,
and these planets all exert slight forces on each other.
As they often say in textbooks, a full discussion of These slight forces cause slight perihelion precessions.
Einstein’s relativistic theory of gravity is “beyond the
scope of this text”. However we can look at at least one In the 250 years from the time of Newton’s discovery
of the predictions. As far as satellite orbit calculations of the law of gravity, to the early 1900s, astronomers
are concerned, we can think of Einstein’s theory as a carefully worked out the predicted orbits of the planets,
slight modification of the Newtonian theory. We have including the effects of the forces between the planets
seen that any modification of the factors G, ms or mein themselves. This work, done before the development
the Newtonian gravitational force law would not have of computers, was an extremely laborious task. A good
a detectable effect. The only thing we could notice is fraction of one’s lifetime work could be spent on a
some change in the exponent of r. single calculation.

With a few of quick runs of the Orbit 1 program, you

will discover that the satellite orbit is very sensitive to
the exponent of r. In Figure (33) we have changed the
exponent from – 2 to – 1.9. This simply requires
changing

G * ms * me R ∧ 2

to

G * ms* me R ∧ 1.9

in the formula for Fg . The result is a striking change

in the orbit. When the exponent is – 2, the elliptical
orbit is rock steady. When we change the exponent to
– 1.9, the ellipse starts rotating around the earth. This Figure 33
rotation of the ellipse is called the precession of the Planetary orbit when the gravitational
perihelion, where the word “perihelion” describes the force is modified to a 1 / r 1.9 force.
line connecting the two focuses of the ellipse.
 8-31

The orbit of the planet Mercury provided a good test of (The Eddington expedition verified more than just the
these calculations because its orbital ellipse is more fact that light is deflected by the gravitational attraction
extended than that of the other close-in planets. The of a star. You can easily construct a theory where the
more extended an ellipse, the easier it is to observe a energy in the light beam is related to mass via the
2
precession. (You cannot even detect a precession for formula E = mc , and then use Newtonian gravity to
a circular orbit.) Mercury’s orbit has a small but predict a deflection. Einstein’s General Relativity
observable precession. Its orbit precesses by an angle predicts a deflection twice as large as this modified
that is slightly less than .2 degrees every century. This Newtonian approach. The Eddington expedition ob-
is a very small precession which you could never detect served the larger prediction of General Relativity,
in one orbit. But the orbit of Mercury has been providing convincing evidence that General Relativity
observed for about 3000 years, or 30 centuries. That is rather than Newtonian gravity was the more correct
over a 5 degree precession which is easily detectable. theory of gravity.)

When measuring small angles, astronomers divide the

Exercise 18
degree into 60 minutes of arc, and for even smaller
angles, divide the minute into 60 seconds of arc. One Start with your Orbit 1 program, modify the exponent in
the gravitational force law, and see what happens.
second of arc, 1/3600 of a degree, is a very small angle.
Begin with a small modification so that you can see how
A basketball 30 miles distant subtends an angle of
to plot the results. (If you make a larger modification, you
about 1 second of arc. In these units, Mercury’s orbit will have to change the plotting window to get interest-
precesses about 650 seconds of arc per century. ing results.)
By 1900, astronomers doing Newtonian mechanics
calculations could account for all but 43 seconds of arc (To get the 43 seconds of arc per century precession of
per century precession of Mercury’s orbit as being Mercury’s orbit, using a modified gravitational force
caused by the influence of neighboring planets. The 43 law, the force should be proportional to 1/ r 2.00000016
seconds of arc discrepancy could not be explained. instead of 1/ r 2 .)
One of the important predictions of Einstein’s relativ-
istic theory of gravity is that it predicts a 43 second of
arc per century precession of Mercury’s orbit, a preces-
sion caused by a change in the gravitational force law
and not due to neighboring planets. Einstein used this
explanation of the 43 seconds of arc discrepancy as the
main experimental foundation for his relativistic theory
of gravity when he just presented it in 1915. The
importance of the Eddington eclipse expedition in
1917 is that a completely new phenomena, predicted
by Einstein’s theory, was detected.
8-32 Newtonian Mechanics

CONSERVATION OF angular momentum. In this demonstration the instruc-

ANGULAR MOMENTUM tor uses only one dumbbell. After a student assists the
With the ability to work with realistic satellite orbits instructor in getting his rotation started, the instructor
extends the dumbbell out to full arm’s reach, for
rather than just the circular orbits, we will be able to
instance, when he is facing the class, and pulls his arm
make significant tests of the laws of conservation of
angular momentum and of energy, as applied to satel- in when he is facing away as shown in Figure (34).
Some practice is needed to maintain this pattern and not
lite motion. In this section, we will first see how
lose one’s balance.
Kepler’s second law of planetary motion is a direct
consequence of the conservation of angular momen- The rather expected result of this demonstration is that
tum, and then do some calculations with the Orbit 1 the instructor rotates more slowly when his arm is far
program to see that a satellite’s angular momentum is out, and more rapidly when his arm is in close. If we
in fact conserved—does not change as the satellite goes associate the dumbbell with a satellite orbiting the
around the earth. In the next section we will first take earth, we see the same speeding up as the lever arm
a more general look at the idea of a conservation law, about the axis of rotation is reduced, and slowing down
and then apply this discussion to the conservation of as the lever arm is increased.
energy for satellite orbits.
A fairly simple geometrical construction demonstrates
Recall that Kepler’s second law of planetary motion that the rule about the radius vector sweeping out equal
states that a line from the sun to the planet, the radius areas in equal times is precisely what is required for
vector, sweeps out equal areas in equal times. For this conservation of angular momentum. In Figure (35a)
to be true when the planet is in an elliptical orbit, the we have plotted an elliptical satellite orbit showing the
planet must move faster when in close to the sun and the position of the planet for two different equal time
radius vector is short, and slower when far away and the intervals. The time intervals ∆t are short enough that
radius vector is long. we can fairly accurately represent the displacement of
the satellite by short, straight lines of length v1 ∆t in the
To intuitively see that this speeding up and slowing
upper triangle and v2 ∆t in the lower triangle. With this
down is a consequence of the conservation of angular
approximation we can represent the areas swept out by
momentum, one can modify the three dumbbell experi-
the radius vector by triangles as shown by the shaded
ment we used to demonstrate the conservation of
areas in Figure (35a).

equal areas
sun

r2
2
v2 ∆t

h=r
(a)

Figure 35
b = v∆t
(a) (b) Calculating the area swept out by the
planet during a short time interval ∆ t . (b)
Figure 34
One dumbbell experiment.
 8-33

Now the area of a triangle is one half the base times the As a direct check of the conservation of angular mo-
altitude. If you look at the lower triangle in Figure mentum in the satellite orbit program, note that if a
(35a), and take the side v2 ∆t as the base, then the particle is located a distance x from an axis of rotation
distance labeled r2⊥ is the altitude, as seen in the sketch and is moving in the y direction with a velocity vy as
in Figure (34b). Thus the area of the triangle at position shown in Figure (36a), the lever arm about the origin is
2 is simply x, and the particle’s angular momentum about
the origin a is
area swept
out at position 2 = 1 (base) × (altitude)
2 particle's angular
in a time ∆t a = mxvy momentumin Figure (36a)
(24)
(21)
= 1 (v2 ∆t) × r 2⊥
2 Using the right hand convention illustrated in Figure
(7-14), we see that this particle has angular momentum
When the satellite is at position 2 in Figure (35a),
directed up, out of the paper. We will call this positive
moving at a velocity v2, the distance of closest ap-
angular momentum. (You can think of m as a small
proach if it continued at the same velocity v2 would be
piece of the bicycle wheel shown in Figure 7-14.)
the distance r2⊥ . Thus r2⊥ is the “lever arm” for the
motion of the satellite at this point in the orbit. Now consider a particle of mass m located a distance y
from the origin traveling in the – x direction as shown
We get a similar formula for the area of the triangle at
in Figure (36b). By the right hand convention the
position 1. Using Kepler’s second law which says that
angular momentum is still positive (you could think of
these areas should be equal for equal times ∆t, we get
this m as another part of the same bicycle wheel), but
1 1 the x velocity is now negative. Thus the formula for this
v1 ∆t r1⊥ = v2 ∆t r2⊥ (22) particle’s angular momentum is
2 2

Dividing Equation 22 through by ∆t and multiplying b = – myvx (25)

both sides by 2m, where m is the mass of the satellite,
We have to put in the minus (–) sign to counteract the
gives
fact that vx is negative but b is positive.
m 1v1 r1⊥ = m 2v2 r2⊥ (23)
It turns out that if a particle is in the xy plane at some
arbitrary position R = (x,y) , and has some arbitrary
Recall that the definition of a particle’s angular mo-
velocity v = (v x,vy) in the xy plane, then the formula
mentum about some axis is the linear momentum
for the angular momentum 0 of the particle about the
p = mv times the lever arm r⊥ (see Equations 7–15,
origin is
16). Thus the left side of Equation 23 is the satellite’s
angular momentum at position 1, the right side at o = m xvy – yvx (26)
position 2. The statement that the satellite sweeps out y
equal areas in equal times is thus equivalent to the vx
statement that the satellite’s angular momentum mvr⊥ y m
has the same value all around the orbit. Like the vy y
dumbbell in Figure (34), the satellite moves faster
when r⊥ is small, and slower then r⊥ is large, in order m x x
to conserve angular momentum. 0 x 0

Figure 36a Figure 36b

Here = mx vy . Here = my vx.
8-34 Newtonian Mechanics

You can see that this general result is just a combina- Exercise 19
tion of the two special cases we considered in Figures
Add lines (27) and (28) to your Orbit 1 program and
(36) and Equations 24 and 25. (Equation 26 also comes
check that angular momentum is conserved. Use
from the formula = m r × v where r × v is the vector several different initial conditions so that you can check
cross product of r and v. We will discuss vector cross conservation of angular momentum for different ellipti-
products in detail later in Chapter 11. For now Equa- cal orbits. (Make sure that Lz is calculated within the
tion 26 is all we need.) calculational loop so that the latest values of Rx, Ry, Vx
and Vy are used for each calculation.) Also, if you set the
With Equation 26, we can easily test whether angular satellite mass m equal to 1, the values for Lz will be
momentum is in fact conserved in our satellite orbit easier to interpret. (The value of the constant m does not
calculations. By the end of the calculational loop, we matter since you are simply checking that Lz is constant
have already calculated new values of the satellite’s x during the satellite’s orbital motion.)
and y coordinates R x and R y , and x and y velocity
components vx and vy . Thus to calculate the satellite’s
Exercise 20
angular momentum, all we need is the line
The fact that angular momentum is conserved in Exer-
LET Lz = M * Rx*Vy – Ry*Vx (27) cise 19 should not be too surprising because you have
already checked in earlier exercises that the elliptical
where we are using the name L z because we are orbit obeys Kepler’s second law, and as we have just
observing the z component of the satellite’s angular seen, Kepler’s second law implies conservation of
momentum, as indicated in Figure (37). angular momentum. In this exercise, see if angular
momentum is also conserved if we modify the gravita-
To check that angular momentum is conserved, we tional force law as we did in Exercise 19. Take your
could add a print line at the end of the calculational loop program from Exercise 19, the one that prints out the
like values of the angular momentum, change the exponent
of r in Newton’s law of gravity, and see if angular
IF MOD (I, 40) = 0 THEN PRINT Rx, Ry, Lz momentum is conserved while the ellipse is precessing.
(28)
By printing the values of R x and R y as well as L z , we
can see where the satellite is in its orbit as well as the
value of the angular momentum at that point.

x rotating
wheel
Figure 37
Angular momentum vector of a rotating wheel.
 8-35

CONSERVATION OF ENERGY balance, there is a convention that if your balance is

In addition to angular momentum, there is another positive, the bank owes you money, and if the balance
quantity that is conserved during a satellite’s orbital is negative, you owe the bank money. A zero balance
motion. In Chapter 10, which is completely devoted to indicates that neither owes each other anything. If the
the topic of energy, we will discuss techniques for bank is not worried about your credit, it does not make
deriving formulas for various forms of energy. But it much difference whether your balance is positive,
is not necessary to be able to derive energy formulas in negative or zero, you can still write checks, make
order to be able to appreciate and use the concept. deposits, and go about your normal business.

The fundamental idea behind the concept of energy is In the way we deal with energy, what we call the zero
that energy is a conserved quantity. To study the of energy does not make much difference either. We
conservation of energy is often a more difficult job than can think of a power company borrowing energy from
studying the conservation of linear or angular momen- a coal company just as it borrows money from a bank.
tum, because there are many forms that energy can In this sense the power company can have a negative
take, and not all the forms are easy to recognize. But in energy balance just as it has a negative bank balance.
certain simple examples like the motion of an earth The fact that energy is conserved means that the power
satellite, there are only two forms of energy we have to company cannot create energy out of nothing to repay
deal with, and the conservation of energy is easy to the debt. The difference between the power company
observe. and physical systems like satellites in orbit is that we let
power companies pay their energy debt with cash, a
Unlike linear and angular momentum, energy does not physical system can increase its energy balance only by
point anywhere. Energy is represented by a number, getting energy from somewhere else.
not a vector. You get a bill from your electric company
for the amount of electrical energy you used the previ- In our accounting scheme for energy, some terms are
ous month. The electric company has a formula, based positive and some are negative. The term called kinetic
on the reading of your electric meter, for the amount of energy is always positive. In most circumstances,
electrical energy you used. Because energy is con- kinetic energy is given by the formula 1/2 mv2 where
served, the power company could not create the energy m is the mass of the object and v the object’s speed.
they sold you out of nothing, they probably got the Kinetic energy is positive because neither m or
energy either from a nuclear power plant or by burning 1/2 m v2 can become negative.
fossil fuels. If they got the energy from fossil fuels, that To observe conservation of energy for satellite motion,
energy originally came from the sun, from the combin- it is necessary to account for two forms of energy. One
ing of hydrogen nuclei to form helium nuclei. If the is kinetic energy 1/2 v2, the other is what is called
electricity came from a nuclear power plant, the energy gravitational potential energy. Our formula for gravi-
came from the splitting of large uranium or plutonium tational potential energy will be – Gmsme/r where G
nuclei into smaller nuclei. The uranium and plutonium is the gravitational constant, ms and me the masses of
nuclei were formed by getting their energy from a the satellite and earth respectively, and r the separation
supernova explosion that must have occurred over between them. This formula looks much like the
five billion years ago. gravitational force formula, except that it is propor-
2
In our discussion of energy in Chapter 10, we will see tional to 1/r rather than 1/r .
that there is a close analogy between keeping track of
your checkbook balance in a bank and keeping a record
of the amount of energy a system has. With a bank
8-36 Newtonian Mechanics

What is often upsetting to students when they first Exercise 21

encounter the gravitational potential energy formula is Using the steps described above, check that the
the minus sign. How can energy be negative? This is satellite’s total energy Etot is conserved. (You will notice
essentially a result of our accounting procedure. The slight variations in the value of Etot, the values are not as
important feature of energy is that it is conserved. If the steady as they were in the printout of angular momen-
gravitational potential energy in some part of an orbit tum. Exercise 22 suggests a way of improving the
becomes more negative, then the kinetic energy has to energy calculation and getting better results.)
become more positive so that the total is conserved, i.e.,
stays constant. As far as energy conservation is con- As a variation, print out the values of the kinetic energy,
cerned, it does not make any difference what the total potential energy and Etot. You will see big changes in
the kinetic and potential energy, while the sum Etot
energy is, as long as it is constant.
remains nearly constant. Start the satellite with different
At this point we have made no effort to explain where initial conditions and check for energy conservation for
the formulas 1/2 mv2 for kinetic energy and different elliptical orbits.
–Gm s m e /r for gravitational potential energy came
from. That is a subject for Chapter 10. What we are Exercise 22
concerned with now is to see if the Total Energy, the We can obtain a more accurate calculation of the
sum of these two, is conserved as the satellite moves satellite’s total energy by slightly modifying the value
around its orbit. of v used in the kinetic energy formula. When we put
the calculation of Etot at the end of the calculational
total energy gravitational loop, we are using the value of v at the end of the time
of a satellite = kinetic
energy + potential step dt. It turns out that we get a more accurate
in orbit energy
energy calculation if we use a value of v that is the
(29)
G m sm e average of the value we had when we entered the
E tot = 1 mv 2 – r calculational loop and the value a time dt later when
2
we left. This averaging is easily accomplished using
We will check for conservation of energy in much the the following commands inserted into your calcula-
same way we checked for conservation of angular tional loop.
momentum using our Orbit 1 program. Near the end LET Vold = V new line saving old value of v
of the calculational loop, after we have calculated the LET Vx = . . .
latest values of the satellite position r and velocity v, your old lines calculating
LET Vy = . . .
and have also calculated the corresponding magni- the next new value of v
tudes r and v, we can add the line LET V = SQR ( Vx * Vx + Vy * Vy)
LET Vnew = V saving the new value of V
LET Etot = Ms*V*V/2 – G*Ms*Me/R (30) LETV = ( Vold+ Vnew ) /2 setting V to the average value
LET Etot = ( Ms * V * V )/2 – G * Ms * Me/R
Then we can add a print line like
The steps above using the average of Vnew and Vold for
IF MOD(I, 40) = 0 THEN PRINT Rx, Ry, Etot V in the calculation of the kinetic energy represents the
kind of specialized computer trick we have tried to avoid
By looking at the printed values of E tot we can see in this text. However, the trick works so well, the
whether this formula for E tot is conserved as the improvement in the value of the total energy is so great
satellite moves around. that it is worth the effort. This is particularly true for
project work where a check for conservation of energy
is the main check of the validity of the calculation. (You
can usually spot computer errors by printing out the total
energy, because computer errors almost never con-
serve energy.
 8-37

Exercise 23 (optional, more like a project)

It turns out that if we modify the formula for the
gravitational force, for example changing the expo-
nent of r from + 2 to – 1.9, we also have to modify the
formula for the gravitational potential energy in order
to observe energy conservation. You will learn in
Chapter 10 that the formula for the gravitational
potential energy is the integral of the magnitude of the
force. We can, for example, obtain our formula for
gravitational potential energy from the gravitational
force formula by the following integration
r
Gmsme Gms me
2 dr = – r (31)
r
∞
If you modify the gravitational force formula, you can do
the same kind of integration to get the corresponding
potential energy formula. (In Chapter 10 we will have a
lot more to say about this integration. For now you can
treat the integration as a convenient device for obtain-
ing the potential energy formula. Since the important
feature of energy is that it is conserved, if you find from
running your Orbit 1 program that the total energy turns
out to be conserved, you know you have the correct
potential energy formula no matter how it was derived.)

For this exercise, start by modifying the gravitational

force law by changing the exponent of r from + 2 to – 1.9.
Then run your Orbit 1 program using the formula
– Gm s m e /r for potential energy to see that this formula
does not work. (Use the accurate version of the
program from Exercise 22 so that you can be more
confident of the results.)

Then integrate Gm s m e /r1.9 to find a new potential

energy formula. See if energy is conserved with your
new formula. Once this is successful, try some other
modification.
Chapter 9
Applications of
Newton’s Second
Law

CHAPTER 9 APPLICATIONS OF
NEWTON’S SECOND LAW

In the last chapter our focus was on the motion of

planets and satellites, the study of which historically
lead to the discovery of Newton’s law of motion and
gravity. In this chapter we will discuss various appli-
cations of Newton’s laws as applied to objects we
encounter here on earth in our daily lives. This chapter
contains many of the examples and exercises that are
more traditionally associated with an introductory
physics course.
9-2 Applications of Newton’s Second Law

ADDITION OF FORCES In other words the vector ma , the ball’s mass times its
The main new concepts discussed in this chapter are acceleration, is equal to the vector sum of the forces
how to deal with a situation in which several forces are acting upon it. More formally we can write this
acting at the same time on an object. We had a clue for statement in the form
how to deal with this situation in our discussion of
projectile motion with air resistance, where in Figure the vector sum of more general
(1) reproduced here, we saw that the acceleration a of
the Styrofoam projectile was the vector sum of the
ma = Σ Fi = the forces acting form of Newton's
i on the object second law
acceleration g produced by gravity and the accelera-
tion a air produced by the air resistance (4)

a = g + a air Equation 4 forms the basis of this chapter. The basic

(1) rule is that, to predict the acceleration of an object, you
If we multiply Equation 1 through by m, the mass of the first identify all the forces acting on the object. You
ball, we get then take the vector sum of these forces, and the result
is the object’s mass m times its acceleration a .
ma = mg + ma air (2) When we begin to apply Equation 4 in the laboratory,
we will be somewhat limited in the number of different
We know that mg is the gravitational force acting on
forces that we can identify. In fact there is only one
the ball, and it seems fairly clear that we should identify
ma air as the force Fair that the air is exerting on the ball. force for which we have an explicit and accurate
formula, and that is the gravitational force mg that acts
Thus Equation 2 can be written
on a mass m. Our first step will be to identify other
forces such as the force exerted by a stretched spring,
ma = Fg + Fair so that we can study situations in which more than one
force is acting.
3

ind"
a3 "w
g
a air
Fs

m
v3

a 3 = g + a air

mg
Figure 2
Figure 1
Spring force balanced by the gravitational force.
Vector addition of accelerations.
 9-3

SPRING FORCES another 50 gm mass, the spring stretched to a length of

The simplest way to study spring forces is to suspend 54.8 centimeters. We added up to five 50 gram masses
a spring from one end and hang a mass on the other as and plotted the results shown in Figure (4).
shown in Figure (2). If you wait until the mass m has Looking at the plot in Figure (4) we see that the points
come to rest, the acceleration of the mass is zero and lie along a straight line. This means that the spring force
you then know that the vector sum of the forces on m is linearly proportional to the distance the spring has
is zero. In this simple case the only forces acting on m been stretched.
are the downward gravitational force mg and the
upward spring force Fs . We thus have by Newton’s To find the formula for the spring force, we first draw
second law a line through the experimental points and note that the
line crosses the zero force axis at a length of 35.9 cm.
Σ Fi = Fs + mg = ma = 0 (5) We will call this distance the unstretched length So .
i Thus the distance the spring has been stretched is
S – So , and the spring force should be linearly pro-
and we immediately get that the magnitude Fs of the portional to this distance. Writing the spring force
spring force is equal to the magnitude mg of the formula in the form
gravitational force.

As we add more mass to the end of the spring, the spring

Fs = k S – So
stretches. The fact that the more we stretch the spring, (6)
the more mass it supports, means that the more we
stretch the spring the harder it pulls back, the greater Fs all we have left is determine the spring constant k.
becomes.
Mass (in Grams)
To measure the spring force, we started with a spring 300
suspended from a nail and hung 50 gm masses on the
end, as shown in Figure (3). With only one 50 gram
250
mass, the length S of the spring, from the nail to the
hook on the mass, was 45.4 cm. When we added
200 (200 gm, 73.7 cm)

150

Fs = K(S – S0)
100
S

50

Length of spring S

50 gm 20 40 60 80 100 120 cm
S0 = 35.9cm
50 gm Figure 4
Figure 3
Plot of the length of the spring as
Calibrating the spring force.
a function of the force it exerts.
50 gm
9-4 Applications of Newton’s Second Law

The easy way to find the value of k is to solve Equation The Spring Pendulum
6 for k and plug in a numerical value that lies on the The spring pendulum experiment is one that nicely
straight line we drew through the experimental points. demonstrates that an object’s acceleration is propor-
Using the value Fs = 200 gm × 980 cm /sec tional to the vector sum of the forces acting on it . In this
= 19.6 × 10 4 dynes when the spring is stretched to a experiment, shown in Figure (5), we attach one end of
distance S = 73.7 cm gives a spring to a nail, hang a ball on the other end, pull the
ball back off to one side, and let go. The ball loops
Fs 19.6 × 10 4 dynes around as seen in the strobe photograph of Figure (6).
k = =
S – So 73.7 – 35.9 cm The orbit of the ball is improved, i.e., made more open
and easier to analyze, if we insert a short section of
dynes
= 5.18 × 10 3 cm
string between the end of the spring and the nail, as
indicated in Figure (5).
Equation 6, the statement that the force exerted by a
spring is linearly proportional to the distance the spring This experiment does not appear in conventional text-
is stretched, is known as Hooke’s law. Hooke was a books because it cannot be analyzed using calculus—
contemporary of Isaac Newton, and was one of the first there is no analytic solution for this motion. But the
to suspect that gravitational forces decreased as 1/r2. analysis is quite simple using graphical methods, and a
There was a dispute between Hooke and Newton as to computer can easily predict this motion. The graphical
who understood this relationship first. It may be more analysis most clearly illustrates the point we want to
of a consolation award that the empirical spring force make with this experiment, namely that the ball’s
“law” was named after Hooke, while Newton gets acceleration is proportional to the vector sum of the
credit for the basic gravitational force law. forces acting on the ball.

Hooke’s law, by the way, only applies to springs if you In this experiment, there are two forces simultaneously
do not stretch them too far. If you exceed the “elastic acting on the ball. They are the downward force of
limit”, i.e., stretch them so far that they do not return to gravity Fg = mg , and the spring force Fs . The spring
the original length, you have effectively changed the force Fs always points back toward the nail from which
spring constant k. the spring is suspended, and the magnitude of the

nail

string

spring

ball

Figure 5 Figure 6
Experimental setup. Strobe photograph of a spring pendulum.
 9-5

spring force is given by Hooke’s law Fs = k S – So . spring, is 93.0 cm. You can check this for yourself by
Since we can calibrate the spring before the experiment marking off the distance from the edge of the ball to the
to determine k and So , and since we can measure the nail on a piece of paper, and then measuring the
distance S from a strobe photograph of the motion, we separation of the marks using the graph paper (as we
can determine the spring force at each position of the did back in Figure (1) of Chapter 3). We measure to the
ball in the photograph. edge of the ball and not the center, because that is where
the spring ends, and in calibrating the spring we mea-
In Figure (7) we have transferred the information about sured the distance S to the end of the spring. (If we
the positions of the ball from the strobe photograph to measured to the center of the ball, that would introduce
graph paper and labeled the first 17 positions of the ball
from – 1 to 15. Consider the forces acting on the ball Nail (50,130)
when it is located at the position labeled 0. The spring
force Fs points from the ball up to the nail which in this
photograph is located at a coordinate (50, 130). The
distance S from the hook on the ball to the nail, the
distance we have called the stretched length of the 93.0 cm

Experimental Coordinates 10 20
-1) ( 91.1, 63.1) 0 10 30 40 50 60 70 80 90 100
0) ( 88.2, 42.8)
1) ( 80.2, 24.4) 9 11
2) ( 68.0, 12.0)
3) ( 52.9, 8.6) 90 90
4) ( 37.4, 14.7)
5) ( 24.0, 28.8) 8
6) ( 14.2, 47.5) 80
12 80
7) ( 9.0, 67.0)
8) ( 8.2 , 83.9)
9) ( 11.1, 95.0) 70 70
10) ( 16.7, 98.8)
11) ( 23.9, 94.1)
7
12) ( 32.2, 81.5)
13 –1
60 60
13) ( 41.9, 62.1)
14) ( 52.1, 39.9)
15) ( 62.2, 19.4)
16) ( 70.3, 6.0) 50 50
17) ( 75.3, 2.8) 6
40 14 0 40

30 30
5
1
20 15 20

Figure 7
Spring pendulum 10 4 10
data transferred 2
to graph paper. 3
0 10 20 30 40 50 60 70 80 90 100
9-6 Applications of Newton’s Second Law

an error of about 1.3 cm, which produces a noticeable The vector addition is done graphically in Figure (8a),
error in our results.) giving us the total force acting on the ball when the ball
is located at position 0. On the same figure we have
It turns out that we used the same spring in our repeated the steps discussed above to determine the
discussion of Hooke’s law as we did for the strobe total force acting on the ball when the ball is up at
photograph in Figure (6). Thus the graph in Figure (4) position 10. Note that there is a significant shift in the
is our calibration curve for the spring. (The length of total force acting on the ball as it moves around its orbit.
string added to the spring is included in the unstretched
length So). Using Equation 6 for the spring force, we According to Newton’s second law, it is this total force
get Ftotal that produces the ball’s acceleration a. Explicitly
the vector m a should be equal to Ftotal . To check
Fs = k S – S o Newton’s second law, we can graphically find the
ball’s acceleration a a at any position in the strobe
dynes photograph, and multiply the mass m to get the vector
= 5.18 × 10 3 cm × 93.0 – 35.9 cm
m a.
= 29.6 × 10 4dynes (8) In Figure (8b) we have used the techniques discussed
The direction of the spring force is from the ball to the in Chapter 3 to determine the ball’s acceleration vector
nail. Using a scale in which 104 dynes = 1 graph paper a∆t2 at positions 0 and 10. (Recall that for graphical
square, we can draw an arrow on the graph paper to work from a strobe photograph, we had
represent this spring force. This arrow, labeled Fs , a = (s 2 – s 1)/∆t2 or a∆t2 = (s 2 – s 1) , where s 1 is
starts at the center of the ball at position 0, points toward the previous and s 2 the following displacement vec-
the nail, and has a length of 29.6 graph paper squares. Nail (50,130)

Throughout the motion, the ball is subject to a gravita-

tional force Fg which points straight down and has a 93.0 cm

magnitude mg. For the strobe photograph of Figure

Fs
(6), the mass of the ball was 245 grams, thus the 10 40 50 60 70 80 90 100

gravitational force has a magnitude

90
Fg = mg = 245 gm × 980 cm2
sec mg Ftot 80

= 24.0 × 10 dynes
4 (9)
70 70
Fs
This gravitational force can be represented by a vector –1
60 60

labeled mg that starts from the center of the ball at

50 50
position 0, and goes straight down for a distance of 24
graph paper squares (again using the scale 1 square = Ftot 0
40 40
104 dynes.)
30 30
The total force Ftotal is the vector sum of the individual mg
1
forces Fs and Fg 20 20

10 10
Ftotal = Fs + mg (10)
0 10 20 30 40 50 60 70 80 90 100

Figure 8a
Force vectors.
 9-7

tors and ∆t the time between images.) At position 0 the as the direction of the vector a ∆t 2 at position 0 in
vector a ∆t 2 has a length of 5.7 cm as measured directly Figure (8b). In a similar way we have constructed the
from the graph paper. Since ∆t = .1sec for this strobe vector ma at position 10 .
photograph, we have, with ∆t 2 = .01,
As a comparison between theory and experiment, we
a ∆t 2 = a × .01 = 5.7 cm have drawn both the vectors Ftotal and ma at positions
0 and 10 in Figure (8c). While the agreement is not
a = 5.7 cm2 = 570 cm2 exact, it is the best we can expect, considering the
.01 sec sec accuracy with which we can read the strobe photo-
graphs. The important result is that the vectors Ftotal
Using the fact that the mass m of the ball is 245 gm, we and ma can be seen to closely follow each other as the
find that the length of the vector ma at position 0 is ball moves around the orbit. (In Exercise 1 we ask you
to compare Ftotal and ma at a couple of more positions
cm to see these vectors following each other.) (I once
ma at position 0
= 245 gm × 570 showed a figure similar to Figure (8c) to a mathemati-
sec2 cian, who observed the slight discrepancy between the
gm cm vectors Ftotal and ma and said, “Gee, it’s too bad the
= 14.0 × 104
sec2 experiment didn’t work.” He did not have much of a
= 14.0 × 104 dynes feeling for experimental errors in real experiments.)

Exercise 1
In Figure (8c) we have plotted the vector m a at position
0 using the same scale of one graph paper square Using the data for the strobe photograph of Figure (6),
4 as we have been doing above, compare the vectors
equals 10 dynes . Since m a has a magnitude of
4
14.0 × 10 dynes , we drew an arrow 14.0 squares Ftotal and ma at two more locations of the ball.
long. The direction of the arrow is in the same direction

10 10 40 50 60 70 80 90 100
40 50 60 70 80 90 100

9 11
a 10 ∆t 2 90 90

80 80
Ftot
ma
70
m a 10 70 70 70

–1
60 60 60 60
a 0 ∆t 2
50 50
ma 50
ma0 50

40
0 40 40
0 40
Ftot

30 30 30 30

1 20 20
20 20

10 10 10 10

0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100

Figure 8b Figure 8c
Acceleration vectors. Comparing Ftotal and ma .
9-8 Applications of Newton’s Second Law

Computer Analysis To apply this general structure to the spring pendulum

of the Ball Spring Pendulum problem, we first have to be able to describe the
It turns out that using the computer you can do quite a direction of the spring force Fs . This is done using the
good job of predicting the motion of the ball bouncing vector diagram of Figure (9). The vector Z represents
on the end of the spring. A program for predicting the the coordinate of the nail from which the spring is
motion seen in Figure (7) is listed in the appendix of this suspended, S the displacement from the nail to the ball,
chapter. Here all we will discuss are the essential and R the coordinate of the ball. From Figure (9) we
features that you will find in the calculational loop of immediately get the vector equation
that program.
Z+S = R (12)
The main features of any program that predicts the
motion of an object are the following lines, written out which we can solve for the spring length S
in English
S = R–Z (13)
! Calculational Loop From Figure (7) we see that the nail is located at the
coordinate (50,130) thus
Let R new = R old + Vold * dt

Let F1 = . . .
find forces acting Z = (50,130)
on the object
Throughout the motion of the ball, the spring force
Let F2 = . . .
points in the – S direction as indicated in Figure (10),
find the vector
Let Ftotal = F1 + F2 + . . . sum of the forces
thus the formula for the spring force can be written

Let a = Ftotal/m Newton's Fs = –S k S – So (14)

second law

Let Vnew = Vold + a * dt where k S – So is the magnitude of the spring force

Loop Until . . . ! Repeat calculation determined in Figure (4).
(11)

nail at (50,130) nail

S
Fs
Z

R mg

R = Z+ S
Figure 9
Vector diagram. Z is the coordinate Figure 10
of the nail, R the coordinate of the Force diagram, showing the two
ball, and S the displacement of the forces acting on the ball.
ball from the nail.
 9-9

Using Equation 6 for the spring force, the English The rest of the program, discussed in the Appendix, is
calculational loop for the spring pendulum becomes much like our earlier projectile motion programs, with
a new calculational loop. In Figure (11) we have
plotted the results of the spring pendulum program,
! Calculational loop for spring pendulum
where the crosses represent the predicted positions of
Let R new = R old + Vold * dt the ball and the squares are the experimental positions.
If you slightly adjust the initial conditions for the
Let S = R – Z
motion of the ball, you can make almost all the crosses
Let Fs = – S * k * S – S o fall within the squares. How much adjustment of the
initial conditions you have to do gives you an indication
Let Fg = mg of the size of the errors involved in determining the
Let Ftotal = Fs + Fg positions of the ball from the strobe photograph.

Let A = Ftotal / m Analytic Solution

If you pull the ball straight down and let go, the ball
Loop Until . . . (15) bounces up and down in a periodic motion that can be
analyzed using calculus. The resulting motion is called
A translation into BASIC of the lines for calculating S a sinusoidal oscillation which we will discuss in
and Fs would be, for example, considerable detail in Chapter 14. You will see that if
you can use calculus to obtain an analytic solution,
LET Sx = Rx – Zx there are many ways to use the results. The oscillatory
LET Sy = Ry – Zy spring motion serves as a model for describing many
phenomena in physics.
LET S = SQR (Sx * Sx + Sy * Sy)
LET Fsx = (– Sx / S) * k * (S – So)
LET Fsy = (– Sy / S) * k * (S – So) (16)

Figure 11
Output from the ball spring
program. The crosses are the
points predicted by the computer
program, while the black squares
represent the experimental data
points. The program in the
Appendix illustrates how the data
points can be plotted on the same
diagram with your computer plot.
9-10 Applications of Newton’s Second Law

THE INCLINED PLANE What makes the analysis of this problem different from
Galileo discovered the formulas for projectile motion the motion of the spring pendulum discussed in the last
by using an inclined plane to slow the motion down, section is the fact that the cart is constrained to move
making it easier to measure positions and velocities. He along the air track. This tells us immediately that the
studied rolling balls, whereas we wish to study sliding cart accelerates along the track, and has no acceleration
objects using a frictionless inclined plane. perpendicular the track. If there is no perpendicular
acceleration, there must be no net force perpendicular
The frictionless inclined plane was more or less a to the track. From this fact alone we can determine the
figment of the imagination of the authors of introduc- magnitude of the force Fp exerted by the track.
tory textbooks, at least until the development of the air
track. And even with an air track some small effects of Before we do any calculations, let us set up the problem
friction can be observed. We will discuss the inclined in such a way that we can take advantage of our
plane here because it illustrates a useful technique for knowledge that the cart moves only along the track.
analyzing the forces on an object, and because it leads Without thinking, we would likely take the x axis to be
to some interesting laboratory experiments. in the horizontal direction and the y axis in the vertical
direction. But with this choice the cart has a component
As a simple experiment, place a book, a floppy disk, or of velocity in both the x and y directions.
some small object under one end of an air track so that
the track is tilted at an angle θ as shown in Figure (12). The analysis is greatly simplified if we choose one of
If you keep the angle θ small, you can let the air cart the coordinate axes to lie along the plane. In Figure
bounce against the bumper at the end of the track (14), we have chosen the x axis to lie along the plane,
without damaging anything. and decomposed the downward gravitational force
into an x component which has a magnitude mgsin θ
To analyze the motion of the air cart, it helps to and a – y component of magnitude mgcosθ .
exaggerate the angle θ in our drawings of the forces
involved as we have done this in Figure (13). The first Now the analysis of the problem is easy. Starting with
step in handling any Newton’s law problem is to Newton’s law in vector form, we have
identify all the forces involved. In this case there are
two forces acting on the air cart; the downward force of ma = Σ Fi = mg + Fp (17)
gravity mg and the force Fp of the plane against the
cart. Separating Equation 17 into its x and y components, we
get
The main feature of a frictionless surface is that it can
exert only normal forces, i.e., forces perpendicular to
the surface. (Any sideways forces are the result of
Fp t
“friction”.) Thus Fp is perpendicular to the air track, car
inclined at an angle θ away from the vertical direction.
air cart m

m
air track
θ
θ
Figure 12 Figure 13 mg
Tilted air cart. Forces on the air cart.
 9-11

max = mg = mgsin θ (18a) Exercise 2

x
A one meter long air track is set at an angle of
θ = .03 radians . (This was done by placing a 3 millimeter
may = 0 thick floppy disk under one end of the track.
= Fp – mg y (a) From your knowledge of the definition of the radian,
explain why, to a high degree of accuracy, the sin θ and
= Fp – mgcosθ (18b) θ are the same for these small angles.

where we set ay = 0 because the cart moves only in the (b) The cart is released from rest at one end of the track.
x direction. How long will it take to reach the other end. (You can
consider this to be a review of the constant acceleration
From Equation 18b we immediately get formulas.)

Fp = mgcosθ (19)

as the formula for the magnitude of the force the plane

exerts on the cart.

Of more interest is the formula for ax which we imme-

diately get from Equation 18a

ax = g sin θ (20)

We see that the cart has a constant acceleration down

the plane, an acceleration whose magnitude is equal to
the acceleration due to gravity, but reduced by a factor
sin θ . It is this reduction that slows down the motion, Portrate of Galileo
and allowed Galileo to study motion with constant
acceleration using the crude timing devices available to
him at that time.

Fp

m
θ
cos
mg
θ θ
x mg Galileo’s Inclined plane

mg sin θ Above photos from the informative web page

Figure 14
http://galileo.imss.firenze.it/museo/b/egalilg.html
Choosing the x axis to lie along the plane.
9-12 Applications of Newton’s Second Law

FRICTION Friction, however, plays too important a role in our

If you do the experiment suggested in Exercise 2, lives to be ignored. Remember the first time you tried
measuring the time it takes the cart to travel down the to skate and did not have a surface with enough friction
track when the track is tilted by a very small angle, the to support you. To handle friction, a number of
results are not likely to come out very close to the empirical rules have been developed. One of the more
prediction. The reason is that for such small angles, the useful rules is “if it squeaks, oil it”. At a slightly higher
effects of “friction” are noticeable even on an air track. level, but not much, are the formulas for friction that
appear in introductory physics text books. Our lack of
In introductory physics texts, the word “friction” is respect for these formulas comes from the experience
used to cover a multitude of sins. With the air track, of trying to verify them in the laboratory. There is some
there is no physical contact between the cart and track. truth to them, but the more accurately one tries to verify
But there are air currents that support the cart and come them, the worse the results become. With this state-
out around the edge of the cart. These air currents ment in mind about the friction formulas, we will state
usually slow the air cart down, giving rise to what we them, and provide one example. Hundreds of ex-
might call friction effects. amples of problems involving friction formulas can be
found in other introductory texts.
In common experience, skaters have as nearly a fric-
tionless surface as we are likely to find. The reason that
you experience little friction when skating is not be- Inclined Plane with Friction
cause ice itself is that slippery, but because the ice melts In our analysis of the air cart on the inclined track, we
under the blade of the skate and the skater travels along mentioned that a frictionless surface exerts only a
on a fine ribbon of water. The ice melts due to the normal force on an object. If there is any sideways
pressure of the skate against the ice. Ice is a peculiar force, that is supposed to be a friction force Ff . In
substance in that it expands when it freezes. And Figure (15) we show a cart on an inclined plane, with
conversely, you can melt it by squeezing it. If, how- a friction force Ff included. The normal force Fn is
ever, the temperature is very low, the ice does not melt perpendicular to the plane, the friction force Ff is
at reasonable pressures and is therefore no longer parallel to the plane, and gravity still points down.
slippery. At temperatures of 40° F below zero, roads
on ice in Alaska are as safe to drive on as paved roads.
Fn
When two solid surfaces touch, the friction between Ff
them is caused by an interaction between the atoms in
the surfaces. In general, this interaction is not under- m
stood. Only recently have computer models shed some
light on what happens when clean metal surfaces
interact. Most surfaces are quite “dirty” at an atomic θ
scale, contaminated by oxides, grit and whatever. It is
unlikely that one will develop a comprehensive theory mg
of friction for real surfaces. Figure 15
Friction force acting on the cart.
 9-13

To analyze the motion of the cart when acted on by a Coefficient of Friction

friction force, we write Newton’s second law in the To go any further than Equation 24, we need some
usual form values for the magnitude of the friction force Ff . It is
traditional to assume that Ff is proportional to the force
ma = ΣFi = mg + Fn + Ff (21) Fn between the surfaces. Such a proportionality can be
written in the form
The only change from Equation 13 is that we have
added in the new force Ff . Ff = µFn (25)
Since the motion of the cart is still along the plane, it is where the proportionality constant µ is called the
convenient to take the x axis along the plane as shown coefficient of friction.
in Figure (16). Breaking Equation 21 up into x and y
components now gives Equation 25 makes the explicit assumption that the
friction force does not depend on the speed at which the
ma x = ΣFx = mg sin θ – Ff (22a) object is moving down the plane. But it is easy to show
that this is too simple a model. It is harder to start an
may = ΣFy = –mg cos θ + Fn = 0 (22b) object sliding than to keep it sliding. This is why you
should not jam on the brakes when trying to stop a car
From 22b we get, suddenly. You should keep the tires rolling so that there
Fn = mg cos θ (23) is no sliding between the surface of the tire and the
surface of the road.
which is the same result as for the frictionless plane.
The difference between non slip or static friction and
The new result comes when we look at motion down sliding friction is accounted for by saying that there are
the plane. Solving 22a for ax gives two different coefficients of friction, the static coeffi-
cient µs which applies when the object is not moving,
ax = g sin θ – Ff /m (24) and the kinetic coefficient µk which applies when the
objects are sliding. For common surfaces like a rubber
Not surprisingly, the friction force reduces the accel-
tire sliding on a cement road, the static coefficient µs
eration down the plane.
is greater than the sliding or kinetic coefficient µk .

Let us substitute Equation 25 into Equation 24 for the

y motion of an object down an inclined plane, and then
see how the hypothesis that Ff is proportional to Fn
Fn can be tested in the lab. Using Equation 25 and 24 gives
Ff ax = g sin θ – Ff /m
m = g sin θ – µFn /m
θ
cos
mg Using Fn = mg cos θ gives
θ
x mg
ax = g sin θ – µg cos θ
mg sin θ
Figure 16 = g sin θ – µ cos θ (26)
9-14 Applications of Newton’s Second Law

Equation 26 clearly applies only if sin θ is greater than If you try this experiment in the lab, you may encounter
µ cos θ because friction cannot pull the object back up various difficulties. If you try to slide a block down a
the plane. reasonably smooth board, you may get fairly consistent
results and obtain values for µs and µk . But if you try
If we have a block on an inclined plane, and start with to improve the experiment by cleaning and smoothing
the plane at a very small angle, so that sin θ is much the surfaces, the results may become inconsistent be-
less than µ cos θ , the block will sit there and not slide. cause clean surfaces have a tendency to stick rather
If you increase the angle until sin θ = µ cos θ, with µ than slide.
the static coefficient of friction, the block should just
start to slide. Thus µs is determined by the condition The idea that friction forces can be described by two
coefficients µs and µk allows the authors of introduc-
sin θ = µs cos θ tory physics texts to construct all kinds of homework
problems involving friction forces. While these prob-
or dividing through by cos θ lems may be good mental exercises, comparable to
solving challenging crossword puzzles, they are not
µs = tan θs (27) particularly appropriate for an introductory physics
course. The reason is that the formula Ff = µFn is an
where θs is the angle at which slipping starts. over simplification of a complex phenomena. A decent
treatment of friction effects belongs in a more ad-
After the block starts sliding, µ is supposed to revert to vanced engineering oriented course where there is time
the smaller coefficient µk and the acceleration down to study the limitations and applicability of such a rule.
the plane should be

ax = g sin θ – µk g cos θ (26a)

Supposedly one can then determine the kinetic coeffi-

cient µk by measuring the acceleration ax and using
Equation 26a for µk .
 9-15

STRING FORCES If we now assume that the rope is effectively massless,

Another favorite device of the authors of introductory we get
texts is the massless string (or rope). The idea that a
string has a small mass compared to the object to which F1 + F2 = 0 (28)
it is attached is usually a very good approximation.
And strings and ropes are convenient devices for Thus F1 and F2 are equal in magnitude and oppositely
transferring a force from one object to another. directed. (Note that if there were a net force on a
massless rope, the rope would have an infinite accel-
In addition, strings have the advantage that you can eration.)
immediately tell the direction of the force they trans-
mit. The force has to be along the direction of the string A convenient way to analyze the effects of a taut rope
or rope, for a string cannot pull sideways. We used this or string is to say that there is a tension T in the rope, and
idea when we discussed the motion of a golf ball that this tension transmits the force along the rope. In
swinging in a circle on the end of a string. The string Figure (18) we have redrawn the tug of war and
could only pull in along the direction of the string included the tension T. The point where child 1 is
toward the center of the circle. From this we concluded holding the rope is subject to the left directed force F1
that the force acting on the ball was also toward the exerted by the child and the right directed force caused
center of the circle, in the direction the ball was by the tension T in the rope. The total force on this point
accelerating. of contact is F1 + T1 . Since the point of contact is
massless, we must have F1 + T1 = 0 and therefore the
To see how to analyze the forces transmitted by strings tension T on the left side of the rope is equal to the
and ropes, consider the example of two children pulling magnitude of F1. A similar argument shows that the
on a rope in a game of tug of war show in Figure (17). tension force T2 exerted on the second child is equal to
Let the child labeled 1 be pulling on the rope with a the magnitude of F2 . And since the magnitude of F1
force F1 and child labeled 2 pulling with a force F2 . and F2 are equal, the tension forces must also be equal.
Assuming that the rope is pulled straight between them,
the forces F1 and F2 will be oppositely directed. Isaac Newton noted that when a force was transmitted
via a massless medium, like our massless rope, or the
Applying Newton’s second law to the rope, and assum- force of gravity, the objects exerted equal and opposite
ing that the force of gravity on the rope is much smaller forces (here T1and T2) on each other. He called this the
than either F1or F2 and therefore can be neglected, we Third Law of Motion. We will have more to say about
have Newton’s third law in our discussion of systems of
particles in Chapter 11.)
m rope a rope = F 1 + F2

F1 T1 T2 F2
F1 F2

T1 = T2 = T

(1) (2) (1) (2)

Figure 17 Figure 18
Tug of war. Tension T in the rope.
9-16 Applications of Newton’s Second Law

THE ATWOOD’S MACHINE = h1 + h2 (30)

As an example of using a string to transmit forces,
consider the device shown in Figure (19) which is does not change. Differentiating Equation 30 with
respect to time and setting d /dt = 0 gives
called an Atwood’s Machine. It simply consists of two
masses at the ends of a string, where the string runs over
d dh1 dh2
a pulley. We will assume that the pulley is massless and 0 = = + = v1 + v2
the bearings in the pulley frictionless so that the only dt dt dt
effect of the pulley is to change the direction of the
where v1 = dh1 /dt is the velocity of mass 1, etc.
string.
Differentiating again with respect to time gives
To predict the motion of the objects in Figure (19) we
start by analyzing the forces on the two masses. Both dv1 dv2
0 = + = a1 + a2 (31)
masses are subject to the downward force of gravity, dt dt
m1g and m2g respectively. Let the tension in the
string be T. As a result of this tension, the string exerts Thus the desired relationship is
an upward force T on both blocks as shown. (We saw
in the last section that this force T must be the same on a1 = – a2 (32)
both masses.)
(You might say that it is obvious that a 1 = – a 2 ,
Applying Newton’s second law to each of the masses, otherwise the string would have to stretch. But if you
noting there is only motion in the y direction, we get are dealing with more complicated pulley problems, it
is particularly convenient to write down a formula for
m 1a 1y = T – m 1 g the total length of the string, and differentiate to obtain
(29) the needed extra relationship between the accelera-
m 2a 2y = T – m 2 g tions.)

In Equations 29, we note that there are three unknowns Using Equation 32 in 29 we get
T, a 1y and a 2y , and only two equations. Another
relationship is needed. This other relationship is sup- m 1a 1y = T – m 1 g
plied by the observation that the length of the string, (32b)
–m 2a 1y = T – m 2 g
given from Figure (19a) is

T
h1 h2 T
T m2
T m1
m2
m1
m 2g
m 2g m 1g
m 1g Figure 19b
Figure 19a Forces involved.
An Atwood’s machine consists of two masses suspended
from a string looped over a pulley. The acceleration is
proportional to the difference in mass of the two objects.
 9-17

Solving 32a for T to get T = m1 a1y + g and using this Exercise 3

in Equation 32b gives In a slight complication of the Atwood’s Machine, we
–m 2a 1y = m 1a 1y + m 1g – m 2 g use two pulleys instead of one as shown in Figure (20).
We can treat this problem very much like the preceding
or example except that the length of the string is h1 + 2h2
plus some constant length representing the part of the
m –m
a 1y = g m 1 + m 2 (33) string that goes over the pulleys and the part that goes
1 2 up to the ceiling. Calculate the accelerations of masses
m1 and m2. For what values of m1 and m2 is the system
From Equation 33 we see that the acceleration of mass balanced?
m1 is uniform, and equal to the acceleration due to
gravity, modified by the factor m1 – m2 / m1 + m2 . Exercise 4
If you want something a little more challenging than
When you solve a new problem, see if you can check Exercise 3, try analyzing the setup shown in Figure (21),
it by seeing if the limiting cases make sense. In or construct your own setup. For Figure (21), it is enough
Equation 30, if we set m2 = 0, then a 1 = g and we have to set up the four equations with four unknowns.
a freely falling mass as expected. If m 1 = m 2 , then the
masses balance and a1y = 0 as expected. When a
formula checks out in its limiting cases, as this one did,
there is a good chance that the result is correct.

The advantage of an of the Atwood’s Machine is that

by choosing m1 close to , but not equal to m2 , you can
reduce the acceleration, making the motion easier to
observe, just as Galileo did by using inclined planes. If
you reduce the acceleration too much by making m1too
nearly equal to m2 , you run the risk that even small
friction in the bearings of the pulley will dominate the
results.

h1 m3
h2

m1 m1

m2 m2
Figure 20 Figure 21
Pulley arrangement for Exercise 3. Pulley arrangement for Exercise 4
9-18 Applications of Newton’s Second Law

THE CONICAL PENDULUM max = Tx

Our final example in this chapter is the conical pendu- (34)
may = 0 = Ty – mg
lum. This is one of our favorite examples because it
involves a combination of Newton’s second law, circu-
The special feature of the conical pendulum is the
lar motion, no noticeable friction, and the predictions
fact that, because the ball is travelling in a circle, we
can be checked using an old boot, shoelace and wrist-
know that it is accelerating toward the center of the
watch.
circle with an acceleration of magnitude a = v2 /r. At
For a classroom demonstration of the conical pendu- the instant shown in Figure (23), the x direction
lum, we usually suspend a relatively heavy ball on a points toward the center of the circle, thus a = ax and
thin rope, with the other end of the rope attached to the we have
ceiling as shown in Figure (22). The ball is swung in ax = v2 /r (35)
a circle so that the path of the rope forms the surface of
a cone as shown. The aim is to predict the period of the The rest of the problem simply consists of solving
ball’s circular orbit. Equations 34 and 35 for the speed v of the ball and
using that to calculate the time the ball has to go
The distances involved and the forces acting on the ball around. The easy way to solve these equations is to
are shown in Figure (23). The ball is subject to only two write them in the form
forces, the downward force of gravity mg , and the
tension force T of the string. If the angle that the string mv2
Tx = max = (36a)
makes with the vertical is θ , then the force T has an r
upward component Ty = Tcosθ and a component di- Ty = mg (36b)
rected radially inward of magnitude Tx = Tsin θ . (We
are analyzing the motion of the ball at the instant when Dividing Equation 36a by 36b and using
it is at the left side of its orbit, and choosing the x axis Tx Ty = Tsin θ Tcosθ = tan θ , we get
to point in toward the center of the circle at this instant.)
Tx mv2 v2
Applying Newton’s second law to the motion of the = tan θ = =
Ty mgr gr
ball, noting that ay = 0 since the ball is not moving up
and down, gives

θ θ

T h

θ
Ty
Tx
r
Figure 22
The conical pendulum.
mg
Figure 23
Forces acting on the ball.
 9-19

Next use the fact that tan θ = r/h to get Exercise 5 Conical Pendulum
2 Construct a pendulum by dangling a shoe or a boot
r v
tan θ = = from a shoelace.
h gr
(a) Verify that for small angles θ, you get the same period
g
v = r (37) if you swing the shoe in a circle to form a conical
h pendulum, or back and forth to form a simple pendulum.
Finally we note that the period is the distance traveled (b) Time 10 swings of your shoe pendulum and verify
in one circuit, 2 π r , divided by the speed v of the ball Equation 38 or 39. (You can get more accurate results
period 2πr 2πr using a smaller, more concentrated mass, so that you
= = can determine the distance more accurately.) Try
of orbit v r g/h several values of the shoe string length to check that
the period is actually proportional to .
period = 2π h (38)
g Exercise 6
This is what we like to call a clean desk problem. Clear
The prediction of Equation 38 is easily tested, for off your desk, leaving only a pencil and a piece of paper.
example, by timing 10 rotations of the ball and Then starting from Newton’s second law, derive the
dividing the total time by 10. Note that if the angle θ is formula for the period of a conical pendulum.
kept small, then the height h of the ball is essentially
What usually happens when you do such a clean desk
equal to the length of the rope, and we get the formula problem is that since you just read the material, you think
you can easily do the analysis without looking at the text.
period ≈ 2 π (39) But if you are human, something will go wrong, you get
g
stuck somewhere, and may become discouraged. If
Equation 39 is the famous formula for the period of you get stuck, peek at the solution and finish the
what is called the simple pendulum, where the ball problem. Then a day or so later clean off your desk
again and try to work the problem. Eventually you
swings back and forth rather than in a circle. Equation
should be able to work the problem without peeking at
39 applies to a simple pendulum only if the angle θ is the solution, and at that point you know the problem well
kept small. For large angles, Equation 38 is exact for and remember it for a long time.
a conical pendulum, but Equation 39 has to be replaced
by a much more complicated formula for the simple When you are learning a new subject like Newton’s
pendulum. (We will discuss the analysis of the simple second law, it is helpful to be fully familiar with at least
pendulum in Chapter 11 on rotations and oscillations.) one worked out example for each main topic. In that
way when you encounter that topic again in your work,
Note that the formula for the period of a simple in a lecture, or on an exam, you can draw on that
pendulum depends only on the strength g of gravity and example to remember what the law is and how it is
the length of the pendulum, and not on the mass m or applied.
the amplitude of the swing. As a result you can
At various points in this course, we will encounter
construct a clock using the pendulum as a timing problems that serve as excellent examples of a topic in
device, where the period depends only on how long you the course. The conical pendulum is a good example
make the pendulum. because it combines Newton’s second law with the
formula for the acceleration of a particle moving in a
circle; the prediction can easily be tested by experi-
ment, and the result is the famous law for the period of
a pendulum. When we encounter similarly useful ex-
amples during the course, they will also be presented
as clean desk problems.
9-20 Applications of Newton’s Second Law

APPENDIX ! --------- Plot data

DO
THE BALL SPRING PROGRAM READ Rx,Ry
CALL BOX
! --------- Plotting window LOOP UNTIL END DATA
! (x axis = 1.5 times y axis)
SET WINDOW -40,140,-10,110 DATA 88.2, 42.8
DATA 80.2, 24.4
! --------- Draw & label axes DATA 68.0, 12.0
BOX LINES 0,100,0,100 DATA 52.9, 8.6
PLOT TEXT, AT -3,0 : "0" DATA 37.4, 14.7
PLOT TEXT, AT -13,96: "y=100" DATA 24.0, 28.8
PLOT TEXT, AT 101,0 : "x=100" DATA 14.2, 47.5
DATA 9.0, 67.0
! ---------- Experimental constants DATA 8.2, 83.9
LET m = 245 DATA 11.1, 95.0
LET g = 980 DATA 16.7, 98.8
LET K = 5130 DATA 23.9, 94.1
LET So = 35.9 DATA 32.2, 81.5
LET Zx = 50 DATA 41.9, 62.1
LET Zy = 130 DATA 52.1, 39.9
DATA 62.2, 19.4
! --------- Initial conditions
LET Rx = 88.2 ! --------- Subroutine "CROSS" draws
LET Ry = 42.8 ! a cross at Rx,Ry.
LET Vx = (80.2 - 91.1)/(2*.1) SUB CROSS
LET Vy = (24.4 - 63.1)/(2*.1) PLOT LINES: Rx-2,Ry; Rx+2,Ry
LET T = 0 PLOT LINES: Rx,Ry-2; Rx,Ry+2
CALL CROSS END SUB

! --------- Subroutine "BOX" draws

! --------- Computer Time Step ! a cross at Rx,Ry.
LET dt = .001 SUB BOX
LET i = 0 PLOT LINES: Rx-1,Ry+1; Rx+1,Ry+1
PLOT LINES: Rx-1,Ry-1; Rx+1,Ry-1
! --------- Calculational loop PLOT LINES: Rx-1,Ry+1; Rx-1,Ry-1
DO PLOT LINES: Rx+1,Ry+1; Rx+1,Ry-1
LET Rx = Rx + Vx*dt END SUB
LET Ry = Ry + Vy*dt
END
LET Sx = Rx - Zx
LET Sy = Ry - Zy
LET S = Sqr(Sx*Sx + Sy*Sy)

Let Fs = K*(S - So) The new feature is the READ statement at the top of this
LET Fx = -Fs*Sx/S column. Each READ statement reads in the next
LET Fy = -Fs*Sy/S - m*g values of Rx and Ry from the DATA lines below. We
then call BOX which plots a box centered at Rx,Ry.
LET Ax = Fx/m
LET Ay = Fy/m The LOOP statement has this plotting continue until
we run out of data. (In Figure 11, we filled in the boxes
LET Vx = Vx + Ax*dt with a paint program to make them stand out.)
LET Vy = Vy + Ay*dt

LET T = T + dt
LET i = i+1

IF MOD(i,100) = 0 THEN CALL CROSS

PLOT Rx,Ry

LOOP UNTIL T > 1.6

 10-1

Chapter 10
Energy

CHAPTER 10 ENERGY

In principle, Newton's laws relating force and accel- – Gmme / r the gravitational potential energy of the
eration can be used to solve any problem in mechanics satellite. We saw that Etotal did not change its value
involving particles whose size ranges from that of as the satellite went around its orbit.
specks of dust to that of planets. In practice, many
mechanics problems are too difficult to solve if we try It turns out that energy is a much more complex subject
to follow all the details and analyze all the forces than we might suspect from the discussion of satellite
involved. For instance f = ma presumably applies motion. There are many forms of energy, such as
to the motion of the objects involved in the collision of electrical energy, heat energy, light energy, nuclear
two automobiles, but it would be an enormous task to energy and various forms of potential energy. Some-
study the details of the collision by analyzing all the times there is a simple formula for a particular form of
forces involved. energy, but sometime it may be hard even to figure out
where the energy has gone. Despite the complexity,
In a complicated problem, we cannot follow the one simple fact remains, if we look hard enough we find
motion of all the individual particles; instead we look that energy is conserved.
for general principles that follow from Newton's laws
and apply these principles to the system of particles as If, in fact, it were not for the conservation of energy, we
a whole. We have already discussed two such general would not have invented the concept in the first place.
principles: the laws of conservation of linear and Energy is a useful concept only because it is conserved.
angular momentum. We have found that if two cars What we are going to do in this chapter is first take a
traveling on frictionless ice collide and stick together, more general look at the idea of a conservation law,
we can use the law of conservation of linear momen- and then see how we can use energy conservation to
tum to calculate their resulting motion. We do not develop formulas for the various forms of energy we
have to know how they hit or any other details of the
encounter. We will see, for example, where the formula
collision.
1 / 2 mv2 for kinetic energy comes from, and we will
In our discussion of satellite motion, we saw that there show how the formula – Gmme / r for gravitational
was another quantity, which we called energy, that potential energy reduces to a much simpler formula
was conserved. Our formula for the total energy of the when applied to objects falling near the surface of the
satellite was Etotal = 1 / 2 mv2 – Gmme / r where earth.
2
1 / 2 mv was called the kinetic energy and
10-2 Energy

CONSERVATION OF ENERGY 25 blocks. However, there is a box in the room, a toy

Because energy comes in different forms, it is more box, and the mother goes to open the toy box, but the
difficult to state how to compute energy than how to boy says, 'No, do not open my toy box,' and screams.
compute linear momentum. But, as we shall see, it is Mother is not allowed to open the toy box. Being
not necessary to state all the formulas for all the extremely curious, and somewhat ingenious, she in-
different forms of energy. If we know the formula for vents a scheme! She knows that a block weighs 3
some forms of energy, we can use the law of conserva- ounces, so she weighs the box at a time when she sees
tion of energy to deduce the other formulas as we need 28 blocks, and it weighs 16 ounces. The next time she
them. How a conservation law can be used in this way wishes to check, she weighs the box again, subtracts 16
is illustrated in the following story, told by Richard ounces and divides by 3. She discovers the following:
Feynman in The Feynman Lectures on Physics (Vol. I, number of weight of box – 16 oz
Addison-Wesley, Reading, Mass., 1963). +
blocks seen 3 oz
"Imagine a child, perhaps 'Dennis the Menace,' who = constant
has blocks that are absolutely indestructible, and can- There then appear to be some gradual deviations, but
not be divided into pieces. Each is the same as the careful study indicates that the dirty water in the
other. Let us suppose that he has 28 blocks. His mother bathtub is changing its level. The child is throwing
puts him with his 28 blocks into a room at the beginning blocks into the water, and she cannot see them because
of the day. At the end of the day, being curious, she it is so dirty, but she can find out how many blocks are
counts the blocks very carefully, and discovers a phe- in the water by adding another term to her formula.
nomenal law—no matter what he does with the blocks, Since the original height of the water was 6 inches and
there are always 28 remaining! This continues for a each block raises the water a quarter of an inch, this
number of days, until one day there are only 27 blocks, new formula would be
but a little investigating shows there is one under the
rug—she must look everywhere to be sure that the number of weight of box – 16 oz
+
number of blocks has not changed. blocks seen 3 oz
One day, however, the number appears to change— height of water – 6 inches
+ (1)
there are only 26 blocks. Careful investigation indi- 1/4 inch
cates that the window was open, and upon looking = constant
outside, the other two blocks are found. Another day
In the gradual increase in the complexity of her world,
careful count indicates that there are 30 blocks! This
she finds a whole series of terms representing ways of
causes considerable consternation, until it is realized
calculating how many blocks are in places where she
that Bruce came to visit, bringing his blocks with him,
is not allowed to look. As a result of this, she finds a
and he left a few at Dennis' house. After she had
complex formula, a quantity which has to be com-
disposed of the extra blocks, she closes the window,
puted, which always stays the same in her situation."
does not let Bruce in, and then everything is going
along all right, until one time she counts and finds only
 10-3

Similarly, we will find a series of terms representing MASS ENERGY

ways of calculating various forms of energy. Unlike On earth, the greatest supply of useful energy ulti-
the story, where some blocks are actually seen, we mately comes from the sun, mainly as sunlight, which
cannot see energy; all of the terms in our equation for is a form of radiant energy. The energy we obtain from
energy must be computed. But if we have included fossil fuel, such>é” coal and wood, and the energy we
enough terms and have not neglected any forms of get from hydroelectric dams came originally from the
energy, the numerical value of all the terms taken sun. On a clear day, the sun delivers as much energy to
together will not change; that is, we will find that half a square mile of tropical land as was released by the
energy is conserved. first atomic bomb. In about 1 millionth of a second, the
It is not necessary, however, to start with the complete sun radiates out into space an amount of energy equal
energy equation. We will begin with one term. Then, to that used by all of mankind during an entire year.
as the complexity of our world increases, we will add The sun emits radiant energy at such an enormous
more terms to the equation so that energy remains rate that if it burned like a huge lump of coal, it would
conserved. last about 5000 years before burning out. Yet the sun
has been burning at nearly its present rate for over 5
billion years and should continue burning for an-
other 5 billion years. How the sun could emit all of
this energy was explained in 1905 when Einstein
discovered that mass and energy are related through
the well-known equation

E = mc 2
(2)
where E is energy, m mass, and c the speed of light.

The sun's source of energy is the tiny fraction of its mass

that is being converted continually to radiant energy
through nuclear reactions. Similar processes occur
when the hydrogen bomb is exploded. To indicate the
amount of energy that is in principle available as mass
energy, imagine that the mass of a 5–cent piece (5 gm)
could be converted entirely into electrical energy. This
electrical energy would be worth several million dol-
lars. The problem is that we do not have the means
available to convert mass completely into a useful form
of energy. Even in the nuclear reactions in the sun or
in the atomic or hydrogen bombs, only a few tenths of
1% of the mass is converted to energy.

Since most of the energy in the universe is in the form

of mass energy, we shall begin to develop our equation
for energy with Einstein's formula E = mc2. As we
mentioned, we will add terms to this equation as we
discover formulas for other forms of energy.
10-4 Energy

Ergs and Joules Exercise 1

Our first step will be to use the Einstein energy (a) Use dimensions to determine how many ergs there
formula to obtain the dimensions of energy. In the are in a joule. (Check your answer against the statement
CGS system of units we have that a 100-watt bulb uses 100 joules or 109 ergs of
2 gm cm 2 energy per second.)
E = m gm × c 2 cm 2 = mc 2
sec sec 2 (b) As you may have guessed, a 1 watt light bulb uses
The set of dimensions gm cm2 /sec2 is called an erg. 1 joule of energy per second. How many joules of
energy does a 1000 watt bulb or heater use in one hour.
gm cm
2 (This amount of energy is called a kilowatt hour
1 2
= 1 erg CGS units (abbreviated kwh) and costs a home owner about 10
sec cents when supplied by the local power company.)
In the MKS system of units, we have (c) If a 5-cent piece (which has a mass of 5 grams) could
2 m2 be converted entirely to energy, how many kilowatt
E = m kg × c2 m 2 = mc 2 kg hours of energy would it produce? What would be the
sec sec 2
value of this energy at a rate of 10¢ per kilowatt hour?
where the set of dimensions of kg m 2 /sec2 is called
a joule.
2
kg m
1 2
= 1 joule MKS units
sec
It turns out that for many applications the MKS joule is
a far more convenient unit of energy than the CGS erg.
A 100-watt light bulb uses 100 joules of energy per
second, or 1 billion ergs of energy per second. The erg
is too small a unit of energy for many applications, and
it is primarily for this reason that the MKS system of
units is more often used than the CGS system. This is
particularly true when dealing with electrical phenom-
ena.
 10-5

KINETIC ENERGY Solution: The first step is to calculate 1 – v 2 /c2 for

From the recoil definition of mass (Chapter 6), we the muons. An easy way to do this is as follows:
saw that the mass of an object increases with speed, v = .995 c
becoming very large when the speed of the object
approaches the speed of light. The formula for the v = .995 = 1 – .005
c
increase in mass with speed was simply
v 2 = 1 – .005 2
m0
m = (6-14) c2
1 – v 2 /c2
= 1 – 2 .005 + .005 2
where m 0 is the mass of the particle at rest (the rest
= 1 – .01 + .000025
mass). When we combine this formula with Einstein's
equation E = mc2, we get as the equation for the We have neglected .000025 compared to .01 be-
energy of a moving particle cause it is so much smaller. We now have
m 0c2 2
E = mc2 = (3) 1 – v2 ≈ 1 – 1 – .01 = .01
1– v 2 /c2 c
According to Equation (3), when a particle is at rest 2
(v = 0), its energy is given by 1 – v2 ≈ .01 = .1 = 1
c 10
E0 = m 0c2 rest energy (4) (This procedure is discussed in more detail in the
This energy m 0 c2 is called the rest energy of the section on approximation formulas in Chapter 1.)
particle. As a particle begins to move, its mass, and Now that we have 1 – v 2 /c2 = 1/10 for these
therefore its energy, increases. The extra energy muons, we can calculate their relativistic mass
that a particle acquires as a result of its motion is
called kinetic energy. If mc2 is the total energy, then m0 m0
the formula for the particle's kinetic energy is m = = = 10 m 0
1– v 2 /c2 1/10
kinetic = total – rest
energy energy energy Thus the mass of the muons has increased by a factor
of 10. The total energy of the muons is
KE = mc2 – m 0c2 (5)
E = mc2 = 10m 0 c2 = 10 m 0 c2
Example 1
The muons in the motion picture Time Dilation of the Hence, their total energy is also 10 times their rest
µ–Meson (Muon) Lifetime moved at a speed of .995c. energy. Their increase in energy, or their kinetic
By what factor did their mass increase and what is their energy, is
kinetic energy? 2 2
KE = mc – m0 c = 10m0 c – m0c
2 2

2
KE = 9 m0 c
This kinetic energy 9m 0c2 is the amount of additional
energy that is required to get muons moving at a speed
v = 0.995c .
10-6 Energy

Exercise 2 ing much slower than the speed of light, for instance,
Assume that an electron is traveling at a speed
1000 mi/sec or less, there is an easier way to calcu-
v = .99995c. late the energy of the object than by evaluating
1 – v 2 /c2 directly.
2 2
(a) What is 1 – v /c for this electron? In the section on approximation formulas in Chapter
(b) By what factor has its mass increased over its rest
1, it was shown that when v/c is much less than 1,
mass? then we can use the approximate formula

(c) By what factor has its total energy increased over its
1 ≈ 1 + α (1-25)
1–α 2
rest energy?
to get
(d) The rest mass of an electron is
m 0 = 0.911 x 10– 27 gm . What is its rest energy ( in 2
1 ≈ 1 + v2 (6)
ergs)? 1 – v 2 /c2 2c
(e) What is the total energy (in ergs) of this electron? The approximate formula 1 + v 2 /2c2 is much easier
to use than 1/ 1 – v 2 /c2 . Moreover, if v/c is a small
(f) What is the kinetic energy of this electron in ergs?
number, then the formula is quite accurate, as illus-
trated in Table 1. It should be noted however that when
Slowly Moving Particles v becomes larger than about .1c, the approximation
In Example 1, where the particle (muon) was mov- becomes less accurate. When we reach v = c, the exact
ing at nearly the speed of light, we determined its formula is 1/ 1 – v 2 /c2 = ∞ but the approximate
increase in mass and its kinetic energy by calculat- formula gives 1 + v 2 /2c2 = 1.5. At this point the
ing 1 – v 2 /c2 . However, when a particle is mov- approximate formula is no good at all!

2
Table 1 Numerical check of the Approximation Formula 1 ≈ 1+ v2
1 – v 2 /c 2 2c

valueof exact formula valueof approximate formula

v
1 2
1 + v2
1 – v 2 /c2 2c

.01c 1.000050003 1.000050000

.1c 1.005037 1.005000
.2c 1.0206 1.0200
.3c 1.048 1.045
.5c 1.148 1.125
.7c 1.41 1.25
.9c 2.30 1.40
.99c 7.1 1.49

c ∞ 1.5
 10-7

If we use Equation (6), the total energy of a particle It is worth noting that, at one time, only the kinetic
becomes energy term 1/2m 0 v 2 in Equation 7 was recognized as
2
a form of energy. Before 1905, it was not known that
E = mc m 0c2 should be included in the equation for conserva-
1 tion of energy, because no one had ever observed the
2 exact
= m0 c rest mass of an object to change. The first evidence that
formula
1 – v 2 /c2 the rest energy had to be included came from the study
of nuclear reactions. In these reactions enormous
2 amounts of energy were released, producing a detect-
2 v approximate
≈ m0c 1 + able change in the nuclear rest masses.
2 formula
2c
2 So long as an object is moving at a speed of .1c or less,
2 2 v the kinetic energy of that object will be far less than its
≈ m0c + m0c 2
2c rest mass energy. For example, let us compare the
kinetic energy to the rest mass energy of a 10–gm pistol
The factor c2 cancels in the second term, and we are
bullet that travels with a speed of about 300 m/sec.
left with the approximate formula
Using MKS units, we find that the bullet's kinetic
approximateformula energy (KE) is
1 for particlesmoving
E ≈ m0c2 + m0 v 2
2 at speedsless than (7) 1
about .1c KE = m0v 2
2
Since Equation (7) contains the approximation made in 1 2
Equation (6), it is not valid for particles traveling faster = × .01 kg × 300 m/sec
than about one tenth of the speed of light. For particles 2
traveling at nearly the speed of light, we must use = 450 joules
E = m 0c2 / 1 – v 2 /c2 . But for particles traveling as
slowly as a few thousand miles an hour or less, Equa- This is enough to allow a bullet to penetrate a plank.
tion (6) is so accurate that any error would be difficult The rest mass energy E0 of the bullet is
to detect.
E0 = m0 c2
For all but the last section of this chapter, we will 2
confine our discussion to the energy of objects travel- = .01 kg 3 × 10 8 m/sec
ing at slow speeds, where Equation (7) is not only
accurate, but is the simplest equation to use. When we = 9 × 10 14 joules
look at this equation, we can see that the mass energy This is the amount of energy released in a moderate-
E = mc2 is now written in two distinct parts m 0c2, sized atomic bomb.
which is the rest mass energy, and 1/2m 0 v 2, which is
the energy of motion or kinetic energy

E = m 0c 2 rest 1 2 kinetic
energy + 2 m 0 v energy (7a)

Written in this way, our equation for total energy is

beginning to resemble Equation (1), which was used to
determine the number of blocks in Dennis' room. We
now have two terms representing two different kinds of
energy.
10-8 Energy

Exercise 3 GRAVITATIONAL
For the preceding example of a 10 gram bullet: POTENTIAL ENERGY
Let us continue our search for terms to add to our
a) at 10 cents per kilowatt hour, what is the value of the
equation for energy. Suppose that a ball of mass m is
bullet's kinetic energy?
dropped from a height h above the floor, as shown in
b) what is the value of its rest energy? Figure (1). Immediately before the ball hits the floor,
it has a rest energy m 0 c2, and a kinetic energy 1 2m
2 0v 2 .
c) how fast would the bullet be traveling if it had twice
Immediately before the ball was dropped, however, it
as much kinetic energy?
had the same rest energy m 0 c2 but no kinetic energy.
Where did the kinetic energy that it possessed just
before it hit the floor come from?

If we were observant, we might have noted that some

effort was needed to lift the ball from the floor to a
height h. As the ball was lifted a new kind of energy
was being stored. This new form of energy, which was
released when the ball was dropped, is called gravita-
tional potential energy. When it is included, our
equation for energy becomes
gravitational
Etotal = m 0 c2 1 2
+ m 0v + potential (8)
2 energy

at rest m kinetic energy = 0 To find the formula for the gravitational potential
energy, we will assume that energy is conserved and
that the total energy of the ball, immediately before it
is released, is equal to the total energy of the ball
h immediately before it hits the ground.

When a ball is dropped from a height h, it accelerates

downward with a constant acceleration g until it hits the
m kinetic energy = 1 mv 2
floor. Thus we can use the constant acceleration
2 formulas (see Appendix 1 in Chapter 4.)
1 2
s = vit + at
2
v
Figure 1
vf = vi + at
Falling Weight. When a weight is dropped
it gains kinetic energy. This kinetic energy with vi = 0, a = g, and s = h we get
comes from the energy we stored in the
object when we lifted it up to a height h. h = 1 gt 2 (12)
2
vf = gt (13)
 10-9

Substituting t = vf / g from Equation 13 into Equa- Exercise 4

tion (12) gives Call v2 the speed of the ball when it has fallen halfway
to the floor.
v2 vf 2
h = 1 g f2 = (a) Explain why the ball's total energy, when it has fallen
2 g 2g
halfway to the floor, is
1 v 2 = gh (14) halfway
2 f Etotal down = m 0c2 + 1 m 0v22 + m 0g h
2 2
Multiplying Equation 14 through by m 0 gives
(b) Calculate v2 (just as we calculated vf ) and show
1 m v 2 = m gh (15) that the total energy of the ball when halfway down is the
2 0 f 0
same as when it was released, or just before it hit the
Suppose that we use m 0gh as the formula for gravita- floor.
tional potential energy. (The greater h, the higher we
have lifted the ball, the more potential energy we have Exercise 5
stored in it.) Show that the formula for gravitational potential energy
has the dimensions of joules (in the MKS system) and
ergs (in the CGS system).
formula for near the
m 0 gh = gravitational surface of (16) Exercise 6
potential energy the earth
What is the gravitational potential energy (in joules and
ergs) of a 100–gm ball at a height of 2 meters above the
floor? (Measure h starting from the floor.)
Before the ball is released, its total energy is in the form
of rest energy and gravitational potential energy
What happens to the energy after the ball has hit the
before floor and is lying at rest? At this point, it no longer has
E total release = m 0c 2 + m 0gh (17) kinetic energy or gravitational potential energy. Now
what should we add to our equation to maintain conser-
Just before the ball hits the floor, where it has kinetic vation of energy? In this case, we have to look "under
energy but no potential energy (since h = 0), the total the rug," in the "dirty water," and "out the window" all
energy is at once. When the ball hit the floor, we heard a thump;
just before thus, some of the ball's energy has been dissipated as
E total hitting floor = m 0c 2 + 1 m 0 v f 2 (18)
2 sound energy. We find that there is a dent in the floor;
hence we know that some of the energy has gone into
At first, Equations 17 and 18 for total energy look rearranging the molecules in that part of the floor. Also,
different; but since 1/2m 0vf 2 = m 0gh (Equation 15), because the bottom of the ball and the floor underneath
they give the same numerical value for the ball's total became slightly warmer after the ball hit the floor, we
energy. Thus, we conclude that we have chosen the conclude that some of the energy was converted into
correct formula for calculating gravitational potential heat energy. (In some collisions, such as when a
energy. mining pick strikes a stone, we see what looks like a
spark, which shows us that some of the kinetic energy
has been changed into radiant energy, or light.)
10-10 Energy

After the ball hits the floor, the formula for total energy Notation
becomes as complicated as Since our discussion for the remainder of this chapter
2 1 2
will deal with objects moving at speeds much less than
Etotal = m 0 c + m0 v + m0 gh the speed of light, objects whose mass m is very nearly
2
equal to the rest mass m 0 , we will stop writing the
+ sound energy subscript 0 for the rest mass. With this notation, our
+ energy to cause a dent (19) formulas for kinetic energy and gravitational potential
energy are simply 1/2mv 2 and mgh. Only when we
+ heat energy + light energy discuss objects like atomic particles whose speeds can
Because energy can appear in so many forms that are become relativistic, will we be careful to distinguish
often difficult to detect, it was not until many years after the rest mass m 0 from the total mass m.
Newton that conservation of energy was established as
a general law. The law of conservation of energy is
Example 2
used to solve only those problems where very little
energy "escapes" in a form that is difficult to detect. In Consider a simple pendulum consisting of a ball swing-
a complicated collision problem we can calculate only ing on the end of a string, as shown in Figure (2). When
how much energy is "lost," that is, changed to other the ball is released from a height h it has a potential
forms of energy. energy m 0gh. As the ball swings down toward the
bottom, h decreases and the ball loses potential energy
On an atomic scale, however, we do not have to think but gains kinetic energy. At the bottom the original
of energy as being "lost" because the various forms of potential energy mgh has been entirely converted into
energy are more easily detected. For example, we will kinetic energy 1/2mv 2. Then the ball climbs again,
see in Chapter 17 that the heat energy and sound energy gaining potential energy but losing kinetic energy.
are primarily the kinetic energy of the atoms and
molecules; thus, these do not appear as separate forms pivot
of energy. It is on this small scale that the law of
conservation of energy may be most accurately veri-
fied.

On the other hand, if we can neglect the effects of

friction and air resistance, the law of conservation of
energy can be used to solve mechanics problems that C A (at rest)
would otherwise be difficult to solve. We will illustrate
this with two examples in which gravitational potential
energy m 0gh is converted into kinetic energy 1/2m 0v 2 h
and vice versa. B
Figure 2
Application of conservation of energy
to pendulum motion. The speed at B
can be found by equating the kinetic
energy at B 1/ 2mv2 to the potential
energy lost in going from A to B (mgh).
 10-11

Finally, at position C, the ball has swung back up to a Example 3

height h and all the kinetic energy has been changed to It should be noted that we are able to calculate the
potential energy. The ball stops momentarily at posi- speed of the ball in the preceding example without an
tion C, and the swing is reversed. Eventually, however, analysis of the forces involved. An even more
the pivot becomes warm and air currents are set up by striking example of conservation of energy that would
the swinging pendulum; thus, the pendulum itself be nearly impossible to analyze in terms of forces is
gradually loses energy and finally comes to rest. that of a skier traveling down a very icy hill. If he is
not an experienced skier, he may not know how to
As long as we can neglect air resistance and friction in dissipate some of his kinetic energy as heat and sound
the pivot we can use the conservation of energy equa- by scraping the edges of his skis against the ice. If he
tion to calculate the speed of the ball at position B. is not able to dissipate energy, then no matter how he
Before the ball is released turns, no matter how twisted a path he takes, when he
Etotal A = m 0c2 + mgh reaches bottom, all his potential energy m 0gh will
have been converted to kinetic energy 1/2m 0 v 2, in
At position B, where h = 0 which case his speed at the bottom of the hill will be
Etotal B = m 0c2 + 1 mvB2 2gh . To see why an inexperienced skier should
2 not try icy hills, consider that if the hill has a 500–ft
If energy is conserved rise, his speed at the bottom will be 179 ft/sec or 122
mi/hr. This result is computed not from the details of
Etotal A = Etotal B the skier's path, but from the knowledge that he was
not able to dissipate energy. As we mentioned at the
m 0c2 + mgh = m 0c2 + 1 mv 2
2 beginning of the chapter, the conservation of energy
is one of the general principles of mechanics that can
Note that since m 0c2 did not change, it does not enter be applied successfully without knowing all the de-
into this calculation. Here we could apply the conser- tails involved in the physical situation.
vation of energy equation without considering the rest
energy. We now have
Exercise 7
mgh = 1 mvB 2 A car coasts along a road that leads from the top of a
2 300–ft–high hill, down through a valley, and up over a
vB 2 = 2gh 200 ft high hill. Assume that the car does not dissipate
energy through friction and air resistance.
vB = 2gh
(a) If the car starts at rest from atop the higher hill, how
fast will it be traveling when it reaches the top of the
2
lower hill ( g = 32 ft/sec ) ?

(b) If the car is initially moving at 80 ft/sec (55 mi/hr)

when it starts coasting at the top of the higher hill, how
fast will the car be moving when it reaches the top of
the lower hill?
10-12 Energy

WORK We will begin the discussion with the introduction of a

Let us take another look at the example where we new term which we will call work. In some ways this
dropped a ball of mass m from a height h above the floor is an unfortunate choice of a word, for everyone has
as shown in Figure (3). At the height h, the ball had a their own idea of what “work” is, and it seldom
gravitational potential energy mgh. Just before hitting coincides with the physicist's definition. In the physicist's
the floor, all this gravitational potential energy had definition, a force does work on an object when it adds
been converted to kinetic energy 1/2 mv2. We know energy to the object. More explicitly, the work a force
that the ball speeded up, accelerated, because gravity does is equal to the energy that the force supplies. In the
was exerting a downward force mg on the ball as it fell. case of the falling ball the gravitational force supplied
an amount of energy mgh, therefore that is the work that
There appears to be a coincidence in this example. the gravitational force did as the ball fell.
Gravity pulls down on the ball with a force of magni-
tude mg, the ball falls a distance h, and the ball gains a work done by the
kinetic energy equal to (mg)× h . In this example the force of gravity = mgh (20)
energy that gravity supplies to the ball by pulling down as the ball fell
on it is equal to the gravitational force (mg) times the
distance h over which the force acted. Is this a From Equation (20), we see that for the case where we
coincidence, or does this example provide a clue as to have a constant force, and the object moves in the
the way in which forces supply energy? direction of the force, the work done is equal to the
magnitude of the force times the distance moved.
In this case, where we have a constant force mg, and the
If the force is
ball moves in the direction of the force for a distance h,
constant and the
the increase in energy is the force times the distance. Work = Force × Distance distance is in
the direction of
In more general examples, however, the situation can the force
be more complex. If the object is not moving in the (21)
direction of the force, then only the component of the
force in the direction of motion adds energy to the Exercise 8
object. And if the force is not constant, we have to break Show that force times distance has the same dimen-
the problem into many small steps, and calculate the sions as energy. (Get the dimensions of energy from
energy gained in each step. We shall see that calculus E = mc2 .)
provides powerful techniques to handle these situa-
tions. As the first complication, or correction to our definition
of work, suppose that the force is not in the same
direction as the motion. Suppose, for example, a
hockey puck slides for a distance S along frictionless
ice as shown in Figure (4). During this motion a
mg h gravitational force mg is acting and the puck moves a
distance S. But the puck coasts along at constant speed;
it does not gain any energy at all. In this case the
gravitational force does no work.
S
Figure 3 v
A ball, subject to a gravitational force mg,
falling a distance h, gains a kinetic energy mgh.
mg
Figure 4
The force of gravity does no work
on the sliding hockey puck.
 10-13

The problem with the hockey puck example is that the When we throw the ball up, the angle between the
gravitational force is down and the motion is sideways. downward force and upward motion is θ = 180°,
In this case the – y directed gravitational force has no cos θ = – 1, and we get
component along the x directed motion of the puck. In
order for the puck to gain energy, it must accelerate in W = F ⋅ S = F S cos θ
the x direction, but there is no x component of force to = mgh(–1) = –mgh
produce that acceleration.
We now predict that gravity is taking energy from the
Now let us consider an example where the force is ball, which is also correct.
acting opposite to the direction of motion. If we throw
Finally, in the case of a hockey puck, the angle θ
a ball up in the air, the ball starts out with the kinetic
between the – y directed force and the x directed
energy 1/2 mv02 that we gave it. As the ball rises,
motion is 90°. We have cos θ = 0, so that F ⋅ S = 0
gravity acts against the motion of the ball and removes
and the gravitational force does no work. Again the
kinetic energy. When the ball has risen to a height h
formula W = F ⋅ S works.
given by mgh = 1/2 mv02, all the kinetic energy is gone
and the ball stops. The ball has reached the top of the
trajectory. This example tells us that when the force is Exercise 9
directed opposite to the direction of motion, the work A frictionless plane is inclined at an angle θ as shown
is negative—the force removes rather than adds en- in Figure (5). A hockey puck initially at a height h above
ergy. the ground, slides down the plane. When the puck gets
to the bottom, it has moved a distance S = h / cosθ as
shown. (This comes from h = S cos θ )
The Dot Product
a) Verify the formula S = h/cos θ for the two cases θ = 0
This is where our discussion has lead so far. We have
and θ = 90°. I.e., what are the values for h / cosθ for
a quantity called “work” which is a form of energy. It these two cases, and are the answers correct?
is the energy supplied by a force acting on a moving
object. Now energy, given by formulas like E = mc2, b) Show that the work W = Fg ⋅S, done by the gravi-
is a scalar quantity; it is a number that does not point tational force as the puck slides down the plane, is
anywhere. But our formula for work = force times mgh no matter what the angle θ is.
distance involves two vectors, the force F and the
c) Explain the result of part (b) from the point of view of
distance S. What mathematical way can we combine conservation of energy.
the two vectors F and S to get a number for the work W?
One possibility, that we discussed back in the chapter
on vectors, is the scalar or dot product.
θ
S

S
W = F⋅S = F S cos θ θ (22)
=

F
h/

h
co
s

Mathematically the dot product turns the vectors F

θ

and S into a scalar number W. Let us see if W = F⋅S

is the correct formula for work. If F and S are in the
same direction, θ = 0°, cos θ = 1, and we get Figure 5
Diagram for Exercise 9.
W = F⋅S = F S cos θ = F S

Applied to the case of a falling ball, F = mg , S = h

and we get W = mgh which is correct.
10-14 Energy

Work and Potential Energy Let us consider another example where we store poten-
In the discussion of energy, physicists tend to use a lot tial energy by doing work against a force. Suppose I tie
of words like work, potential energy, kinetic energy, one end of a spring to a post and pull on the other end
etc. What we are doing is building a conceptual picture as shown in Figure (6). As I stretch the spring, I am
to help us organize a number of physical phenomena exerting a force Fme and moving the end of the spring
and related mathematical equations. You will find that in the same direction. Therefore I am doing positive
when you see this picture, are familiar with the “jar- work on the spring, and this energy is stored in what we
gon”, these concepts become easy to use and powerful can call the “elastic potential energy” of the stretched
in their applications. Much of this chapter is to intro- spring. (We know that a stretched spring has some
duce the jargon and develop the picture. form of potential energy, for a stretched spring can be
used to launch a ball up into the air.)
The ideas of work and potential energy are closely
related and play critical roles in the picture of energy.
Let us discuss some examples simply from the point of Non-Constant Forces
view of getting used to the jargon. Our example above, of storing energy in a spring by
stretching it, introduces a new complication. We
Suppose I pick a ball of mass m off the floor and slowly cannot calculate the work I do Wme in stretching the
lift it up to a height h. While lifting the ball, I have to spring by writing Wme = Fme ⋅ S . The problem is that,
just barely overcome the downward gravitational force the farther I stretch the spring, the harder it pulls back
mg. Therefore I exert an upward directed force of (Hooke’s law). If I slowly pull the spring out, I have to
magnitude mg, and I do this for a distance h. Since my apply an increasingly stronger force. If we try to use the
upward force and the upward displacement are in the formula Wme = Fme ⋅ S , the problem is what value of
same direction, the work I do, call it Wme, is my force Fme to use. Do we use the weak Fme at the beginning
mg times the distance h, or Wme = mgh. Using the of the pull, the strong one at the end, or some average
ideas of potential energy discussed earlier, we can say value.
that all the energy Wme = mgh that I supplied lifting the
ball went into gravitational potential energy mgh. We could use an average value, but there is a more
general way to calculate the work I do. Suppose I wish
While I was lifting the ball, gravity was pulling down. to pull the spring from an initial position xi to a final
The downward gravitational force and the upward position xf. Imagine that I break this span from xi to xf
displacement were in opposite directions and therefore into a bunch of small intervals of width ∆x , ending at
the work done by the gravitational force was negative. points labeled x0, x1, ... xn as shown in Figure (7).
While we are storing gravitational potential energy, During each small interval the spring force does not
gravity does negative work. When we let go of the ball, change by much, and I can stretch the spring through
gravity releases potential energy by doing positive that interval by exerting a force equal to the strength of
work.
Fme

Figure 6
Doing work on a spring.
 10-15

the spring force at the end of the interval. For example 23 becomes the definite integral of Fs (x) from the
in stretching the spring from position x0 to x1, I apply initial position x i to the final position x f :
a force of magnitude Fs (x1 ) for a distance ∆x and xf
therefore do an amount of work
( W me ) total = – Fs(x) dx (24)
(∆Wme)1 = Fs (x1 )∆x
xi
To get out to position x2, I increase my force to Fs (x2 )
and apply that force over another interval ∆x to do an The statement of the work we did, Equation 24, can be
amount of work written more formally by noting that the spring force
Fs (x) is actually a vector which points opposite to the
(∆Wme)2 = Fs (x2 )∆x direction I pulled the spring. In addition, we should
If I keep repeating this process until I reach the final think of each ∆x or dx as a small vector displacement
position xf, the total amount of work I have done is ∆x or dx in the direction I pulled. Since my force was
directed opposite to Fs, the work I did during each
(Wme )total = (∆Wme )1 + (∆Wme )2 + ... interval dx can be written as the dot product

+ (∆Wme )n dWme = Fme ⋅ dx = – Fs ⋅ dx

= Fs(x1 )∆x + Fs (x2 )∆x + ... and the formula for the total work I did becomes

+ Fs (x n )∆x xf
n
( W me ) total = – Fs ⋅ dx (25)
= ∑ Fs(x i )∆x (23)
xi
i=1

In Equation 23, we still have an approximate calcula- Equation 25 is more general but a bit clumsier to use
tion as long as the intervals ∆x are of finite size. We than 24. To use Equation 25, we would first note that
get an exact calculation of the work I do if we take the I was pulling along the x axis, and thus dx = dx . Then
limit as ∆x goes to zero, and the number of intervals I would note that the spring force was opposite to the
goes to infinity. In that limit, the right side of Equation direction I was pulling, so that – Fs(x) ⋅ dx = +Fs (x)dx

∆x1 ∆x 2

x0 x1 x2 xn
xi xf
Figure 7
I can stretch the spring through a series of
small intervals of length ∆x. In each interval I
apply a constant force that is just strong enough
to get the spring to the end of the interval.
10-16 Energy

where Fs(x) is the formula for the strength of the spring connecting the cart to each end of the track as shown in
force. That gets me back to Equation (24) and the Figure (8). Mark the center of the cart with an arrow,
problem of evaluating the definite integral. and choose a coordinate system where x = 0 is at the
equilibrium position as shown in Figure (8a).
Potential Energy Stored in a Spring With this setup, the spring force is always a restoring
Springs are useful in physics demonstrations and prob- force that is pushing the cart back to the equilibrium
lems because of the simple force law (Hooke’s law) position x = 0. If we give the cart a positive displace-
which is quite accurately obeyed by real springs. In our ment as in Figure (8b), we get a left directed or negative
study of the motion of a ball on the end of a spring in spring force. A negative displacement shown in (8c)
Chapter 9, we saw that the formula for the strength of produces a right directed or positive spring force. And
the spring force was to a high degree of accuracy, the strength of the spring
Fs = K(S – S0 ) (9–6) force is proportional to the magnitude of the displace-
ment from equilibrium.
where S is the length of the spring and S0 the unstretched
length (the length at which Fs goes to zero in All of these results can be described by the formula
Figure 9–4).
Fs(x) = –Kx (26)
We can simplify the spring force formula, get rid of the
S0, by considering a situation where an object is held in
an equilibrium position by spring forces. Suppose for where the minus sign tells us that a positive displace-
example we have a cart on an air track with springs ment x produces a negative directed force and vice
versa. There is no S0 or x0 in Equation 26 because we
chose x = 0 to be the equilibrium position where Fs = 0.
Equation 26 is what one usually finds as a statement of
(a) Hooke’s law, and K is called the spring constant.
x = 0 equilibrium
Equation 26 allows us to easily calculate the potential
Fs energy stored in the springs. If I start with the cart at rest
at the equilibrium position as shown in Figure (8a), and
(b) pull the cart to the right a distance xf, the work I do is
x x=xf x=xf

Fs Wme = Fme dx = (–Fs) dx

x=0 x=0
(c) x=xf

x = Kx dx (27)
x=0
Figure 8
The spring force Fs is always opposite to the
where I have to exert a force Fme = –Fs to stretch the
displacement x. If the spring is displaced spring.
right, Fs points left, and vice versa.
 10-17

In Equation 27, the constant K can come outside the Exercise 11

integral, we are left with the integral of xdx which is
With a little bit of cleverness, we can use energy
x2 /2, and we get
conservation to predict the speed of the cart at any point
x=xf xf along the air track. Suppose you release the cart from
x2 x 2f rest at a distance xf , and want to know the cart’s speed
Wme = K x dx = K = K
2 0
2 at, say, xf /2. First calculate how much potential energy
x= 0 the cart loses in going from xf to xf /2, and then equate
Noting that all the work I do is stored as “elastic that to the kinetic energy 1/2 mv2 that the cart has
potential energy of the spring”, we get the formula gained at xf /2.

2
x
Spring potential energy = K (28)
2

In Equation 28, we replaced xf by x since the formula

applies to any displacement xf I choose.

Exercise 10
If you pull the cart of Figure (8) back a distance xf from
the equilibrium position and let go, all the potential
energy you stored in the cart will be converted to kinetic
energy when the cart crosses the equilibrium position x
= 0. Use this example of conservation of energy to
calculate the speed v of the cart when it crosses x = 0.
(Assume that you release the cart from rest.)

Exercise 10, which you should have done by now,

illustrates one of the main reasons for bothering to
calculate potential energy. It is much easier to predict
the speed of the ball using energy conservation than it
is using Newton’s second law. We can immediately
find the speed of the ball by equating the kinetic energy
at x = 0 to the potential energy at x = xf where we
released the cart. To make the same prediction using
Newton’s second law, we would have to solve a
differential equation and do a lot more calculation.
10-18 Energy

WORK ENERGY THEOREM The next step is a standard calculus trick that you may
The reason that it is easier to apply energy conserva- or may not remember. We will first move things
tion than Newton’s second law is that when we have around a bit in the integral on the right side of Equation
a formula for potential energy, we have already done 30:
much of the calculation. We can illustrate this by f f
deriving what is called the “Work Energy Theorem” dvx dx
where we use Newton’s second law to derive a m dx = m dvx (31)
dt dt
relation between work and kinetic energy. We will i i
first derive the theorem for one dimensional motion, Next note that dx/dt = vx, the x component of the
and then see that it is easily extended to motion in velocity of the particle. Thus the integral becomes
three dimensions.
f vf
Suppose a particle is moving along the x axis as shown dx
in Figure (9). Let a force Fx (x) be acting on the particle. m dvx = m vx dv x (32)
dt
Then by Newton’s second law i vi
dvx(x) After this transformation, we can do the integral be-
Fx(x) = ma x(x) = m (29) cause everything is now expressed in terms of the one
dt
variable vx . Using the fact that the integral of vx dvx is
Multiplying by dx and integrating to calculate the work v2x /2, we get
done by the force Fx, we get
f
vf vf
f
dvx(x) vx2
Fx(x)dx = m dx m vxdvx = m
dt (30) 2 v
i vi i
i
In Equation (30), we are integrating from some initial (33)
1 2 1 2
position x i where the object has a speed vxi , to a = mvfx – mvix
2 2
position x f where the speed is vxf.
Using Equations (31) through (33) in Equation (30)
Fx (x) gives
v
xf
x Fx(x)dx =
1 2
mvfx
1 2
– mvix
Figure 9 2 2
An x directed force acting on a xi
particle moving in the x direction. (34)
The left side of Equation 34 is the work done by the
force Fx as the particle moves from position xi to
position xf. The right side is the change in the kinetic
energy. Equation 34 tells us that the work done by the
force Fx equals the change in the particle’s kinetic
energy. This is the basic idea of the work energy
theorem.
 10-19

To derive the three dimensional form of Equation 34, Several Forces

start with Newton’s second law in vector form Suppose several forces F1, F2, ... are acting on the
F = ma (35) particle as the particle moves from position i to position
f. Then the vector F in Equations 35 through 41 is the
Take the dot product of Equation 35 with dx and
total force Ftot which is the vector sum of the individual
integrate from i to f to get
forces:
f f
F = Ftot = F1 + F2 + ... (42)
F ⋅ dx = ma ⋅dx (36)
Our formula for the work done by these forces becomes
i i
f f
Writing
F ⋅ dx = (F1 + F2 + ...) ⋅dx
a ⋅ dx = a x dx + a y dy + a z dz (37) i
i
we get
f f
f
f
dvx dvy dvz
= F1 ⋅ dx + F2 ⋅ dx + ... (43)
F ⋅ dx = m dx + dy + dz (38) i i
dt dt dt
i
i and we see that the work done by several forces is just
Following the same steps we used to get from Equation the numerical sum of the work done by each force
31 to 33, we get acting on the object. Equation 41 now has the interpre-
tation that the total work done by all the forces acting
f on a particle is equal to the change in the kinetic
1 1
F ⋅ dx = 2
mvfx 2
– mvix energy of the particle.
2 2
i
+ 1 mv 2fy – 1 mv 2iy
2 2
+ 1 mv 2fz – 1 mv 2iz (39)
2 2
Finally noting that by the Pythagorean theorem
vi2 = vix
2 + v2 + v2
iy iz

vf2 = vfx
2 + v2 + v2
fy fz (40)
we get, using (40) in (39)

f
1 1
F ⋅ dx = mvf2 – mvi2
2 2
i
(41)
which is the three dimensional form of the work energy
theorem.
10-20 Energy

Conservation of Energy Spring forces have the property that the energy stored
To see how the work energy theorem leads to the idea in the spring depends only on the length of the spring,
of conservation of energy, suppose we have a particle and not on how the spring was stretched. This means
subject to one force, like the spring force Fs acting on that the change in the spring’s potential energy does not
an air cart as shown in Figure (8). If the cart moves from depend upon whether I moved the cart, or I let go and
position i to position f, then the work energy theorem, the spring moves the cart. We should remove Fme from
Equation 41 gives Equation 45 and simply express the spring potential
energy in terms of the spring force
f
1 1 f
F ⋅ dx = mvf2 – mvi2 (44) change in spring
2 2 = – Fs ⋅dx (46)
i potential energy
i
In our analysis of the spring potential energy, we saw
that if I slowly moved the cart from position i to position Equation 46 says that the change in potential energy is
f, I had to exert a force Fme that just overcame the spring minus the work done by the force on the object as the
force Fs, i.e., Fme = –Fs. When I moved the cart object moves from i to f. There is a minus sign because,
slowly, the work I did went into changing the potential if the force does positive work, potential energy is
energy of the cart. Thus the formula for the change in released or decreases. We will see that Equation 46 is
the cart’s potential energy is a fairly general relationship between a force and it’s
associated potential energy.
change in the f
potential energy of We are now ready to convert the work energy theorem
the cart when the = F ⋅dx into a statement of conservation of energy. Rewrite
cart moves from Equation 44 with the work term on the right hand side
position i to position f i
and we get
(45)
f f
=– Fs ⋅dx 0 = – Fs ⋅dx +
1 1
mvf2 – mvi2 (47)
2 2
i i
Equation 45 is essentially equivalent to Equation 25 The term in the first curly brackets is the change in the
which we derived in our discussion of spring forces. particle’s potential energy, the second term is the
change in the particle’s kinetic energy. Equation 47
says that the sum of these two changes is zero
change in change in
0 = + (47a)
potential energy kinetic energy
If we define the total energy of the particle as the sum
of the particle’s potential energy plus its kinetic energy,
then the change in the particle’s total energy in moving
from position i to position f is the sum of the two
changes on the right side of Equation 47a. Equation
47a says that this total change is zero, or that the total
energy is conserved.
 10-21

Conservative and f
work I do
Non-Conservative Forces while moving =– Fme ⋅dx
We mentioned that the potential energy stored in a the eraser i
spring depends only on the amount the spring is
stretched, and not on how it was stretched. This means f
(49)
that the change in potential energy depends only on the = – Ffriction ⋅dx
initial and final lengths of the spring, and not on how we
i
stretched it. This implies that the integral
There are two problems with this example. The
f
integrals in Equation 49 do depend on the path I take.
– Fs ⋅dx If I move the eraser around in circles I do a lot more
i
work than if I move it in a straight line between the two
has a unique value that does not depend upon how the points. And when I get to position f, there is no stored
particle was moved from i to f. potential energy. Instead all the energy that I supplied
overcoming friction has probably been dissipated in
Gravitational forces have a similar property. If I lift an the form of heat.
object from the floor to a height h, the increase in
gravitational potential energy is mgh. This is true Physicists divide all forces in the world into two
whether I lift the object straight up, or run around the categories. Those forces like gravity and the spring
room five times while lifting it. The formula for the force, where the integral
change in gravitational potential energy is f

f F ⋅dx
change in
gravitational = – Fg ⋅dx i

potential energy depends only on the initial and final positions i and f,
i
are called "conservative" forces. For these forces there
h is a potential energy, and the formula for the change in
= – Fgy dy potential energy is minus the work the force does when
the particle goes from i to f.
0
h All the other forces, the ones for which the work
= – – mg dy integral depends upon the path, are called non-conser-
0 vative forces. We cannot use the concept of potential
energy for non-conservative forces because the for-
= mgh (48) mula for potential energy would not have a unique or
Again we have the change in potential energy equal to meaningful value. The non-conservative forces can do
minus the work done by the force. work and change kinetic energy, but as we see in the
case of friction, the work ends up as something else like
Not all forces, however, work like spring and gravita- heat rather than potential energy.
tional forces. Suppose I grab an eraser and push it
around on the table top for a while. In this case I am It is interesting that on an atomic scale, where energy
overcoming the friction force between the table and the does not disappear in subtle ways like heat, we almost
eraser, and we have Fme = – Ffriction . The total work always deal with conservative forces and can use the
done by me as I move the eraser from an initial position concept of potential energy.
i to a final position f is
10-22 Energy

GRAVITATIONAL POTENTIAL Zero of Potential Energy

ENERGY ON A LARGE SCALE Another difference is that the formula mgh for a ball in
In our computer analysis of satellite motion, we saw the room measures changes in gravitational potential
that the quantity E tot, given by energy starting from the floor where h = 0. In a rather
arbitrary way,we have defined the gravitational poten-
1 GM em tial energy to be zero at the floor. This is a convenient
Etot = mv 2 – (50) choice for people working in this room, but people
2 r
working upstairs or downstairs would naturally choose
was unchanged as the satellite moved around the earth. their own floors rather than our floor as the zero of
As shown in Figure (10), m is the mass of the satellite, gravitational potential energy for objects they were
v its velocity, R its distance from the center of the earth, studying.
and M e is the mass of the earth. This was our first non
trivial example of conservation of energy, where Since conservation of energy deals only with changes
1/2 mv2 is the satellite’s kinetic energy, and –GMem/R in energy, it does not make any difference where you
must be the formula for the satellites's gravitational choose your zero of potential energy. A different
potential energy. Our discussion of the last section choice simply adds a constant to the formula for total
suggests that we should be able to obtain this formula energy, and an unchanging or constant amount of
for gravitational potential energy by integrating the energy cannot be detected. The most famous example
gravitational force F g = GMem/r 2 from some initial of this was the fact that a particle’s rest energy m0c2 was
to some final position. unknown until Einstein introduced the special theory
of relativity, and undetected until we saw changes in
Here on the surface of the earth, the formula for rest energy caused by nuclear reactions. In the case of
gravitational potential energy is mgh. This simple the gravitational potential energy of a ball, if we use the
result arises from the fact that when we lift an object floor downstairs as the zero of gravitational potential
inside a room, the strength of the gravitational force mg energy, we add the constant term (mg)h floor to all our
acting on it is essentially constant. Thus the work I do formulas for E tot (where h floor is the distance between
lifting a ball a distance h is just the gravitational force floors in this building). This constant term has no
mg times the height h. Since this work is stored as detectable effect.
potential energy, the formula for gravitational potential
energy is simply mgh. In finding a formula for gravitational potential energy
of satellites, planets, stars, etc., we should select a
In the case of satellite motion, however, the strength of convenient floor or zero of potential energy. For the
the gravitational force was not constant. In our first motion of a satellite around the earth, we could choose
calculation of satellite motion in Chapter 8, the satellite gravitational potential energy to be zero at the earth’s
started 1.1 earth radii from the center of the earth and surface. Then the satellite’s potential energy would be
went out as far as r = 5.6 earth radii. Since the positive when its distance r from the center of the earth
gravitational force drops off as 1 r2 , the gravitational is greater than the earth radius r e , and negative should
force was more than 25 times weaker when the satellite r become less than r e . Such a choice would be
was far away, than when it was launched. reasonable if we were only going to study earth satel-
v lites, but the motion of a satellite about the earth is very
m closely related to the motion of the planets about the sun
and the motion of moons about other planets. Choos-
ing r = r e as the distance at which gravitational poten-
r tial energy is zero is neither a general or particularly
convenient choice.

Me
Figure 10
Earth satellite.
 10-23

In describing the interaction between particles, for (In the analogy between potential energy and a check-
example an electron and a proton in a hydrogen atom, ing account, the discovery of rest energy m0c2 would
the earth and a satellite, the sun and its planets, or the be like discovering that you had inherited the bank. The
stars in a galaxy, the convenient choice for the zero of checks still work the same way even though your total
potential energy is where the particles are so far apart assets are vastly different.)
that they do not interact. If the earth and a rock are a
hundred light years apart, there is almost no gravita- Let us now return to Equation (50) and our formula for
tional force between them, and it is reasonable that they gravitational potential energy of a satellite
do not have any gravitational potential energy either. GMem
gravitational
= – (50a)
Now suppose that the earth and the rock are the only potential energy r
things in the universe. Even at a hundred light years
there is still some gravitational attraction, so that the First we see that if the satellite is very far away, that as
rock will begin to fall toward the earth. As the rock gets r goes to infinity, the potential energy goes to zero.
closer to the earth it will pick up speed and thus gain Thus this formula does give zero potential energy when
kinetic energy. It was the gravitational force of attrac- the earth and the satellite are so far apart that they no
tion that caused this increase in speed, therefore there longer interact. In addition, the potential energy is
must be a conversion of gravitational potential energy negative, as it must be if the satellite falls in to a distance
into kinetic energy. r, converting potential energy into kinetic energy.

This gives rise to a problem. The rock starts with zero What we have to do is to show that Equation (50a) is in
gravitational potential energy when it is very far away. fact the correct formula for gravitational potential
As the rock approaches the earth, gravitational poten- energy. We can do that by calculating the work gravity
tial energy is converted into kinetic energy. How can does on the satellite as it falls in from r = ∞ to r = r. This
we convert gravitational potential energy into kinetic work, which would show up as the kinetic energy of a
energy if we started with zero potential energy? falling satellite, must be the amount of potential energy
spent. Thus the potential energy balance must be the
Keeping track of energy is very much a bookkeeping negative of this work. Since the work is the integral of
scheme, like keeping track of the balance in your bank the gravitational force times the distance, we have
account. Suppose you begin the month with a balance
of zero dollars, and start spending money by writing
checks. If you have a trusting bank, this works because
your bank balance simply becomes negative.

In much the same way, the rock falling toward the earth
started with zero gravitational potential energy. As the
rock picked up speed falling toward the earth, it gained
kinetic energy at the expense of potential energy. Since
it started with zero potential energy, and spent some, it
must have a negative potential energy balance. From
this we see that if we choose gravitational potential
energy between two objects to be zero when the objects
are very far apart, then the potential energy must be
negative when the objects are a smaller distance apart.
When we think of energy conservation as a bookkeep-
ing scheme, then the idea of negative potential energy
is no worse than the idea of a negative checking account
balance.
10-24 Energy

gravitational R Carrying out the integral in Equation 51 gives

potential energy = – Fg ⋅ dr R R
at position R GM em dr
∞ dr = GM em
r2 r2
∞ ∞
R
GMm R
= – dr (51) GM em 1 1
2 =– = GM em –
r r ∞
∞ R
∞
where we used the fact that the integral of 1 r 2 is – 1 r .
Equation 51 may look a bit peculiar in the way we have Thus we get
handled the signs. We have argued physically that the R
gravitational potential energy must be negative, and GM em GM em
we know that it must be equal in magnitude to the dr =
integral of the gravitational force from r = ∞ to r = R. r2 R
∞
By noting ahead of time what the sign of the answer
must be, we can do the integral easily without keeping As a result the gravitational potential energy of the
track of the various minus signs that are involved. (One satellite a distance R from the center of the earth is
minus sign is in the formula for potential energy, – GMem/R as given in Equation 50a.
another is the dot product since Fg points in and dr out,
a third in the integral of r – 2, and more come in the
evaluation of the limits. It is not worth the effort to get
all these signs right when you know from a simple
physical argument that the answer must be negative.)
 10-25

Gravitational Potential Since h/Re is a very small number compared to one, we

Energy in a Room can use our small number approximation
Before we leave our discussion of gravitational poten- 1
tial energy, we should show that the formula – GMem/r ≈ 1–α if α << 1
1+α
leads to the formula mgh for the potential energy of a
ball in a room. To show this, let us use the formula to write
– GMem/r to calculate the increase in gravitational 1
potential energy when I lift a ball from the floor, a ≈ 1– h
h
1 + Re Re
distance Re from the center of the earth, up to a height
h, a distance Re+ h from the center of the earth, as so that
shown in Figure (11). 1 ≈ 1 1 – h = 1 – h2 (53)
Re + h Re Re Re Re
We have
Using Equation 53 in (52) gives
GMem
PE at height h = –
Re + h
Increase 1 1 h
= – GM em – –
GMem in PE Re Re R 2
e
PE at floor = –
Re
h
Increase = GM em
in PE
= PE at h – PE at floor R e2

GM em GM em GM e
= – – – = mh
Re + h Re R e2

= GM em 1 – 1 (52)
Re Re + h
Finally noting that GMe/ Re2 = g, the acceleration
To evaluate the right side of Equation 52, we can write due to gravity at the surface of the earth, we get

1 = 1 1
Re + h Re 1 + h R Increase = mgh
e in PE

h which is the expected result.

Re Re +

Figure 11
A height h above the surface of the earth.
10-26 Energy

SATELLITE MOTION The gravitational potential energy of the satellite is

AND TOTAL ENERGY always negative. Since the satellite is a distance r from
the center of the earth, its potential energy is
Consider a satellite moving in a circular orbit about the
earth, as shown in Figure (12). We want to calculate the potential GMm
kinetic energy, potential energy, and total energy (sum energy = – r
of the kinetic and potential energy) for the satellite. To The total energy of the satellite is
find the kinetic energy, we analyze its motion, using
Newton's laws. The only force acting on the satellite is potential
the gravitational force Fg given by Etotal = kinetic
energy + energy

Fg directed = 1 GMm + – GMm

Fg = GMm toward the earth 2 r r
r2
Etotal = – 1 GMm (54)
where we now let M = mass of the earth and m = mass 2 r
of the satellite. Since the satellite is moving at constant The total energy of a satellite in a circular orbit is
speed v in a circle of radius r, its acceleration is v 2 /r negative.
toward the center of the circle
2 Now consider a satellite in an elliptical orbit. In particu-
a = vr a directed
toward the earth lar, suppose that the orbit is an extended ellipse, as
shown in Figure (13). At apogee, the farthest point
Since a and Fg are in the same direction, by Newton's from the earth, the satellite is moving very slowly
second law (explain why by using Kepler's law of equal areas). For
all practical purposes, the satellite drifts out, stops at
Fg = m a apogee, then falls back toward the earth. At apogee, the
satellite has almost no kinetic energy; at this point its
GMm = mv 2
r2 r total energy is nearly equal to its negative potential
energy
From this last equation we find that the kinetic energy
1 2mv 2 of the satellite is Etotal = – GMm
rapogee
1 mv 2 = 1 GMm kinetic
2 2 r energy
r apogee
The kinetic energy, as always, is positive.

apogee earth

v
r Figure 13
Satellite in a very eccentric orbit. By Kepler's
m
M Fg law of equal areas, a satellite with the above
orbit would almost be at rest at apogee.

Figure 12
Satellite in a circular orbit.
 10-27

Since the total energy is conserved, Etotal remains Even at enormous distances, where the negative poten-
negative throughout the orbit. If similar satellites are tial energy – GMm /r is about zero, some kinetic
placed in different orbits, the one that goes out the energy would still remain, and the satellite would
farthest (has the greatest rapogee) is the one with the least escape from the earth!
negative total energy, but all the satellites in elliptical
orbits will have a negative total energy. By choosing potential energy to be zero when the
satellite is very far out, the total energy becomes a
Suppose an extra powerful rocket is used and a satellite meaningful number in itself. If the total energy is
is launched with a positive total energy. In such a case, negative, the satellite will remain bound to the earth; it
the positive kinetic energy must always exceed the does not have sufficient energy to escape. If a satellite
negative potential energy. No matter how far out the launched with positive total energy, it must escape
satellite goes, headed for apogee, it will always have since the negative gravitational potential energy is not
some positive kinetic energy to carry it out farther. sufficiently great to bind the satellite to the earth. If the
satellite's total energy is zero, it barely escapes.

The orbits of comets about the sun are interesting

examples of orbits of different total energies. It can be
shown that when a satellite's total energy is positive, its
hyperbola a) orbit will be in the shape of a hyperbola, which is an
open-ended curve, as shown in Figure (14a). In this
orbit the comet has a positive total energy and never
returns.

If the total energy of the comet is zero, the orbit will be

in the shape of an open curve, called a parabola (Figure
14b). A comet in this kind of orbit will not return either.
parabola b) When the comet's total energy is slightly less than zero,
it must return to the sun. In this situation the comet's
orbit is an ellipse, even though it may be a very
extended ellipse. A comparison of an extended ellipse
and a parabola is shown in Figure (14c). From this
figure we can see that near the sun there is not much
difference in the motion of a comet with zero or slightly
negative total energy. The difference can be seen at a
great distance, where the zero-energy comet continues
ellipse c)
to move away from the sun, but the slightly negative-
energy comet returns.

The circular, or nearly circular, motion of the planets is

a limiting case of elliptical motion. The small circular
circle d)
orbits (Figure 14d) are occupied by planets that have
Figure 14 large negative total energies. Thus the planets are
a) Hyperbolic orbit of comet with positive total tightly bound to the sun.
energy.
b) Parabolic orbit of comet with zero total energy.
c) Elliptical orbit of comet with slightly negative total
energy. (Dashed lines show parabolic orbit for
comparison.)
d) Nearly circular orbits of the tightly bound (large
negative energy) planets.
10-28 Energy

Example 4 Escape Velocity Converting this to more recognizable units, such as mi/
At what speed must a shell be fired from a super cannon sec, we have
in order that it escapes from the earth? Does it make any 1
difference at what angle the shell is fired, so long as it vescape = 1.12 × 10 6 cm/sec ×
1.6 × 10 5 cm/mi
clears all obstructions? (Neglect air resistance.)
= 7 mi/sec (11.2 km / sec )
Solution: If the shell is fired at a sufficiently great initial
This is also equal to 25,200 mi/hr, which is far faster
speed so that its total energy is positive, it will eventu-
than the initial velocity required to put a satellite in an
ally escape from the earth, regardless of the angle at
orbit 100 mi high.
which it is fired (so long as it clears obstructions). To
calculate the minimum muzzle velocity at which the
shell can escape, we will assume that the shell has zero Exercise 12
total energy, so that it barely escapes. When Etotal = 0 Calculate the escape velocity required to project a
we have just after the shell is fired shell permanently away from the moon
( mmoon = 7.35 × 1025 gm, rmoon = 1.74 × 108 cm ) .
GMe m
0 = 1 mv 2 – re
2 Exercise 13
which gives Once a shell has escaped from the earth, what must its
speed be to allow it to escape from our solar system?
2GMe
v2 = re (55)
Exercise 14
Putting in numbers Find the escape velocities from the earth and the moon,
using the planetary units given on page 8-14.
G = 6.67 × 10 –8 cm3 /gm sec2
Me = 5.98 × 10 27 gm
re = 6.38 × 10 8 cm
we get

2× 6.67 × 10 –8 cm3 /gm sec2 5.98 × 10 27 gm

v2 =
6.38 × 10 8 cm

10 –8 × 10 27
= 2× 6.67 × 5.98 cm3 gm/gm cm sec2
6.38 10 8
= 1.25 × 10 12 cm2 /sec2

vescape = 1.12 × 10 6 cm/sec

 10-29

BLACK HOLES We now finish our discussion of satellite motion by

A special feature of satellite motion we have just seen going to the opposite extreme, and consider the behav-
is that we can tell whether or not a satellite can escape ior of particles whose kinetic energy is much greater
simply by comparing kinetic energy with the gravita- than their rest energy. Such a particle must be moving
tional potential energy. If the satellite's positive kinetic at a speed very close to the speed of light. We
energy is greater in magnitude than the negative gravi- considered such a particle in Exercise 7 of Chapter 6.
tational potential energy, then the satellite escapes, There we saw that electrons emerging from the Stanford
never to return on its own. This is true no matter how linear accelerator travelled at a speed v =
or from where the satellite is launched (provided it does .9999999999875 c, and had a mass 200,000 times
not crash into something.) greater than the rest mass. For such a particle, almost
all the energy is kinetic energy. In the formula E = mc 2 ,
So far we have limited our discussion to slowly moving only one part in 200,000 represents rest energy.
objects where the approximate formula 1/2 mv 2 is
adequate to describe kinetic energy. We got the Actually we wish to go one step farther, and discuss
formula 1/2 mv 2 back in Equation 7 by expanding particles with no rest energy, particles that move at the
E = mc 2 to get speed of light. The obvious example, of course, is the
photon, the particle of light itself.
m 0c 2
E = mc 2 = From one point of view there is not much difference
1 – v 2/c 2 between an electron travelling at a speed
≈ m 0c 2 + 1/2 m 0v 2 (7) .9999999999875 c with only 1 part in 200,000 of its
rest energy kinetic energy
energy in the form of rest energy, and a photon travel-
ing at a speed c and no rest energy. Taking this point
The basic idea is that Einstein's formula E = mc 2 gives of view, we will take as the formula for the energy of
us a precise formula for the sum of the rest energy and a photon E = mc 2 , and assume that this is pure kinetic
the kinetic energy. In the special case the particle is energy.
moving slowly, we can use the approximate formula
for 1 – v 2/c 2 to get the result shown in Equation 7. Applying the formula E = mc 2 to a photon implies
that a photon has a mass m p = E p/c 2 . We will now
For familiar objects like bullets, cars, airplanes, and make the assumption that this mass m photon is gravi-
rockets, the kinetic energy is 1/2 m 0v 2 , much, much, tational mass, and that photons have a gravitational
much smaller than the rest energy m 0c 2 . The kinetic potential energy – GMm p/r like other objects. Our
energy of a rifle bullet, for example, is enough to allow assumption, which is slightly in error, is that Newtonian
the bullet to penetrate a few centimeters into a block of gravity, which is a non relativistic theory, applies to
wood. The rest energy of the bullet, if converted into particles moving near to or at the speed of light. It turns
explosive energy, could destroy a forest. In fact, one out that Einstein's relativistic theory of gravity gives
way to tell whether or not the approximate formula almost the same answers, that we are seldom off by
1/2 m 0v 2 is reasonably accurate, is to check whether more than a factor of 2 in our predictions.
the kinetic energy is much less than the rest energy. If
it is, you can use the approximate formula; if not, you
can't.
10-30 Energy

Suppose we have a photon a distance r from a star of From Equation 58 we see that when a photon is as close
mass M s . If the photon has a mass m p , then the as it can get to the surface of the sun, the gravitational
formula for the total energy of the photon, its kinetic potential energy contributes very little to the total
energy m pc 2 plus its gravitational potential energy energy of the photon, only 2 parts in a million.
– GM s m p/r is
However, suppose that the a star had the same mass as
GM s m p the sun but a much, much smaller radius. If it's radius
E tot = m pc 2 – r R s were small enough, the factor 1 – GM s/R sc 2 in
Since m p appears in both terms, we can factor it out Equation 58 would become negative, and a photon
(and also take out a factor of c 2 ) to get grazing the surface of this star would have a negative
total energy. The photon could not escape from the star.
GM s No photons emerging from the surface of such a star
E tot = m pc 2 1 – (56)
rc 2 could escape, and the star would cease to emit light.

Equation 56 applies only when the photon is outside the Let us see how small the sun would have to be in order
star, i.e., when the distance r is greater than the radius that it could no longer radiate light. That would happen
R of the star. when the factor 1 – GM s /R s c 2 is zero, when photons
emerging from the surface of the sun have zero total
In most cases, the gravitational potential energy is energy. Putting in numbers we get
much less than the kinetic energy of a photon, and
GM s
gravity has little effect on the motion of the photon. For = 1 (59)
example, if a photon were grazing the surface of the sun R sc 2
(if r in Equation 56) were equal to the sun's radius R sun )
GM s
we would have Rs =
c2
GM s 3
E tot = m pc 2 1 – (57) 6.67 × 10 – 8 cm 2 × 1.99 × 10 33cm
Rs c2 = cm sec
2
3 × 10 10 cm 2/sec 2
Putting in numbers M s = 1.99 × 10 33gm ,
R s = 6.96 × 10 10cm we have
R s = 1.48 × 10 5cm
3
6.67 × 10 – 8 cm 2 × 1.99 × 10 33gm = 1.48 kilometer (60)
GM s gm sec
=
2 2 2 Equation 60 tells us that an object with as much mass
Rs c 6.96 × 10 10cm × 3 × 10 10 cm 2
sec as the sun, confined to a sphere of radius less than 1.48
kilometers, cannot radiate light. Although we used the
= .00000212
non relativistic Newtonian gravity in this calculation,
Thus Einstein's relativistic theory of gravity makes the same
prediction.
E tot = m pc 2 1 – .00000212 (58)
 10-31

In discussions of black holes, one often sees a reference A Practical System of Units
to the radius of the black hole. What is usually meant In the CGS system of units, where we measure distance
is the radius given by Equation 59, the radius at which in centimeters, mass in grams and time in seconds, the
light can no longer escape if a mass M s is contained unit of force is the dyne ( 1 dyne = 1 gm cm/sec 2 ) and
within a sphere of radius R s . the unit of energy is the erg ( 1 erg = 1 gm cm 2 /sec 2 ).
We have found the CGS system quite convenient for
Do black holes exist? Can so much mass be concen- analyzing strobe photographs with 1 cm grids. But
trated in such a small sphere? The question has been when we begin to talk about forces and particularly
difficult to answer because black holes are hard to energy, the CGS system is often rather inconvenient. A
observe since they do not emit light. They have to be force of one dyne is more on the scale of the force
detected indirectly, from the gravitational pull they exerted by a fly doing push-ups than the kind of forces
exert on neighboring matter. In the sky there are many we deal with in the lab. A baseball pitched by Roger
binary star systems, systems in which two stars orbit Clemens has a kinetic energy of over a million ergs and
about each other. In some examples we have observed a 100 watt light bulb uses ten million ergs of electrical
a bright star orbiting about an invisible companion. energy per second. The dyne and particularly the erg
Careful analysis of the orbit of the bright star suggests are much too small a unit for most every day situations.
that the invisible companion may be a black hole.
There is recent evidence that gigantic black holes, with In the MKS system of units, where we measure dis-
the mass of millions of suns, exists at the center of many tance in meters, mass in kilograms and time in seconds,
galaxies, including our own. the unit of force is the newton and energy the joule. The
force required to lift a 1 kilogram mass is 9.8 newtons
That a black hole cannot radiate light is only one of the (mg), and the energy of a Roger Clemens’ pitch is over
peculiar properties of these objects. When so much 10 joules. When working with practical electrical
matter is concentrated in such a small volume of space, phenomena, the use of the MKS system is the only
the gravitational force becomes so great that other sensible thing to do. The unit of power, the watt, is one
forces cannot resist the crushing force of gravity, and as joule of energy per second. Thus a 100 watt light bulb
far as we know, the matter inside the black hole consumes 100 joules of electrical energy per second.
collapses to a point—a zero sized or very, very small Volts and amperes are both MKS units, the corre-
sized object. At the present time we do not have a good sponding CGS units are statvolts and esu, which are
theory for what happens to the matter inside a black almost never used.
hole. (We will have more to say about black holes in
later chapters.) Where CGS units are far superior is in working with the
basic theory of atoms, as for the case of the Bohr theory
Exercise 15 discussed in Chapter 36. This is because the electric
Studies of the motion of the stars in our galaxy suggests force law has a much simpler form in CGS units. What
that at the center of our galaxy is a large amount of mass we will do in the text from this point on is to use MKS
concentrated in a very small volume. For this problem, units almost exclusively until we get through the
assume that a mass of 100 million suns is concentrated chapters in electrical theory and applications. Then we
in the small volume. If this massive object is in fact a will go back to the CGS system in most of our discus-
black hole, what is the radius from which light can no sions of atomic phenomena.
longer escape?
Chapter 11
Systems of Particles

CHAPTER 11 SYSTEMS OF
PARTICLES

So far in our applications of Newton’s second law, we Diver - Movie

have treated objects as individual point particles. In For a student project, Tobias Hays was videotaped
some problems, this appeared to be a reasonable doing a series of dives, a few of which are shown in
approximation. But in others, like the motion of an this movie. Working frame by frame, you can see
apple falling from a tree, it appeared to be a remark- that the diver's center of mass follows a parabolic
able result that the entire earth could be treated as a trajectory. (It was actually more instructive to use a
parabolic trajectory to locate the diver's center of
point particle located at the center of the earth. Newton mass in various positions.)
supposedly invented calculus to show that one could do
this.

In this chapter we will look at ways to handle systems

consisting of many particles. In particular, we will see
that often the concept of center of mass allows us to
treat a group of interacting particles as a single par-
ticle. This can lead to an enormous simplification of the
analysis and a clearer understanding of the result.
11-2 Systems of Particles

CENTER OF MASS sents circular motion about a moving center. It looks

To introduce the concept of center of mass, let us begin like the drawing of epicycles from a text on ancient
with some examples of mechanical systems that at first Greek astronomy.
appear fairly complex, but which are greatly simplified For a similar but much more difficult problem, con-
when we focus our attention on the motion of the center sider the globular cluster shown in Figure (2). Globular
of mass. clusters are fairly common objects in our galaxy. A
Consider the motion of the earth about the sun. To typical globular cluster is a swarm of several million
analyze the problem, we could first treat the sun as a stars all attracting each other to form a single gravita-
point mass fixed at the origin and apply Newton’s tionally bound body. We can think of it as a confined
second law to the earth to determine the earth’s ellipti- gas of stars, confined not by a bottle or a rubber balloon
cal orbit. On closer inspection, however, we note that but by the gravitational attraction between the stars.
the earth is not a point particle but an earth-moon The globular clusters in our galaxy lie outside the main
system. A more accurate treatment of the problem body of stars of the galaxy, typically orbiting around or
requires us to consider two interacting particles both through the galaxy. If we wished to calculate the orbit
orbiting the sun. of a globular cluster, our first approximation would be
With a computer program it is not too difficult to set up to treat it as a point particle. To do better than that with
the earth-moon-sun system and directly solve for the a computer program, we might try to analyze the
motion of the earth and moon. Because the sun is so motion of each star as we did in the earth-moon-sun
massive, it is still a good approximation to place the sun problem above. But the futility of doing this would
at rest at the origin of the coordinate system. We then soon become obvious. Each of the millions of stars
have the earth subject to the gravitational force of the interact not only with the gravitational force of the
sun and the moon, and the moon experiencing the galaxy, but with each other. Each star is subject to
gravitational force of the sun and the earth. If we millions of forces, and any direct computer calculation
include all of these forces in the program, and start off becomes impossibly complex. One might try to sim-
with reasonable initial conditions, we will get the plify the problem by considering a cluster of only a few
expected result that the earth goes about the sun in an
elliptical orbit with a slight wobble, and the moon goes
around the earth in a nearly circular orbit. If we plot the
moon’s orbit, exaggerating the orbit radius as in Figure
(1), the orbit looks somewhat peculiar because it repre-

Moon
Sun
Earth

Figure 1
Motion of the moon around the sun. In this
sketch, we have greatly exaggerated the size
of the moon's orbit about the earth in order Figure 2
to show the epicycle like motion of the Globular cluster (NCG 5272).
moon.
 11-3

hundred stars, but even then a lot of time on a super Center of Mass Formula
computer is needed for a meaningful prediction. Suppose that we have a collection of n particles
m 1 , m 2 , m n , as shown in Figure (4). The coor-
Despite all the forces involved, the motion of the cluster dinate vector for particle m1 is R1, that for m 2 is R2, etc.
can easily be analyzed if we focus on the motion of the We define the total mass M of the collection of
center of mass of the cluster. When we calculate the particles as simply the sum of the individual masses
motion of the center of mass, all internal forces cancel,
and we have to consider only the net force of the galaxy n
on the total mass of the stars. In the earth-moon-sun
problem, the center of mass of the earth-moon system
M = m1 + m2 + mi = ∑ mi (1)
i=1
travels in an elliptical orbit around the sun. The earth
and moon each orbit about this center of mass point. The coordinate vector R com of the center of mass point
The calculation of the motion of the center of mass of is then defined by the formula
the earth-moon is the same as calculating the motion of
a point planet about the sun. formula for
The idea of the center of mass is a familiar concept for MR com = ∑ m iR i centerof mass (2)
i coordinate
it is a balance point. The center of mass of a long thin
rod is the point where the rod balances on your finger.
If you mark the balance point and throw the rod in the Since this is a vector equation, it can be written as the
air, giving it a spin, the balance point follows a smooth collection of the three scalar equations
parabolic trajectory while the rest of the rod rotates
about the balance point as shown in Figure (3). The MXcom = ∑ m ix i (2a)
balance point or center of mass moves as if all the mass i
of the rod were concentrated at that point.

Although the idea of a center of mass or balance point MYcom = ∑ m iy i (2b)

i
is fairly straightforward, the formula for the center of
mass of a collection of particles looks a bit peculiar at
first. But when you get used to the formula, you will MZ com = ∑ m i z i (2c)
find that it is fairly easy to remember and leads to i
impressive simplifications when used in Newton’s m3
laws. m1

R3

ck
m2
sti
er
met
me Acme meter stick
Ac
k
stic
ter

R1
e me

M
Acm

Acm
e
m
et
er
st
ick

R2
m4
er stick
e met
Acm

R com
Acme
meter
stick

Ac
m
e
m
et
er
stic
k

R4
Ac

m5
me
me
ter
stick

R5
Acme meter stick

Figure 4
Figure 3
To calculate the center of mass of a collection of
A meter stick tossed in the air rotates about its center particles, you start by constructing the coordinate
of mass (balance point). The center of mass itself vector for each particle. You then use Equation 2 to
travels along the parabolic path of a point projectile. calculate the center of mass coordinate vector.
11-4 Systems of Particles

where Xcom, Ycom and Z com are the x, y and z Exercise 1

coordinates of the center of mass, and x i , yi , and z i The formula for center of mass looks like it depends on
are the x, y, z coordinates of the i-th particle. where you place the origin of the coordinate system. To
see that this is not true recalculate X com for the two
To see that Equation 2 does give the balance point of a
masses in Figure (5), placing the origin at m2 , and show
collection of particles, let us consider the simple case of that the pivot point comes out at the same place on the
a horizontal massless rod of length l with two masses rod.
m1 and m2 at each end, as shown in Figure (5). If m1
is placed at the origin of the coordinate system, then Exercise 2
Equation 2a gives Show that the center or mass of the earth-moon system
is located inside the earth.
M X com = m1 × 0 + m2 l
Exercise 3
X com = m 2 l = m+2 m
l
M m 1 2 An ammonia molecule consists of a nitrogen atom and
three hydrogen atoms located on the corners of a
If m1 and m 2 are the masses of two children on a tetrahedron, as shown in Figure (6). Locate the center
seesaw of length l, then the pivot should be placed a of mass of the ammonia molecule.
distance X com from m1 for the children to balance. (As
a quick check, if m1 = m 2 , then X com = l/2 which is Dynamics of the Center of Mass
obviously correct.)
Let us now see how the concept of center of mass can
[If you want to calculate the center of mass of a be used to handle the dynamic behavior of a system of
continuous object like an irregularly shaped sheet of particles. Suppose we have a system with n particles.
plywood, you can mark the plywood off into small The formula for their center of mass coordinate R com
sections of mass dm i located at r i. Then M is the mass is from Equation 2
of the whole sheet of plywood, and the x coordinate of M R com = m1 R 1 + m2 R 2 + + mn R n (2)
the center of mass is
If we differentiate Equation 2 with respect to time, and
MX com = Σ x i dm i →
i
x dm note that the velocity Vcom is given by dR com/dt , we
get
where x i is the x coordinate of r i. You can then replace MVcom = m 1v1 + m 2v2 + + m nvn (3)
the sum over the dm i by an integral. This is a typical
problem treated in an introductory calculus course where vi = dR i /dt is the velocity of the i-th particle.
and will not be discussed further here.] N

l
m1 m2

X com
H H
Figure 5
Calculating the center of mass for two particles. H
Figure 6
Structure of the ammonia molecule.
 11-5

Equation 3 already has an interesting interpretation. Let us now write out Equation 5 as it might be applied
Since m1v1 is the linear momentum of particle 1, m2v2 to the globular cluster:
that of particle 2, etc., we see that M Vcom is equal to the
vector sum of the linear momenta of all the particles MA com = FG1 + F21 + F31 +
under consideration. We will come back to this point
when we discuss the concept of linear momentum in + FG2 + F12 + F32 + (6)
more detail.
+
Differentiating Equation 3 with respect to time, noting
that A com = dVcom /dt is the acceleration of the center where FG1 is the force of the galaxy on star #1, F21 is
of mass point, we get the force of star #2 on star #1, F31 the force of star #3
on star #1, etc. In the next collection of terms we have
MA com = m1 a 1 + m2 a 2 + + mn a n (4) FG2 , the force of the galaxy on star 2, F12 the force of
star #1 on star #2, etc.
where a i = dvi /dt is the acceleration of the i-th par-
ticle. Rearranging the order in which we write the forces, we
get
The final step is to use Newton’s second law to replace
MA com = FG1 + FG2 + FG n
ma i by the vector sum of the forces acting on particle
i. Calling this sum Σ Fi , we get
+ F21 + F12
MA com = Σ F1 + Σ F2 + + Σ Fn (5) (7)
+ F31 + F13 +
Equation 5 tells us that MA com is equal to the sum of
every force acting on all the particles. In Equation 7, we have separated the external forces
FG1 + FG2 + exerted by the galaxy on the indi-
If we wish to apply Equation 5 to the motion of
vidual stars from the internal forces like F12 and F21
something as complex as a globular cluster, it looks like
between stars in the cluster. All the internal forces can
we are still in trouble because, as we have mentioned,
be grouped in pairs, like F12 + F21 , the force of star
each of the millions of stars in the cluster interacts with
#2 on star #1 plus the force of star #1 on star #2.
all the other stars in the cluster. Acting on each star is
Because these are equal and opposite forces, all the
the gravitational force of the galaxy plus the millions of
pairs of internal forces cancel (over a trillion pairs for
forces exerted by the other stars.
an average cluster), and we are simply left with the
But Newton’s law of gravity provides an enormous vector sum of the external forces. In our cluster
simplification. When two objects interact gravitation- example, the vector sum of all the external forces is just
ally, they exert equal and opposite forces on each other the net force Fext the galaxy exerts on the cluster, and
as shown in Figure (7). In that case Fg1 = –Fg2 or if we are left with the fantastically simple result
we add these two forces of interaction we get
equation for
Fg1 + Fg2 = 0 MA com = F center of (8)
ext
mass motion

Fg1 Fg2 Equation 8 tells us that the center of mass of the

m1 m2
globular cluster moves exactly as if the cluster were a
r single mass point of mass M equal to the total mass of
the cluster subject to a single force Fext equal to the
Fg1 = – Fg2 total gravitational force exerted by the galaxy on all the
Figure 7 stars in the cluster. This remarkable result explains
Two objects exert equal and
opposite forces on each other.
11-6 Systems of Particles

why we can often represent a complex system by a NEWTON’S THIRD LAW

single mass point in the analysis of the system’s In our analysis of the motion of a globular cluster, the
behavior. great simplification came when all the internal forces
When this result is applied to the earth-moon system canceled in pairs in Equation 7, and we were left with
shown in Figure (8), we have the following picture. the result that the acceleration of the center of mass
When we calculate the motion of the center of mass of point was determined solely by the external forces
the earth and moon about the sun, the internal forces acting on the swarm of stars. The cancellation occurred
between the earth and moon cancel, and we are left with because the gravitational attraction between two ob-
jects is equal in magnitude and oppositely directed.
Mearth +Mmoon A com = Fsun on earth +Fsun on moon Fg1 Fg2
m1 m2
where Fsun on earth is the force of the sun on the earth,
and Fsun on moon that of the sun on the moon. The center What about other forces? What if two stars are in
of mass moves with the same acceleration as a point collision? Is the force between them still equal and
particle of mass (Mearth + Mmoon ) subject to the total opposite? If not, then we would have to take internal
force the sun exerts on the two. This results in an forces into account when predicting the motion of the
elliptical orbit for the center of mass. center of mass? Unable to believe that this would
happen, Newton proposed that when two bodies inter-
Exercise 4 act, then the force between them is always equal in
Two air carts of equal mass are connected by a spring magnitude and oppositely directed, no matter what
as shown in Figure (9). A small black marker is placed forces are involved. This assumption is known as
at the center of the spring which remains at the center Newton’s Third Law of Mechanics. The third law
of mass of the carts. One of the carts is given a shove guarantees that internal forces cancel in pairs and that
to the right so that the two carts move off to the right in
center of mass motion is determined only by external
an undulating drift. Describe the motion of the black
forces.
marker.
[Newton’s first law, as we have mentioned, is that in the
absence of any external forces, an object will move with
Figure 9
uniform motion. Although there is a direct conse-
quence of the second law F = ma , Newton explicitly
stated the result, because it was not such an obvious
idea in Newton’s time, when a horse and buggy was the
smoothest ride available.
earth In a traditional course in Newtonian mechanics,
Newton’s three laws are presented at the beginning as
moon
basic postulates and everything else is derived from
them. This was a logical approach for over 200 years
during which time there were no known exceptions to
Fsun on moon these laws. But with the discovery of special relativity
Fsun on earth in 1905 and quantum mechanics in 1923, we now know
that Newton’s laws are an approximate set of equa-
tions which apply with great accuracy to objects like
stars, planets, cars and baseballs, but which have to be
significantly modified when we consider objects mov-
Figure 8 ing at speeds near the speed of light, and which fail
Forces on the earth and moon, as they go around completely on a subatomic scale.]
the sun. (The force of the sun on the earth is
much larger than the other three forces shown.)
 11-7

CONSERVATION OF LINEAR MOMENTUM Now use Newton’s third law to cancel all the internal
In Chapter 6, after introducing the recoil definition of forces, and we are left with
mass, we saw that the quantity m1v1 + m2v2 remained
unchanged when the two objects were recoiled from dPtotal
= Fext Newton's second law (12)
each other. dt for a system of particles

where Fext is the vector sum of all the external forces

m1 m2 V1 = V2 = 0 acting on the system.

V1 V2 If we have an isolated system with no external forces

m1 m2 acting on it (if our globular cluster were drifting
m1V1 + m2V2 = 0 through empty space), then
(13)
dPtotal for an
We proposed that this was one example of a more = 0 isolated
general conservation law—the conservation of linear dt system
momentum. Now, using Newton’s third law, we can
explicitly demonstrate how the law of conservation of Equation 13 is our desired statement of the law of
momentum applies to objects obeying Newton’s laws. conservation of linear momentum. In words it says that
the total linear momentum of an isolated system is
Our discussion begins with Equation 3, reproduced
conserved—does not change with time. (Linear mo-
below as Equation 9, that was obtained by differentiat-
mentum is also conserved if there are external forces
ing the formula for the center of mass of a system of
but their vector sum is zero. For example, a cart on an
particles. The result was
air track experiences the downward force of gravity
MVcom = m 1v1 + m 2v2 + + m nvn and the upward force of the air, but these forces exactly
(9) cancel.)
= P total
where P total = m 1v1 + + m n vn is the vector sum In deriving Equation 13, we had to use Newton’s
of the momenta of all the particles in the system. We hypothesis that all internal forces cancel in pairs. And
will call this the total momentum of the system. We see we also needed Newton’s second law to relate d mvi /dt
from Equation 9 that this total momentum is simply the to the vector sum of the forces acting on the i-th particle.
total mass M times the velocity of the center of mass. However, the law of conservation of linear momentum
is known to apply on a subatomic scale where the
Differentiating Equation 9 with respect to time, as we concepts of force and acceleration loose their meaning.
did in going from Equation 3 to 4 we get This suggests that our derivation of momentum conser-
vation is somehow backwards. A more logical route is
dP total to assume conservation of linear momentum as a basic
= m 1a1 + m 2 a 2 + + m na n (10) principle and derive the consequences.
dt
This is essentially Equation 4, except that we are
replacing MAcom by dPtotal/dt .

Using the fact that m1a1 is the vector sum of all the
forces acting on particle 1, m2a2 the sum acting on
particle 2, etc., we get

dPtotal sum of all the forces

= (11)
dt acting on all the particles
11-8 Systems of Particles

[To see how such a derivation would look, consider an Newton recognized that a slight generalization of the
isolated system where there are no external forces. We second law would make it unnecessary to assume that
get from Equation 11 a particle’s mass was constant. He actually expressed
the second law in the form

vector sum of all the dp d

dPtotal F = = mv (16)
= internal forces acting (14) dt dt
dt
on the particles where F is the total force acting on a particle of mass
m, moving with a velocity v . In the special case that
But conservation of linear momentum requires that the m is constant, then Newton’s second law becomes
linear momentum of an isolated system be conserved. d mv
Thus d(PTotal)/dt must be zero, and therefore the F = = m dv = ma (m = constant)
dt dt
vector sum of all the internal forces must be zero. This (17)
must be true no matter what kind of forces are involved.
Thus we should view the Equation F = ma as a
Newton’s third law is a bit more restrictive in that it
special case of the more general law F = dp/dt .
requires the internal forces to cancel in pairs. The
cancellation in pairs is the simplest picture, but not The momentum form of Newton’s second law is
necessarily required for conservation of linear mo- advantageous if we are considering problems like that
mentum.] in the following exercise where momentum is being
transferred to an object at a known rate and we wish to
Momentum Version determine the effective force. A more basic application
of Newton’s Second Law is for relativistic problems where mass changes with
In our discussion of Newton’s second law, we have velocity. Then we must use the momentum form of the
consistently assumed that the mass of a particle was law in order to account for the mass change. It is
constant, so that we could take the m outside the interesting that Newton had the insight to present the
derivative. For example, in going from Equation 3 to second law in a form that would handle Einstein’s
Equations 4 and 5 in the center of mass derivation, we theory 200 years later.
used the following steps to relate the rate of change of
momentum of the i-th particle d(m ivi)/dt to the total Exercise 5
force Fi acting on the particle A boy is washing the door of his father’s car by squirting
a hose at the door. Assume that the water comes out of
d m ivi the hose at a rate of 20 kilograms (liters) per minute at
= m idvi = m i a i = F i (15) a speed of 12 meters/second. When the water hits the
dt dt
door it dribbles down the side. What force, in newtons,
If we take F = ma to be the basic form of Newton’s does the water exert on the door?
second law, we have to assume m is constant in order
to get ma and then F . (To solve this, simply calculate the amount of momen-
tum per second that the water brings to the door.)
 11-9

COLLISIONS Impulse
In our every day experience, collisions are something An overview of the air cart force detector experiment
we usually try to avoid, whether it is running into a door is shown in Figure (10). Focusing on the momentum
or an automobile accident. Hitting a baseball is an involved, we see that the cart is initially moving down
obvious exception. In physics, collisions turn out to the track with a momentum p i as shown in Figure
play an extremely important role, particularly in the (10a). In Figure (10b) it collides with the force detector,
study of elementary particles. For example, the atomic and in (10c) is moving back up the track with a
nucleus was discovered as a result of experiments momentum p f . The net effect of the collision is to
involving the collision between α particles and atoms change the cart’s momentum from p i to p f .
in a gold foil.

Collisions generally happen rapidly, and one is not force

used to observing what happens during a collision. It detector
is usually a before and after scene, what was the
situation before the collision, and what did things look pi
like after. In most physics experiments like those
mB
involving elementary particles, that is all we can ob-
serve. However we will begin our discussion of
collisions with an experiment that is explicitly de- a) before collision
signed to allow us to study the situation during the
collision.

In this experiment, an air cart moving down an air track

collides with a force detector mounted at the end of the mB
track. Rather than have the metal cart bounce off the
metal arm of the force detector, we slow the collision b) during collision
down by mounting a stretched rubber band on the end
of the cart. With the rubber band colliding with the
force detector, it takes several milliseconds for the
collision to occur and the cart to reverse directions. pf
During this time we can record both the force the cart mB
exerts on the force detector and the velocity of the cart.
Using the momentum form of Newton’s second law,
c) after collision
we will find that there is a particularly simple way to
analyze the collision in terms of the concept of impulse.
Figure 10
When collisions are either too rapid or on too small a Collision of an aircart with the force detector.
scale to be observed directly, we can almost always During the collision, the force detector first
apply the law of conservation of momentum to analyze has to push on the cart to stop it, and then
the results. In elementary particle collisions, both give an essentially equal push to move it out.
energy and momentum may be conserved. In some
introductory physics lab experiments, momentum is
conserved during the collision and energy after. We
will see how this lets us make detailed predictions
about the behavior of the objects involved.
11-10 Systems of Particles

During the collision itself, the force detector is exerting In the air cart experiment, we will record the force F t
a force F t on the cart. The force F t acts for only a from the output of the force detector, and directly
short time, but can be measured in detail by the force compare the integral of that force with the change in the
detector. To relate this force to the observed change in momentum of the cart.
the momentum of the cart, we start with Newton’s law
in the form Note that both pf and – pi are directed to the right. Thus
the magnitude of pf – pi is equal to the numerical sum
dp pf + pi
Ft = (16 repeated)
dt
pf – pi = pf + pi
Multiplying through by dt gives
As a result, the impulse supplied by the force detector
F t dt = dp (18) has a magnitude pf + pi or about 2 pi if the cart comes
out at the same speed it went in. The force detector
Now integrate both sides of this equation from a time supplies 2 pi because one pi is required to stop the cart,
t i before the collision, when the cart momentum was and the other is required to shove it out again.
pi , to a time t f after the collision, when the cart
momentum was pf . We get
Calibration of the Force Detector
tf pf The force detector is designed to put out a voltage that
is proportional to the force exerted on the detector
F t dt = dp = pf – pi (19)
beam. To convert this voltage reading to a force
ti pi measurement, we have to calibrate the force detector.
This is easily done by running a string from the air cart,
Since p f – p i is the change in the momentum of the over a pulley and down to some weights as shown in
cart as a result of the collision, we get Figure (11). As we add weights to the string we
increase the force the cart exerts on the beam.
tf change in the
impulseof the Figure (12) shows the output of the force detector as we
F t dt = momentum of (20)
force F t the air cart added a series of 20 gm weights. Adding 3 weights
ti
changes the voltage by 42.8 millivolts (mV), thus each
This integral of F t over the time of the collision is weight changes the output voltage by 42.8/3 = 14.3
called the impulse of the force F t . The force exerted mV. (One millivolt = 10 – 3 volts.) Each added weight
by the force detector alters the momentum of the cart, corresponds to an increase of the force by
and how much it alters it is equal to the impulse of the ∆F = mg = 20 gm × 980 cm/s2 = 19600 dynes. Thus
force. the factor for converting from millivolts output for this
force detector, to dynes of force is
force
detector

mB

Figure 11 Figure 12
Calibrating the force detector. Voltage output when 20 gram weights are added.
 11-11

conversion 19600 newtons have a formula for the curve, we cannot do the integral
=
factor 14.3 millivolts analytically. Instead, we have to use some graphical
(21) technique to find the area under that curve. One way is
dynes
= 1370 to superimpose the curve on graph paper and count the
millivolts
squares underneath. A slightly less accurate way we
The Impulse Measurement will use is to construct a triangle whose area is, to our
One way to set up the collision experiment is shown in best estimate, equal to the area under the curve. We
Figure (13). To soften the collision we have added a have done this with the dashed line triangle seen in
metal bracket to the cart and stretched a rubber band Figure (15). We have adjusted the triangle so that the
across the open end. Adjusting the tension in the rubber extra area at the top matches the area lost at the sides.
band allows us to change the length of time during
which the collision occurs. The area of a triangle is (1/2 base × altitude). In Figure
(15), the base of the triangle is 3.00–2.60 = 0.40
Figure (14) shows a fairly typical output of the force seconds. The height is seen to be close to 56800 dynes.
detector. There is no force until the rubber band reaches Thus the area is
the detector. The force then increases and then de-
area of
creases symmetrically, and becomes 0 when the cart = 1 × 0.40 sec × 56800 dynes
leaves. Using our calibration factor of 1370 dynes/mV, triangle 2
we have graphed the force, in dynes, as a function of (22)
= 11360 dyne seconds
time in Figure (15).
Equation 22 is our result for the impulse of the force
The impulse of this force is the integral of the force F(t).
curve from the time t 1 that the rubber band gets to the
force detector, to t 2 when it leaves. Since we do not Although the force detector measures the magnitude of
rubber band the force F(t) that the cart exerts on the detector, by
Newton’s third law, this should be equal in magnitude
top view of aircart but oppositely directed to the force exerted by the
detector on the cart. Thus the magnitude of the impulse
force calculated in Equation 22 should be equal to the
detector beam magnitude of the change in the momentum of the cart,
Figure 13 as a result of the collision.
A rubber band is used to soften the collision.

dynes
54800

41100

27400

13700

Figure 15
Force versus time graph. The area under the force curve
Figure 14
is about equal to the area of the triangle we drew.
Output of the force detector. There is zero force
before the cart arrives and after it leaves.
11-12 Systems of Particles

Exercise 6 On the way back, we find from Figure (17c) that the sail
(a) What is the direction of the force exerted by the force
took 412 milliseconds or .412 seconds to pass the
detector on the cart? photodetector. Thus the final speed of the cart was

(b) What is the direction of the vector ∆p = p f – p i ? vf = 10 cm = 24.27 sec

cm (24)
.412 sec
(c) How do these two directions compare? The cart, sail and rubber band apparatus had a total
mass of m cart which was measured to be
Change in Momentum m cart = 227 gm (25)
To measure the momentum of the cart before and after
the collision, we mounted a 10 cm long sail on the top
of the cart as shown in the side view of Figure (16).
Mounted above the track is a light source and a photo-
detector seen in the top view. When the sail on the cart
interrupts the light beam, there is an abrupt change in
the voltage output by the photodetector. The lower,
dashed curves in Figures (14) and (17) are from the
output of the photodetector. In Figure (17a) we are
measuring the length of time the sail took to pass by the
photocell on the way down to the force detector. We
see that this time was 400 milliseconds or .400 seconds.
Thus the velocity of the cart on the way down was

vi = 10 cm = 25 sec
cm (23)
.400 sec Figure 17a
Simultaneous recording of both the voltage
output from the force detector (solid line)
photo and the photo detector (dashed line).
detector
10 cm long card
Figure 17b,c
v Measuring the
length of time the
sail took to go past
the photodetector.
a) side view The 10 cm sail took
light 400 milliseconds to
pass on the way in
source (upper curve) and
10 cm long card
412 milliseconds on
the way out. We see
it slowed down a bit.
photo b) top view (Another way to
detector determine the speed
of the aircart is to tilt
Figure 16
the airtrack and
One way to measure the velocity of the aircart is to release the cart from
mount a 10 cm long sail is mounted on top of the aircart. a known height.)
While the sail is interrupting the light beam, there is a
change in the output voltage of the photo detector.
 11-13

Thus the initial and final momenta have magnitudes Momentum Conservation
cm during Collisions
pi = mvi = 227 gm × 25.0 sec
In our force detector experiment we used a rubber band
gm cm to slow down the collision so that we could do a more
= 5680 sec
accurate analysis of the impulse. Even so the impulsive
pf = mvf = 227 gm × 24.27 seccm force F t acted for only a very short time. The most
important point of the experiment is that, no matter how
gm cm short the time is, the impulse, the time integral of F t is
= 5510 sec
equal to the total change in momentum. If we had let the
The magnitude ∆p of the change in momentum is the metal end of the cart strike the force detector, the
sum of these two values collision would have taken much less time, but the force
gm cm gm cm would have been much greater. The integral of the
∆p = pi + pf = 5680 sec + 5510 sec
larger force over the shorter time would still equal the
change in the momentum of the cart.
change in the
gm cm
linear momentum = ∆p = 11200 sec Suppose that, instead of an air cart colliding with a force
of the aircart detector, we had two air carts colliding with each other.
During the collision they would by Newton’s third law,
(26) exert equal and opposite forces F t on each other. Thus
From Equation 22, we saw that the total impulse they would exert upon each other equal and opposite
supplied by the force detector was impulses F t dt . As a result, the momentum gained
by one cart would precisely be equal to the momentum
total impulse tf lost by the other. The net result is conservation of
from the = F t dt = 11360 dyne sec momentum during the collision.
force detector ti Now consider a slightly more complex situation. Sup-
pose I throw a red billiard ball up in the air, and you
(27) throw a blue billiard ball, and the two balls collide before
We see that to within a quite reasonable experimental landing.
error, the total impulse supplied by the force detector
equals the change in linear momentum of the aircart. During this collision, more forces are involved. There
is the force of the red billiard ball on the blue one, the
Exercise 7 force of the blue billiard ball on the red one, and there are
Explain why ∆p in Equation 26 is the sum of the magni- the gravitational forces acting on both. To study the
tudes pi and pf . change in the momenta of the balls, it appears that we
must now account for all four forces at once.
Exercise 8
However there is something special about the impulsive
Show that the dimensions of impulse (dyne seconds) forces found in collisions. These forces are usually very
are the same as momentum (gm cm/sec). large but act for a very short time. During this short
time the collision forces are usually much larger than
Exercise 9
external forces like gravity. So much larger, in fact, that
How much energy, in joules, did the cart lose in the we can usually neglect external forces during a colli-
collision with the force detector? What percentage of sion. Since the collision forces conserve linear momen-
the cart’s initial kinetic energy was this? tum, we get the result that linear momentum is con-
served during a collision even if external forces are
present. The only exception would be if the collision is
so slow that the external forces have time to act and
change the system’s momentum during the collision.
This is usually not the case.
11-14 Systems of Particles

Collisions and Energy Loss Solving for the final speed vf , we get
While we can use conservation of linear momentum in vf = m v (31)
the analysis of a collision, often energy conservation is m+M i
not applicable. If the objects are deformed or give off We can now calculate the amount of energy that must
heat, light or sound, energy escapes in ways that are be dissipated in this collision. The initial energy E i is
difficult to measure. the kinetic energy of the putty ball
In some situations, however, we can use momentum E i = 1 mvi2 (32)
2
conservation to figure out how much energy must be
The final energy E f is the kinetic energy of the two
“lost” to deformation, heat, sound, etc. Suppose, for
together
example, we hang a steel ball of mass M from a string
as shown in Figure (18), and throw a putty ball of mass Ef = 1 m+M vf 2 (33)
2
m at the steel ball. The putty ball is initially moving at
The energy “lost”, which must have gone into deform-
a speed vi , then hits and sticks to the steel ball. The two
ing the putty ball, is
move off together at speed vf .
Elost = Ei – Ef
Even though energy is “lost” when the putty ball
squashes up against the steel ball, momentum is con- = 1 mvi2 – 1 m+M vi2 (34)
served during the collision. The initial momentum is all 2 2
carried by the putty ball, and is thus Using Equation 31 for vf in Equation 34 gives
initial mvi 2
p i = mvi (28)
momentum E lost = 1 mvi2 – 1 m+M
2 2 m+M
After the collision the two move off together at a speed
vf and thus have a total momentum 2
= 1 mvi2 – 1 m vi2
pf = m+M vf (29) 2 2 m+M

Since momentum is conserved, we have = 1 mvi2 1 – m

2 m+M
pi = pf
mvi = m+M vf (30) E lost = 1 mvi2 M (35)
2 m+M
It may be somewhat surprising that even though we
may not have the slightest idea how energy is lost
during the deformation of the putty, we can calculate
precisely how much energy this uses.

Exercise 10
Check that Equation 35 gives reasonable results for the
two special cases M = 0 and M = ∞ (for m ≠ 0).

m vi M m M vf m Exercise 11
putty ball steel ball A 500 gram steel ball is suspended from a string as
before collision just after collision shown in Figure (18). It is struck by a putty ball that
Figure 18
sticks, and half the putty ball’s initial kinetic energy is lost
Collision of a putty ball with a steel ball. in the collision. What is the mass of the putty ball?
 11-15

After the putty ball has collided with the steel ball in our You can assume that momentum is conserved during
preceding example, the two will rise together to a the collision, energy afterward, and from a measure-
maximum height h and final angle θ f before swinging ment of the height h, determine the speed vi of the
back down again. This is illustrated in Figure (19). We bullet before it hit the block. This provides a rather
can predict this height h by applying energy conserva- simple, inexpensive way to measure the speed of
tion after the collision. The kinetic energy Ef just after bullets.
the collision is transformed into gravitational potential
energy (m + M)gh at the top of the swing, giving From the prospective of an introductory physics course,
the ballistic pendulum experiment clearly distinguishes
1 m+M v 2 = m+M gh the use of momentum conservation and energy conser-
2 f
vation. During the collision, momentum is conserved
or
but energy is not. During the rise up to a height h,
vf 2 energy is conserved but momentum is not. To analyze
h = (36)
2g such problem, you must develop an understanding of
If you want to calculate the final angle θ f , you use the when you can apply the conservation laws and when
fact that h = – cos θf as can be seen from Figure you cannot. (For a safer ballistic pendulum demonstra-
(19). tion, see Figure (21) on the next page.)

This example of the steel and putty ball is essentially Exercise 12

equivalent to the ballistic pendulum discussed in Exer-
(a) A bullet of mass mb traveling at a speed vb is fired
cise 12. In that problem a block of wood of mass M is into a block of wood of mass M hanging at rest as
suspended from two strings as shown in Figure (20). A shown in Figure (20). The combined block and bullet
bullet of mass m is fired into the block and the block rise to a height h. Find a formula for the speed vb of
with the bullet stuck inside rises to a height h as shown. the bullet.

(b) For part (a), suppose the bullet's mass is mb = 10 gm,

the block of wood has a mass M = 1000 gm, and the final
θf height h is 12 cm. What was vb in cm/sec?
cos θf

M
M
m m V
h
vf M h
m M m
Figure 20
just after collision Ballistic pendulum experiment.

Figure 19
After the collision, energy is conserved. This
allows us to calculate how high the ball rises.
11-16 Systems of Particles

Exercise 13 On an atomic and subatomic scale we do not have the

In a lecture demonstration that is safer to perform than
usual sound and friction, and small deformations may
the above ballistic pendulum experiment, a wastebas- not be allowed. In these circumstances there is no way
ket is suspended from two cords and a pillow is placed for energy to become “lost” and the resulting collisions
inside the wastebasket, as shown in Figure (21). are truly elastic. Thus in the study of the collisions of
atomic and subatomic particles, we have examples
Various members of the class are selected to throw a where both energy and linear momentum are con-
softball into the wastebasket. A scale is constructed to served.
indicated how fast the ball was thrown. If the mass m of
the softball is 200 gm, the mass M of the wastebasket Figure (22), shown previously in Chapter 6 as Figure
and pillow is 1000 gm, and the length of the suspension (6-3), shows the track of a proton as it moves through
cords are 2 meters from the ceiling to the center of the a hydrogen bubble chamber. The incoming track ends
wastebasket, determine the distance (x) that the end of when the proton strikes a hydrogen nucleus (another
the basket travels if the ball is thrown at 40 miles/hour. proton) that is part of the liquid hydrogen. Three
dimensional stereoscopic photographs show that the
Collisions that Conserve two protons recoil from each other at an angle of 90°.
Momentum and Energy Here we are looking at the behavior of matter on a
subatomic scale where Newtonian mechanics does not
When a bullet plows into a block of wood, a consider-
apply, and we would like to find out what we can learn
able amount of energy goes into deforming the wood
from this collision.
and bullet. On the other hand if two hardened steel ball
bearings collide, almost all the energy stays in the form Does this photograph show us that momentum and
of kinetic energy of the particles and very little is “lost” energy are still conserved on the subatomic scale? To
as heat, sound, and the deformation of the objects. find out we will analyze the collision of larger objects
When no energy is lost this way, we say that the like steel ball bearings, where momentum is conserved
collision is elastic. If energy is lost, then the collision and energy nearly so, and see if the results explain what
is called inelastic. we see in Figure (22).

Cord of length l

V
m M

20 40 60 80 mph

X
Pi
Pf1
Figure 21
A safer ballistic pendulum experiment
for classroom demonstrations. Pf2

Figure 22
An incoming proton collides with a proton at
rest. The two protons recoil at right angles.
 11-17

Elastic Collisions That ball 1 stops dead in its tracks explains the common
We will start with the simplest elastic collision we can toy where two or more steel balls are suspended from
think of—a ball of mass is traveling at a velocity vi a string as shown in the end view—Figure (24a). One
strikes an identical ball at rest, as shown in Figure (23). of the balls is pulled back as shown in Figure (24b) and
We will assume that the collision is straight on so that released. At the collision, it comes to rest and the
the two balls go off in the same direction at speeds v1 second ball goes on. Then the second ball comes back,
and v2 as shown. The idea is to apply conservation of strikes the first one and stops, and the first ball goes
momentum and energy in order to predict the final back up. For good hard steel balls, this process goes on
speeds v1 and v2. for a long time before we see motion decrease.

From conservation of momentum we get If energy were not conserved, if some were lost in the
collision, then both balls would move forward after the
momentum
mvi = mv1 + mv2 conservation
(27) collision. In the extreme case the balls would stick and
move off together. This is a completely inelastic
Canceling the m’s we have collision where the maximum energy is lost (consistent
vi = v1 + v2 (28) with conservation of momentum). If you perform this
From conservation of energy we have experiment and notice that the incoming ball really
comes to rest, that is experimental proof that both
1 1 1 energy
mvi2 = mv12 + mv22 conservation
(29) energy and momentum were in fact conserved in the
2 2 2 collision.
The 1/2 m’s cancel and we have
vi 2 = v1 2 + v2 2 (30)
If we square Equation (28) we get
vi 2 = v1 2 + 2v1 v2 + v2 2 (31)
(a) (b)
Comparing this with Equation (30) we see that
end view side view
v1 v2 = 0
Figure 24
I.e., either v2 = 0, which means there was no collision When the balls collide, the moving ball stops and
at all, or v1 = 0 which means that ball 1 stops dead in the stationary ball moves on at the same speed.
its tracks and ball 2 goes on at the initial speed vi.

Vi V1 V2
m m m m
Figure 23
Collision of two identical steel balls.
We want to calculate v1 and v2 .
11-18 Systems of Particles

In the next simplest example, suppose the two balls of To state this result another way, if two equal masses
Figure (23) collide but bounce off at an angle as shown collide and one is originally at rest, they always
in Figure (25), ball 1 coming off at a velocity v1 , and emerge at right angles (or the incoming one stops).
ball 2 at a velocity v2 as shown. In this case we have This is experimental evidence that both energy and
the same formula for conservation of energy, but momentum are conserved in the collision.
conservation of momentum must now be written as the
vector equation It is this way that we learn that energy and momentum
are both conserved during the collision of two protons
momentum
mvi = mv1 + mv2 (32) in a hydrogen bubble chamber. This is a remarkable
conservation
result considering that we do not have to know any-
Again the m’s cancel and we are left with the following thing about what kind of forces were involved in the
vector equation for the velocity vectors collision.
vi = v1 + v2 (32a) If we have elastic collisions between objects of differ-
The equation if pictured in Figure (26). ent masses, moving at each other with different speeds,
as in Figure (27), we still have the conservation of
Recall that energy conservation gave momentum
2 2 2 energy
vi = v1 + v2 (30) m1v1 i + m2v2 i = m1v1 f + m2v2 f (33)
conservation
and conservation of energy
which is simply the Pythagorean theorem when ap-
plied to the triangle in Figure (26). Thus the incoming 1 m1v2 + 1 m2v2 = 1 m1v2 + 1 m2v2 (34)
1i 2i 1f 2f
2 2 2 2
speed vi must be the hypotenuse and v1 and v2 the sides
of a right triangle. Energy conservation requires that The only problem is that the algebra quickly becomes
messy.
the two balls emerge from the collision at right angles.
V1
If you are in the business of working with collision
problems, you will find it much easier to go to a
Vi m coordinate system where the center of mass of the
m m colliding particles is at rest. In this coordinate system
the particles go in and come out symmetrically and the
m
V2 equations are easy to solve. Then you transform back
Figure 25 to your original coordinate system to see what the
If the collision is not straight on, particle velocities should be in the laboratory.
the balls come off at an angle.

V1 V2 V2f
m 2 V2i
m2
Vi
Figure 26 V1i
Equation 30 requires that the
velocity vectors form a right triangle. m1 m1 V1f
Figure 27
Arbitrary collision of two balls.
 11-19

DISCOVERY OF THE ATOMIC NUCLEUS We have seen from our analysis of the elastic collision
In his early experiments with radioactivity, Ernest of two equal mass particles, that the incoming particle
Rutherford found that radioactive atoms emitted three stops and the struck particle continues on. Only if the
kinds of radiation which, as we have mentioned, he mass of the struck particle is greater than the mass of the
called α rays, β rays and γ rays. The α rays turned out incoming particle, will the incoming particle bounce
to be heavy positively charged particles, later identified back. And only if the mass of the struck particle is
as helium nuclei. The β rays were beams of negatively much greater than the mass of the incoming particle
charged particles later determined to be electrons, and will the incoming particle rebound with nearly the
the neutral γ rays turned out to be high energy particles same energy that it had coming in. Thus Rutherford
of light (photons). and Geiger’s observation that some of the α particles
bounced right back out of the gold foil, indicated that
In the early 1900s, before 1912, it was not clear how they struck a solid object much more massive than an
these particles were emitted or what the structure of the α particle.
atom was. Since J. J. Thomson’s experiments with
electron beams in 1895, it was known that atoms Most α particles passed through the gold foil without
contained electrons, and it was also known that com- much deflection, indicating that most of the volume of
plete atoms were electrically neutral and much more the gold foil was devoid of mass. The few collisions
massive than an electron. Thus the atom had to have that did result in a recoil indicated that the mass in a gold
mass and positive charge in some form or other, but no foil was concentrated in incredibly small regions of
one knew what form. By 1912 the plum pudding model space. A more detailed analysis showed that the
was quite popular. This was a picture in which mass scattering was caused by an electric force, thus they
and positive charge was spread throughout the atom knew that both the mass and positive charge were
like the pudding, and the electrons were located at located in a tiny region of the atom. In this way the
various points, like the plums. A rather vague picture atomic nucleus was discovered.
at best.

In 1912 Rutherford and Hans Geiger began a series of

experiments using beams of radioactive particles to
probe the structure of matter. These experiments could
begin after Geiger had developed a tube to detect
radioactive particles. This device later became known
as a Geiger counter, and is still used through the world
to monitor radiation.

In the first set of experiments, a beam of α particles

were aimed at a gold foil. It was expected that some of
the α particles would be slightly deflected as they
passed through the positive matter in the gold atoms, or
came near electrons. To the utter amazement of both
Rutherford and Geiger, some of the α particles bounced
straight back out of the gold foil, with essentially the
same kinetic energy they had going in.
11-20 Systems of Particles

NEUTRINOS If Pauli’s particle existed, it would need the following

The discovery of the neutrino, or at least the prediction properties: (1) it had to be electrically neutral because
of its existence, is another important event in the history no electric charge was lost in β decays; (2) it would
of physics that is related to the conservation of energy have to have a very small rest mass because the electron
and linear momentum. or β particle sometimes carried out almost all the
available energy, leaving none for creating the new
After Rutherford’s discovery of the nucleus, it became particle’s rest mass; (3) the new particle had to have
clear that the high energy radioactive emissions, α, β, almost no interaction with matter, otherwise someone
and γ rays, must be coming from the nucleus of the would have seen it.
atom. Thus a study of these rays should give valuable
information about the nature of the nucleus itself. Initially there was not much enthusiasm for Pauli’s idea
of an undetectable particle. At that point no one had
After a number of years of experimentation it was seen an electrically neutral particle, and the fact that it
determined that whenever an α particle or γ ray was did not interact with matter made it seem too specula-
emitted, the energy carried out by the α particle or γ ray tive.
was precisely equal to the energy lost by the nucleus.
But decays involving β particles were different. In In 1932 the neutron was discovered which demon-
studying β decays, one always got a spread of energies strated that neutral particles did exist. Shortly after that
of the β particle. Sometimes the β particle carried out Enrico Fermi developed a detailed theory of the weak
almost all the energy lost by the nucleus, and some- interaction in which neutrinos played a significant role.
times only the relatively small rest energy of the (Fermi called Paul’s particle the neutrino, or “little
electron. By the late 1920s it was clear that energy was neutral one” to distinguish it from the more massive
apparently not conserved in β decay reactions. Neils neutron.) Detailed verification of Fermi’s weak inter-
Bohr proposed that the law of conservation of energy action theory convinced the physics community that
had to be modified for nuclear reactions involving β neutrinos should exist.
decays. The new rule was that the final energy was The neutrino, actually detected in 1956 and now com-
always less than or equal to the initial energy. monly seen in numerous experiments, is a remarkable
In 1930 Wolfgang Pauli, one of the founders of quan- particle in that it is subject to only the weak and
tum theory, objected to the idea that energy was not gravitational interactions. All other known particles
conserved in β decay events. Pauli noted that the are subject to the electric and nuclear forces. For
conservation of energy, linear momentum, and angular example π mesons are subject to the nuclear force, and
momentum are all apparently violated at the same time can travel only a short distance through matter before
in β decay. Either the entire structure of physical law colliding with a nucleus and being stopped.
was being violated, or there was another explanation. The muon discussed in the relativity chapter, does not
Pauli’s other explanation was that the energy, the linear feel the strong nuclear force and therefore can travel
momentum, and the angular momentum were all being much farther through matter (hundreds of meters)
carried out at the same time by an unseen particle. before being stopped. Muons are electrically charged
and therefore are stopped by the weaker electric inter-
action. Photons also interact through the electric
interaction, and therefore have a limited range travel-
ing through matter. (An X-ray is an example of a
photon passing through matter.)
 11-21

The weak interaction is so weak compared to the Neutrino Astronomy

nuclear or electric force, that a particle like the neutrino An exciting development involving neutrinos is the
which feels only the weak interaction force, can travel birth of neutrino astronomy. In the fusion reaction that
incredible distances through matter before being powers our sun, where four hydrogen nuclei (protons)
stopped. To have a good chance of stopping a single end up as a helium 4 nucleus, the weak interaction and
neutrino with a stack of lead, one would need a pile of the β decay process comes into play in the conversion
lead, light years thick. of two of the protons into the neutrons of the helium
nucleus. Thus the emission of neutrinos must accom-
This does not mean that neutrinos are impossible to pany the fusion reaction, and the neutrinos themselves
detect. Instead of using a detector light years across to must carry off a significant amount of the energy
detect one neutrino, one can use a source that produces liberated by the fusion reaction.
an incredible number of neutrinos and use a reasonable
sized detector so that one has some chance of stopping In a star like the sun, the fusion reaction takes place
a few neutrinos. In 1956 Cowan and Rines placed a down in the core of the star where the temperatures are
tank car full of carbon tetrachloride cleaning fluid in highest. Any light emitted by the fusion reaction
front of a nuclear reactor that was estimated to emit should take the order of about 10,000 years to work its
about 1015 neutrinos per square centimeter per second. way out. Thus if the fusion reaction in the sun were shut
They observed that about two chlorine atoms per off today, it would be roughly 10,000 years before the
month in the tank car of carbon tetrachloride were sun dimmed.
converted by neutrino interactions into argon atoms
which were counted individually. Neutrinos, however, escape from the core of the sun
without delay. If the fusion reaction stopped and we
In modern experiments carried out using high energy were monitoring the neutrinos from the sun, we would
particle accelerators, neutrino reactions are routinely know about it within 8 minutes. As a result there is
seen. The reason that more neutrinos are detected in considerable incentive to observe the solar neutrinos,
these experiments is that the weak interaction becomes for that gives us a picture of what is happening in the
less weak as the energy of the particles involved sun’s core now.
increases. The high energy accelerators produce neu-
trinos with great enough energy that they are not too To study solar neutrinos, and do other experiments like
difficult to detect. As a result, the neutrino interactions look for decay of the proton, several large neutrino
has become an important research tool in the study of detectors have been set up around the world. Solar
the basic interactions of matter. Neutrinos make a neutrinos have been monitored fairly carefully for over
particularly clean tool for these studies because they a decade, and there is an unexplained, perhaps disturb-
have no nuclear or electric interactions. Neutrino ing result. Only about one third as many neutrinos are
experiments are not contaminated by effects of the being emitted by the sun as we expect from what we
nuclear and electric forces. think the fusion reaction should produce. Perhaps we
are not detecting all we should, but the detectors are
getting better and the number remains at 1/3. This is
one of the major puzzles of astronomy.
11-22 Systems of Particles

That neutrino astronomy is really here was dramati-

cally illustrated with the supernova explosion of 1987.
On the average, supernovas occur about once per
century per galaxy. Kepler saw the last supernova
explosion in our galaxy 400 years ago. In 1987 a
graduate student spotted the sudden appearance of a
bright star in the large Magellanic cloud, a close small
neighboring galaxy. This was the first supernova
explosion in the local region of our galaxy in 400 years.

In a supernova explosion, huge quantities of neutrinos

should be emitted. In fact a fair fraction of the energy
of the explosion should be carried out by neutrinos.

Theoretical models of supernova explosions suggest

that light should take about three hours to work its way
out through the expanding envelope of gas before it
starts its trek through space at the speed c. Neutrinos,
on the other hand, should escape without being slowed
down, and have a three hour head start on the light. If
neutrinos have no rest mass, and therefore travel at the
speed of light, they should have reached the earth about
three hours before the light. Two of the major neutrino
detectors, one in the US and one in a tunnel in the Alps,
detected significant pulses of neutrinos about three
hours before the flare-up of the star was seen. (This was
determined by a later analysis of the neutrino data.)
That event marks the birth of neutrino astronomy on a
galactic scale.
Chapter 12
Rotational Motion

CHAPTER 12 ROTATIONAL MOTION

Our discussion of rotational motion begins with a angular force is equal to the time rate of change of
review of the measurement of angles using the concept angular momentum. (The angular force is more com-
of radians. We will refer to an angle measured in monly known as torque).
radians as an angular distance. If we are discussing
an object that is rotating, we will describe the rotation The angular analogy to Newton’s second law looks a
in terms of the increase in angular distance, namely an bit peculiar at first. It involves lever arms and vectors
angular velocity. And if the speed of rotation is that point in funny directions. After some demonstra-
changing, we will describe the change in terms of an tions to show that the equation appears to give reason-
angular acceleration. able results, we apply the equation to predict the
motion of a gyroscope. The prediction appears to be
In Chapter 7, linear momentum and angular momen- absurd, but we find that that is the way a gyroscope
tum were treated as distinctly separate topics. The behaves.
main point of this chapter is to develop a close analogy
between the two concepts. The linear momentum of an Our focus in this chapter is on angular momentum
object is its mass m times its linear velocity v . We will because that concept will play such an important role
see that angular momentum can be expressed as an in our later discussions of atomic physics and electrons
angular mass times an angular velocity. (Angular and nuclear magnetic resonance. There are other
mass is more commonly known as moment of inertia). important and interesting topics such as rotational
Then, using the formalism of the vector cross product kinetic energy and the calculation of moments of
(mentioned in Chapter 2), we will see that angular inertia which we discuss in more detail in the appendix.
momentum can be treated as a vector quantity, which These topics are not difficult and lead to some good
explains the bicycle wheel experiments we discussed in lecture demonstrations and laboratory experiments.
Chapter 7. We put them in an appendix because they do not play
the essential role that angular momentum does in our
The fundamental concept of Newtonian mechanics is later discussions.
that the total force F acting on an object is equal to the
time rate of change of the object’s linear momentum;
F = dp/dt . Using the vector cross product formalism,
we will obtain a complete angular analogy to this
equation. We will find that a quantity we call an
12-2 Rotational Motion

RADIAN MEASURE To relate radians to degrees, we use the fact that there
From the point of view of doing calculations, it is more are 360 degrees/cycle and dimensional analysis to find
convenient to measure an angle in radians than the the number of degrees/radian
more familiar degrees. In radian measure the angle θ degrees cycles
360 cycle × 1 radian
shown in Figure (1) is the ratio of the arc length s to the 2π
(3)
radius r of the circle degrees degrees
= 360 radian = 57.3 radian
2π
θ ≡ sr radians (1) Fifty seven degrees is a fairly awkward unit angle for
purposes of drafting and
Since s and r are both distances, the ratio s/r is a navigation; no one in his or
dimensionless quantity. However we will find it her right mind would mark o
convenient for the angular analogy to keep the name a compass in radians. How- 1 radian = 57.3
radians as if it were the actual dimension of the angle. ever, in working with the
For example we will measure angular velocities in dynamics of rotational mo-
radians per second, which is analogous to linear tion, radian measure is the only reasonable choice.
velocities measured in meters per second.

Since the circumference of a circle is 2πr , the number Angular Velocity

of radians in a complete circle is The typical measure of angular velocity you may be
= 2π
familiar with is revolutions per minute (RPM). The
r = 2π
complete r
θ circle
tachometer in a sports car is calibrated in RPM; a
In discussing rotation, we will often refer to going typical sports car engine gives its maximum power
around one complete time as one complete cycle. In around 5000 RPM. Engine manufacturers in Europe
one cycle, the angle θ increases by 2π. Thus 2π is the are beginning to change over to revolutions per second
number of radians per cycle. We will find it conve- (RPS), but somehow revving an engine up to 83 RPS
nient to assign these dimensions to the number 2π: doesn't sound as impressive as 5000 RPM. (Tachom-
eters will probably be calibrated in RPM for a while.)
2π radians (2) In physics texts, angular velocity is measured in radians
cycle per second. Since there are 2π radians/cycle , 83
revolutions or cycles per second corresponds to
2π × 83 = 524 radians/second. Few people would
know what you were talking about if you said that you
s should shift gears when the engine got up to 524 radians
θ per second.
r
Exercise 1
What is the angular velocity, in radians per second,
of the hour hand on a clock?
Figure 1
The angle θ in radians is defined as the ratio
of the arc length s to the radius r: θ = s/r .
 12-3

Our formal definition of angular velocity is the time Angular Analogy

rate of change of an angle. We almost always use the At this point we have a complete analogy between the
Greek letter ω (omega) to designate angular velocity rotation of a motor shaft and one dimensional linear
motion. This analogy becomes clear when we write out
angular the definitions of position, velocity, and acceleration:
velocity
ω ≡ dθ radians
dt second
(4) Linear Angular
motion motion
When thinking of angular velocity ω picture a line
marked on the end of a rotating shaft. The angle θ is Distance x meters θ radians
the angle that the line makes with the horizontal as
shown in Figure (2). As the shaft rotates, the angle θ t Velocity v = dx meters ω = d θ radians
dt second dt second
increases with time, increasing by 2π every time the
shaft goes all the way around. dv meters dω radians
Acceleration a = α=
dt second 2 dt second 2
Angular Acceleration
d 2x d2 θ
When we start a motor, the angular velocity of the shaft = = (7)
d t2 dt 2
starts at ω = 0 and increases until the motor gets up to
its normal speed. During this start-up, ω t changes As far as these equations go, the analogy is precise.
with time, and we have an angular acceleration α Therefore any formulas that we derived for linear
defined by motion in one dimension must also apply to angular
motion. In particular the constant acceleration formu-
angular
α ≡ dω radians2 (5) las, derived in Chapter 3, must apply. If the linear and
acceleration dt second angular accelerations a and α are constant, then we get
The angular acceleration α has the dimensions of Constant Acceleration Formulas
radians/sec2 since the derivative gives us another Linear motion (a = const) Angular motion ( α = const)
factor of time in the denominator. Combining Equa-
1 2 1 2
tion 4 and 5 relates α to θ by x = v0 t + at θ = ω0t + αt (8)
2 2
v = v0 + at ω = ω 0 + αt (9)
d2θ
α = (6)
dt2
Exercise 2
ω An electric motor, that turns at 3600 rpm (revolutions per
minute) gets up to speed in 1/2 second. Assume that the
angular acceleration α was constant while the motor
was getting up to speed.
θ (t)
a) What was α (in radians/ sec2 )?

b) How many radians, and how many complete cycles,

did the shaft turn while getting up to speed?

Figure 2
End of a shaft rotating at an angular velocity ω .
12-4 Rotational Motion

Tangential Distance, The speed of the particle along the path, which we can
Velocity and Acceleration call the tangential speed vt, is the time derivative of the
So far we have used the model of a rotating shaft to tangential distance s(t)
illustrate the concepts of angular distance, velocity and ds(t) dθ(t)
acceleration. We now wish to shift the focus of our vt = = d r θ(t) = r
dt dt dt
discussion to the dynamics of a particle traveling along
a circular path. For this we will use the model of a small where r comes outside the derivative since it is con-
mass m on the end of a massless stick of length r shown stant. Since dθ(t) /dt is the angular velocity ω , we get
in Figure (3). The other end of the stick is attached to tangential
vt = rω velocity (11)
and is free to rotate about a fixed axis at the origin of our
coordinate system. The presence of the stick ensures
that the mass m travels only along a circular path of The tangential acceleration a t , the acceleration of the
radius r. The quantity θ(t) is the angular distance particle along its path, is the time derivative of the
travelled and ω(t) the angular velocity of the particle. tangential velocity
dvt(t) d[rω(t)] dω(t)
When we are discussing the motion of a particle in a at = = = r = rα
circular orbit, we often want to know how far the dt dt dt
particle has travelled, or how fast it is moving. The tangential
distance s along the path (we could call the tangential a t = rα acceleration (12)
distance) travelled is given by Equation 1 as
tangential where again we took the constant r outside the deriva-
s = rθ distance
(10) tive, and used α = dω /dt .

r m
k v
s stic
ssles
ma
r
θ
2
pivot a r = vr
Figure 3
Mass rotating on the end of a massless stick.

Figure 4
Particle moving at a constant speed in a circle of
radius r accelerates toward the center of the circle
with an acceleration of magnitude a r = v 2/r.
 12-5

Radial Acceleration Bicycle Wheel

If the angular velocity ω is constant, if we have a For much of the remainder of the chapter, we will use
particle traveling at constant speed in a circle, then a bicycle wheel, often weighted with wire wound
α = d ω /dt = 0 and there is no tangential acceleration around the rim, to illustrate various phenomena of
a t . However, we have known from almost the begin- rotational motion. Conceptually we can think of the
ning of the course that a particle traveling at constant bicycle wheel as a collection of masses on the ends of
speed v in a circle of radius r has an acceleration massless rods as shown in Figure (6). The massless
directed toward the center of the circle, of magnitude rods form the spokes of the wheel, and we can think of
v 2 /r , as shown in Figure (4). We will now call this the masses m as fusing together to form the wheel.
center directed acceleration the radial acceleration a r When forming a wheel, all the masses have the same
radius r, same angular velocity ω and same angular
v t2 radial acceleration α . If we choose one point on the wheel
ar = (13)
r acceleration from which to measure the angular distance θ, then as
far as angular motion is concerned, it does not make any
Exercise 3 difference whether we are discussing the mass on the
Express the radial acceleration a r in terms of the orbital end of a rod shown in Figure (3) or the bicycle wheel
radius r and the particle’s angular velocity ω. shown in Figure (6). Which model we use depends
upon which provides a clearer insight into the phenom-
If a particle is traveling in a circular orbit, but its speed ena being discussed.
v t is not constant, then it has both a radial acceleration ω
a r = v t2 / r , and a tangential acceleration a t = rα . The
radial acceleration is always directed toward the center
of the circle and always has a magnitude v 2 /r . The
tangential acceleration, if it exists, is tangential to the
circle, pointing forward (counterclockwise) if α is r
positive and backward if α is negative. These accelera-
tions are shown in Figure (5).
a t = rα
Figure 6
Bicycle wheel as a collection of masses
v2 on the end of massless rods.
ar = r
= rω2

Figure 5
Motion with radial and tangential acceleration.
12-6 Rotational Motion

ANGULAR MOMENTUM Angular Momentum

In Chapter 7, we defined the angular momentum of of a Bicycle Wheel
a mass m traveling at a speed v in a circle of radius r as We will begin our discussion of the angular momen-
= mvr (7-11) tum of a bicycle wheel using the picture of a bicycle
wheel shown in Figure (6), i.e., a collection of balls on
As we saw, in Figure the end of massless rods or spokes. If the wheel is
(7-9) reproduced here, rotating with an angular velocity ω , then each ball has
the quantity = mvr a tangential velocity vt given by Equation 11a
did not change when
we had a ball moving vt = rω (11repeated)
in a circle on the end of
a string, and we pulled If the i-th ball in the wheel (identified in Figure 7) has
in on the string. The a mass m i , then its angular momentum i will be given
radius of the circle de- by
creased, but the speed Figure 7–9 i = m ivt r = m i (rω)r
increased to keep the Ball on the end of a string,
product vr constant. swinging in a circle. = (m ir 2 ) ω (14)
i
This was our introduction to the concept of the conser-
vation of angular momentum. Assuming that the total angular momentum L of the
bicycle wheel is the sum of the angular momenta of
After that, we went on to consider some rather interest- each ball (we will discuss this assumption in more
ing experiments where we held a rotating bicycle detail shortly) we get
wheel while standing on a freely turning platform. We
Σi Σi
found that these experiments could be explained quali-
L = i = m ir 2 ω (15)
tatively if we thought of the angular momentum of the
bicycle wheel as being a vector quantity which
pointed along the axis of the wheel, as shown in Figure Since each mass m i is at the same radius r and is
(7-15) reproduced below. What we will do now is traveling with the same angular velocity ω , we get

Σi mi r 2ω
develop the formalism which treats angular momen-
tum as a vector. L =

Noting that M = Σi mi is the total mass of the bicycle

2 wheel, we get
angular momentum
L = Mr 2ω of a bicycle wheel (16)
vt
mi
r

Movie

Figure 7–15 Figure 7

When the bicycle wheel is turned over and its angular
momentum points down, the person starts rotating with The angular momentum of the i’th ball is mivt r i .
twice as much angular momentum, pointing up.
 12-7

Angular Velocity as a Vector ANGULAR MASS OR

To explain the bicycle wheel experiment discussed in MOMENT OF INERTIA
Chapter 7, we assumed that the angular momentum L Equation 17 expresses the angular momentum L of a
was a vector pointing along the axis of the wheel as bicycle wheel as a numerical quantity Mr 2 times the
shown in Figure (8a). We can obtain this vector vector angular velocity ω . This is not very different
concept of angular momentum by first defining a from linear momentum p which is the mass (M) times
vector angular velocity ω as shown in Figure (8b). the linear velocity vector v
We will say that if a wheel is rotating with an angular
velocity ωrad/sec , the vector ω has a magnitude of p = Mv (18)
ωrad/sec , and points along the axis of rotation as shown
We obtain an analogy between linear and angular
in Figure (8b). Since the axis has two directions, we use
momentum if we call the quantity Mr 2 the angular
a right hand convention to select among them. Curl the
mass of the bicycle wheel. Designating the angular
fingers of your right hand in the direction of the
mass by the letter I, we get
direction of the rotation, and the thumb of your right
hand will point in the direction of the vector ω . L = Iω (19)
angular mass
Angular Momentum as a Vector
I = Mr 2 (moment of inertia) (20)
Since the vector ω points in the direction we want the of a bicycle wheel
angular momentum vector L to point, we can obtain a
vector formula for L by simply replacing ω by ω The quantity I is usually called moment of inertia
in Equation 16 for the angular momentum of the rather than angular mass, but angular mass provides a
bicycle wheel better description of what we are dealing with. We will
use either name, depending upon which seems more
vector formula for the appropriate.
L = Mr 2 ω angular momentum (17)
of a bicycle wheel

Figure 8a
The angular L
momentum
vector.

Figure 8b
The angular ω
velocity
vector.
12-8 Rotational Motion

Calculating Moments of Inertia Since all the mass in the hollow cylinder is out at a
Equation 20 is not the most general formula for calcu- radius r, just as it is for a bicycle wheel, the hollow
lating moments of inertia. The bicycle wheel is special cylinder has a moment of inertia dI given by
in that all the mass is essentially out at a single radius
r. If, instead, we had a solid wheel where the mass was dI = dm × r 2
spread out over different radii, we would have to
= M × 2r dr 2Mr 3 dr
2 ×r =
2 (22)
conceptually break the wheel into a number of separate
R R2
rim-like wheels of radii r i and mass m i , calculate the
moment of inertia of each rim, and add the results The total moment of inertia of the cylinder is the sum
together to get the total moment of inertia. of the moments of inertia of all the hollow cylinders.
This addition is done by integrating the formula for dI
In Appendix A we have relatively complete discussion from r = 0 out to r = R.
of how to calculate moments of inertia, and how
r=R
moment of inertia is related to rotational kinetic energy.
There you will see that rotational kinetic energy is I solid
cylinder = dI
1/2 Iω 2 , which is analogous to the linear kinetic energy r=0
1/2 Mv 2 . This material is placed in an appendix, not r=R
because it is difficult, but because we do not wish to = 2Mr 3dr
digress from our discussion of the analogy between R2
r=0
linear and angular momentum. At this point, one
example and one exercise should be a sufficient intro- r=R
duction to the concept of moment of inertia. = 2M r 3dr
R2
r=0
Example 1 R
4
Calculate the moment of inertia, about its axis, of a = 2M r4 = 2M R2
2 4
cylinder of mass M and outside radius R. Assume that R 4R
the cylinder has uniform density. 0

Solution: We conceptually break the cylinder into a I solid

cylinder
= 1 MR 2 (23)
2
series of concentric cylinders of radius r and thickness
dr as shown in Figure (9). Each hollow cylinder has a
mass given by
end area of hollow cylinder
dm = M × total end area
dr

r
= M × 2πr dr
2 = M×
2r dr (21)
πR R2

Figure 9
Calculating the moment of inertia of
a cylinder about its axis of rotation.
 12-9

Two points are made in Example 1, The first is that VECTOR CROSS PRODUCT
calculating the moment of inertia of an object usually The idea of having the angular velocity ω being a
requires an integration, because different parts of the vector pointing along the axis of rotation gave us a nice
object are out at different distances r from the axis of analogy between linear momentum p = Mv and angu-
rotation. Secondly we see that the moment of inertia of lar momentum L = Iω . But to obtain the dynamical
a solid cylinder is less than the moment of inertia of a equation for angular momentum, the one analogous to
bicycle wheel of the same mass and outer radius Newton’s second law for linear momentum, we need
( 1/2 MR 2 for the cylinder versus MR 2 for the bicycle the mathematical formalism of the vector cross product
wheel). This is because all the mass of the bicycle defined back in Chapter 2. Since we have not used the
wheel is out at the maximum radius R, while most of the vector cross product before now, we will briefly review
mass of the solid cylinder is in at smaller radii. the topic here.
A considerable amount of time can be spent discussing If we have two vectors A and B like those shown in
the calculation of moments of inertia of various shaped Figure (11), the vector cross product A × B is defined
objects. Rather than do that here, we will simply to have a magnitude
present a table of the moments of inertia of common
objects of mass M and outer radius R, about an axis that
passes through the center. A × B = A B sin θ (24)

Object Moment of Inertia where A and B are the B

cylindrical shell 1 MR 2
magnitudes of the vectors A θ
solid cylinder 1 2 MR 2 and B, and θ is the small A
2 3 MR 2 angle between them. Note Figure 11
spherical shell
that when the vectors are par- The vectors A and B.
solid sphere 2 5 MR 2 allel, sin θ = 0 and the cross
product is zero. The cross product is a maximum when
Exercise 4
the vectors are perpendicular. This is just the opposite
As shown in Figure (10) we have a thick-walled hollow from the scalar dot product which is a maximum when
brass cylinder of mass M, with an inner radius R i and the vectors are parallel and zero when perpendicular.
outer radius R o. Calculate its moment of inertia about
Conceptually you can think of the dot product as
its axis of symmetry. Check your answer for the case
R i = 0 (a solid cylinder) and for R i = R 0 (which corre-
measuring parallelism while the cross product mea-
sponds to the bicycle wheel). sures perpendicularity.

The other major difference between the dot and cross

product is that with the dot product we end up with a
number (a scalar), while with the cross product, we end
up with a vector. The direction of A × B is the most
Ri peculiar feature of the cross product; it is perpendicu-
lar to the plane defined by the vectors A and B . If we
draw A and B on a sheet of paper as we did in Figure
(11), then the directions perpendicular to both A and
B are either up out of the paper or down into the paper.
Ro To decide which of these two directions to choose, we
use the following right hand rule. (This is an arbitrary
Figure 10 convention, but if you use it consistently in all of your
Thick-walled hollow cylinder. calculations, everything works out OK).
12-10 Rotational Motion

Right Hand Rule for Cross Products If you did the exercise (5b) correctly, you found that
To find the direction of the vector A × B , point the B × A points in the opposite direction from the vector
fingers of your right hand in the direction of the first A × B . In all previous examples of multiplication you
vector in the product (namely A ). Then, without have likely to have encountered, the order in which you
breaking your knuckles, curl the finger of your right did the multiplication made no difference. For ex-
hand toward the second vector B . Curl them in the ample, both 3 x 5 and 5 x 3 give the same answer 15. But
direction of the small angle θ . If you do this correctly, now we find that A × B = – B × A and the order of the
the thumb of your right hand will point in the direction multiplication does make a difference. Mathemati-
of the cross product A × B . Applying this to the cians say that cross product multiplication does not
vectors in Figure (11), we find that the vector A × B commute.
points up out of the paper as shown in Figure (12).
There is one other special feature of the cross product
worth noting. If A and B are parallel, or anti parallel,
Exercise 5 then they do not define a unique plane and there is no
(a) Follow the steps we just mentioned to show that unique direction perpendicular to both of them. Vari-
A × B from Figure (11) does point up out of the paper.
ous possibilities are indicated in Figure (13). But when
(b) Show that the vector B × A points down into the the vectors are parallel or anti parallel, sin θ = 0 and
paper. the cross product is zero. The special case where the
cross product does not have a unique direction is when
A B the cross product has zero magnitude with the result
that the lack of uniqueness does not cause a problem.

A B
B
B B
A
θ
Figure 13
A A If the vectors A and B are either parallel or
antiparallel, then as shown above, there is a whole
plane of vectors perpendicular to both A and B.
Figure 12
Right hand rule for vector cross
product A × B. Point the fingers of your
right hand in the direction of the first
vector A and then curl them in the direction
of the second vector B (without breaking
your knuckles). Your thumb will then point
in the direction of the cross product A × B.
 12-11

CROSS PRODUCT DEFINITION Exercise 6

OF ANGULAR MOMENTUM a) Look at Figure (14c) show- vi
Let us now see how we can use the idea of a vector cross ing the vectors ri (which point
i
product to obtain a definition of angular momentum into the paper) and vi . Point the axis of ri
mi
vectors. To explain the bicycle wheel experiments, we fingers of your right hand in the rotation
wanted the angular momentum to point along the axis direction of ri and then curl
Figure 14c
them toward the vector vi . Does
of the wheel as shown in Figure (14a). Since there are The three vectors
your thumb point in the direc- r i , vi and i
two directions along the axis, we have arbitrarily
tion of the vector i shown? (If it
chosen the direction defined by the right hand conven- does not, you have peculiar
tion shown. (Curl the fingers of your right hand in the knuckle joints or are not follow-
direction of the rotation and your thumb will point in ing instructions).
the direction of L ).
b) Choose any other mass that forms the bicycle wheel
shown in Figure (14b). Call that the mass m i . Show that
the vector i = m i r i × vi also points down the axis,
Figure 14a
Right hand parallel to i. Try this for several different masses, say
rule for L one at the top, one at the front, and one at the bottom of
angular the wheel.
momentum.
If you did Exercise 6 correctly, you found that all the
angular momentum vectors i = m i r i × vi were par-
allel to each other, all pointing down the axis of the
In Figure (14b) we went to the masses and spoke model wheel. We will define the total angular momentum of
of the bicycle wheel, and selected one particular mass the wheel as the vector sum of the individual angular
which we called m i . This mass is located at a coordi- momentum vectors i
nate vector r i from the center of the wheel, and is
traveling with a velocity vi . According to our defini- total angular
tion of angular momentum in Chapter 7, using the L momentum
of wheel
= Σi i = Σi mi r i × vi (26)
formula = mvr , the ball’s angular momentum should
be It is easy to add the vectors i because they all point in
= m ir ivi (7-11again) the same direction, as shown in Figure (15). Thus we
can add their magnitudes numerically. (It is just the
What we want to do now is to turn this definition of numerical sum we did back in Equation 15).
angular momentum into a vector that points down the
axis of the wheel. This we can do with the vector cross
7
product of r i and vi . We will try the definition of the Figure 15 6
vector i as Since all the 5
angular
i = m i r i × vi (25) momentum 4
vectors i point
vi in the same 3

Figure 14b i direction, we 2

Angular can add them up 1
momentum of one mi numerically.
of the balls in the ri
ball-spoke model of
a bicycle wheel.
12-12 Rotational Motion

To do the sum starting from Equation (26) we note that Ther × p Definition of
for each mass m i , the vectors r i and vi are perpen- Angular Momentum
dicular, thus A slight rewriting of our definition of angular momen-
vi tum, Equation 25, gives us a more compact, easily
r i × vi = r v sin θ remembered result. Noting that the linear momentum
θ
= rv for θ = 90° p of a particle is p = mv , then a particle’s angular
ri
momentum can be written
Then note that for a rotating wheel, the speed v of the
rim is related to the angular velocity ω by = mr × v
v = rω (11repeated) = r × mv
so that
= r ×p (30)
r i × vi = r v = r rω = r 2 ω (27)
In Chapter 7, we saw that the magnitude of the angular
Finally note that the vector ω points in the same
momentum if a particle was given by the formula
direction as r i × vi , so that Equation 27 can be written
as the vector equation = r⊥ p (7-15)

r i × vi = r 2ω
for all
(28) where r⊥ was the lever arm or perpendicular distance
mass mi
from the path of the particle to the point O about which
Using Equation 28 in Equation 26 gives we were measuring the angular momentum. This was
illustrated in Figure (7-10) (reproduced here), where a
L = Σi i = Σi mi r i × vi ball of momentum p , passing by an axis O, is caught by
a hook and starts rotating in a circle.
= Σi mi r 2ω p = mv

path of b
all
= Mr 2ω a)

angular momentum ball heading

r = pe
Mr 2 ω
for hook rpendicu
L = of a rotating (29) distance lar
fro
bicycle wheel of ball to m path
point O
O
where M is the sum of the individual masses m i .
p = mv
Equation 29 is the desired vector version of our original
Equation 16.
b)
ball catches
The important point to get from the above discussion is on hook r
that by using the vector cross product definition of
angular momentum i = m i r i × vi , all the i for each O
mass in the wheel pointed down the axis of the wheel,
and we could thus calculate the total angular momen-
tum by numerically adding up the individual i .
p = mv

c) r
ball swinging in circle,
with angular momentum
O
= mvr
Figure 7-10
As the ball is caught by the hook, its
angular momentum, about the point O,
remains unchanged. It is equal to (r⊥ p) .
 12-13

After the ball is caught it is traveling in a circle with an Exercise 7

angular momentum = r⊥ mv = r⊥p . By defining the
Using the vectors r and p in Figure (16), does the
angular momentum as r⊥p even before the ball was
vector = r × p point up out of the paper or down into
caught, we could say that the ball had the same angular the paper?
momentum r⊥p before it was caught by the hook as it
did afterward; that the angular momentum was un- The intuitive point you should get from this discussion
changed when the ball was grabbed by the hook. is that the magnitude of the vector cross product r × p
The idea that the angular momentum is the linear is equal to the magnitude of p times the perpendicular
momentum times the perpendicular lever arm r⊥ fol- lever arm r⊥ . We will shortly encounter the cross
lows automatically from the cross product definition of product r × F where F is a force vector. We will
angular momentum = r × p . To see this, consider a immediately know that the magnitude of r × F is r⊥ F
ball with momentum p moving past an axis O as shown where again r⊥ is a perpendicular lever arm.
in Figure (16a). At the instant of time shown, the ball
is located at a coordinate vector r from the axis. The
angle between the vectors r and p is the angle θ
shown in Figure (16b). The vector cross product r × p
is given

= r × p = rp sin θ (31)

However we note that the lever arm or perpendicular

distance r⊥ is given from Figure (16a)

r⊥ = r sin θ (32)
Combining Equations 31 and 32 gives

= r × p = (r sinθ)p = r⊥ p (33)

which is the result we used back in Chapter 7.

Figure 16a pa
th
The coordinate of
ba
vector r and the r ll
lever arm r⊥ are θ
related by Ο θ
r
r⊥ = r sin θ . axis p

Figure 16b r
The angle between θ
r and p is θ .
p
12-14 Rotational Motion

ANGULAR ANALOGY TO This product is zero because the vectors v and p = mv

NEWTON’S SECOND LAW are parallel to each other, and the cross product of
parallel vectors is zero. Thus Equation 36 becomes
We now have the mathematical machinery we need to
formulate a complete angular analogy to Newton’s d r × p = r × dp (38)
second law. We do this by noting that to go from dt dt
linear momentum p to angular momentum , we took
With this result, let us return to Newton’s law for linear
the cross product with the coordinate vector r
momentum
= r ×p (30repeated) dp
F = (39)
The origin of the coordinate vector r is the point about dt
which we wish to calculate the angular momentum. As long as we do the same thing to both sides of an
equation, it is still a correct equation. Taking the vector
To obtain a dynamical equation for angular momen- cross product r × on both sides gives
tum , we start with Newton’s second law which is a dp
dynamical equation for linear momentum p r ×F = r × (40)
dt
dp Using Equation 38 in Equation 40 gives
F = (11-16)
dt
d
where F is the vector sum of the forces acting on the r ×F = r ×p (41)
dt
particle.
Finally note that r × p is the particle’s angular mo-
With one mathematical trick, we can reexpress mentum , thus
Newton’s second law in terms of angular momentum.
The mathematical trick involves evaluating the expres- d
r×F = (42)
sion dt
d r ×p (34)
dt Equation (39) told us that the net linear force is equal to
In the ordinary differentiation of the product of two the time rate of change of linear momentum. Equation
functions a(t) and b(t), we would have 42 tells us that something, r × F , is equal to the time
d ab = da b + a db rate of change of angular momentum. What should we
dt dt dt
(35) call this quantity r × F ? The obvious name, from an
angular analogy would be an angular force. Then we
The same rules apply if we differentiate a vector cross
could say that the angular force is the time rate of
product. Thus
change of angular momentum, just as the linear force
d r × p = dr × p + r × dp (36) is the time rate of change of linear momentum.
dt dt dt
The world does not use the name angular force for
Equation 36 can be simplified by noting that
r × F . Instead it uses the name torque, and usually
v = dr designates it by the Greek letter τ (“tau”)
dt
so that
definition
torque τ ≡ r × F of torque
(43)
dr
× p = v × p = v × mv = 0 (37)
dt
With this naming, the angular analogy to Newton’s
second law is

torque = rate
d
τ = of change of (44)
dt angular momentum
 12-15

ABOUT TORQUE Exercise 8

To gain an intuitive picture of the concept of torque In Figure (17) we have separately drawn the vectors F
τ = r × F, imagine that we have a bicycle wheel with a and r for each diagram. Using the right hand rule for
fixed axis, and push on the rim of the wheel with a force cross products, find the direction of τ = r × F for each of
F as shown in Figure (17). In (17a) the force F is these three diagrams.
directed through the axis of the wheel, in this case the
force has no lever arm r⊥ . In (17b), the force is applied If you did Exercise 8 correctly, you found that r × F = 0
above the axis, while in (17c) the force is applied below for Figure (17a), that τ = r × F pointed up out of the
the axis. paper in (17b), and down into the paper in (17c). Thus
we find that when we apply a zero torque as in (17a), we
Intuitively, you can see that the wheel will not start get zero change in angular momentum. In (17b) we
turning if you push right toward the axis. When you applied an upward directed torque, and saw that the
push above the axis as in (17b), the wheel will start to wheel would start to turn to produce an upward directed
rotate counter clockwise. By our right hand convention angular momentum. In (17c), the downward directed
this corresponds to an angular momentum directed up torque produces a downward directed angular momen-
out of the paper. tum. These are all results we would expect from the
equation τ = d /dt .
In (17c), where we push below the axis, the wheel will
start to rotate clockwise, giving it an angular momen- In our discussion of angular momentum, we saw that
tum directed down into the paper. = r × p had a magnitude = r⊥ p where r⊥ was the
perpendicular lever arm. A similar result applies to
torque. By the same mathematics we find that the
r magnitude of the torque τ produced by a force F is
F
a) τ = r⊥ F (45)
r F where r⊥ is the perpendicular lever arm seen in Figures
(17b,c).

Intuitively, the best way to remember torque is to think

of it as a force times a lever arm. To turn an object, you
need both a force and a lever arm. In Figure (17a), we
r⊥ F
r had a force but no lever arm. The line of action of the
r force went directly through the axis, with the result that
b) θ
the wheel did not start turning. In both cases (17b) and
F (17c), there was both a force and a lever arm r⊥ , and the
wheel started turning.

To get the direction of the torque, to determine whether

τ points up or down (and thus gives rise to an up or down
F angular momentum), use the right hand rule applied to
θ the vector cross product τ = r × F . A convention,
c) r
r which we will use in the next chapter on Equilibrium, is
r⊥ to say that a torque that points up out of the paper is a
F
positive torque, while a torque pointing down into the
paper is a negative one. With this convention, we see that
Figure 17
the force in Figure (17b) is exerting a positive torque
Both a force F and a lever arm r⊥ are needed to
turn the bicycle wheel. The product r⊥ F is the (and creating positive angular momentum), while the
magnitude of the torque τ acting on the wheel. force in Figure (17c) is producing a negative torque (and
creating negative angular momentum).
12-16 Rotational Motion

CONSERVATION OF Now break the net force Fi into the sum of the external
ANGULAR MOMENTUM forces Fi external and the sum of the internal forces
Fi internal . This gives
In our discussion of a system of particles in Chapter 11,
we saw that if we had a system of many interacting
particles, with internal forces Fi internal between the
dL =
dt Σi r i × Fi external + Σi r i × Fi internal
(49)
Σi τi external + Σ τi internal
particles, as well as various external forces Fi external ,
=
we obtained the equation

Fexternal = dP (11-12) Next assume that all the internal forces are equal and
dt opposite as required by Newton’s third law, and are
where Fexternal is the vector sum of all the external directed toward or away from each other. In Figure
forces acting on the system, and P is the vector sum of (19) we consider a pair of such internal forces and note
all the momenta p i of the individual particles. This that both coordinate vectors r 1 and r 2 have the same
result was obtained using Newton’s third law and perpendicular lever arm r ⊥ . Thus the equal and
noting that all the internal forces cancel in pairs. In the opposite forces F1,2 external and F2,1 internal create equal
case where there is no net external force acting on the and opposite torques which cancel each other in Equa-
system, then dP/dt = 0 and the total linear momentum tion (49). The result is that all torques produced by
of the system is conserved. internal forces cancel in pairs, and we are left with the
general result
We can obtain a similar result for angular momentum
by starting with the definition of the total angular dL
momentum L of a system as being the vector sum of τ external =
dt
the angular momentum of the individual particles i (50)
definition of the where τ external is the vector sum of all the external
L ≡ Σi i total angular momenta (46) torques acting on the system of particles, and L is the
of a system of particles vector sum of the angular momentum of all of the
particles.
Differentiating Equation (46) with respect to time
gives

Σi dt i
dL = d i th particle
(47) Figure 18
dt Coordinate ri
For an individual particle i, we have vector for the
i th particle.
d i Equation 44
= τ i = r i × Fi applied to (48)
dt particle i

where Fi is the vector sum of the forces acting on the

particle i. As shown in Figure (18), we can take r i to
be the coordinate vector of the i-th particle. For this
discussion, we can locate the origin of the coordinate 1
F12 internal
system anywhere we want.
Figure 19
F21 internal
Both coordinate r1
Substituting Equation (48) into Equation (47) gives
vectors r 1 and r
r 2 have the same
Σi Σi r i × Fi
dL = d i 2
= perpendicular
dt dt lever arm r ⊥ r2
 12-17

In order to define torque or angular momentum, we Another consequence of Equation 50 is that if we have
have to choose an axis or origin for the coordinate an isolated system of particles with no net external
vectors r i . (Both torque and angular momentum torque acting on it, then the total angular momentum
involve the lever arm r ⊥ about that axis.) Equation 50 will be unchanging, will be conserved. This is one
is remarkably general in that it applies no matter what statement of the law of conservation of angular mo-
origin or axis we choose. In general, choosing a mentum. Our derivation of this result relied on the
different axis will give us different sums of torques and assumption of Newton’s third law that all internal
a different total angular momentum, but the new torques forces are equal and opposite and directed toward each
and angular momenta will still obey Equation 50. other. Since angular momentum is conserved on an
atomic, nuclear and subnuclear scale of distance, where
In some cases, there is a special axis about which there Newtonian mechanics no longer applies, our deriva-
is no external torque. In the bicycle wheel demonstra- tion is in some sense backwards. We should start with
tions where we stood on a rotating platform, the freely the law of conservation of angular momentum as a
rotating platform did not contribute any external torques fundamental law, and show for large objects which
about it own axis, which we called the z axis. As long obey Newtonian mechanics, the sum of the internal
as we did not touch another person or some furniture, torques must cancel. This is the kind of argument we
then the z component of the external torques were zero. applied to the conservation of linear momentum in
Since Equation 50 is a vector equation, that implies Chapter 11 (see Equation 11-14).
dL z
τz external = = 0 (51)
dt
and we predict that the z component of the total angular
momentum (us and the bicycle wheel) should be
unchanged, remain constant, no matter how we turned
the bicycle wheel. This is just what we saw.

Movie

Figure 7–15 repeated

Since the platform is completely free to rotate about the z axis, there are no z directed external
torques acting on the system consisting of the platform, person and bicycle wheel. As a result the
z component of angular momentum is conserved when the bicycle wheel is turned over.
(Note: when the wheel is being held up, we are looking at the under side.)
12-18 Rotational Motion

GYROSCOPES The force F, in Figure (20), produces a torque

The gyroscope provides an excellent demonstration of τ = r × F that also points to the right as shown. (The
the predictive power of the equation τ = dL dt. dt right hand convention used here is to point your fingers
Gyroscopes behave in peculiar, non intuitive ways. in the direction of the first vector r , curl them in the
The fact that a relatively straightforward application of direction of the second vector F, and your thumb points
the equation τ = dL dt predicts this bizarre behavior, in the direction of the cross product r × F = τ.)
provides a graphic demonstration of the applicability When we start with the bicycle wheel at rest, and apply
of Newton’s laws from which the equation is derived. the right directed torque shown in Figure (20), we get
a right directed angular momentum L. Thus the torque
Start-up τ and the resulting angular momentum L point in the
For this discussion, a bicycle wheel with a weighted same direction. In addition, the longer we apply the
rim will serve as our example of a gyroscope. To torque, the faster the wheel spins, and the greater the
weight the rim, remove the tire and wrap copper wire angular momentum L. Thus both the direction and
around the rim to replace the tire. The axle needs to be magnitude of L are consistent with the equation
extended as shown in Figure (20). τ = dL dt. dt

As an introduction to the gyroscope problem, start with

the bicycle wheel at rest, hold the axle fixed, and apply
a force F to the rim as shown in Figure (20). The force
shown will cause the wheel to start spinning in a
direction so that the angular momentum L points to the
right as shown. (Curl the fingers of your right hand in
the direction of rotation and your thumb points in the
direction of L .)

τ=rxF

F r

axel
L

ω
Figure 20
Spinning up the bicycle wheel. Note that the
resulting angular momentun L points in the
same direction as the applied torque τ .
Figure 25 Movie
The gyroscope really works!
 12-19

Precession The formula for this gravitational torque τg is

When we apply the equation τ = dL dt to a gyro-
torque about point O
scope that is already spinning, and apply the torque in produced by the
a direction that is not parallel to L, the results are not so τ g = r × mg gravitational force (52)
obvious. on the bicycle wheel

Suppose we get the bicycle wheel spinning rapidly so The new feature of the gyroscope problem, which we
that it has a big angular momentum vector L, and then have not encountered before, is that the torque τ does
suspend the bicycle wheel by a rope attached to the end not point in the same direction as the angular momen-
of the axle as shown in Figure (21). tum L of the bicycle wheel. If we look at Figure (21),
point the fingers of our right hand in the direction of the
To predict the motion of the spinning wheel, the first vector r, and curl our fingers in the direction of the
step is to analyze all the external forces acting on it. vector mg, then our thumb points down into the paper.
There is the gravitational force mg which points This is the definition of the direction of the vector cross
straight down, and can be considered to be acting at the product r × mg. But the angular momentum L of the
center of mass of the bicycle wheel, which is the center bicycle wheel points along the axis of the wheel to the
of the wheel as shown. Then there is the force of the right in the plane of the paper. In order to view both the
rope which acts along the rope as shown. No other angular momentum vector L and the torque vector τ
detectable external forces are acting on our system of in the same diagram, we can look down on the bicycle
the spinning wheel. wheel from the celing as shown in Figure (22).
One thing we know about the force Frope is that it acts When we started the wheel spinning, back in Figure
at the point labeled O where the rope is tied to the axle. (20), the torque τ and angular momentum L pointed in
If we take the sum of the torques acting on the bicycle the same direction, and we had the simple result that the
wheel about the suspension point O, then Frope has no longer we applied the torque, the more angular mo-
lever arm about this point and therefore contributes no mentum we got. Now, with the torque and angular
torque. The only torque about the suspension point O momentum pointing in different directions as shown in
is produced by the gravitational force mg whose lever Figure (22), we expect that the torque will cause a
arm is r, the vector going from point O down the axle change in the direction of the angular momentum.
to the center of the bicycle wheel as shown in Figure
(21).
τ = r x mg
rope

mg points down
Frope τ = r x mg
r
axel
L O L
O r rope axel
axis for r
torque

looking down on wheel

mg
mg
Figure 21 Figure 22
Suspend the spinning bicycle wheel by a rope Looking down from the ceiling, the vector mg
attached to the axle. The gravitational force mg points down into the paper and τ = r × mg
has a lever arm r about the axis O. This creates points to the top of the page. In this view we
a torque τ = r × mg pointing into the paper. can see both the vectors L and τ .
12-20 Rotational Motion

To predict the change in L , we start with the angular Since the torque τ is in the horizontal plane, the vector
form of Newton’s second law L new = L old + τ dt is also in the horizontal plane.
And since τ and L old are perpendicular to each other,
τ = dL L new has essentially the same length as L old . What is
dt
and multiply through by the short (but finite) time happening is that the vector L is starting to rotate
interval dt to get counter clockwise (as seen from above) in the horizon-
tal plane.
dL = τ dt (53)
One final, important point. For this experiment we
Equation 53 gives us dL, which is the change in the
were careful to spin up the bicycle wheel so that before
bicycle wheel’s angular momentum as a result of
we suspended the wheel from the rope, the wheel had
applying the torque τ for a short time dt.
a big angular momentum pointing along its axis of
To see the effect of this change dL, we will use some rotation. When we apply a torque to change the
of the terminology we used in the computer prediction direction of L, the axis of the wheel and the angular
of motion. Let us call L old the old value of the angular momentum vector L move together. As a result the axis
momentum that the bicycle wheel had before the time of the bicycle wheel also starts to rotate counter clock-
interval dt, and L new the new value at the end of the wise in the horizontal plane. The bicycle wheel, instead
time interval dt. Then L new will be related to L old by of falling as expected, starts to rotate sideways.
the equation
Once the bicycle wheel has turned an angle d θ side-
L new = L old + dL (54) ways, the axis of rotation and the torque τ also rotate
by an angle d θ, so that the torque τ is still perpendicular
Using Equation 53 for dL gives to L as shown in Figure (24). Since τ always remains
perpendicular to L, the vector τ dt cannot change the
L new = L old + τdt (55) length of L. Thus the angular momentum vector L
A graph of the vectors L old , L new , and τ dt is shown remains constant in magnitude and rotates or “pre-
in Figure (23). In this figure the perspective is looking cesses” in the horizontal plane. This is the famous
down on the bicycle wheel, as in Figure (22). precession of a gyroscope which is nicely demon-
strated using the bicycle wheel apparatus of Figure
τ = r x mg (21).

Figure 22 repeated
Looking down from mg points down
the ceiling.
t
r τ 2d
O L
rope axel
L3
τ 1dt
L2
looking down on wheel
dθ
dθ L1 τ0dt
dθ
L0
L new
τ dt Figure 24
dθ After each time step dt, the angular
L old momentum vector L (and the bicycle
wheel axis) rotates by another angle dθ.
Figure 23
The vectors Lold, Lnew and τ dt as seen
from the top view of Figure 22.
 12-21

To calculate the rate of precession we note from If you try the bicycle wheel demonstration that we
Figures (23) or (24) that the angle d θ is given by discussed, the results come out close to the prediction.
Instead of falling as one might expect, the wheel
dθ = τdt (56) precesses horizontally as predicted. There is a slight
L
where we use the fact that τdt is a very short length, and drop when you let go of the wheel, which can be
thus sin dθ and d θ are equivalent. Dividing both compensated for by releasing the wheel at a slight
sides of Equation 56 through by dt, we get upward angle.

dθ = τ (57) If you look at the motion of the wheel carefully, or study

dt L the motion of other gyroscopes (particularly the air
But d θ/dt is just the angular velocity of precession, bearing gyroscope often used in physics lectures) you
measured in radians per second. Calling this preces- will observe that the axis of the wheel bobs up and
sional velocity Ωprecession ( Ω is just a capital ω down slightly as it goes around. This bobbing, or
“omega”), we get epicycle like motion, is called nutation. We did not
predict this nutation because we made the approxima-
τ precessional tion that the axis of the wheel exactly follows the
Ωprecession = angular velocity (58) angular momentum vector. This approximation is very
L of a gyroscope good if the gyroscope is spinning rapidly but not very
good if L is small. Suppose, for example we release the
Exercise 9 wheel without spinning it. Then it simply falls. It starts
A bicycle wheel of mass m, radius r, is spun up to an to rotate, but along a different axis. As it starts to fall
angular velocity ω. It is then suspended on an axle of it gains angular momentum in the direction of τ. A
length h as shown in Figure (21). Calculate more accurate analysis of the motion of the gyroscope
can become fairly complex. But as long as the gyro-
(a) the angular momentum L of the bicycle wheel. scope is spinning fast enough so that the axis moves
(b) the angular velocity of precession. with L , we get the simple and important results dis-
cussed above.
(c) the time it takes the wheel to precess around once
(the period of precession). [You should be able to
obtain the period of precession from the angular
velocity of precession by dimensional analysis.]

(d) A bicycle wheel of total mass 1kg and radius 40cm,

is spun up yo a frequency f = 2πω = 10 cycles/sec. The
handle is 30cm long. What is the period of precession
in seconds? Does the result depend on rhe mass of the
bicycle wheel?
12-22 Rotational Motion

ROTATIONAL KINETIC ENERGY

APPENDIX Let us go back to our example, shown in Figure (3)
Moment of Inertia and repeated here, of a ball of mass m, on the end of a
massless stick of length r, rotating with an angular
Rotational Kinetic Energy velocity ω . The speed v of the ball is given by Equation
11 as
In the main part of the text, we briefly discussed
v = rω (11repeated)
moment of inertia as the angular analogy to mass in the
formula for angular momentum. As linear momentum and the ball’s kinetic energy will be
p of an object is its mass m times its linear velocity v
kinetic = 1 mv 2 = 1 m rω 2
energy 2 2
p = mv linear momentum (A1)
the angular momentum is the angular mass or = 1 mr 2 ω 2 (A4)
2
moment of inertia I time the angular velocity ω
Since the ball’s moment of inertia I about the axis of
= Iω angular momentum (A2) rotation is mr 2 , we get as the formula for the ball’s
kinetic energy
In the simple case of a bicycle wheel, where all the mass analogous
kinetic = 1 Iω 2
is essentially out at a distance (r) from the axis of the energy 2 to 1/2 mv2 (A5)
wheel, the moment of inertia about the axis is
We see the angular analogy working again. The ball
moment of inertia has a kinetic energy, due to its rotation, which is
I = Mr 2 of a bicycle wheel (A3)
analogous to 1/2 mv 2 , with the linear mass m replaced
where M is the mass of the wheel. by the angular mass I and the linear velocity v replaced
by the angular velocity ω .
When the mass of an object is not all concentrated out
at a single distance (r) from the axis, then we have to ω
calculate the moment of inertia of individual parts of
the object that are at different radii r, and tie together the
various pieces to get the total moment of inertia. This r m
usually involves an integration, like the one we did in k
stic
Equations 21 through 23 to calculate the moment of ss
ssle
inertia of a solid cylinder. ma
θ
For topics to be discussed later in the text, the earlier
discussion of moment of inertia is all we need. But pivot
there are topics, such as rotational kinetic energy and its Figure 3 repeated
connection to moment of inertia, which are both inter- Mass rotating on the end of a massless stick.
esting, and can be easily tested in both lecture demon-
strations and laboratory exercises. We will discuss
these topics here.
 12-23

If we have a bicycle wheel of mass M and radius r In most of our examples we will consider objects like
rotating at an angular velocity ω, we can think of the bicycle wheels or hollow cylinders where the mass is
wheel as being made up of a collection of masses on the essentially all at a distance r from the axis of rotation,
ends of rods as shown in Figure (6) repeated here. For and we can use the formula Mr 2 for the moment of
each individual mass m i , the kinetic energy is 1/2 m iv 2 inertia. But often the mass is spread out over different
where v = r ω is the same for all the masses. Thus the radii and we have to calculate the angular mass. An
total kinetic energy is example is a rotating shaft shown back in Figure (9),
where the mass extends from the center where r = 0 out
kinetic energy
of bicycle wheel
= Σi 12 m i r 2 ω2 to the outside radius r = R.

Suppose we have an arbitrarily shaped object rotating

= 1 r 2 ω2 Σ m i an angular velocity ω about some axis, as shown in
2 i Figure (A1). To find the moment of inertia, we will
calculate the kinetic energy of rotation and equate that
= 1 r 2 ω2 M
2 to 1/ 2Iω 2 to obtain the formula for I. To do this we
where the sum of the masses Σ m i is just the mass M conceptually break the object into many small masses
of the wheel. The result can now be written dm i located a distance ri from the axis of rotation as
shown. Each dmi will have a speed v i = ri ω , and thus
kinetic energy a kinetic energy
= 1 Mr 2 ω 2
of bicycle wheel 2

= 1 I ω2 (A6)
kinetic energy
of object
= Σi 1m v 2
2 i i
2
If we call Mr 2 the angular mass, or moment of inertia = Σi 1 m r 2 ω2
2 i i
I of the bicycle wheel, we again get the formula
1 2 I ω 2 for kinetic energy of the wheel. Thus we see = 1 ω 2Σ m i r i 2
2 i
that, in calculating this angular mass or moment of
inertia, it does not make any difference whether the = 1 ω2 I (A7)
2
mass is concentrated at one point as in Figure (3), or
spread out as in Figure (6). The only criterion is that the From Equation A7, we see that the general formula for
mass or masses all be out at the same distance r from the moment of inertia is
axis of rotation.
I = Σi m i r i2 (A8)
ω
ω
mi

r ri

axis of
rotation

Figure 6 repeated Figure A1

Bicycle wheel as a collection of masses Calculating the moment of inertia of
on the end of massless rods. an object about the axis of rotation.
12-24 Rotational Motion

In example 1, Equations 21 through 23, we showed you COMBINED

how to calculate the moment of inertia of a solid TRANSLATION AND ROTATION
cylinder about its axis of symmetry. In that example we
In our discussion of the motion of a system of particles,
broke the cylinder up into a series of concentric shells we saw that the motion was much easier to understand
of radius r i and mass dm i , calculated the moment of
if we focused our attention on the motion of the center
inertia of each shell dm i r i2 , and summed the results
of mass of the system. The simple feature of the motion
as required by Equation A7. As in most cases where we of the center of mass, was that the effects of all internal
calculate a moment of inertia, the sum is turned into an
forces cancelled. The center of mass moved as if it were
integral.
a point particle of mass M, equal to the total mass of
In Exercise 3 which followed Example 1, we had you the system, subject to a force F equal to the vector sum
calculate the moment of inertia, about its axis of of all the external forces acting on the object.
symmetry, of a hollow thick-walled cylinder. The
When the system is a rigid object, we have a further
calculation was essentially the same as the one we did
simplification. The motion can then be described as the
in Example 1, except that you had to change the limits motion of the center of mass, plus rotation about the
of integration. The following exercise gives you more
center of mass. To see that you can do this, imagine that
practice calculating moments of inertia, and shows you
you go to a coordinate system that moves with the
what happens when you change the axis about which object’s center of mass. In that coordinate system, the
the moment of inertia is calculated.
object’s center of mass point is at rest, and the only
Exercise A1
thing a rigid solid object can do is rotate about that
point.
Consider a uniform rod of mass M and length L as
shown in Figure (A2). A key advantage of viewing the motion of a rigid object
a) Calculate the moment of inertia of the rod about the this way is that the kinetic energy of a moving, rotating,
center axis, labeled axis 1 in Figure (A2). solid object is simply the kinetic energy of the center of
mass motion plus the kinetic energy of rotation. Ex-
b) Calculate the moment of inertia of the rod about an plicitly, if an object has a total mass M, and a moment
axis that goes through the end of the rod, axis 2 in Figure of inertia Icom about the center of mass (parallel to the
(A2). About which axis is the moment of inertia greater? axis of rotation of the object) then the formula for the
Explain why. kinetic energy of the object is

axis 2 axis 1 kinetic energy

of moving and = 1 MVcom
2
+ 1 I com ω 2 (A9)
2 2
rotating object

where Vcom is the velocity of the center of mass and ω

m the angular velocity of rotation about the center of
mass.
L
More important is the idea that motion can be separated
into the motion of the center of mass plus rotation about
the center of mass. To emphasize the usefulness of this
concept, we will first consider an example that can
Figure A2 easily be studied in the laboratory or at home, and then
Calculating the moment of inertia of a long thin rod. go through the proof of the equation.
 12-25

Example—Objects Rolling Thus we get for magnitudes

Down an Inclined Plane ωr = vcom ; ω = vcom /r (A12)
Suppose we start with a cylindrical object at the top of
an inclined plane as shown in Figure (A3), and measure Using Equation A12 in A10 gives
the time the cylinder takes to roll down the plane. Since
2
we do not have to worry about friction for a rolling 1 2 1 vcom
object, we can use conservation of energy to analyze mgh = mvcom + I 2
2 2 r
the motion.
1 I 2 (A13)
If the cylinder rolls down so that its height decreases by = m + 2 vcom
2 r
h as shown, then the loss of gravitational potential
energy is mgh. Equating this to the kinetic energy
Let us take a look at what is happening physically as the
gained gives
cylinder rolls down the plane. In our earlier analysis
1 1 2 of a block sliding without friction down the plane, all
mgh = m v2com + I ω (A10)
2 2 the gravitational potential energy mgh went into ki-
where m is the mass of the cylinder, vcom the speed of netic energy 1/2 mv2com . Now for a rolling object, the
the axis of the cylinder, I the moment of inertia about gravitational potential has to be shared between the
2
the axis and ω the angular velocity. kinetic energy of translation 1/2 mvcom and the kinetic
energy of rotation 1/2 Iω 2. The greater the moment of
If the cylinder rolls without slipping, there is a simple inertia I, the more energy that goes into rotation, the less
relationship between vcom and ω. We are picturing the available for translation, and the slower the object rolls
rolling cylinder as having two kinds of motion— down the plane.
translation and rotation. The velocity of any part of the
cylinder is the vector sum of vcom plus the velocity due In our discussion of moments of inertia, we saw that for
to rotation. two cylinders of equal mass, the hollow thin-walled
cylinder had twice the moment of inertia as the solid
At the point where the cylinder touches the inclined one. Thus if you roll a hollow and a solid cylinder down
plane, the rotational velocity has a magnitude vrot = ω r, the plane, the solid cylinder will travel faster because
and is directed back up the plane as shown in Figure less gravitational potential energy goes into the kinetic
(15). If the cylinder is rolling without slipping, the energy of rotation. You get to figure out how much
velocity of the cylinder at the point of contact must be faster in Exercise A2.
zero, thus we have
ω
rolling without
vcom + ω r = 0 slipping (A11)

r
vcom ωr
h

θ Figure A4
The velocity at the point of contact is the sum of the
Figure A3 center of mass velocity and the rotational velocity.
Calculating the speed of an object rolling down a plane. This sum must be zero if there is no slipping.
12-26 Rotational Motion

Before you work Exercise A2, think about this ques- PROOF OF THE
tion. The technician who sets up our lecture demon- KINETIC ENERGY THEOREM
strations has a metal sphere, and does not know for sure
We are now ready to prove the kinetic energy theorem
whether the sphere is solid or hollow. (It could be a for rotational motion. If we have an object that is
solid sphere made of a light metal, or a hollow sphere
rotating while it moves through space, its total kinetic
made from a more dense metal.) How could you find
energy is the sum of the kinetic energy of the center of
out if the sphere is solid or hollow? mass motion plus the kinetic energy of rotational
motion about the center of mass. The proof is a bit
Exercise A2 formal, but shows what you can do by working with
You roll various objects down the inclined plane shown vector equations.
in Figure (A3).
Consider a solid object, shown in Figure (A5), that is
(a) a thin walled hollow cylinder
moving and rotating. Let R com be the coordinate
(b) a solid cylinder vector of the center of mass of the object. We will think
of the object as being composed of many small masses
(c) a thin walled sphere m i which are located at R i in our coordinate system,
(d) a solid sphere and a displacement r i from the center of mass as
shown. As we can see from Figure (A5), the vectors
and for comparison, you also slide a frictionless block R com , R i and r i are related by the vector equation
down the plane:

(e) a frictionless block R i = R com + r i (A14)

For each of these, calculate the speed vcom after the We can obtain an equation for the velocity of the small
object has descended a distance h. (It is easy to do all mass m i by differentiating Equation A14 with respect
cases of this problem by writing the object’s moment of to time
inertia in the form I = αMR 2, , where α = 1 for the hollow
cylinder, 1/2 for the solid cylinder, etc.) What value of α dR i dR com d ri
= + (A15)
should you use for the sliding block? dt dt dt
Writing your results in the form vcom = β 2gh summa-
rize your results in a table giving the value of β in each
case. ( β = 1 for the sliding block, and is less than 1 mi
for all other examples.)

Exercise A3 A Potential Lab Experiment ri

In Exercise A2 you calculated the speed vcom of various
objects after they had descended a distance h. A block
sliding without friction has a speed v given by com
2 Ri
mgh = 1/ 2 mv , or v = 2gh . The rolling objects were
moving slower when they got to the bottom. For all
heights, however, the speed of a rolling object is slower
R com
than the speed of the sliding block by the same constant
factor. Thus the rolling objects moved down the plane
with constant acceleration, but less acceleration than
the sliding block. It is as if the acceleration due to gravity
were reduced from the usual value g. Using this idea, Figure A5
and the results of Exercise A2, predict how long each of Analyzing the motion of a small piece of an object.
the rolling objects take to travel down the plane. This
prediction can be tested with a stop watch.
 12-27

which can be written in the form Now the quantity m i vi that appears in the last term of
Equation A18 is the linear momentum of mi as seen in
Vi = Vcom + vi (A16) a coordinate system where the center of mass is at rest.
where Vi = dR i /dt is the velocity of mi in our coordi- To evaluate the sum of these terms, let us choose a new
nate system, Vcom = dR com /dt is the velocity of the coordinate system whose origin is at the center of mass
center of mass of the object, and vi = dr i /dt is the of the object as shown in Figure (A6). In this coordinate
velocity of m i in a coordinate system that is moving system the formula for the center of mass of the small
with the center of mass of the object. masses mi is

The kinetic energy of the small mass m i is r com = Σi m i r i = 0 (A19)

1 m V2 = 1 m V • V Differentiating Equation 19 with respect to time gives

2 i i 2 i i i
Σi m i dti Σi m i v i
dr
= = 0 (A20)
= 1 m i Vcom + vi • Vcom + vi
2
Equation A20 tells us that when we are moving along
= 1 m i Vcom 2
+ 2Vcom • vi + vi2 with the center of mass of a system of particles, the total
2
linear momentum of the system, the sum of all the m i vi ,
= 1 m i Vcom
2
+ 1 m i vi2 + m i Vcom • vi is zero.
2 2
(A17)
Using Equation A20 in A18 gets rid of the last term. If
we let M = Σ m i be the total mass of the object, we get
The total kinetic energy of the object is the sum of the
kinetic energy of all the small pieces m i i

total total
kinetic
energy
= Σi 12 mi Vi2 kinetic
energy
= 1 MVcom
2
2
+ Σi 12 m i v i2 (A21)

= 1 Vcom
2
2
Σi mi + Σi 12 m i vi2 + Vcom • Σi mi vi Equation A21 applies to any system of particles, whether
(A18) the particles make up a rigid object or not. The first
2
In two of the terms, we could take the common factor term, 1 2MV
2 com is the kinetic energy of center of mass
Vcom outside the sum. motion, and Σ 1 2 m i vi2 is the kinetic energy as seen
by someone moving along with the center of mass. If
the object is solid, then in a coordinate system where the
center of mass is at rest, the only thing the object can do
mi is rotate about the center of mass. As a result the kinetic
energy in that coordinate system is the kinetic energy
ri of rotation. If the moment of inertia about the axis of
rotation is I com , then the total kinetic energy is
com + 1 2 I com ω where ω is the angular ve-
2 2
1 2MV
2
locity of rotation. This is the result we stated in
com Equation A9.

Figure A6
Here we moved the origin of the coordinate
system to the center of mass.
Chapter 13
Equilibrium

CHAPTER 13 EQUILIBRIUM

When does a structure fall over, when does a bridge

collapse, how do you lift a weight in a way that
prevents serious injury to your back? We begin to
answer such questions by applying Newton’s laws to
an object that has neither linear nor angular accel-
eration. The most interesting special case is when an
object is at rest and will stay that way, when it is not
about to tip over or collapse.
13-2 Equilibrium

EQUATIONS FOR EQUILIBRIUM Example 1 Balancing Weights

If the center of mass of an object is not accelerating, As our first example, suppose we have a massless rod
then we know that the vector sum of the external forces of length L and suspend two masses m1and m2from the
acting on it is zero. If the object has no angular ends of the rod as shown in Figure (1). The rod is then
acceleration, then the sum of the torques about any axis suspended from a string located a distance x from he
must be zero. These two conditions left end of the rod. What is the distance x and how
strong a force F must be exerted by the string?
sum of
external Σ Fi external = 0 (1) Solution: The first step is to sketch the situation and
forces i draw the forces involved, as we did in Figure (1). Our
system will be the rod and the two masses. The external
sum of forces acting on this system are the two gravitational
torques
about Σ τ i external = 0 (2) forces m 1g and m 2g , and the force of the string F .
Since all the forces are y directed, when we set the
i
any axis vector sum of these external forces to zero we have
are what we will consider to be the required conditions ΣFy = F – m1 g – m2 g = 0 (3)
for an object to be in equilibrium.
Thus we get for F
Equations 1 and 2 are a complete statement of the basic
F = m1 + m2 g (4)
physics to be discussed in this chapter. Everything else
will be examples to show you how to effectively apply and we see that F must support the weight of the two
these equations in order to understand and predict when masses.
an object will be in equilibrium. In particular we wish
to show you some techniques that make it quite easy to To figure out where to suspend the rod, we use the
apply these equations. condition that the net torque produced by the three
external forces must be zero. Since a torque is a force
times a lever arm about some axis, you have to choose
F an axis before you can calculate any torques. The
important point in equilibrium problems is that you can
choose the axis you want. We will see that by
intelligently selecting an axis, we can simplify the
x L–x problem to a great extent.

Our definition of a torque τ caused by a force F is

O τ = r ×F (5)
m1 m2
where r is a vector from the axis O to the point of
m2 g application of the force F as shown in Figure (2).

In this chapter we do not need the full vector formalism

for torque that we used in the discussion of the gyro-
m1 g scope. Here we will use the simpler picture that the
magnitude of a torque caused by a force F is equal to
Figure 1
Masses m1 and m2 suspended from a massless
the magnitude of F times the lever arm r⊥ , which is the
rod. At what position x do we suspend the rod distance of closest approach between the axis and the
in order for the rod to balance? line of action of the force F as shown in Figure (2). If
the torque tends to cause a counter clockwise rotation,
as it is in Figure (2), we will call this a positive torque.
If it tends to cause a clockwise rotation, we will call that
a negative torque.
 13-3

(In Figure (2), the vector r × F points up out of the We obtained Equation 7 by setting the torques about
page. If the vector F were directed to cause a clockwise the balance point equal to zero. This choice had the
rotation, then r × F would point down into the paper. advantage that the suspending force F had no lever arm
[It is good practice to check this for yourself.] Thus we and therefore did not appear in our equations. We
are using the convention that torques pointing up are mentioned earlier that the condition for equilibrium
positive, and those pointing down are negative.) was that the sum of the torques be zero about any axis.
In Exercises 1 and 2 we have you select different axes
Returning to our problem of the rod and weights shown about which to set the torques equal to zero. With these
in Figure (1), let us take as our axis for calculating other choices, you will still get the same answer for x,
torques, the point of suspension of the rod, labeled namely Equation 7, but you will have two unknowns,
point O in Figure (1). With this choice, the force F F and x, and have to solve two simultaneous equations.
which passes though the point of suspension, has no You will see that we simplified the work by choosing
lever arm about point O and therefore produces no the suspension point as the axis and thereby eliminating
torque about that point. F from our equation.
The gravitational force m 1g has a lever arm x about
point O and is tending to rotate the rod counter clock- Exercise 1
wise. Thus m 1g produces a positive torque of magni- If we choose the left end of the rod as our axis, as shown
tude m1 gx. The other gravitational force m 2g has a in Figure (3), then only the forces F and m2g produce
lever arm (L – x) and is tending to rotate the rod a torque
clockwise. Thus m 2g is producing a negative torque (a) Is the torque produced by F positive or negative?
magnitude – m2 g L – x . Setting the sum of the
torques about point O equal to zero gives (b) Is the torque produced by m2g positive or negative?

m1 gx – m2 g L – x = 0 (6) (c) Write the equation setting the sum of the torques
about the left end equal to zero. Then combine that
equation with Equation 4 for F and solve for x. You
m2 should get Equation 7 as a result.
x = L (7)
m1 + m2
Exercise 2
Let us check to see that Equation 7 is a reasonable Obtain two equations for x and F in Figure (1) by first
result. If m1 = m2 , then we get x = L/2 which says setting the torques about the left end to zero, then by
that with equal weights, the rod balances in the center. setting the torques about the right end equal to zero.
If, in the extreme, m1 = 0 , then we get x = L, which Then solve these two equations for x and see that you
tells us that we must suspend the rod directly over m 2 , get the same result as Equation 7
also a reasonable result. And if m2 = 0 we get x = 0 as
expected.
F = m1 g + m2 g

r
o τo = r x F
axis x
r axis
F
L
m2
Figure 2
Figure 3
The torque τ = r × F has a magnitude τ = r⊥ × F .
Torques about the left end of the rod. m2 g
13-4 Equilibrium

GRAVITATIONAL FORCE ACTING But the sum Σ m ix i is by definition equal to M times

AT THE CENTER OF MASS the x coordinate of the center of mass of the object

Σi mixi
When we are analyzing the torques acting on an
MXcom ≡ (10)
extended object, we can picture the gravitational force
on the whole object as acting on the center of mass
point. To prove this very convenient result, let us Using Equation 10 in 9 gives
conceptually break up a large object of mass M into τO = MgXcom (11)
many small masses m i as shown in Figure (4), and
calculate the total gravitational torque about some Equation 11 says that the gravitational torque about any
arbitrary axis O. An individual particle m i located a axis O is equal to the total gravitational force Mg times
distance xi down the x axis from our origin O produces the horizontal coordinate of the center of mass of the
a gravitational torque τi given by object. Thus the gravitational torque is just the same as
if all of the mass of the object were concentrated at the
τi = m i g x i (8) center of mass point.
where mig is the gravitational force and xi the lever
arm. Adding up the individual torques τi to obtain the Exercise 3
total gravitational torque τO gives A wheel and a plank each have a mass M. The center
of the wheel is attached to one end of a uniform beam
τO = Σ τi = Σ migxi of length L. A nail is driven through the center of mass
of the plank and nailed into the other end of the beam as
i i
(9) shown in Figure (5). Where do you attach a rope around
= gΣ mixi the beam so that the beam will balance? Explain how
i you got your answer.

rope
mi

O
axis plank mass m
xi
mi g
beam
Figure 4 length L
Conceptually break the large object of mass M into
many small pieces of mass m i , located a distance x i wheel
down the x axis from our arbitrary origin O. mass m
Figure 5
A wheel and a plank are attached
to the ends of a uniform beam.
 13-5

TECHNIQUE OF SOLVING Example 3 Wheel and Curb

EQUILIBRIUM PROBLEMS A boy is trying to push a wheel up over a curb by
In our discussion of the balance problem shown in applying a horizontal force Fboy as shown in Figure
Figure (1), we saw that there were several ways to solve (6a). The wheel has a mass m, radius r, and the curb a
the problem. We always have the condition that for height h as shown. How strong a force does the boy
equilibrium the vector sum of the forces is zero have to apply?
ΣiFi = 0 and sum of the torques τ i about any axis is Solution: We will consider the wheel to be the object
zero Σi τ i = 0 . By choosing various axes we can in equilibrium, and as a first step sketch all the forces
easily get enough, or more than enough equations to acting on the wheel as shown in Figure (6b). We can
solve the problem. If we are not careful about the way treat this as an equilibrium problem by noting that as the
we do this however, we can end up with a lot of wheel is just about to go up over the curb, there is no
simultaneous equations that are messy to solve. force between the bottom of the wheel and the road.
Our first solution of the equilibrium condition for There is, however, the force of the curb on the wheel,
Figure (1) suggests a technique for simplifying the labeled Fcurb in Figure (6b). We know the point at
solution of equilibrium problems. In Equation 6 we set which Fcurb acts but we do not know off hand either the
to zero the sum of the torques about the balance point magnitude or direction of Fcurb, nor are we asked to find
O shown in Figure (1) reproduced here. We wanted to Fcurb.
calculate the position x of the balance point, and were
not particularly interested in the magnitude of the force
F . By taking the torques about the point O where F has r
Fboy
no lever arm, F does not appear in our equation. As a
result the only variable in Equation 6 is x, which can be
immediately solved to give the result in Equation 7. As
we saw in Exercises 1 and 2, if we chose the torques h
about any other point, both variables x and F appear in
our equations, and we have to solve two simultaneous mg
equations. Figure 6a
A boy, exerting a horizontal force on the axle of a
We will now consider some examples and exercises wheel, is trying to push the wheel up over a curb.
that look hard to solve, but turn out to be easy if you take How strong a force must the boy exert?
the torques about the correct point. The trick is to find
a point that eliminates the unknown forces you do not
want to know about.
F Fboy
Fcurb

x L–x O

O mg no force
m1 m2
Figure 6b
m2 g Forces on the wheel as the wheel is
just about to go up over the curb.
m1 g
Figure 1 (Repeated)
We can eliminate any force by a proper choice of axis.
13-6 Equilibrium

We can eliminate the unknown force Fcurb by setting to The final slip in solving this problem is to relate the
zero the torques about the point O where the curb distance to the wheel radius r and curb height h. As
touches the wheel. Since Fcurb has no lever arm about shown in Figure (6d), the right triangle from the axle to
this point, it will not appear in the resulting equation. the curb has sides , (r – h) and hypotenuse r. By the
Pythagorean theorem we get
In Figure (6c) we have sketched the geometry of the 2
2 2
problem. About the point O the force Fboy has a lever r = + r– h
arm (r – h) and is tending to cause a clockwise rotation or
about point O. Thus Fboy is producing (by our conven-
tion) a negative torque, of magnitude Fboy (r – h). The = 2 rh– h2 (13)
gravitational force mg has a lever arm shown in which finishes the problem.
Figure (6c), and is tending to produce a counter clock-
wise rotation. Thus it is producing a positive torque of Exercise 4
magnitude + mg about point O. Since there are no
The direction of Fcurb is slightly off in Figure (6b). Explain
other torques about point O, setting the sum of the
what would happen to the wheel if Fcurb pointed as
torques equal to zero gives
shown.
– Fboy(r – h) + mg = 0
(12)
Fboy = mg
r–h
r 2r
We immediately see that if the curb is as high as the
axle, if r = h, there is no finite force that will get the
wheel over the curb.

Figure 7a
A frictionless rod is placed in a
Fboy hemispherical frictionless bowl. What is
the equilibrium position of the rod?
r (r – h)
O

mg
Figure 6c
Geometry of the problem.

axel
r
(r – h)
curb
Figure 7a
Figure 6d We simulated a frictionless rod in a hemispherical
frictionless bowl by placing ball bearing rollers at
one end of the rod and the edge of the “bowl”. The
rod always comes to rest at this angle.
 13-7

Example 4 Rod in a Frictionless Bowl surface of the bowl is normal to the bowl and points
We include the problem here, first because it gives toward the center of the circle defining the bowl. The
some practice with what we mean by a frictionless force Fr between the rim of the bowl and the friction-
surface, but more importantly it is an example where less rod must be perpendicular to the rod as shown.
we can gain considerable insight without solving any
equations. Off hand we know nothing about the magnitude of the
forces Fb and Fr , only their directions. If we extend the
You place a frictionless rod of length 2r in a frictionless lines of action of Fb and Fr they will intersect at some
hemispherical bowl of radius r. Where does the rod point above the rod as shown in Figure (7c). If we set
come to rest? (Put in just enough friction to have it to zero the sum of the torques about this intersection
come to rest.) The situation is diagramed in Figure (7a). point, where neither Fb or Fr has a lever arm, then
In Figure (7b), we have made a reasonably accurate neither Fb or Fr will contribute. The only remaining
simulation of the problem by using a semi circular torque is that produced by the gravitational force mg .
piece of plastic for the bowl and placing small rollers on If the rod is in equilibrium, then the torque produced by
one end of the rod and one rim of the bowl to mimic the mg about the intersection point must also be zero, with
frictionless surfaces. In Figure (7c) we have sketched the result that the line of action of mg must also pass
the forces acting on the rod. There is the downward through the intersection point as shown in Figure (7d).
force of gravity mg that acts at the center of mass of the Thus the rod will come to rest when the center of mass
rod, the force Fb exerted by the bowl on the end of the of the rod lies directly below the intersection point of
rod, and the force Fr exerted by the rim. Fb and Fr . This result is nicely demonstrated by
comparing the prediction, Figure (7d), with the experi-
The idealization that we have a frictionless surface is ment, Figure (7b).
equivalent to the statement that the surface can only
exert normal forces, forces perpendicular to the sur-
face. Thus the force Fb exerted by the frictionless

Fr Fr

Fb Fb

mg mg
Figure 7c Figure 7d
Forces acting on the rod. Because the bowl is For equilibrium, the center of mass must lie directly
frictionless, Fb is perpendicular to the surface of below the intersection point of Fb and Fr .
the bowl. Because the rod is frictionless Fr is
perpendicular to the rod.
13-8 Equilibrium

Exercise 5 Exercise 6 Ladder Problem

A spherical ball of mass m, radius r, is suspended by a A ladder is leaning against a frictionless wall at an angle
string of length attached to a frictionless wall as shown θ as shown in Figure (9). Assume that the ladder is
in Figure (8). massless, and that a person of mass m is on the ladder.

(a) show that the line of action of the tension force T (the The force between the ground and the bottom of the
line of the string) passes through the center of the ball. ladder can be decomposed into a normal component
Fn , and a horizontal component Ff that can exist only
(b) find the tension T. if there is friction between the ladder and the ground.

It is traditional in introductory texts to say that the ladder

will start to slip if the friction force Ff exceeds a value of
µFn where µ is called the “coefficient of static friction”.
This idea is reasonable in that as the normal force Fn
increases, so does the gripping or friction force Ff .
However the coefficient µ depends so much upon the
circumstances of the particular situation, that the theory
T is not particularly useful. What, for example, should you
use for the value of µ if the ends of the ladder sink down
into the ground?

However for the sake of this problem, assume that the

ladder will just start to slip when Ff = µFn . Assume that
r
µ has the value µ = 1/ 3 = .557 .
(a) at what angle θ would you place the ladder so that it
will not start to slip until the person climbing it just
reaches the top?
Figure 8
Ball suspended from a frictionless wall. (b) at what angle θ would you place the ladder so that
it will not start to slip until the person has gone half way
up?

frictionless surface

θ
person of
mass m
massless ladder

mg
Fn

Ff = µF
Fn
Figure 9
Ladder leaning against a frictionless wall.
 13-9

Example 5 A Bridge Problem In Equation 13 we have two unknowns, F2 and Td Thus

A bridge is constructed from massless rigid beams of length we need another equation. If we take the bridge as a whole,
. The ends of the beams are connected by a single large bolt and calculate the torques on about the point a (to eliminate
that acts more or less like a big hinge. As a result the only the force F1 ), we get
forces you can have in each beam is either tension or 5
compression (i.e. each beam either pulls or pushes along the ΣT about a = F2 3 – mg = 0 (14)
length of the beam.) The idea is to be able to calculate the 2
tension or compression force in any of the beams when a Solving Equation 14 for F2 gives
load is placed on the bridge.
5
F2 = mg (15)
In this example, we will place a mass m in the center of the 6
right most span as shown in Figure (10a). To illustrate the Using this value in Equation 13 gives
process of calculating tension or compression in the beams,
we will calculate the force in the upper left hand beam. For – mg
now we will assume that the beam is under tension and Td = (18)
3 3
exerts a force Td on joint d as shown. If it turns out that the
beam is under compression, then the magnitude of Td will
turn out to be negative. Thus we do not have to know ahead The minus (–) sign indicates that the beam is under com-
of time whether the beam is under tension or compression. pression.

Solution: When you have a statics problem involving an Exercise 7

object with a lot of pieces, and you want to calculate the
force in one of the pieces, the first step is to isolate part of the Find the tension (or compression in the beam that goes
object and consider that as a separate system with external from point (d) to point (e) in the bridge problem of Figure
forces acting on it. In Figure (10b) we have chosen as our (10a).
isolated system the part of the bridge made from the girders
that have been drawn in heavy lines. The external forces Td
acting on this isolated system are the gravitational force mg b d f
acting on the mass m, the supporting force F2 that holds up
the right end of the bridge, (we will assume that the ends of
the bridge are free to slide back and forth, so that the
a m g
supporting forces F1 and F2 point straight up). In addition
the tension (or compression) forces we have labeled Td ,
c e
Tc1 and Tc2 are also acting on our isolated section of the
bridge. mg
Looking at Figure (10b), it is immediately clear that the
Figure 10a
forces we do not want to know anything about are the
Bridge with a truck on the last span.
tension forces Tc1 and Tc2 . We can eliminate these forces
by setting to zero the torques about the joint labeled c. Using
our convention that counter clockwise torques are positive
and clockwise ones are negative, we get b Td d f

3 3 3
ΣTabout c = F2 2 – mg + Td = 0 h= 2
2 2 a Tc2 g
m
(13) Tc1 c e 2 2
where we used the fact that Td's lever arm h is the altitude F1
of an equilateral triangle. F2
mg

Figure 10b
Finding the tension in the span from b to d.
13-10 Equilibrium

Exercise 8 Working with Rope

As most sailors know, if you use rope correctly, you can
create very large tension forces without exerting that
strong a force yourself. Suppose, for example, you wish
to make a raft out of two long logs with two short spacer
planks between them as shown in Figure (11a). You
wish to hold the raft together with a rope around the
center as shown.

The first step is to tie, as tightly as you can, the logs

together as shown in the end view of Figure (11b). Then
take another piece of rope and tie it as tightly as you can Figure 11a
as shown in Figure (11c). If you do a reasonably good Constructing a raft by tying two logs together, with
job, you can create a large tension in the rope holding wood spacers.
the logs together.

To analyze the problem, let T1 be the tension in the rope

holding the logs together, and T2 the tension in the line
between the ropes as shown in Figure (11d). For this
problem, assume that angle θ in Figure (11d) is 5
degrees, and the tension T2 that you could supply in
winding the line around the ropes was 200 newtons
(enough force to lift a 20 kilogram mass). What is the
tension T1 in the ropes holding the logs together?
Figure 11b
End view of raft.

Figure 11c
Tightening the rope.

θ
T1 T1

T2
Figure 11d
Tensions in the ropes.
 13-11

LIFTING WEIGHTS We see that the back muscle has to pull down with a
AND MUSCLE INJURIES force that is a factor of L / times greater than the
The previous exercise on tying a raft together illustrates weight mg of the shopping bag. With = 2 cm , you
the fact that with some leverage, you can create large see that if you hold the shopping bag out at arms length,
tensions in a rope. Similar large forces can exist in your say L = 80 cm, then Fm is L / = 40 times as great as
muscles when you lift weights, particularly if you do the weight of the shopping bag. For you to lift a 10 kg
not lift the weights properly. bag at arms length, your back is essentially lifting a 10
kg × 40 = 400 kg mass, which has a weight of almost
To illustrate the importance of lifting heavy objects half a ton! If instead you pulled your arm in so that L
correctly, consider the sketch of Figure (12) showing a was only 20 cm, then Fm drops 1/4 of its original value.
shopper holding a funny looking 10 kg shopping bag Do your back a favor and do not lift heavy objects out
out at arms length. We wish to determine the forces that at arms length.
must be exerted on the backbone and by the back
muscles in order to support this extra weight. Exercise 9

To analyze the forces, think of the upper body and arm Write a single equation that allows you to solve for the
compressional force Fb exerted on the backbone, as
as essentially a rigid object supported by the backbone
shown in Figure (12).
and back muscles as shown. Since we are interested in
the extra forces required to lift the weight, we will = 2 cm
ignore the weight of the upper body itself. L = 80 cm

The external forces acting on the upper body are the

upward compressional force Fb acting on the
backbone, and the downward force Fm acting at the
point where the back muscle is attached to the thigh-
bone. (This is in reaction to the upward force exerted
back
on the thigh bone by the contacting back muscle.) muscle
There is also the weight mg of the shopping bag. We funny
looking
are letting L be the horizontal distance from the back- shopping b
bone to the shopping bag, and the separation between bag
the point where the thigh bone pushes up on the mg
Fb
backbone and pulls down on the back muscle.

If we want to solve for the force Fm exerted by the back Fm

muscle we can eliminate Fb from our equation by
setting to zero the sum of the torques about point b, the Figure 12
Forces on the backbone .
point where the thigh bone contacts the backbone. We
get
Σ τ about b = F m – MgL = 0

F m = mg L (9)
13-12 Equilibrium

Exercise 10 calf muscle

Figure (13) shows the structure of the lower leg and foot.
Assume that this person is raising her heel a bit off the
ground, so that her foot is touching the floor at only one
place, namely the point labeled P in the diagram. Also
assume that the calf muscle is attached to the foot leg bone
bones at the point labeled (a), and that the leg bone acts
at the pivot point (b). If her mass is 60 kilograms, what
must be the forces exerted by the calf muscle and the
leg bone? Compare the strength of these forces with her
weight. (This and the next problem adapted from Halliday
& Resnick.)
b
Exercise 11 a
The arm in Figure (14) is holding a 20 kilogram mass.
The arm pivots around the points marked with a small P
black circle. What is the compressional force on the 5 15 cm
cm
bones in this joint? (Neglect the mass of the arm.)
Figure 13
Foot

muscle

m
pivot
point

30 cm
3.5 cm

Figure 14
Arm
Chapter 14
Oscillations
and Resonance
CHAPTER 14 OSCILLATIONS AND
RESONANCE

Oscillations and vibrations play a more significant

role in our lives than we realize. When you strike a bell,
the metal vibrates, creating a sound wave. All musical
instruments are based on some method to force the air
around the instrument to oscillate. Oscillations from
the swing of a pendulum in a grandfather’s clock to the
vibrations of a quartz crystal are used as timing
devices. When you heat a substance, some of the
energy you supply goes into oscillations of the atoms.
Most forms of wave motion involve the oscillatory
motion of the substance through which the wave is
moving.

Despite the enormous variety of systems that oscillate,

they have many features in common, features exhibited
by the simple system of a mass on a spring. As a result
we will focus our attention on the analysis of the motion
of a mass on a spring, describing ways in which other
forms of oscillation are similar.
14-2 Oscillations and Resonance

OSCILLATORY MOTION Now pull the mass down so that the white stripe lines
Suspend a mass on the end of a spring as shown in up with the lowest position of the stick the same height
Figure (1), gently pull the mass down and let go. The as the stick when the stick is at the bottom position (6
mass will oscillate up and down about the equilibrium o’clock). Start the motor rotating at the correct fre-
position. How do we describe the kind of motion we quency and release the mass when the stick is at the
are looking at? bottom. If you do this just right, some practice may be
required, you will see in the television picture that the
One of the best ways to see what kind of motion we are white stripe and the stick move up and down together
dealing with is to perform the demonstration illustrated as if they were a single object.
in Figure (2a). In that demonstration, we place a
rotating wheel beside an oscillating mass, and view the From this demonstration we conclude that the up and
two objects via a TV camera set off to the side as shown. down oscillatory motion of the mass is the same as
A white tape is placed around the mass, and a short stick circular motion viewed sideways. As a result we can
is mounted on the rotating wheel as seen in the edge use what we know about circular motion to understand
view, Figure (2b). This edge view is the one displayed oscillatory motion. As a start, we will say that the
by the TV camera. oscillatory motion has an angular frequency ω that is
the same as the angular velocity ω of the rotating wheel
The wheel is mounted on a variable speed motor, which when the mass and the stick go up and down together.
allows us to adjust the angular velocity of the ω of the
wheel so that the wheel goes around once in precisely
the same length of time it takes the mass to bob up and ω
down once. The height of the wheel is adjusted so that
when the mass is at rest at its equilibrium position, the
white stripe on the mass lines up with the axis of the
wheel. As a result if the stick is in a horizontal position
(3 o’clock or 9 o’clock) and the mass is at rest, the stick
and the white stripe line up on the television image. TV Camera mass on rotating
a spring wheel
Figure 2a
Lecture setup for comparing the oscillation of a
mass on a spring with a rotating wheel. A stick
is mounted on the rotating wheel, and a TV
camera off to the side provides a side view of
the oscillating mass and rotating wheel.

rotating
m wheel

Figure 1
Mass suspended from a spring. If you gently pull
the mass down and release it, it will oscillate up white
and down about the equilibrium position. tape motor

stick
Figure 2b
Side view of the oscillating mass and rotating wheel,
as seen by the TV camera. When the motor is
adjusted to the correct frequency, the mass and the
stick are observed to move up and down together.
 14-3

THE SINE WAVE the distance x = vt = ωt that the truck has travelled is
There is another way to picture the sideways view of the same as the angular distance θ = ωt that the stick
circular motion. As more or less a thought experiment, has travelled. Finally let the radius of the circle around
suppose that we take our rotating wheel with a stick which the stick is travelling be r = 1, so that the
shown in Figure (2a) and shine a parallel beam of light undulating curve goes up to a maximum value y = + 1
at it, sideways, as shown in Figure (3). Now picture a and down to a minimum value y = – 1. With these
truck with a big billboard mounted on the back, driving conditions, the curve traced out is the mathematical
away from the light source as shown on the right side function
of Figure (3). The image of the stick will move up and y = sin θ = sin (ωt) (1)
down on the billboard as the truck moves forward.
Let us remove the truck and billboard and look at the
Finally picture the line traced out in space by the image sine curve itself more carefully as shown in Figure (4).
of the stick on the moving billboard. The image is The horizontal axis of the sine curve is the angular
going up and down with a frequency ω and moves distance θ = ωt that the rotating stick has travelled.
forward at a speed v, the speed of the truck. The result Starting at 0 when θ = 0 , the sine curve completes one
is an undulating curve we call a sine wave. full cycle or undulation just when the wheel has gone
around once and θ = 2π . Thus one cycle of a sine wave
To make this definition of the sine wave more specific, goes from 0 to 2π as shown in Figure (4). The sine
assume the truck crosses the point x = 0 just when the wave reaches a maximum height at θ = π/2 , goes back
angle θ of the stick is zero as shown in Figure (3). Let to zero again at θ = π , has a minimum value at 3π/2
us imagine that the truck drives at a speed v = ω , so that and completes the cycle at θ = ωt = 2π .

rotating
image of stick
π/2 wheel
stick on billboard
1
light source

y θ
π 2π 3π
π 0 0 π 3π 5π
θ = ωt
2 2 2

ω −1
3π/2
billboard v= ω
on truck
Figure 3
Project the image of the stick onto the back of a truck moving
at a speed v = ω t, and the image traces out a sine wave.

π/2
1

θ 3π
π 2
π 0
0 π 2π
θ = ωt
2

ω −1
3π/2

Figure 4
The sine curve y = sin θ = sin ω t .
14-4 Oscillations and Resonance

Exercise 1 The basic question for the dynamic picture is how long
does one oscillation take. The time for one oscillation
Somewhere back in the dim past, you learned that sin θ
is called the period T of the oscillation. We can
was the ratio of the opposite side to the hypotenuse in
therefore ask what is the period T of an oscillation
a right triangle. Applied to Figure (5), this is
whose angular velocity, or angular frequency is ω.
sin θ ≡ y/ r (2)
The solution to this problem is to note that the sine wave
Show that this older definition of sin θ , at least for angles completes one cycle when θ = 2π . But θ is just the
θ between 0 and π/ 2 , is the same as the definition of angular distance ωt . Thus, if t = one period T, we have
sin θ we are using in Figure (3) and (4).
θ = ωT = 2π
As you can see from Exercise 1, our rotating stick 2π period of
picture of the sine wave is mathematically equivalent T = a sinusoidal (3)
ω
to the definition of sin θ you learned in your first oscillation
trigonometry class. What may be new conceptually is To remember formulas like Equation 3, we can use the
the dynamic aspect of the definition. Figures (3) and (4) same set of dimensions we used in our discussion of
connect rotational motion to oscillatory motion and to angular motion. If we remember the dimensions
the shape of a sine wave. The relationship between the
static picture and the dynamic one is that the angular
distance θ is equal to the angular velocity ω times the ω radians
second
angular frequency
elapsed time t.
2π radians
rotating π/2 cycle
wheel stick
T seconds
cycle
period
r
y
π θ
0 cycles
f frequency
second
ω
3π/2 then we can go back and forth between the quantities ω,
T and f simply by making the dimensions come out
sin θ y/r right.
r y

θ
x
Figure 5
Our old definition of sin θ .
 14-5

For example Exercise 3

2π radians As shown in Figure (6), an air cart sitting on an air track
sec cycle
T = = 2π sec
ω cycle
has springs attached to the ends of the track as shown.
cycle radians
ω sec We are taking x = 0 to be the equilibrium position of the
cart. It turns out that the car oscillates back and forth
with the same kind of sinusoidal motion as the mass on
ω radians the end of a spring shown in Figures (1) and (2).
= ω sec
cycles sec cycles
f sec =
2π radians 2π Assume that the mathematical formula for the coordi-
cycle
nate x of the cart is

T sec = 1 = 1 sec x = x 0 sin ωt

cycle cycles f cycle
f sec
where
We repeated these dimensional exercises, because it is radians
x 0 = 4 cm ; ω =3
essential that you be able to easily go back and forth sec
between quantities like frequency, angular frequency, (a) Figure (7) is an x–t graph of the position of the air cart
and period. in Figure (6). We have drawn in the sine curve, and
drawn tick marks at important points along the x and t
axis. On the x axis, the tick marks are at the maximum
Exercise 2
and minimum values of x. On the t axis, the tick marks
(a) A spring is vibrating at a rate of 2 seconds per cycle. are at 1/4, 1/2, 3/4 and 1 complete cycle. Insert on the
What is the angular velocity ω of this oscillation? graph, the numerical values that should be associated
with these tick marks.
(b) What is the period of oscillation, in seconds, of an
oscillation where ω = 1 ? (b) Where will the cart be located at the time
t = 300 π seconds ?

x = 0 equilibrium
t
ω

x Figure 7
The x-t graph for the motion
Figure 6
of the air cart in Figure 6.
Mass with springs on an air cart. We take
the equilibrium position to be x = 0.
14-6 Oscillations and Resonance

Phase of an Oscillation In Figure (9) we have sketched the sine wave for
Our sine waves, by definition, begin at 0 when θ = 0. several different phases. At a phase φ = π/2 , the wave
This is equivalent to saying that in Figure (2), the truck starts at 1 for θ = 0 and goes down to 0 at
crossed x = 0 at the same instant that the rotating wheel θ = π /2 = 90°. This is just what cos θ does, and we
crossed θ = 0 . We did not have to make this choice, the have what is called a cosine wave. We have
rotating wheel could have been at any angle φ when the cos ωt = sin ωt + π/2 (6)
truck crossed t = 0 as shown in Figure (8). If the stick
were up at an angle φ when the truck crossed the zero When the phase angle gets up to π or 180°, the sine
of the horizontal axis, we say that the resulting sine wave is reversed into a – sin ωt wave. At φ = 2π we
wave has a phase φ. The formula for the resulting sine are back to the sine wave we started with.
curve is
φ is the sine wave
y = sin ωt + φ (4)
phaseangle
0 π 2π
φ=0
You can see from this equation that at t = 0, the angle
of the sine wave is θ = ωt + φ = φ
ω

cosine wave
π 2π
φ π 2π
t φ= π
2
}

φ
phase
angle
Figure 8 t=0
The phase angle φ . –sine wave
2π
φ=π π

– cosine wave
π 2π

φ= 3π
2

sine wave
π 2π

φ = 2π

Figure 9
Various phases of the sine wave. When φ = π / 2, the
wave is called a cosine wave. (It matches the
definition of a cosine, which starts out at 1 for θ = 0.
At a phase angle φ = π or 180°, the sine wave has
reversed and become – sin (ω t). At φ = 2π , we are back
to a sine wave again.
 14-7

Exercise 4 MASS ON A SPRING;

In trigonometry class, or somewhere perhaps, you were ANALYTIC SOLUTION
given the trigonometric identity Let us now apply Newton’s laws to the motion of a
mass on a spring and see how well the results compare
sin a + b = sin a cos b + sin b cos a
with the sinusoidal motion we observed in the demon-
Use this result to show that stration of Figure (2). In our analysis of the spring
pendulum in Chapter 9 (see Figure 9-4) we saw that the
sin ω t + π/2 = cos ω t (7a) spring exerted a force whose strength was linearly
sin ω t + π = – sin ω t
proportional to the amount the spring was stretched, a
(7b)
result known as Hooke’s law.
sin ω t + 3 π/2 = – cos ω t (7c)
When we have the one dimensional motion of a mass
sin ω t + 2π = sin ω t (7d) oscillating about its equilibrium position, as illustrated
in Figure (10), then we get a very simple formula for the
These are the results graphed in Figure (9). net force on the object. If we displace the cart of Figure
(10) by a distance x from equilibrium, there is a
F restoring force whose magnitude is proportional to x
pushing the cart back toward the equilibrium position.

We can completely describe this restoring force by the

x formula
Hooke's law
F = – kx (8)
F (restoring force) restoring force

If x = 0, the cart is at its equilibrium position and there

is no force. If x is positive, as in Figure (10), the
x restoring force is negative, pointing back to the equilib-
x
(distance rium position. And if x is negative, the restoring force
stretched) points in the positive direction. All these cases are
–kx handled by the formula F = –kx.

Figure 10
Fs = –kx For a mass bobbing up and down on a spring, shown in
If the cart is displaced a distance x from Figure (11), there is both a gravitational and a spring
equilibrium, there is a restoring force F = - kx. force acting on the mass. But if you measure the net
force starting from the equilibrium position, you still
get a linear restoring force. The net force is always
directed back toward the equilibrium position and has
a strength proportional to the distance x the mass is
away from equilibrium. Thus Equation 8 still describes
F x the net force on the mass.
equilibrium
position
Figure 11
Restoring force for a mass on a spring.
14-8 Oscillations and Resonance

Exercise 5 When working with differential equations, there is a

Describe experiments you could carry out in the labo-
traditional form in which to write the equation. The
ratory to measure the force constant k for the air cart highest derivative is written to the far left, all terms with
setup of Figure (10). To do this you are given the air cart the unknown variable are put on the left side of the
setup, a string, a pulley, and some small weights. equation, and the coefficient of the highest derivative is
set to one. (A reason for this tradition is that only a few
Sketch the setup you would use to make the measure- differential equations have been solved. If you write
ments, and include some simulated data to show how them all in a standard form, you may recognize the one
you would obtain a numerical value of k from your data. you are working with.) Putting Equation 11a in the
(This is the kind of exercise you would do ahead of time standard form by dividing through by m and rearrang-
if you planned to do a project studying the oscillatory
ing terms gives
motion of the air cart and spring system.
2
dx k
To apply Newton’s laws to the problem of the oscillat- 2
+ x = 0 (11)
ing mass, let x(t) be the displacement from equilibrium dt m
of either the air cart of Figure (10) or the mass of Figure
A standard way to solve a differential equation is to
(11). The velocity vx and the acceleration a x of either
guess the answer, and then plug your guess into the
the cart or mass is
equation to see if the guess works. A course in differ-
dx t ential equations basically teaches you how to make
vx ≡ (9)
dt good guesses. In the absence of such a course, we have
d 2x t to use whatever knowledge we have about the system
dvx
ax ≡ = (10) in order to make as good a guess as we can. That is why
dt dt 2 we did the demonstration of Figure (2). In that demon-
The x component of Newton’s second law becomes stration we saw the oscillating mass moved the same
Fx = ma x way as a stick on a rotating wheel, when the wheel is
viewed sideways. We then saw that this sideways view
d2 x of rotating motion is described by the mathematical
– kx = m (11a)
dt2 function sin ωt . From this we suspect that a good
guess for the function x(t) may be
where we used Hooke’s law, Equation 8 for Fx .
x t = sin ωt initial guess (12)
The result, Equation 11a, involves both the variable
x(t) and its second derivative d2x/dt2. An equation In order to see if this guess is any good, we need to
involving derivatives is called a differential equation, substitute values of x and d 2 x/dt2 into Equation 11.
and one like Equation 11a, where the highest derivative To do this, we need derivatives of x = sin ωt .
is the second derivative, is called a second order
differential equation. Differential equations are harder From your calculus course you learned that
to solve than algebraic equations like x2 = 4, because d
the answer is a function or a curve, rather than simple sin ωt = ω cos ωt (13)
dt
numbers like ± 2.
d
cos ωt = – ω sin ωt (14)
dt
 14-9

Thus if we start with Equation 18 is easily solved with the choice

x = sin ω t (12) k angular frequency
ω = (19)
we have m of oscillating cart
dx = ω cos ωt
dt
(15) Thus we have shown not only that ( x = sin ωt ) is a
solution of Newton’s second law, but we have also
d 2 x = d dx = d ω cos ωt solved for the frequency of oscillation ω. Newton’s
dt 2 dt dt dt second law predicts that the air cart will oscillate at a
frequency ω = k/m . This result is easily tested by
= ω – ω sin ωt experiment.
d 2 x = – ω 2 sin ωt (16)
dt 2 Exercise 6
(a) In the formula ( x = sin ω t ), ω is the angular fre-
To check our guess x = sin ω t , we substitute the quency of oscillation, measured in radians per second.
values of x from Equation 12 and d2 x/dt2 from Equa- Using the formula ω = k/m and dimensional analysis,
tion 16 into the differential Equation 11. We get find the predicted formula for the period T of the
oscillation, the number of seconds per cycle.
d2 x kx
+m = 0 (b) A mass m = 245 gms is suspended from a spring as
dt 2
shown in Figure (12). The mass is observed to oscillate
– ω 2 sin ωt k sin ω t
+m = 0 (17) up and down with a period of 1.37 seconds. From this
determine the spring constant k. How could this result
have been used in the spring pendulum experiment
We put the question mark over the equal sign, because
? x = sin ωt discussed in Chapter 9 (page 9-3)?
the question we want to answer is whether
can be a solution to our differential equation. Can the (You can check your answer, since the ball and spring
sum of these two terms be made equal to zero as of Figure (12) are the same ones we used in the spring
required by Equation 11? pendulum experiment.)

The first thing we note in Equation 17 is that the

function ( sin ωt ) cancels. This is encouraging, be-
cause if we ended up with two different functions of
time, say a sin ωt in one term and a cos ωt in the other,
there would be no way to make the sum of the two terms
to be zero for all time, and we would not have solved the
equation. However because the ( sin ωt ) cancels, we
are left with m
2 k ?
–ω + = 0 (18) Figure 12
m The spring constant can be determined by
measuring the period of oscillation.
14-10 Oscillations and Resonance

The guess we made in Equation 12, ( x = sin ωt ) is not Exercise 8

the only possible solution to our differential Equation (a) Show that the guess
11. In the following exercises, you show that
x = A cos (ω t) (12b)
x = A sin (ωt) (12a)
is also a solution, where A is an arbitrary constant. is also a solution to our differential Equation 11. This
should be an expected result, because the only differ-
Since the function sin (ωt) oscillates back and forth
ence between a sine wave and a cosines wave is the
between the values + 1 and – 1, the function A sin (ωt)
choice of the time t = 0 when we start measuring the
oscillates back and forth between + A and – A. Thus oscillation.
A represents the amplitude of the oscillation. The fact
that Equation 12a, with arbitrary A, is also a solution to (b) The sine and cosine waves are only special cases
Newton’s second law, means that a sine wave with any of the more general solution
amplitude is a solution. (This is true as long as you do x = A sin (ωt + φ) (12c)
not stretch the spring too much. If you pull a spring out
too far, if you exceed what is called the elastic limit, the where φ is the phase of the oscillation discussed in
spring does not return to its original shape and its spring Figure (9). Show that Equation 12c is also a solution of
constant changes.) our differential Equation 11. [Hint: the derivative of
sin (ωt + φ) is – ω cos (ωt + φ) . You can, if you want,
prove this result using Equations 13 and 14 and the
Exercise 7
trigonometric identities
As a guess, try Equation 12a as a solution to the
differential Equation 11. Follow the same kind of steps sin (a + b) = sin a cos b + cos a sin b
we used in checking the guess x = sin (ωt ) , and see cos (a + b) = cos a cos b – sin a sin b
why Equation 12a is a solution for any value of A.
Remember that φ is a constant.]

1 Exercise 9
We do not want you to think every function is a solution
to Equation 11. Try as a guess
x = e – αt (20)
– αt
e which represents an exponentially decaying curve
shown in Figure (13). To do this you need to know that

d – αt
e = – α e – αt (21)
dt
When you try Equation 20 as a guess, what goes wrong?
Figure 13 Why can’t this be a solution to our differential equation?
The exponential decay curve e – α t . [Or, by what crazy way could you make it a solution?]
 14-11

Conservation of Energy The total energy E tot of the cart at any time t is
Back in Chapter 10 we calculated the formula for the total kinetic potential
potential energy stored in the springs when we pulled energy = energy + energy
the cart of Figure (10) a distance x from equilibrium. (25)
E tot = 1 mω 2 cos2 ωt + 1 k sin 2 ωt
The result was 2 2
spring At first sight, Equation 25 does not look too promising.
potential U spring = 1 k x 2 (10-28) It seems that E tot is some rather complex function of
energy 2
time, hardly what we expect if energy is conserved.
We then used the law of conservation of energy to However remember that the frequency ω is related to
predict how fast the cart would be moving when it the spring constant k by ω = k/m , thus we have
crossed the x = 0 equilibrium line if it were released
2
from rest at a position x = x 0 . The idea was that the 1 mω 2 = 1 m k
potential energy 12 k x 02 the springs have when the cart 2 2 m
is released, is converted to kinetic energy 12 m v02 the
= 1mm k = 1k (26)
cart has when it is at x = 0 and its speed is v = v0 . 2 2
Thus the two terms in our formula for E tot have the
Exercise 10 same coefficient 12 k , and E tot becomes (using Equa-
See if you can derive Equation 10-28 without looking tion 26 in 25)
back at Chapter 10. If you cannot, review the derivation
now. E tot = 1 k cos2 ωt + sin 2 ωt (27)
2
Equation 27 can be simplified further using the trigo-
Using conservation of energy to predict the speed of the
nometric identity
air cart was particularly useful back in Chapter 10
because at that time we did not have the analytic cos 2 a + sin 2 a = 1 (28)
solution for the motion of the cart. Now that we have
solved Newton’s second law to predict the motion of for any value of a. Thus the term in square brackets in
the cart, we can turn the problem around, and see if Equation 27 has the value 1, and we are left with
energy is conserved by the analytic solution. k
E tot = (29)
An analytic solution for the position x(t) and velocity 2
v(t) of the cart is The total energy of the mass and spring system is
x(t) = sin ωt constant as the oscillator moves back and forth. Energy
is conserved after all !
dx (22)
v(t) = = ω cos ωt
dt
Exercise 11
For this solution, the kinetic and potential energies are What is the total energy of an oscillating mass whose
kinetic 1 mv 2 = 1 mω 2 cos2 ωt amplitude of oscillation is A? [Start with the solution
energy (23)
2 2 x(t) = A sin ω t , calculate v(t), and then calculate

potential 1 kx 2 = 1 k sin 2 ωt E tot = 1 kx2 + 1 mv2 .

energy (24) 2 2
2 2
14-12 Oscillations and Resonance

THE HARMONIC OSCILLATOR The Torsion Pendulum

The sinusoidal motion we have been discussing, which One example of simple harmonic motion is provided
results when an object is subject to a linear restoring by part of the apparatus used by Cavendish to detect the
force F = – kx, is called simple harmonic motion and gravitational force between two lead balls. The appa-
the oscillating system is often called a harmonic ratus, illustrated back in Figure (8-8), contains two
oscillator. These general names are used because there small lead balls mounted on a light rod, which in turn
are many examples in physics of simple harmonic is suspended from a glass fiber as shown in Figure
motion. In some cases the sine wave solution sin ωt (14a). (Such glass fibers are easy to make. Heat the
is an exact solution of a Hooke’s law problem. In many center of a glass rod in a Bunsen burner until the glass
other cases, the solution is approximate, valid only for is about to melt, and then pull the ends of the rod apart.
small amplitude oscillations where the displacements The soft glass stretches out into a long thin fiber.)
x do not become too big. In the following sections we
If you let the rod with two balls come to rest at its
will consider examples of both kinds of problems.
equilibrium position, then rotate then rod by an angle θ
in the horizontal plane as shown in the top view of
Figure (14b), the glass fiber exerts a torque tending to
rotate the rod back to its equilibrium position. Careful
experiments have shown that the restoring torque
exerted by the glass fiber is proportional to the angular
distance θ that the rod has been rotated from equilib-
glass fiber

rium. The rod is acting like an angular spring, produc-

ing a restoring torque τr given by an angular version of
Hooke’s law
angular version
τ r = – kθ of Hooke's law (30)
m m
In the Cavendish experiment two large balls of mass M
Figure 14a are placed near the small balls as shown in Figure (15).
Side view of the torsion pendulum
used in the Cavendish experiment.
M
Fg
r m
θ θ

equilibrium m
position
Figure 14b Fg
Top view of the torsion pendulum. The light drawing M
shows the equilibrium position of the pendulum, the
dark drawing shows the pendulum displaced by an Figure 15
angle θ. In this displaced position, the glass fiber In the Cavendish experiment, a torque is exerted
exerts a restoring torque τrestoring = – k θ . on the torsion pendulum by the gravitational force
of the large lead spheres. In the new equilibrium
position, the gravitational torque just balances the
restoring torque of the torsion pendulum.
 14-13

The gravitational forces Fg between the large and directly measure the value of k. To determine k by
small balls produce a net torque τg on the suspended direct measurement would mean applying known forces
rod of magnitude of magnitude Fg , but the only forces around that are
τg = 2 rFg (31) sufficiently weak are the gravitational forces you are
trying to measure.
where r is the distance from the center of the rod to the
small mass (the lever arm), and the factor of 2 is from Fortunately there is an easy way to obtain an accurate
the fact that we have a gravitational force acting on each value of the restoring constant k. Remove the large lead
pair of the balls. balls, displace the rod from equilibrium by some
reasonable angle θ as shown in Figure (14b), and let go.
The suspended rod will finally come to rest at an angle You will observe the rod to swing back and forth in an
θ0 where the gravitational torque τg just balances the oscillatory motion. The rod, two balls, and glass fiber
glass fiber restoring torque τr , so that there is no net of Figure (14a) form what is called a torsion pendu-
torque the rod. Equating the magnitudes of τr and τg , lum, and the oscillation is caused by the restoring
Equations 30 and 31 gives us torque of the glass fiber. The glass fiber is acting like
τg = τ r an angular spring, creating an angular harmonic mo-
tion in strict analogy to the linear harmonic motion of
2r F g = kθ 0 (32) a mass suspended from a spring.
Equation 32 can then be solved for the gravitational The analogy applies directly to the equations of motion
force τg in terms of the rod length 2r, the restoring of the two systems. For a linear one dimensional
constant k, and the rest angle θ0 . system like a mass on a spring, Newton’s second law
is
The problem the Cavendish experiment has to over- d2 x
come is the fact that the gravitational force between the Fx = max = m 2
two lead balls is extremely weak. You need an appa- dt
ratus where the tiny gravitational torque τg produces The angular version of Newton’s second law, applied
an observable deflection θ0 . That means that the to the simple case of an object rotating about a fixed
restoring torque τr must also be very small. That was axis, is from Equation 30 of Chapter 12
why the long glass fiber was used to suspend the rod,
for it produces an almost immeasurably small restoring d2 θ
τ = Iα = I (12-20)
torque. dt2
In order to carry out the experiment and measure the where τ is the net torque, I the angular mass or moment
gravitational force Fg , you need to know the restoring of inertia, and α the angular acceleration of the object
torque constant k that appears in Equation 32. But the about its axis of rotation.
feature of the glass fiber that makes it good for the
experiment, the small value of k, makes it hard to
14-14 Oscillations and Resonance

For the linear harmonic oscillator (mass on a spring), As a result, by observing the period of oscillation of the
the force Fx is a linear restoring force Fx = – kx, rod and two balls (with the big masses M removed),
which gives rise to the equation of motion and differ- you can determine the restoring constant k of the glass
ential equation fiber, and use that result in Equation 32 to solve for the
2 gravitational force Fg . Because you can measure
Fx = – kx = m d x2 (11a) periods accurately by timing many swings, k can be
dt
measured accurately, and the Cavendish experiment
d2 x k allows you to do a reasonably good job of measuring
+ x = 0 (11)
dt2 m the gravitational force Fg .
For our torsion pendulum, the restoring torque is
Exercise 12
τr = – k θ which gives rise to the equation of motion
Solve the differential Equation 33 by starting with the
and differential equation
guess
τr = – kθ = I d θ2
2
(33a) θ(t) = A sin (ω0t) (35)
dt
Check that Equation 35 is in fact a solution of Equation
d2 θ k 33, and find the formula for the frequency ω0 of the
+ θ = 0 (33)
dt2 I oscillation. Also use dimensional analysis find the
period of oscillation.
Equations 11 and 33 are the same if we substitute the
angular distance θ for the linear distance x, and the Exercise 13
angular mass I for the linear mass m. For the linear In the commercial Cavendish experiment apparatus
motion, we saw that the spring oscillated back and forth shown in Figure 8-8, the small balls each have a mass
at an oscillation frequency ω0 and period T given by of 170 gms, the distance between the small balls is 12
cm, and the observed period of oscillation is 24 minutes.
ω0 = k
m (a) Calculate the value of the restoring constant k of the
glass fiber.
T = 2π = 2π m (19)
ω0 k
(b) How big a torque, measured in dyne centimeters, is
By strict analogy, we expect the torsion pendulum to required to rotate the glass fiber by an angle of one
oscillate with a frequency ω0 and period T given by degree. (Remember to convert degrees to radians.)

ω0 = k
I

T = 2π = 2π I (34)
ω0 k
where I is the moment of inertia of the rod and two balls
about the axis defined by the glass fiber (as shown in
Figure 14).
 14-15

The Simple Pendulum times the tangential acceleration. The tangential com-
Perhaps the most well-known example of oscillatory ponent of the gravitational force is always directed
motion is the simple pendulum which consists of a toward the bottom equilibrium position, thus it is a
mass swinging back and forth on the end of a string or restoring force of the form
rod. The regular swings of this pendulum serve as the Ftangential = – mg sin θ (36)
basic timing device of the grandfather’s clock.
As the mass moves along the arc, the speed of the ball
When we begin to analyze the simple pendulum, we is related to the angle θ by
will find that it is not quite so simple after all. The
restoring force is not strictly a linear restoring force and vtangential = dθ (37)
dt
we end up with a differential equation whose solution
is more complex than the sinusoidal oscillations we a result from the beginning of our discussion of circular
have been discussing. What allows us to include this motion in Chapter 12. (See the discussion before
example in our discussion of simple harmonic motion Equation 12-11.) Differentiating Equation 37, we get
is the fact that, for small amplitude oscillations, the for the tangential acceleration
dvtangential
= d θ
restoring force is approximately linear, and the result- 2
a tangential = (38)
ing motion is approximately sinusoidal. dt dt 2
Figure (16) is a sketch of a simple pendulum consisting Thus Newton’s second law gives
of a small mass m swinging on the end of a string of Ftangential = ma tangential
length . The downward gravitational force mg has a
– mg sinθ = m d θ2
2
component of magnitude mg sinθ directed along the (39)
circular path of the ball. dt
Dividing through by m and rearranging terms gives
Since the ball is constrained to move along the circular us the differential equation
arc, we can analyze the motion of the ball by equating
the tangential forces acting along the arc to the mass d2θ g equation for a
2
+ sin θ = 0 simple pendulum (40)
dt

Equation 40, the differential equation for the simple

θ pendulum, is more complex than the equations we have
been discussing that lead to simple harmonic motion.
If you try as a guess that the motion is sinusoidal and try
the solution θ = sin ωt , it does not work. You are
asked to see why in the following exercise.
mg sin θ
m Exercise 14
Try substituting the guess
mg θ = sin ωt
θ
into Equation 40 and see what goes wrong. Why can’t
Figure 16 you make the left side zero with this guess?
Simple pendulum consisting of a mass
swinging on the end of a string. The
gravitational force has a component mg sin θ
in the direction of motion of the mass.
14-16 Oscillations and Resonance

There is a solution to Equation 40, it is just not the sine The definition of the angle θ, in radian measure, is the
curve we have been discussing. The solution is a curve arc length divided by the radius of the circular arc
called an elliptic integral, a curve generated much as arc length arc length
we generated the sine curve in Figure (3), except that θ ≡ = (42)
radius
the stick whose shadow generates the curve has to
move around an elliptical path rather than around the From Figure (17) we see that for small angles θ the
circular path used in Figure (3). Elliptic integrals carry opposite side x and the arc length are about the same.
us farther into the theory of functions than we want to The smaller the angle θ, the more nearly equal they are.
go in this text, thus we will not discuss the exact If we restrict our analysis to small amplitude swings,
solution of the differential Equation 40. we can replace sin θ by θ in Equation 40, giving us the
differential equation

Small Oscillations d2θ g equation for small

The problem with Equation 40 is the appearance of the + θ = 0 oscillations of a (43)
function sin θ in the second term on the left hand side. dt2 simple pendulum
It is this term that seems to keep us from using the Equation 43 is an equation for simple harmonic mo-
oscillatory solution. tion. If we try the guess θ = sin ω0 t , and plug the
In Figure (17) we look again at the geometry of the guess into Equation 43, we can solve the equation
simple pendulum. In that figure we have a right triangle provided the frequency ω0 and the period of oscilla-
whose small angle is θ, hypotenuse the string length , tion T have the values
and opposite side x. The definition of the sine of the g
angle θ is ω0 =
opposite side x
sin θ ≡ = (41) 2π = 2π
T = ω
period of a
(44)
hypotenuse 0 g simple pendulum

Exercise 15
Substitute the guess θ = A sin ω 0 t into Equation 43
and show that you get a solution provided ω02 = g/ .
θ Then use dimensional analysis to derive a formula for
the period of the oscillation.

From Equation 44, we see that the period of the oscilla-

tion of a simple pendulum depends only on the gravita-
tional acceleration g and the length of the pendulum.
It does not depend on!tXe mass m of the swinging
x object, nor on the amplitude of the oscillation, provided
that the amplitude is kept small. For these reasons the
g th
arc le n simple pendulum makes a good timing device.
Figure 17
x arc length Exercise 16
sin θ = θ =
How long should a simple pendulum be so that it’s
period of oscillation is one second?
 14-17

Simple and Conical Pendulums Exercise 17

In Chapter 9 we analyzed the motion of a conical You can do your own experiments to show that as you
pendulum. The conical pendulum also consists of a increase the amplitude of a simple pendulum, the
mass on a string, but the mass is swung around in a period of oscillation starts to get longer. In contrast,
circle as shown in Figure (18), rather than back and when you increase the radius of the circle for a conical
forth along an arc as for a simple pendulum. pendulum, the height h and the period Tcp become
shorter.
From our analysis of the conical pendulum, we found
that the period of rotation was given by the formula (a) From your own experiments estimate how much
longer the period of a simple pendulum is when the
h period of a maximum angle θmax is 90° than when θmax is small.
Tcp = 2π g conical pendulum (9-34)
(Is it 20% longer, 30% longer? Do the experiment and
where h is the height shown in Figure (18). Consider- find out. Does this percentage depend on the length
of the string?
ing the trouble we went through to get an approximate
solution to the simple pendulum, it seems surprising (b) For a conical pendulum, at what angle θ0 (shown
that Equation 9-34 is an exact solution to Newton’s in Figure 18) is the period half as long as it is for small
second law for any achievable radius x of the circle. angles θ0 ? Give your answer for θ0 in degrees.

For small circles, where x << , the height h and the

string length are approximately the same and we have
g g
Tcp = 2 π ≈ 2π (45)
h
But this is just the period of a simple pendulum if the
oscillations are kept small. Since the two pendulums
have the same period for small oscillations, it makes no
difference, as far as the period is concerned, whether
we swing the balls back and forth or around in a circle.
This prediction is easily checked by experiment.

θ0

m x
Figure 18
The conical pendulum.
 14-18 Oscillations and Resonance

Exercise 18 Exercise 19 A Physical Pendulum

Another way to analyze the simple pendulum is to treat A uniform rod of length is pivoted at one end as shown
the mass and the string as a rigid object that can rotate in Figure (20). It is free to swing back and forth about this
about an axis through the top end of the string as shown axis, forming what is called a physical pendulum. A
in Figure (19). Then use the angular version of Newton’s simple pendulum is one where the mass is all concen-
second law in the form trated at the end as in Figure (19). In a physical
τ = I d θ2
2 pendulum the mass is distributed in some other way, in
(12-20)
dt this case uniformly along the rod.

where τ is the net torque, and (I) the moment of inertia, (a) What is the torque τ about the axis 0 exerted by the
of the mass and string about the axis 0. gravitational force on the rod? (In Chapter 13 near
Equation 13-11, we showed that when calculating the
(a) When the string is at an θ angle as shown in Figure torque exerted by a gravitational force, you may as-
(19), what is the torque τ about the axis 0, exerted by the sume that all the mass is concentrated at the center of
gravitational force mg ? gravity of the object.)
(b) What is the moment of inertia I of the mass and (b) What is the moment of inertia (I) of the rod about the
string about the axis 0? axis at the end of the rod? (See exercise 5 in Chapter
12.)
(c) Show that when you use the above values for τ and
(I) in Equation 12-20 you get the same differential (c) Write the differential equation for the motion of the
Equation 40 that we got earlier for the simple pendulum. rod. (Use the procedure outlined in Exercise 18.)

(d) Find the period of small oscillations of the rod.

axis O axis O

θ θ

mg
m

mg

Figure 19 Figure 20
The simple pendulum A physical pendulum.
treated as a rigid object.
 14-19

NON LINEAR RESTORING FORCES In physics, there are many examples of complex, non
The simple pendulum is an example of an oscillator linear restoring forces which for small amplitudes can
with a non linear restoring force. In Figure (21), we be approximated by a linear restoring force, and which
show the actual restoring force (mg sin θ ) and the therefore lead to small amplitude sinusoidal oscilla-
linear approximation (mg θ ) that we used in order to tions. A rather wild example which we will discuss
solve the differential equation for the pendulum’s shortly, is the collapse of the Tacoma Narrows bridge.
motion. You can see that if the angle θ always remains The bridge undoubtedly started oscillating with small
small, much less than π/2 in magnitude, then the linear amplitude sinusoidal oscillations. What happened was
force (mg θ ) is a good approximation to the non linear that these oscillations were continually driven by the
force (mg sin θ ). Since the linear force gives rise to shedding wind vortices until the amplitude of oscilla-
sinusoidal simple harmonic motion, we expect sinusoi- tion became large and the restoring force was no longer
dal motion for small oscillations of the simple pendu- linear. (There was still a more or less sinusoidal motion
lum. What we are seeing is that a linear restoring force almost up to the point when the bridge collapsed.)
is described by a straight line, and that the non linear
restoring force can be approximated by a straight line
in the region of small oscillations.
θ

Frestoring =
mg sin θ
m

Frestoring
mg
θ
region where
sin θ θ
π
2
θ
−π
2
– mg sin θ

mg θ

Figure 21
The non linear restoring force mg sin θ can be approximated by
the straight line (linear term) mg θ if we keep the angle θ small.
14-20 Oscillations and Resonance

MOLECULAR FORCES While the overall shape of the force curves for the
One of the most important examples of a non linear simple pendulum and the molecular force are quite
restoring force is the molecular force between atoms. different, right in close to the origin both curves can be
Consider, for example, the hydrogen molecule which approximated by a straight line, a linear restoring force.
consists of two hydrogen atoms held together by a As long as the amplitudes of the oscillation remain
molecular force. (We will discuss the origin of the small, we effectively have a linear restoring force and
molecular force in Chapter 17.) any oscillations should be simple harmonic motion.

In the hydrogen molecule, the hydrogen atoms have an In Chemistry texts one often sees molecular forces as
equilibrium separation, and the molecular force pro- being represented by springs as shown in Figure (23).
vides a restoring force to this equilibrium separation. The spring force, given by Hooke’s law, is our ideal
The restoring force, however, is quite non linear. If you example of a linear restoring force. We can now see
try to squeeze the atoms together, you quickly build up that, while the molecular force in Figure (22) does not
a large repulsive force that keeps the atoms from look like a linear spring force, if the amplitude of
penetrating far into each other. oscillation remains small, the spring force provides a
reasonably good approximation to the actual molecu-
If you try to pull the atoms apart, there is an attractive lar force. The chemist’s diagrams are not so bad after
force that pulls the atoms back together. The attractive all.
force never gets too big, and then dies out when the
separation gets much larger than an atomic diameter. In a crystal, like quartz, where you have many atoms
held together by molecular forces, it is possible to get
In Figure (22) we have sketched the molecular force as all the atoms oscillating together. Each atom only
a function of the separation of the atoms, the origin oscillates a very small distance about its equilibrium
being at the equilibrium position. This graph is not too position, but all the oscillations can add up to produce
unlike Figure (21) where we have the force curve for a fairly large, quite detectable oscillation of the crystal
the simple pendulum. For the pendulum, the equilib- as a whole. An advantage of a quartz crystal is that these
rium position is at θ = 0 , thus the origin of both curves oscillations can be both driven and detected by electric
represents the equilibrium position. fields. This vibration or simple harmonic motion of a
small quartz crystal is used as the basic timing device
for digital watches, computers, and almost all forms of
H H modern electronics.

In Galileo’s time we used small oscillations of a non

molecular linear harmonic oscillator, the simple pendulum, as a
force
repulsive
basic time device. Now we use the small oscillations
of a non linear harmonic oscillator, the atoms in a quartz
crystal, as our most convenient timing device. The
region where we have main thing we have changed in the last 300 years is not
an approximately linear the basic physics, but the size and frequency of the
restoring force
separation device.
of atoms
attractive
H H
equilibrium
separation
Figure 22 Figure 23
Sketch of the molecular force between two hydrogen Representation of the molecular
atoms. As long as the atoms stay close to the force by a spring force.
equilibrium position, the force can be represented by
a straight line—a linear restoring force.
 14-21

DAMPED HARMONIC MOTION motion. Starting with the first version of the program
If you start a pendulum swinging or a quartz crystal that has no damping, you can include damping by
oscillating, and do not keep the oscillation going with changing the line
some kind of external force, the oscillation will even- LET F = – k * x
tually die out due to friction forces. Such a dying
oscillation is called damped harmonic motion. to the new line
LET F = – k * x – b * v (49)
The analysis of damped harmonic motion starts out
quite easily. In Newton’s second law add a damping and placing your choice for the damping constant b in
force like the air resistance term we added to our the initial conditions.
analysis of projectile motion. (See Chapter 3, Figure
In contrast, when working with differential equations
31.) We could write, for example
analytically you find that a very small change in the
Ftot = Frestoring + Fdamping equation can make a great deal of difference in the
effort required to obtain a solution. Adding a bit of
Ftot = – kx – bvx (46) damping to a harmonic oscillator changes the curve
where x t is the coordinate of the oscillator, from a pure sinusoidal motion to a dying sine wave. If
vx t = dx/dt its velocity, and we are assuming a you try using a pure sine wave as a guess for the solution
simple linear damping proportional to –v with a to the differential Equation 48 for damped harmonic
strength b. motion, the guess does not work because the pure sine
wave has the wrong shape. The decay of the sine wave
Using Equation 46 in Newton’s second law gives has to be built into your guess before the guess stands
Ftot = – kx – bvx = ma x (47) a chance of working.

With ax = d2x/dt2, this becomes, after dividing through The difficult part about solving differential equations is
by m and rearranging terms that you essentially have to know the answer before
you can solve the equation. You only have to know
d 2x b dx k general features like the fact that in working with
+ + x = 0 (48) Equation 11a you are dealing with a sinusoidal oscilla-
dt2 m dt m
tion. You can then use the differential equation to
determine explicit features like the frequency of the
Equation 48 is our new differential equation for damped oscillation. It is helpful to have a physical example to
harmonic motion. It is like our old differential Equa- tell you what the general features of the motion are, so
tion 11a for undamped oscillation, except that it has the that you can begin the process of solving the equation.
additional term mb dx representing the damping.
dt That is why we begin this chapter with the demonstra-
If we were doing a computer solution of harmonic tion in Figure (2) that the motion of a mass on a spring
motion, adding the damping term represents hardly any is similar to circular motion seen sideways, namely
extra effort at all. In the appendix to this chapter we sinusoidal motion.
discuss a short computer program to handle harmonic
14-22 Oscillations and Resonance

To set up a physical model for damped harmonic section. It is hard to control the damping, just adding
motion is not too difficult. One way to add damping to the pulley in Figure (24) gave us almost too much
the air cart and springs oscillator is to run a string from damping. Worse yet, the damping that we get from
the air cart over a pulley to a small weight as shown in friction in a pulley, or a mass moving up and down in
Figure (24). The idea was to have the weight move up water, is not a simple linear damping force of the form
and down in a glass of water to give us fluid damping. – bv. What is remarkable about these systems is that
But it turned out that there was enough friction in the much more complex forms of damping give us results
pulley itself to give us considerable damping. similar to what we would get with linear damping.
To record the motion of the cart, we used the air cart In Chapter 27 we will study the behavior of basic
velocity detector that we used in Chapter 8 to study the electric circuits made from electrical components called
momentum of air carts during collisions. Figure (25a) capacitors, inductors, and resistors. It turns out that the
shows the velocity of the air cart damped only by the amplitude of the currents in these circuits obey differ-
friction in the pulley. In Figure (25b) water was added ential equations that are exactly like our oscillator
to the glass so that the weight on the string was moving Equations (11) and (45). The damping is caused by the
up and down in water. The result was considerably resistor in the circuit, the damping is accurately given
more damping with the curve almost dying out before by a linear damping term proportional to the amount of
any oscillations take place. resistance in the circuit. (The resistance can be changed
simply by turning a knob on a resistance pot.)
It turns out that mechanical oscillators like a pendulum
or a mass on a spring are not particularly convenient Figure (26), taken from Chapter 27, is an example of
devices for studying damped harmonic motion, or damped harmonic motion in an electric circuit. Here
forced harmonic motion which is the subject of the next we have a curve with enough oscillations so that we can
see how the wave is damped. In Chapter 27 we will see
that the amplitude of the oscillation dies exponentially,
αt
glass
following a mathematical curve of the form e– . As
a result, the wave in Figure (26) has the form
– αt
x = Ae sin ωt
Figure 24 decaying sine wave (50)
Adding damping to the air cart oscillator. amplitude oscillation

Figure 25a Figure 25b

Damping caused by the Resulting motion when water
pulley and weight alone. was added to the glass.
 14-23

It turns out that if we use Equation 50 as our guess for use the following rules of differentiation to calculate
2 2
a solution to our differential Equation 48 for damped dx/dt and d x/ dt
harmonic motion, the guess works, and we can deter- d – αt αt
e = – αe–
mine both the frequency ω and decay rate α in terms dt
of the constants that appear in Equation 48. d da db
a(t)b(t) = b +a
dt dt dt
When we study electric circuits, you will get much
αt
more experience with the exponential function e– , and show that when you try this guess in the differential
and you will have a better laboratory setup for studying Equation 48, you do in fact get a solution, and that
damped and forced harmonic motion. In other words, ω and α are given by
now, with our somewhat crude mechanical experi- 2
k b b
ments and lack of familiarity with exponential damp- ω = – α= (51)
m 4m2 2m
ing, is not the best part of the course to go deeply into
the mathematical analysis of these motions. What we You can see that in the absence of damping, when
will do instead is discuss the motions more or less b = 0, we get back to our old result ω = k/m .
qualitatively and leave the more detailed analysis for
later.
Critical Damping
Exercise 20 Damped Harmonic Motion
In Figure (25b) the damping was so great that the
motion damped out almost before the curve had a
We won’t let you completely off the hook for doing
mathematical analysis of damped harmonic motion. chance to oscillate. It turns out that there is a critical
Start with Equation 5a as a guess for the form for the amount of damping that just kills all oscillations. Any
displacement x(t) for a damped harmonic oscillator further increase in damping and the mass just coasts to
rest.
αt
x(t) = Ae– sin (ωt)
The idea of critical damping can be seen in our analytic
solution for damped harmonic motion obtained Exer-
cise 20. Equation 51 gives us a formula for the
frequency of oscillation ω in terms of the constants k
and b. We can see as the frequency of oscillation goes
to zero, i.e., the period of oscillation becomes infinite
when
k b2
= ; b = 2 mk
m 4m2

critical
b = 4mk (52)
damping
Equation 52 is the condition for critical damping
because if the period of oscillation is infinite there are
Figure 26 no oscillations.
Damped harmonic motion seen in an electric
circuit. Note the difference in time scales. The
electrical oscillations we will study are usually of
much higher frequency than the mechanical ones.
14-24 Oscillations and Resonance

RESONANCE The striking feature of resonance is that a small re-

When you are pushing a child on a swing, you time your peated force can produce a large oscillation. If the
pushes to coincide with the motion of the child. Usu- damping is small then by adding just a little energy with
ally you give a shove just after the child has swung back each shove, the energy accumulates until you end up
and is starting forward again. with a very energetic oscillation. A rather dramatic
consequence of this effect is shown in Figure (27)
If you push forward just as the child is swinging where we see the Tacoma Narrows bridge oscillating
forward, your force F and the child’s velocity v are in wildly and then collapsing.
the same direction, the dot product F • v is positive,
and you are adding energy to the child’s motion. The new bridge was dedicated in April of 1940. Three
Initially the energy you add goes into increasing the months later a reasonably stiff breeze started the bridge
amplitude of the swing. After a while friction effects oscillating, an oscillation that finally destroyed the
become large enough that the energy you add in each bridge’s integrity.
push is dissipated by friction in each swing. (If there is The brute force of the wind itself did not destroy the
not enough friction, or you push too hard, the child will bridge. The bridge was designed to handle far stronger
end up going over the top.) winds. What happened was that as the wind was
The key to getting the child swinging was to time your blowing over the bridge, vortices began to peel off the
shoves so that F • v was always positive. If you pushed bridge. Whenever fluid flows past a cylindrical object
the child at random intervals, so that F • v was some- at the right speed, vortices began to peel off, first on one
times positive, sometimes negative, you would be side of the cylinder, then the other, and are carried
sometimes adding energy and sometimes removing it. downstream, forming a wake of vortices seen in the
The net result would be that your shoves would not be wind tunnel photograph of Figure (28). This vortex
particularly effective in helping the child to swing. structure is called a Karmen vortex street after the
hydrodynamicist Theodore Von Karmen.
To make sure that you are always adding energy to the
child's swing, you want to time your shoves with the In the case of the Tacoma Narrows bridge, vortices
natural frequency of oscillation of the child. When you alternately peeled off the top and bottom of the down
do this, we say that your shoves are in resonance with wind side of the bridge, rocking the bridge at its natural
the oscillation of the child. frequency of oscillation. While no separate jolt by any
one vortex would have much effect on the bridge, the

Figure 27a Figure 27b

Tacoma Narrows bridge oscillating in the After a couple of hours
winds of a mild gale on July 1, 1940. the bridge collapsed.
 14-25

resonance between the peeling of vortices and the where in the driving term, Fd , represents the ampli-
oscillation of the bridge caused the oscillations to grow tude or strength of the sinusoidal driving force. Using
to destructive proportions. Equation 52 in Newton’s second law Fx tot = ma x
gives
The example of the Tacoma Narrows bridge illustrates 2
dx d x
how widely the ideas of simple harmonic motion and – kx – b + Fd sin ωt = m 2 (53)
resonance apply to physical systems. The bridge is far dt dt
more complex than a mass on a spring, and the vortex Dividing through by m and rearranging terms gives
street (line of vortices) exerts a rather complex driving
force. However the bridge had a natural frequency, the
d 2x b dx kx
vortices provided a small driving force at that fre-
2
+ + = Fdsin ωt
quency, and we got a resonant amplification of the dt m dt m
(54)
oscillation.

To apply Newton’s second law to resonant motion, we Equation 54 is the standard form for the differential
have to add an oscillating driving force to the system equation representing forced or resonant harmonic
under study. As we have seen from the Tacoma motion. It is the simplest equation we can write whose
Narrows bridge discussion, we do not need to know the solution has the features we associate with the phenom-
exact form of the driving force, all we need is a ena of resonance.
repetitive force that can be timed with the natural In our study of electric circuits, we can easily create a
oscillation. For the theoretical analysis we can use the circuit whose behavior is accurately described by
simplest mathematical form we can find for the driving Equation 54. We saw that we could use resistors to add
force, which turns out to be a sine wave. linear damping of the form – bv. It is not hard to add a
To write a formula for the driving force, let ω 0 be the purely sinusoidal driving force of the form Fd sin (ωt) ,
natural frequency of oscillation ( ω 0 = k/m if the where we can adjust the driving frequency by turning
damping is small), and let ω be the frequency of the a knob. In other words, with electric circuits we can
driving force. Then the total force, acting on the accurately study the predictions of Equation 54.
oscillating system like a mass on a spring, can be With mechanical systems like a mass on a spring, it is
written hard to get linear damping, and the sinusoidal driving
Fx tot = – kx – bv + Fd sin (ωt) (52) force is usually simulated by some trick such as wig-
gling the supported end of the spring at a frequency ω.
Despite the crudeness of the experiment, the equation
gives a surprisingly good prediction of what we see.

Since we will later have a laboratory setup that accu-

rately matches Equation 54, we will postpone (until
Appendix 1) the mathematical solution of Equation 54.
Instead we will investigate the resonance phenomena
qualitatively, using the simple setup of a mass on a
spring, where we hold the other end of the spring in our
hand and move our hand up and down at a frequency
Figure 28
ω as shown in Figure (29).
Karman vortex street in the flow of water past a
circular cylinder. The vortices peel off of alternate
sides of the cylinder and flow downstream forming a
double line of vortices. (Reynolds number = 140.)
Photograph by Sadatoshi Taneda.
14-26 Oscillations and Resonance

Resonance Phenomena x = x 0 sin (ωt + φ) (57)

We have already seen that if we hold our hand still, pull
the mass down, and let go, the mass oscillates up and where φ is the phase angle of the oscillation (see
down at the natural frequency ω0 = k/m . You can Figures 8 and 9 and Equation 4 at the beginning of this
observe some damping, because the mass finally stops chapter for a discussion of phase angle.) In Equation
oscillating. But the damping is small and has no 56, where ω << ω 0 and there was no phase difference,
noticeable effect on the resonant frequency. (We can the phase angle φ is zero. In Equation 56 where
neglect the term b 2 /4m2 in Equation 51.) ω > > ω 0 , and the motion is completely out of phase,
the phase angle φ is π or 180°.
Now try a very different experiment. Stop the mass
from oscillating, and slowly move your hand up and Equations 55 and 56 represent the two extremes of
down a small distance. If you do this slowly enough driven harmonic motion. The mass moves with a small
and carefully enough, the mass will move up and down amplitude at the same frequency ω as the driving force.
with your hand (just as if the spring were not there). In When the driving frequency is much less than the
this case the formula for the motion of the mass is natural frequency ω 0 , the difference in phase between
the driving force and the response of the mass is zero
x = x 0 sin ωt ω << ω0 (55) degrees. When ω >> ω 0 the phase difference in-
creases to π or 180°.
where the frequency ω of the oscillation of the mass is
the frequency of ω the oscillation of your hand. This As the third experiment, start at the low frequency
only happens if you oscillate your hand at a frequency where the mass is following your hand, and slowly
much much lower than the natural oscillation fre- increase the frequency ω of oscillation of your hand,
quency ω0 . keeping the amplitude of oscillation constant. As ω
approaches ω 0 , the amplitude of oscillation of the
In the next experiment, keep the oscillations of your
mass increases. When you get close to the natural
hand small in amplitude, but start moving your hand up
frequency ω 0 , the oscillation becomes so large that the
and down rapidly, at a frequency ω considerably
mass will most likely jump off the spring. This is the
greater than the natural frequency ω0 . Now, what
phenomena of resonance, the phenomenon that de-
happens is that the mass oscillates at the same fre-
stroyed the Tacoma Narrows bridge.
quency as your hand, but out of phase. When your hand
is going down, the mass is coming up, and vice versa. How big the oscillation of the mass becomes depends
Now the formula for the displacement x of the mass is mainly how close you are to resonance, how close ω is
x = – x 0 sin ωt ω > > ω0 (56) to ω 0 , and how big the damping force is relative to the
driving force. The formula for the amplitude x 0 of the
where the minus sign tells us that the mass is oscillating motion of the mass obtained by solving Equation 54 is
out of phase with our hand.
Fd /m
A way we can write both Equations 55 and 56 is in the x0 =
2
form ω 2 – ω0 2 + b/m ω 2 (58)

ω
where Fd is the strength of the driving force, ω the
Figure 29 driving frequency, ω0 = k/m the natural frequency
Experimental setup for and b the damping constant. In the absence of damping
a qualitative study of (b = 0), Equation 58 predicts an infinite amplitude at the
resonance phenomena.
resonant frequency ω = ω0 . Such an infinite ampli-
tude is prevented either by damping or by the destruc-
m
M tion of the system. (A damping mechanism could have
x saved the Tacoma Narrows bridge.)
 14-27

Exercise 21 Transients
To derive Equation 21, you start with Equation 57 as a There is one more qualitative experiment we want to do
guess with our simple apparatus of the hand held spring and
mass of Figure (29). Instead of gently starting the mass
x = x 0 sin ωt + φ (57)
moving as we had you do in the earlier experiments, let
and substitute that into the differential Equation 54. It the mass fall from some small height and move your
turns out that you can get a solution provided that the hand up and down at the same time.
amplitude x 0 has the value given by Equation 58, and
the phase angle φ is given by If you just let the mass drop from some small height, it
will oscillate up and down at the resonant frequency
b m ω ω0 . If you just start moving your hand slowly at a
tan φ = (59) frequency ω , the mass will move at the same frequency
ω2 – ω 0 2
as your hand, building up to an amplitude given more
Doing the work, actually substituting Equation 57 into 54 or less by Equation 58 and shown in Figure (31), the
and getting Equations 58 and 59 for x 0 and φ is a driven oscillation we have been discussing.
somewhat messy job which we leave to Appendix 1 on
the next page. Here we would rather have you develop If you drop the weight and move your hand at the same
an intuitive feeling for the solutions in Equations 58 and time, you get both kinds of motion at once. You get the
59 by working the following exercises. natural oscillation at a frequency ω0 that eventually
dies out due to damping, and the driven oscillation at
(a) Write the formula for x 0 in the case b = 0. Sketch the the frequency ω that eventually builds up to an ampli-
resulting curve using the axes shown in Figure (30).
tude x 0 . For a while, before the natural oscillation has
Explain what happens as ω → ω0 .
died out, the resulting motion is a mixture of two
(b) When the damping is not zero, find the formula for frequencies of oscillation and can look quite complex.
the amplitude of oscillation x 0 at the resonant fre- The natural oscillation is called a transient because it
quency ω → ω0 . Check the dimensions of your an- eventually dies out. But until the transients do die out,
swer. forced harmonic motion can be fairly complicated to
analyze. In the next chapter we will study a powerful
(c) What is the phase angle φ at resonance? How does
technique called Fourier analysis that allows us to
the phase angle change as we go from ω << ω0 to
study complex motions that involve such a mixture of
ω >> ω0 ?
oscillations.
amplitude
In Figure (31) we have graphed the amplitude x 0 for
a fixed driving force Fd , as a function of ω for several
values of the damping constant b. The main point to get
from this diagram is that the smaller the damping, the
sharper the resonance.
amplitude

Figure 30
Use these axes to
sketch the amplitude
vs. frequency for no
damping.

ω/ω 0 1.0 frequency ω/ω 0 0.6 0.8 1.0 1.2 1.4

Figure 31 frequency
Amplitude of the oscillation for various values
of the damping constant. The amplitude of the
driving force Fd is the same for all curves.
14-28 Oscillations and Resonance

APPENDIX 14–1 In deriving dx/dt we used the fact that

SOLUTION OF THE dx dx dt′ dx
DIFFERENTIAL EQUATION = =
dt dt′ dt dt′
FOR FORCED HARMONIC MOTION where t′ = t – φ/ω so that dt′/dt = 1 . We will also
The equation we wish to solve is Equation 54 use the trigonometric identity
d 2 x + b dx + k x = Fd sin ωt sin a + b = sin a cos b + cos a sin b
(54)
dt 2 m dt m m
to write
where our guess for a solution is Equation 57
F′ sin ωt′ – φ
x = x0 sin ω t + φ (57) = F′ sin ωt′ cos – φ + F′ sin – φ cos ωt′ (64)
= F′ sin ωt′ cos φ – F′ sin φ cos ωt′
The quantities x0 and φ are the unknown amplitude and
phase of the oscillation that we wish to determine. where we used
To simply plug our guess into Equation 54 and grind cos – φ = cos φ , sin – φ = – sin φ
away leads to a sufficiently big mess that we could
easily make a mistake. We will instead simplify things Substituting Equation 63 and 64 into 62, and separately
as much as possible to make the calculation easier. The collecting terms with sin ωt′ and cos (ωt′) , we get
first step is to define the constants
sin ωt′ – ω 2x 0 + ω0 2x 0 – F′ cos φ +
ω0 2 = k/m ; b′ = b/m ; F′ = Fd /m (60)
Next we wish to get the phase angle into the forcing cos ωt′ ωb′x 0 + F′ sin φ
term so that it appears only once in our equation. We
can do this by using a time scale t′ where = 0 (65)

ωt′ = ωt + φ ⇒ ωt = ωt′ – φ (61) Because there are both cos (ωt′) terms and sin ωt′
terms in Equation 65, there is no way to make every-
In terms of the new constants and t′ our differential thing add up to zero for all times unless the coefficients
equation becomes of both cos (ωt′) and of sin ωt′ are separately equal
d 2 x + b′ dx + ω 2 x = F′ sin ωt′ – φ (62) to zero. This gives us the two equations
0
dt 2 dt
F′sin φ = – ωb′x 0 (66)
Our guess, and its first and second derivative are
x = x0 sin ω t′ (63a) F′cos φ = ω0 2 – ω 2 x 0 (67)
dx
= ω x0 cos ω t′ (63b)
dt

d2 x 2
2
= – ω x0 sin ω t′ (63c)
dt
 14-29

If we divide Equation 66 by Equation 67, the F′ and x 0 The transient motion x t r (t) is just damped harmonic
cancel and we get motion that satisfies the equation of motion
sin φ – ωb′ d2x b dx k
= 2
+ + x = 0 (48)
m dt m
cos φ ω02 – ω 2 dt
The question we wish to answer now is whether the
b/m ω (59) forced harmonic motion equation
tan φ =
ω 2 – ω02
d 2x b dx k
+ + x = Fd sin ωt (54)
This is the result we stated earlier, namely Equation 59. dt2 m dt m

To solve for the amplitude x 0 of the oscillation, we can allows us to have both driven and transient motion at
use Equation 66 to get the same time. In other words, is a guess of the form

x 0 = F′ sin φ (68) x = x 0 sin ωt + φ + x tr (70)

ωb′
where x is the sum of the driven motion and an arbitrary
To find the sin φ from the tan φ which we already
amount of transient motion, is this sum also a solution
know, construct the right triangle shown in Figure (32).
of Equation 54?
We have made the opposite and adjacent sides so that
the ratio comes out as tan φ , and the hypotenuse is If you substitute our new guess 70 into 54, the driven
given by the Pythagorean theorem. Thus sin φ is term satisfies the whole equation and the transient
sin φ = ωb′ (69) terms add up to zero because of Equation 48, thus we
ω2 – ω0 2 + ω 2 b′ 2 do get a solution. Transient motions are allowed by
Newton’s second law.
Substituting 69 into 68, and setting b′ = b/m, F′ = Fd /m ,
we get
b' 2
ω2
)+

Fd /m
ω b'
o 2

x0 = (58)
-ω

ω – ω0 + ω b /m
2 2 2 2 2
(ω 2

which is our earlier Equation 58 for x0 .

θ
Transients
ω 2 − ωo2
In our qualitative discussion of forced harmonic mo-
Figure 32
tion, we saw that in addition to the driven oscillation
Triangle to go from tanθ to sinθ.
xdriven = x 0 sin ωt + φ we have just studied , we
could also have transient motion at the natural fre-
quency ω0 . In controlled experiments, you observe
that any transient motion present initially finally dies
out and you are eventually left with just the driven
motion.
14-30 Oscillations and Resonance

APPENDIX 14-2 pendulum. For large amplitude swings, the restoring

COMPUTER ANALYSIS force became non linear which led to a differential
OF OSCILLATORY MOTION equation that is difficult to solve. As the amplitude
increases, there is a lengthening of the period that is
In this appendix, we will use the computer to analyze
easy to measure but difficult to predict using calculus.
the motion of the oscillating air cart shown in Figure
However with the computer, it is as easy to use the exact
(33). For this problem, the computer solution does not
force M g sin θ as it is to use the approximate linear
have the elegance of the calculus solution we have been
force M g θ . Thus with the computer you can predict
discussing. The calculus approach gives us a single
the lengthening of the period and compare your results
solution valid for all values of the experimental param-
with experiment.
eters. With the computer we have to alter the program
and rerun it any time we want to change a parameter In Appendix I, we made a considerable effort to predict
such as the mass of the cart, the spring constant, or the the effects of adding a time dependent driving force to
initial position or velocity. The calculus approach a harmonic oscillator. The work paid off in that we got
gives us a single formula valid for all values of the Equations 58 and 59 which provide a general descrip-
experimental parameters. tion of resonance phenomena. With the computer you
do not get these elegant formulas, but it is much easier
However, the advantage of using the computer is that
to add a time dependent force and see what happens. In
we can easily modify the program to include new
effect the computer solutions can be used as a labora-
physical phenomena. For example, to add damping, all
tory to test the predictions of Equations 58 and 59. This
we have to do is change the command
provides an opportunity for a lot of project work.
LET F = – K * X F
to the command
x
LET F = – K * X – b * V
and rerun the program. To add damping to the calculus
solution, we had to work with a differential equation
(48) that was much more difficult to solve than the
equation for undamped motion (11). F (restoring force)
The computer opens up a number of possibilities for
student project work. For example, in our discussion of x
the simple pendulum shown in Figure (21), we had to x
limit our analysis to small amplitude swings of the (distance
–kx stretched)

Fs = –kx
Figure 33
Reproduction of Figure (10), showing an oscillating
air cart. If the cart is displaced a distance x from
equilibrium, there is a restoring force F = - kx. The
force is measured by adding weights as shown.
 14-31

English Program On the next page we repeat this English program and
In Chapters 5 and 8 our usual approach for solving a show its translation into the computer language
new mechanics problem on the computer was to modify BASIC.
an old working program. But because the harmonic
oscillator is an easy one dimensional problem, we will English Program
start over with a new program. Our general procedure
has been to first write an English program that de-
scribed the steps using familiar notation. Once we ! --------- Experimental constants
checked the steps to see that the program did what we LET M = 191 grams (cart mass)
wanted, we then translated the program into an actual LET K = 3947 dynes/cm (spring constant)
computer language such as BASIC.
! --------- Initial conditions
The English program for the oscillating air cart is LET X 0 = 10 cm
shown in Figure 34. In the first section, we state the
LET V0 = 0 (release from rest)
experimental constants, namely the mass M of the air
cart and the spring restoring constant K. For this LET T0 = 0 (start clock)
particular experiment, the cart has a mass M of 191
! --------- Computer Time Step
grams, and the spring constant K was 3947 dynes/cm.
As indicated in Figure (33), the spring constant K was LET dt = .01
determined by tying a string to the mass, running the
string over a pulley, and hanging weights on the other ! --------- Calculational loop
end. We got a linear force verses the distance curve like DO
the one in Figure (9-4), and used the same method to LET X new = X old + Vold*dt
find K. LET F = – K*X (spring force)
In the next section of the program, we choose an LET A = F/M
explicit set of initial conditions. For this problem we LET Vnew = Vold + A old*dt
start the cart from rest V0 = 0 at a distance 10 cm to LET Tnew = Told + dt
the right of the equilibrium position X 0 = 0 . The cart PLOT X vs T
is released at time T0 = 0 .
LOOP UNTIL T > 15
In the lab we observed that the period of oscillation was
about 1.5 seconds. Thus a calculational time step END
dt = .01 seconds gives us about 150 points for one
oscillation, enough points for a smooth plot. Figure 34
English program for the motion of
The calculational loop is similar to the one in the an oscillating cart on an air track.
projectile motion program of Figure (5-18), page
5-16, except that for one dimensional motion we do not
need vectors, and the old command

LET A = g
is replaced by

LET F = – K * X
LET A = F /M
14-32 Oscillations and Resonance

English Program BASIC Program

! --------- Experimental constants

LET M = 191 grams (cart mass)
LET K = 3947 dynes/cm (spring constant)

! --------- Initial conditions

LET X 0 = 10 cm
LET V0 = 0 (release from rest)
LET T0 = 0 (start clock)

! --------- Computer Time Step

LET dt = .01

! --------- Calculational loop

DO
LET X new = X old + Vold*dt
LET F = – K*X (spring force)
LET A = F/M
LET Vnew = Vold + A old*dt
LET Tnew = Told + dt
PLOT X vs T
LOOP UNTIL T > 15

END
Figure 34 repeated Figure 35a
English program for the motion of an oscillating cart BASIC program for the motion of an oscillating cart
on an air track. on an air track.

The BASIC Program

Because no vectors are involved in the harmonic
oscillator program, the translation into BASIC is al-
most automatic. Drop the subscripts “new” and “old”,
fix up the PLOT statement, add the plotting window
commands, and you have the result shown in Figure
(35a). Select RUN and you get the plot of oscillating
motion shown in Figure (35b).

Figure 35b
Output of the BASIC program, showing the
oscillation of the cart.
 14-33

The plot of Figure (35b) nicely shows the sinusoidal

oscillation, but does not tell us the numerical value of
the period of oscillation. To determine the period, we
modified the program as shown in Figure (36a). The
main change is to replace the PLOT statement by a
PRINT statement. To reduce the output, we included
MOD statement (as described in Exercise 5-5, page
5-9) so that only every tenth calculated point would be
printed. From the output shown in Figure (36b), we see
that the period is close to 1.4 seconds. A more accurate
value of the period can be obtained by not using the
MOD statement and printing every value as shown in
Figure (36c). From this section of data we see that the
period is closer to 1.39 seconds.

Exercise 22
Show that the frequency of oscillation seen in the
computer output of Figure (36) is consistent with the
calculus derived equation

ω = k/M

Figure 36a
Program for numerical output.

Figure 36c
Detailed numerical output. By printing every
calculated numerical value, we can more accurately
determine the period of oscillation.

Figure 36b
Numerical output.
14-34 Oscillations and Resonance

Damped Harmonic Motion

In Figure (37a) we modified the projectile motion
program of Figure (35a) to include damping The only
change, shown in boxes in Figure (37a) is to replace
LET F = – K * X
by
LET F = – K * X – b * V
where we gave b the numerical value of 100 to get the
result shown in Figure (35b).

Exercise 23
(This is more of an introduction to project work)

In our analysis of damped harmonic motion in Exercise

20, we predicted that the frequency for damped har-
monic motion would be
2
ω = k – b (51)
M 4m2
In the special case that
2
k = b ; b = 4 mk (51a)
M 4m2 Figure 37a
BASIC program for the damped harmonic motion.
we get ω = 0 which is the case of critical damping,
where oscillations cease.

Run the damped harmonic oscillator program of Figure

(35a) for values of b near 4 mk and show that oscilla-
tions cease when you get to this critical value.

Figure 37b
Plot of damped harmonic motion.
Chapter 15
One
Dimensional
Wave Motion
CHAPTER 15 ONE DIMENSIONAL WAVE
MOTION

In the last few chapters we have followed the straight- guess just as we did for the simpler equations for
forward procedure of identifying the forces on an oscillatory motion. For oscillatory motion, when we
object, setting the vector sum of the forces equal to the plugged in our guess sin (ω t) , we ended up with a
mass times acceleration, and solving the resulting simple equation ω = k/m for the frequency of the
equation. We began with problems like those involving oscillation. For wave equations, if you plug in a guess
circular motion where we knew the acceleration and representing a wave traveling through the medium,
could solve the equation immediately, the conical you end up with a simple equation for the speed of the
pendulum being an example. With oscillatory motion wave.
we ended up with a differential equation whose solu-
tion had to be guessed. The observation that oscilla- There are some famous wave equations in physics. In
tory motion looks like circular motion viewed sideways 1860 James Clerk Maxwell combined the equations for
helped greatly in this guess. electricity with those for magnetism and, to his sur-
prise, ended up with a wave equation. He initially had
For damped and forced harmonic motion, it was not no idea what the wave was, but he could calculate the
hard to write the differential equations, but the solu- speed of the wave. Whatever wave he was dealing with
tions involved mathematical functions and techniques travelled at a speed of 3× 10 8 meters per second or 1
that may not have been not familiar to the reader. In foot per nanosecond. As he knew of only one thing that
this chapter we are dealing with the subject of wave travelled at that speed—light—he concluded that he
motion, where it turns out that the differential equation had an equation for light waves and that the theory
describing the motion has derivatives in both time and leading to this equation was the theory of light. He had
space. Setting up and solving such an equation re- discovered that light was an electric and magnetic
quires mathematical discussions that are best left to a phenomena.
more advanced level course. Fortunately we can study
the physics of wave motion without working with In 1925, Louis De Broglie explained some baffling
differential equations. phenomena in atomic physics by proposing that elec-
trons have a wave nature. Erwin Schrödinger then
If we went through the effort to derive the differential went further and derived a wave equation for the
equation for wave motion, we would end up with what electron, an equation known as Schrödinger’s equa-
is called a wave equation. Once you have a wave tion that serves as the theoretical foundation for
equation, you can guess a solution and plug in your almost all of chemistry.
15-2 One Dimensional Wave Motion

In 1929 Paul Dirac constructed a relativistic generali- mechanics, but to light and electron waves where
zation of Schrödinger’s equation. The problem with Newtonian mechanics does not apply. Thus by learn-
Dirac’s equation was that it had solutions for two ing these general principles of wave motion, you are
different kinds of wave, one representing the electron developing a foundation in physics that goes beyond
and the other an unknown particle of the opposite Newtonian mechanics.
charge. A particle similar to the electron but opposite
in charge, the positron, was observed in a cloud The two basic principles of wave motion we will discuss
chamber experiment carried out by Carl Anderson in in this text are the principle of superposition and the
1933. Huygens principle. The principle of superposition is
a fancy way of saying that waves add. If two waves are
It turns out that the relativistic wave equation for all moving through each other, they produce a total wave
elementary particles has two solutions, one solution that is the sum of the two waves. Since waves can have
like the electron representing matter, the other, like the negative amplitudes (troughs) this addition of waves
positron, representing antimatter. And the wave equa- can produce cancellation. Two waves running into
tions predict that if a matter particle encounters its each other can, under the right circumstances, cancel
corresponding antimatter particle, the two particles each other out. This cancellation is clearly a wave
can annihilate each other. There is an entire world of phenomena, particles are not expected to do that.
antimatter, the existence of which was predicted by
Dirac’s wave equation. The other general principle of wave motion is Huygens
principle, which tells us how waves spread out in space.
With wave equations playing such an important role in In this and the next chapter we will focus our attention
physics, one might think it is unfortunate that we are not on one dimensional waves which do not spread out.
prepared to derive and solve wave equations. Actually Thus we do not need Huygens principle at this point.
that is a blessing. There are certain general, simple Later in Chapter 33, we discuss two and three dimen-
principles that apply to all forms of wave motion, sional waves and phenomena such as interference and
principles that allow you to understand and predict diffraction. In that chapter we do everything using the
many features of the behavior of waves. These prin- principle of superposition and the Huygens principle.
ciples apply not only to waves like water and sound In that chapter no calculus is used and we obtain results
waves whose behavior can be deduced from Newtonian that apply to a broad spectrum of phenomena, even to
sub atomic particles where concepts like velocity and
acceleration no longer have meaning.
 15-3

WAVE PULSES are moving longitudinally or parallel to the direction of

We begin our discussion of wave motion with the wave motion of the wave pulse.
pulses we described in Chapter 1 on Special Relativity. Sound waves are usually compressional waves travel-
To create a wave pulse on a stretched rope, you flick the ing through matter. A sound wave pulse in air can be
end of the rope and a pulse travels down the rope as viewed as a region of compressed gas where the
shown in Figure (1-4) reproduced here. This is called molecules are closer together as shown in Figure (1). It
a transverse wave because the particles in the rope is the region of compression that moves through the gas
move perpendicular or transverse to the direction of in much the same way as the region of compressed coils
motion of the wave pulse. moves along the Slinky as seen in Figure (1-6).
With a stretched Slinky we were able to observe two To create the compressional wave on the Slinky we
different kinds of wave motion, the transverse wave pulled back on the end of the Slinky and let go. This
seen in Figure (1-5) and a compressional wave seen in gives a small impulse directed down the Slinky. In
Figure (1-6). The compressional wave is also called a much the same way we can use a loudspeaker cone to
longitudinal wave because the particles in the spring create the pressure pulse in the air column of Figure (1).
Here the impulse can be provided by applying a voltage
pulse to the speaker causing the speaker cone to sud-
a) denly jump forward. (If the speaker cone suddenly
jumps back, you get a pulse consisting of a region of
low pressure traveling down the tube.)
b)
A transverse or sideways force in the medium tends to
restore the medium to its original shape. For a trans-
c) verse wave on a stretched rope, the tension on the rope
provides the restoring force. For waves on the surface
of a liquid, gravity or surface tension supplies the
d) restoring force. But for waves passing through the bulk
Figure 1-4 of a liquid or a gas, there are no transverse restoring
Wave traveling down a rope. forces and the only kind of waves we get are the
compressional sound waves.

speaker

V wave

Figure 1-5 Figure 1

Transverse wave on a Slinky A sound wave pulse traveling down through a tube
of air. The pulse consists of a region of compressed
air where the air molecules are closer together.
This region of compression moves through the gas
much as the region of compressed coils moves along
the Slinky in Figure 1-6.

Figure 1-6
Compressional wave on a Slinky.
15-4 One Dimensional Wave Motion

The main difference between a liquid and a solid is that (Building a seismograph is a favorite high school sci-
in a liquid the molecules can slide past each other, while ence fair project. Basically you suspend a large mass
in a solid the molecules are held in place by molecular from springs and have a pen which is attached to the
forces. These forces which prevent molecules from mass draw a line on moving stripchart paper as shown
sliding past each other can also supply a transverse in Figure (2). When the earth shakes, the stripchart
shakes with the earth, but the mass remains more or less
restoring force allowing a solid to transmit both trans-
stationary. The result is a squiggly line on the stripchart
verse and compressional waves. An earthquake, for
whose amplitude is the amplitude of vibration of the
example, is a sudden disruption of the earth that pro- earth.
duces both transverse waves called S waves and longi-
tudinal or compressional waves called P waves. These
waves can easily be detected using a device called a
seismograph which monitors the vibration of the earth. SPEED OF A WAVE PULSE
It turns out that the S and P waves from an earthquake One way to predict the speed of a wave is to set up the
travel at different speeds, and will thus arrive at a differential equation for the wave, plug in a traveling
seismometer at different times. By measuring the wave solution and let the equation tell you the speed.
difference in arrival time and knowing the speed of the Without the wave equation we can in some cases
waves, you can determine how far away the earthquake deduce the speed of the wave using clever tricks. One
was. example is the transverse wave on a rope, whose speed
we will calculate now. Another is the speed of Maxwell’s
Exercise 1 wave of electric and magnetic forces which we will
The typical speed of a transverse S wave through the discuss in Chapter 32.
earth is about 4.5 kilometers per second, while the
compressional P wave travels nearly twice as fast, To calculate the speed of a transverse wave on a rope,
about 8.0 kilometers per second. On your seismo- consider a wave pulse moving down a rope at velocity
graph, you detect two sharp pulses indicating the v as shown in Figure (3a). To analyze the pulse,
occurrence of an earthquake. The first pulse is from the imagine that you are running along with the pulse at the
P wave, the second from the S wave. The pulses arrive same velocity v. From your point of view, shown in
three minutes apart. How far away did the earthquake Figure (3b), the pulse is at rest and the rope is moving
occur? back through the pulse at a speed v.

earth jiggling Now look at the top of the wave pulse. For any
reasonably shaped pulse, the top of the pulse will be
circular, fitting around a circle of radius r as shown in
Figure (3c). This radius r is also called the radius of
curvature of the rope at the top of the pulse.

Finally consider a short piece of rope of length at the

top of the pulse as shown in Figure (3d). If this piece
of rope subtends an angle 2θ on the circle, as shown,
then = 2 rθ and the mass m of this section of rope is
heavy mass rotating drum with
with pen stripchart paper massof short
attached m = µ = µ2rθ (1)
sectionof rope
Figure 2 where µ is the mass per unit length of the rope.
Sketch of a simple seismograph for detecting
earthquake waves. When the earth shakes, the mass The net force on this piece of rope is caused by the
tends to remain at rest, thus the pen records the relative
motion of the stationary mass and shaking earth.
tension T in the rope. As seen in Figure (3d), the ends
 15-5

of the piece of rope point down at an angle θ. Thus the The final step is to note that this section of rope is
tension at each end has a downward component moving at a speed v around a circle of radius r. Thus we
T sin (θ) for a total downward force Fy of magnitude know its acceleration; it is accelerating downward,
toward the center of the circle, with a magnitude v 2 /r .
downward
Fy = 2T sin (θ) ≈ 2Tθ componentof (2) v2 downward
tension force ay = accelerationof (3)
r sectionof rope
If we keep the angle θ small, just look at a very small
section of the rope, then we can approximate sin (θ) by Applying Newton’s second law to the downward com-
θ as we did in Equation 2. ponent of the motion of the section of rope, we get using
Equations 1, 2 and 3
v (pulse) Fy = ma y
(4)
v2
2Tθ = µ2r θ
r
Both the variables r and θ cancel, and we are left with
T = µv2
Figure 3a
Wave pulse, and an observer, moving speed of a wave pulse
v = T
to the right at a speed v. µ on a rope with tension T, (5)
mass per unit length µ

A result we stated back in Chapter 1.

v (rope)
=2rθ

θ θ
T sin θ T sin θ
r
Figure 3b Τ θ θ Τ
From the moving observer’s point of view, the
pulse is stationary and the rope is moving
through the pulse at a speed v.

v
Figure 3d
The ends of the rope point down at an angle θ ,
v (rope) giving a net restoring force Fy = 2 T sin θ .

r
Figure 3c
Assume that the top of the pulse fits
over a circle of radius r.
15-6 One Dimensional Wave Motion

DIMENSIONAL ANALYSIS From these variables we have to construct the velocity.

In the above derivation of the speed of a transverse m
v sec (7)
pulse on a rope, we avoided solving a differential The only variable with the dimensions of seconds in it
equation by observing that the rope at the top of the is the tension T, thus T must be included in our formula
pulse, from the moving observer’s point of view, was
for v. To get rid of kilograms, we must divide T by m
moving with circular motion whose acceleration we to give
know. For other kinds of wave pulses, particularly the
compressional pulse seen in Figure (1-6), we do not kg m 1 = T m
T
have a simple circular motion, and a non calculus sec 2 * m kg m sec 2
derivation of the wave speed becomes even more We are getting there, but we must have the same power
convoluted than the derivation we just went through. of meters and sec in order to get a velocity. If we
We could do it, but it is not worth the effort, especially multiply T/m by L meters, we get
since there are more straightforward ways of predicting
wave speeds when one has the differential equation for T m L m = T L m2
m sec 2 * m sec 2
the wave motion.
Finally we get the correct dimensions by taking the
What we will do instead is use a technique called square root, giving
dimensional analysis to predict the speed of the wave.
With dimensional analysis, you do not work out equa- TL m = T m
tions. Instead you determine what the relevant vari-
v = m sec µ sec (8)
ables are, and then combine those variables in such a where we noted that µ = m /L is the mass per unit
way that the dimensions are correct. If you have length.
selected the correct variables, you get an answer that is
correct to within a constant factor, and sometimes the Equation 8 tells us that no matter what the theory is, if
correct answer. the only relevant variables are T, m and L, the speed of
the wave must be proportional to T µ for the dimen-
To see how dimensional analysis works, let us first sions to work out. We may have missed a factor of
apply it to the example we just worked out—to find the 1/2 or 2π , but the functional dependence must be right.
speed of a transverse wave pulse on a rope. (For clarity,
we will italicize the variable names. We will also use Let us now use dimensional analysis to predict the
MKS units.) The first step is to do some experiments to speed of the compressional Slinky pulse shown in
find out what variables the speed depends upon. You Figure (1-6), or any compressional pulse on a stretched
choose a rope, stretch it, and soon discover that the spring. Since a stretched spring has a tension T and a
speed of the pulse depends upon the tension T. Thus T mass per unit length µ, one might guess that T µ
is one of the variables. Then you try two ropes of the could also be the formula for the compressional wave.
same length but different mass m, and discover that you However compressional and transverse waves do not
get different wave speeds for the same tension. Thus have the same speed. Even more important, you can get
the mass m is one of the relevant variables. Another different wave speeds for the same value of T µ , by
experiment with 2 ropes of the same mass but different using different springs. It turns out that the tension T is
lengths, gives different wave speeds. Thus the rope not a relevant variable.
length L is also important. Further experiments indi-
cate that the speed of the pulse does not depend upon Compressional waves depend upon the stiffness of a
such variables as the color of the rope, the material from material, not the tension. For example a compressional
which it is constructed, or the time of day. Thus you sound pulse will travel down a steel rod whether or not
conclude that the relevant variables and their dimen- the rod is under tension. Pulling on the ends of the steel
sions are rod does not noticeably change the speed of the sound
pulse. Increasing the tension in a spring stretches the
kg m(meter)
T , m kg, L m (6) spring and therefore changes the mass per unit length.
sec2
 15-7

It is the change in mass per unit length, not the change a spring is proportional to kL and not just k, with the
in tension, which affects the speed of the compres- result that the speed of the pulse is related to the
sional pulse. stiffness, as we suspected.

What variable is related to the inherent stiffness of a To see why the inherent stiffness is related to kL,
spring? The one that comes to mind is the spring imagine that we wind a long spring and cut it in half to
constant k that appears in Hooke’s law create two identical springs of length L 1 . As shown in
F = – kx ; Figure (4a), if we apply a force F to one of the springs,
(9) and measure the distance ∆ x that the spring stretches,
k = F newtons
x meter Hooke′s law we can use Hooke’s law to calculate that the spring
constant k 1 is given by
The stiffer the spring, the greater the spring constant k. F
k1 = (11)
Suppose we decide, after enough experimentation, that ∆x
the relevant variables for the compressional pulse on a Now attach the two springs back together and stretch
spring are the spring constant k, spring mass m and the combination with the same force F as shown in
spring length L. We obtain the dimensions of k from Figure (4b). Since each spring feels the same force F,
Hooke’s law, each stretches a distance ∆ x , and the pair stretch a
kg ⋅ m sec 2 kg distance 2 ∆ x . Thus from Hooke’s law the k 2 of the
k newtons
meter = k m = k combination is given by
sec 2
thus we have to construct a quantity with dimensions F 1 F k1
k2 = = = (12)
m/sec from the variables 2∆x 2 ∆x 2
kg
k , M kg, L m where we used k 1 = F/ ∆ x from Equation 11.
sec 2
The only way we can do it is to divide k by m to get rid
of kilograms and multiply by L 2 to get L1
2 (unstretched)
kL m 2
m sec 2
(stretched) F
∆x
Taking the square root gives a quantity with the dimen- F = k 1 ∆ x ; k 1 = F/ x
sions of a velocity
Figure 4a
2 speed of a Measuring the spring constant of a spring.
kL kL
v = m = µ compressional (10)
wave on a spring

where again µ = m /L is the mass per unit length. This 2 L1

is our prediction for the speed of a compressional wave
(unstretched)
on a spring. The actual speed could differ by a constant
factor like 2, but it must have this functional depen- (stretched) F
dence if we are correct in our assumption that the only ∆x 2∆x
relevant variables are k, m and L.
F = k 2 (2 ∆x) ; k 2 = F/2 ∆x k2 = k1 /2
In the formula v = kL/ µ , the appearance of the Figure 4b
product kL, rather than k alone may at first seem Measuring the spring constant
surprising. But it turns out that the inherent stiffness of of two connected springs.
15-8 One Dimensional Wave Motion

When we attach two identical springs together, we end SPEED OF SOUND WAVES
up with a longer spring but we are not changing the The quantity kL that appeared in our formula for the
inherent stiffness of the spring. More importantly we speed of a wave pulse is essentially the stiffness of a unit
do not change the speed of the wave pulse. Connecting length spring. By stiffness we mean the ratio of the
the two springs and keeping the tension F the same force applied to stretch a unit length of spring, to the
merely gives the pulse a longer distance to travel. amount of stretch ∆ x that we get. The same ideas also
Note, however, that when we attach the two springs, the apply to stretching a steel rod or any one dimensional
spring constant is cut in half, but the length is doubled, object that has an elasticity and obeys Hooke’s law.
with the result that the product kL is unchanged. In engineering texts, the force applied to a unit length,
Explicitly we have area, or volume is given the generic name stress, and
the resulting displacement that the stress causes is
k1
k 2L2 = 2L 1 = k 1 L 1 (13) called a strain. The ratio of the stress to the strain, is
2 called the modulus. For our spring, the stress is the
It should now appear more reasonable that the speed of tension force F, and the strain is the change in length per
the wave pulse should be given by kL/ µ . Both the unit length, or ∆ x/L . The ratio of the stress to the strain,
quantity kL, and the mass per unit length µ are inherent F/(∆ x /L) = FL /∆ x is called Young’s modulus. From
properties of the spring that do not depend on the length Hooke’s law, F/∆ x = k , thus Young’s modulus is
of the spring. It is thus reasonable that the speed of the FL /∆ x = kL, the quantity we have been discussing.
wave pulse should also involve only these variables.
When we have a compressional wave in a gas, we can
think of the compression as being caused by a pressure
Project Suggestion pulse that travels through the gas. In the region where
We have spent some time discussing the speed of the gas is compressed, there is a slight excess pressure.
pulses on a stretched spring for two reasons. One is that
The speed of the wave pulse depends upon the response
we used these pulses as our main example of wave
of the gas to this excess pressure.
motion in our introduction to special relativity in Chapter
1. The second is that measuring the speed of pulses on Using the engineering terminology, the excess pres-
a spring makes a nice project, not much equipment is
sure ∆ P represents the stress and the fractional change
needed, and you can fairly easily measure the variables
in volume, ∆ V/V the corresponding strain. In this case
needed to test Equation 10.
the ratio of the stress ∆ P to the strain ∆ V/V is called the
An additional advantage is that Equation 10 might or bulk modulus B.The formula for B is thus
might not be right. Since it was derived by dimensional
analysis, it could be off by a constant factor like 1/2 or B = ∆ P/ ∆ V /V bulk modulus (15)
2π . Therefore you have the challenge of determining
In Chapter 17 we will discuss the concept of a pressure
whether or not there are some missing constant factors.
in a gas, and see how changes in pressure are related to
We expect that the wave speed should be proportional changes in volume. Until we get to that chapter, any
to kL / µ , and if this does not turn out to be correct, we detailed discussion of the concept of bulk modulus is
have made some mistake in our analysis of what vari- premature. What we will do now is assume that it is the
ables are important. For example, in our analysis, we bulk modulus B essentially represents the “stiffness” of
said nothing about the unstretched length L 0 of the the gas and should appear in the formula for the speed
spring. Should L 0 also appear in the formula for vwave ? of a sound wave. We will then use dimensional
The way to find out is to do some experiments.
analysis to figure out what the formula should be.
The experiments are made a bit easier by noting that

kL kL2 k
v = = = L
µ m m
where we used µ = m/L .
 15-9

In the ratio B = ∆ P/ ∆ V /V , the denominator ∆ V /V From this table we see that the speed of sound is
is dimensionless, thus B has the dimensions of pressure considerably higher in the light gasses helium and
which is a force per unit area. hydrogen than in the more dense gas air. (The differ-
kg m kg ence in density is why helium and hydrogen filled
B newtons 2 = B 2 2 = B (15) balloons float in air.) The compressibility or bulk
meter sec m sec 2 m
modulus is usually the same for all gases at a given
To construct a quantity involving B that has the dimen- pressure, thus the higher speed in hydrogen or helium
sions of a velocity, we have to get rid of the kilograms is due to the lower density, as we would expect from the
by dividing by some quantity related to the mass of the formula v = B ρ .
gas. The only reasonable choice is the gas density
ρ kg/m 3 , thus we now have Exercise 2
B kg/(sec 2 m) m2 Steel is much stiffer than aluminum. (You make much
= B
ρ (16)
ρ kg/m 3 sec 2 better springs from steel than aluminum.) Yet the speed
of sound is greater in aluminum than steel. Why?
which is the square of a velocity. Taking the square
root, we get Exercise 3
B speed of a You tap the end of a 10 meter long steel rod with a
v = ρ sound wave (17) hammer. How long before the tap can be detected at
the other end of the rod?
a result we stated back in Chapter 1.
Exercise 4
Equation 17 holds not only for a gas, but also for
A dimensional analysis problem
compressional sound waves in a liquid and a solid.
that you should attempt now.
Liquids and solids, being far more incompressible than
a gas, have a much greater bulk modulus B and When working with the theory of electric and magnetic
therefore higher speeds of sound. The speed of sound phenomena, using the MKS system of units, one en-
in various substances is given in Table 15-1. counters two rather mysterious constants labeled ε 0
(epsilon naught) and µ0 (mu naught). The constant ε 0
Substance Speed of Sound appears in the formula for electric forces, and µ0 in the
in meters/sec formula for magnetic forces. These constants have the
following dimensions
Gases (at atmospheric pressure)
2 2
Air at 0° C 331 coulomb seconds
Air at 20° C 343 ε0 3 (18)
kilogram meter
Helium at 20° C 965
Hydrogen at 20° C 1284
kilogram meter
Liquids µ0 2
(19)
coulomb
Water at 0° C 1402
Water at 20° C 1482 where a coulomb is a unit of electrical charge. The
Sea water at 20° C 1522 numerical values of ε 0 and µ0 are to be found on the
Solids inside cover of this text along with other important
Aluminum 6420 physical constants.
Steel 5941 (a) what combination of the constants ε 0 and µ0 have
Granite 6000 the dimensions of a velocity?
Nuclear matter near c
(b) from the numerical value of this velocity, what do you
Table 15-1 think it is the velocity of?
15-10 One Dimensional Wave Motion

LINEAR AND NONLINEAR Conceptually we can separate all wave motion into two
WAVE MOTION classes. There are the relatively smooth waves that can
pass through each other like the circular ripples of
Few sights are more awesome than the crashing of
Figure (1-2), and the relatively wild waves that crest,
ocean rollers on a rocky beach during a storm. The
crash and change their shapes. The relatively smooth
waves seen in Figure (3) of Chapter 1, produced by
waves all obey what is called a linear wave equation.
hurricane Bertha hundreds of miles out to sea, were
The properties of these waves are well understood and
crashing against the rocky shores of Mt. Desert Island,
their behavior easy to predict. The wild waves obey
Maine, in July 1990. Hundreds of tourists and local
nonlinear wave equations. We know very little about
television station reporters were at the beach to observe
the behavior of nonlinear waves, and in most cases find
the event. At one spot, called ‘Thunder Hole’, the
it very difficult to make predictions about their behav-
crashing waves created a loud boom and a geyser of
ior. (Ocean waves, for example, are linear until they
water that went 40 or 50 feet in the air.
start to crest. When you see whitecaps, the waves have
A very different sight are the circular ripples emerging become nonlinear.)
from where raindrops have hit a puddle of water, seen
In this text we will restrict our discussion to the smooth,
in Figure (1-2), reproduced here. The special feature of
linear waves that behave like the circular ripples.
these ripples is that they maintain their identity as they
Fortunately, most kinds of wave motion we encounter
move through each other. They are still circular waves
in nature, including almost all examples of light waves
even after moving through other waves.
and the probability waves of quantum mechanics, are
linear and therefore relatively easy to analyze. But
there are growing applications for nonlinear waves,
particularly in the field of laser optics.

a)

b)
Figure 1-3
This ocean wave from Hurricane Bertha (July 31, 1990).

c)

d)

Figure 5
Two wave crests running into each other
add up to produce a bigger crest.

Figure 1-2
Rain drops creating circular waves on a puddle.
 15-11

THE PRINCIPLE OF The concept that waves can maintain their identity as
SUPERPOSITION they move through each other, and that they produce a
Figure (1-2) illustrates one aspect of linear wave mo- resultant wave whose height or amplitude is the sum of
the heights or amplitudes of the individual waves, this
tion. The waves can move through each other and
concept is known as the principle of superposition. In
emerge undisturbed. The waves are still circular and
unbroken after they have crossed. more colloquial language, the principle of superposi-
tion says that waves add. The principle of superposi-
There is another simple feature of this wave motion that tion is one of the key concepts of linear wave motion.
is a bit harder to see from that picture. While the waves It distinguishes linear from nonlinear wave motion.
are crossing, they produce a wave whose height is the When nonlinear waves interact, you get something
sum of the heights of the individual waves. If two crests different than the simple sum of the two waves.
are moving through each other, the crests add to
Before leaving our discussion of the principle of super-
produce a higher crest. Two troughs produce a deeper
trough, and a crest and a trough will tend to cancel as position, we wish to take one further look at part (c) of
Figure (6). That is the point where an equal shaped
they move through each other.
crest and trough are right on top of each other, precisely
This adding of the heights of crossing waves is more cancelling each other out. This kind of cancellation of
easily illustrated for the case of one dimensional wave waves is a common feature of wave motion. In fact, it
pulses traveling down a rope. In Figure (5), two similar is what distinguishes wave motion from what we have
crests add together to produce a doubly high crest for an been calling particle motion. If two particles run into
instant. In Figure (6) we see that a similar shaped crest each other, they do not cancel like the waves of Figure
and trough will cancel at the instant they are together. (6). They bounce or crash but not cancel.
Figure (7) illustrates the idea that as any two wave
shapes move through each other, they produce a wave v
shape whose height at any point along the rope is the
v
sum of the heights of the individual waves moving
through each other.

a)

b)

c)

d)
Figure 7
In general, for linear wave motion. We obtain
Figure 6 the shape of the resulting wave by adding the
When a crest meets a trough, there is a amplitudes of the individual waves.
short time when the waves cancel.
15-12 One Dimensional Wave Motion

SINUSOIDAL WAVES down with a sinusoidal oscillation, a sinusoidal shaped

If you ask someone to describe wave motion, they are wave will start traveling down the rope, at a speed
likely to picture water waves and sketch a curve that vwave = T/µ . This creates an example of what is
looks like a sine wave. A sine wave represents just one called a traveling sine wave.
of many possible shapes for a wave. But it is an The problem with creating traveling sine waves on a
important shape because it is often seen in nature and rope, is that the wave reaches the end of the rope,
it is easy to handle mathematically. We will see shortly reflects, and moves back through the incoming wave,
that any arbitrary wave shape can be constructed from complicating the situation. A better example of travel-
sine waves, thus the sine wave can be thought of as a ing sine waves can be seen on the surface of a lake or
basic building block of wave motion. the ocean where there is plenty of room for the waves
To relate the mathematical sine function to wave to move before they strike an object or a shore.
motion, recall our definition of sine function shown There are two distinct ways to view a traveling sine
back in Figure (14-4). The point of that figure is that the wave. One is to move along with the wave. Then all
sine function is the sideways projection of circular you see is a stationary sinusoidal shape. The other is to
motion. As the arrow rotates at an angular velocity ω, stand still and let the wave pass by you. Then you will
the angle θ that the arrow has rotated increases as see the wave oscillate up and down as successive crests
θ = ωt . On the right we have graphed the height of the and troughs pass by you. This is illustrated in Figure (9)
rotating arrow as a function of the angle θ = ωt to where we have sketched a traveling sinusoidal water
obtain a sine curve. wave passing a fixed post in the water. If you move
To actually create the sine wave shape seen in Figure along with the wave, then the shape of the wave does
(14-4), you can start shaking one end of a long rope as not change. But if you look at the post, the level of the
shown in Figure (8). If you move your hand up and water is moving up and down with a sinusoidal oscil-
lation.

π/2
1
Figure 14-4
θ 3π
Definition of the π 2
π 0
0 π 2π
θ = ωt
function
2
sin θ = sin (ω t).
ω −1
3π/2

vwave
ω

Figure 8 x
Sine wave created on a
stretched rope.

λ
 15-13

Wavelength, Period, and Frequency As an example of using dimensions to derive a formula,

We usually describe a wave in terms of its wavelength let us see if we can get a formula for the frequency f of
λ , frequency f or period T. The easy way to remember a wave of wavelength λ . The idea of using dimensions
how to go back and forth between these quantities, is to is to try something, then see if the dimensions match. If
use dimensions. they don't match, change the formula until they do. As
a guess, let us try the formula
When we view the shape of a traveling sine wave, the f = vwaveλ guess
predominant feature is the wavelength, the distance λ
between crests shown in Figure (8). Considering that Putting in dimensions, we have
one full cycle of the wave fits between the crests, we
can assign the dimensions of meters per cycle to λ . cycles
f sec = vwave meters meters
sec × λ cycle
λ meters wavelength 2
cycle = vwave λ meters
sec cycle
When we let the wave pass by us and view the up and
down motion of the surface, we see an oscillation Clearly the dimensions do not match. We have to
whose period is T seconds per cycle and frequency is change the formula so that meters cancel and we get
f cycles per second. cycles upstairs. This can be done if we move λ
downstairs, giving
Since the period T is the length of time it takes one
wavelength λ of the wave to pass by at a speed vwave , cycles vwave meters/sec
we have (distance = speed times time) f sec =
λ meters/cycle

λ meters = vwave meters T second (20) =

vwave cycles
cycle second cycle (21)
λ sec
By assigning the dimensions meter/cycle to λ and which works. Thus the correct formula is f = vwave/λ,
sec/cycle to T, we can get the relationship λ = vwaveT
a result that can be a bit tricky to figure out other ways.
from dimensions without having to memorize formu-
las, or even having to think very much. vwave
post
ωt=0 0
λ
Figure 9
π
Traveling sine wave on the surface ωt= 2
of water. The sine wave shape
moves as a unit along the surface
at a speed vwave . But if we look at a ωt= π
fixed post in the water, the water
level at the post oscillates up and
down with a sinusoidal oscillation. 3π
ωt= 2

ωt= 2π
15-14 One Dimensional Wave Motion

Angular Frequency ω Spacial Frequency k

In Figure (14-4), we reminded ourselves that a sinusoi- When we let a traveling wave pass by us, we observe
dal oscillation is equivalent to the sideways view of a sinusoidal oscillation in time. This oscillation can be
circular motion. If the vector on the left side of Figure described in terms of the number of seconds in each
(14-4) is rotating with an angular velocity ω radians cycle (T seconds/cycle), in terms of the frequency
per cycle, we get one rotation, one period T of the (f cycles/second) or the angular frequency
oscillation, when the angle θ = ωt goes from 0 at t = 0, (ω radians/second) .
to 2π at t = T. Thus at the end of one period,
θ = ωT = 2π . If instead we look at the whole wave at one instant of
time, or move along with the wave, we see a sinusoidal
Again we can avoid memorizing new formulas by oscillation in space. We have described this spacial
using dimensions. We note that 2π is the number of oscillation in terms of the wavelength, the number of
radians in a full circle or cycle. Thus we will assign to meters in each cycle ( λ meters/cycle). What we are
it the dimensions missing is a spacial analogy to frequency, the number
of cycles or radians per meter.
radians
2π (23) By dimensions we immediately see that
cycle
1 cycles
= 1 meter
Then to find the formula, for example, for the wave's
λ meters λ
angular frequency ω in terms of the wave's period T, cycle
we have
is the special analogy to the time frequency f, and that
2π radians / cycle
ω radians
sec =
T sec / cycle 2π radians/cycle 2π radians
=
(23) λ meters/cycle λ meter
2π radians
= sec
T
is the spacial analogy to the angular frequency ω .
Exercise 5 (Do this one now.)
In physics texts, it is not common to use a special
For a traveling sine wave moving at a speed vwave , use
symbol to designate the spacial frequency 1/λ
dimensions to find
(cycles/meter), but it is standard practice to designate
(a) λ in terms of vwave and ω the angular spacial frequency 2π/λ (radians/meter) by
the letter k
(b) vwave in terms of λ and T
spacial
(c) T in terms of λ and ω k radians 2π radians
meter ≡ λ meter frequency
(24)
(d) f in terms of ω
 15-15

The standard name for k is the lackluster expression Exercise 6 (Try this now.)
wave number, which says very little about the quantity.
You have a traveling sine wave moving at a speed vwave .
Instead we will refer to k as the spacial frequency of the
Using dimensions find the formula for vwave in terms of
wave. The higher the spacial frequency k, the more the wave's time frequency ω and spacial frequency k.
radians we get in a meter, just as the higher the time
frequency ω , the more radians we get in one second.

When you first study wave motion, it may be an

irritating complication to have two kinds of frequency,
f cycles/sec and ω radians/sec (or two spacial frequen-
cies 1/λ cycles/meter and k radians/meter). Why not
stick with cycles which are much easier to visualize
than radians? The answer is that in the formulas for sine
waves, the sine function basically has to be expressed y
in terms of an angle, as in sin θ , and radians are an
t=T
angle. To convert time t to an angle, we multiply by
ω as in π 2π ωt

θ radians = ω radians
sec × t sec
ωT = 2π
(one period)
= ωt radians (25a)
a) Time: sin(θ) = sin(ωt)
while to convert the distance x to an angle we multiply
by k as in y
x=λ
θ radians = k radians
meter × x meters
π 2π kx
= kx radians (25b)
kλ = 2π
Using Equations 25a or 25b, we can express the single (one wavelength)
function sin θ either as sin ωt , a sine wave in time
shown in Figure (10a), or as sin kx, a sine wave in space b) Space: sin(θ) = sin(kx)
shown in Figure (10b). From these graphs we can see
that when ωt gets up to 2π , we have one period T, and Figure 10
when kx gets up to 2π we have one wavelength λ . Sine waves in time and space.
15-16 One Dimensional Wave Motion

Traveling Wave Formula Comparing Equations (28) and (29), we see that the
Thus far we have formulas for a time varying sine wave point θ = 0 moves along the x axis at a speed
sin ωt , and a space varying wave sin kx. Now we want
a formula for a traveling sine wave whose amplitude vwave = ω (30)
varies in both space and time. The answer turns out to k
be If you did Exercise 6, you recognize that the quantity
ω/k has the dimensions of a velocity
traveling
y = sin θ = sin kx – ωt sine wave (26)
ω radian/sec = ω meter = v (31)
What we will do is show that this formula represents a k radian/meter k sec wave

sine wave moving down the x axis. Thus the origin does move down the x axis at a speed
vwave , and the formula kx – ωt , is our desired trav-
Figure (11) shows a sinusoidal shape that is moving
eling wave formula.
down the x axis at a speed vwave . If we describe the
wave by the function sin θ , then it is the origin
sin θ = 0 that moves down the axis at a speed vwave . Exercise 7
Thus what we need is a formula for θ so that when we Explain what kind of a wave is represented by the
set θ = 0 , that point does move down the x axis at the formula y = sin kx + ωt .
desired speed.
y = sin(θ)
The answer we gave in Equation 26 suggests that the
correct formula for θ is vwave

θ = kx – ωt (27)
θ =0
Setting θ = 0 we get

θ = 0 = kx – ωt ; kx = ωt vwave

x = ωt (28) θ =0
k
But if the θ = 0 point travels at a speed vwave , then after
a time t, it has traveled a distance x given by vwave t vwave

x = vwavet (29) θ =0

Figure 11
The cycle begins at θ = 0.
 15-17

Phase and Amplitude y = sin (θ)

The Equation 26 for a traveling wave can be general-
ized by noting that the wave can have an arbitrary
(θ) = 0
amplitude A, and an arbitrary constant phase angle φ
to give θ
0
vwave
y = A sin kx – ωt + φ (32)

The amplitude A just makes the sine wave bigger or y = sin (θ + φ)

smaller, and the phase angle φ shifts the sine wave to
the left or right.
(θ + φ) = 0
To see precisely how the phase angle φ shifts the sine θ
wave, we have in Figure (12) compared sin θ and –φ 0
sin θ + φ . The function sin θ + φ crosses zero vwave
when the angle θ + φ = 0 or at θ = – φ . Thus adding
a phase angle φ shifts the sine wave back a distance
– φ radians. If, for example, we set φ = π/2 , the wave Figure 12
is shifted back 1/4 of a wavelength, and we have Adding a phase angle φ shifts
converted a sine wave into a cosine wave. the wave back a distance φ .
15-18 One Dimensional Wave Motion

STANDING WAVES The two kinds of waves on a rope, the traveling wave
In addition to the waves traveling down a rope, another of Figures (8) and (9) and the standing wave of Figure
kind of wave pattern that is easy to achieve are those (13) are closely related to each other. A careful
shown in Figure (13). All you have to do is shake the demonstration shows how traveling waves can turn
end of the rope at the right frequency and one of these into standing waves.
waves will appear. Change the frequency and you can If you start shaking a rope a traveling wave starts down
change to one of the other patterns. the rope as shown in Figure (8). After a while the wave
The waves in Figure (13) are called standing waves reaches the other end of the rope, is reflected, and starts
because the pattern does not move along the rope. The moving back the other way. This reflection is most
points of zero amplitude, the points called nodes of the easily seen if you send a single pulse down the rope so
wave, stay at fixed positions while the rope between that you can see it bounce off the fixed end and come
nodes oscillates back and forth. back to you.

The difference between the wave patterns is character- If you send a series of pulses down the rope, if you
ized by the number of nodes. In Figure (13) all the create a traveling sine wave, then the reflected pulses
waves have nodes at the ends, and there are zero, one, have to move back through the pulses that are still
two and three nodes in between as we go from the left coming in. You now have the superposition of two
to the right pattern. traveling waves moving through each other in opposite
directions.

Figure 13
Standing waves on a rope.
 15-19

You might think that the sum of two traveling waves In the first frame t = 0, the two waves match exactly,
moving through each other could lead to a complex producing a sum that has twice the amplitude of either
pattern, and when you try it in a demonstration it often traveling wave. A quarter of a period later, at t = T/4,
looks that way. The problem with a demonstration is the traveling waves are precisely opposite each other.
that the reflected wave reaching your hand can reflect And the sum is zero all along the wave. This never
again and you begin to build up a mixture of many happens in a traveling wave. A traveling wave is never
waves. completely flat, there is always a crest moving along.
If you do the demonstration carefully, however, you At time t = T/2, half a period later, the traveling waves
can observe a simple result. The sum of the traveling again line up producing a wave of twice the amplitude.
wave and the reflected wave, moving through each Note now that the points in the sum wave that were
other, is a standing wave. The addition of two travel- below the axis at t = 0 are now above the axis at
ing waves of equal amplitude and wavelength moving t = T/2, and vice versa. At time t = 3T/4 the traveling
in opposite directions through each other is illustrated waves are again out of phase and add up to zero. At
in Figure (14). The two traveling waves are shown on t = T, we are back to where we started.
lines (a) and (b) at five different times t = 0, T/4, T/2,
3T/4 and T, where T is the period of the waves. On the From line (c) of Figure (14), we see that the nodes of the
bottom line (c) we have added the amplitudes of the two sum wave remain stationary and the rope between the
traveling waves to get a picture of the amplitude of the nodes oscillates up and down. This is exactly what we
resulting wave. see in the photographs of the standing rope waves in
Figure (13).

t=0 t = T/4 t = T/2 t = 3T/4 t=T

(a)

(b)

(c)

Figure 14
Making a standing wave out of two traveling waves on a rope. In the top line (a) we show a traveling
wave moving to the right. A section of rope is shown at times t = 0, T/4, T/2, 3T/4 and T, where T is the
period of the wave. In (b) we have, in the same section of rope, five views of a traveling wave moving to
the left. In (c), we have added the two traveling waves and get a standing wave with stationary nodes. To
demonstrate the addition, at the center of each time frame we have drawn arrows to show the height of
the wave at that point. The length of the bottom arrow is the sum of the lengths of the upper two arrows.
To add two waves, you add up the heights at each point along the wave.
15-20 One Dimensional Wave Motion

Using a trigonometric identity, we can show math- WAVES ON A GUITAR STRING

ematically that the sum of two traveling waves moving Perhaps the clearest example of standing waves are the
through each other creates a standing wave. The waves on the strings of a stringed instrument such as the
formula for a sine wave traveling to the right is from guitar. The advantage of working with these waves is
Equation 26. that you get to both see the shape of the wave and hear
y moving right = A sin kx – ωt (26) its frequency.

If you worked Exercise 6, you found that the formula The shape of guitar string waves are the same as the
for a similar wave moving left is standing rope waves in Figure (13). In Figure (15), we
have sketched the allowed standing wave patterns on a
y moving left = A sin kx + ωt (33) string of length L. Because the string is fixed at the
ends, we can only have waves with nodes at the ends.
The formula for the sum of the two waves is
bridge string nut
y sum wave = A sin kx – ωt + A sin kx + ωt

The trigonometric identity we will use is

sin a ± b = sin a cos b ± cos a sin b (34)
first
This gives harmonic or λ 1 = 2L
fundamental 1
L
sin kx – ωt = sin kx cos ωt – cos kx sin ωt

sin kx + ωt = sin kx cos ωt + cos kx sin ωt

second λ 2 = 2L
harmonic 2
Add these two sine waves, the cos kx sin ωt terms
cancel and we are left with
y sum wave = 2A sin kx cos ωt (35)
third λ 3 = 2L
harmonic 3
To interpret Equation 35, write it in the form
y sum wave = A x cos ωt (36a)

where the x dependent amplitude A x is fourth λ 4 = 2L

harmonic 4
A x = 2A sin kx (36b)

Equation 36a tells us that the entire wave is oscillating

λ 5 = 2L
in time as cos ωt . However the amplitude of the fifth
harmonic 5
oscillation depends upon the position x long the wave.
Equation (36b) tells us how the amplitude varies with
position. It varies sinusoidally as sin kx, with nodes
permanently located at the points where sin kx = 0. This n th λ n = 2L
harmonic n
sinusoidal variation in the amplitude along the wave is
clearly seen in the photographs of the standing waves
on the rope, Figure (13). Figure 15
Allowed standing waves on a guitar string. The
formula for the wavelength of the nth harmonic
is seen to be λ n = 2 L / n .
 15-21

The wave with no nodes between the ends is called the can be made from two traveling waves moving through
fundamental or first harmonic. Its wavelength is 2L, each other. The resulting standing wave has the same
twice the length of the string. In the second harmonic, frequency and wavelength as the traveling wave, thus
with one node in the middle, a full wave fits on the we can use the traveling wave formulas to determine
string at one time and we have λ 2 = L . Each time we the frequency of the standing wave.
add a node we go up to one higher harmonic. (The
second harmonic is also called the first overtone, etc.) Using dimensions, we see that a traveling wave of
wavelength λ cm/cycle , traveling at a speed v cm/sec,
The general formula for the wavelength of the nth has a frequency f cycles/second given by
harmonic can be seen is we write the progression of
wavelengths in the form cycles
f sec = v meter 1
sec × λ meter/cycle
2L 2L 2L
λ1 = ; λ2 = ; λ3 = , etc. cycles
1 2 3 = v sec (38)
λ
It is clear that the formula for λ n is simply
The speed v of a transverse wave on a string, that has
2L wavelength of a mass per unit length µ, and tension T, is from Equation
λn = the nth harmonic
(37) 5
n

The easiest way to remember Equation 36 is draw a T

sketch of the allowed standing waves and write down vwave = (5)
µ
the progression λ 1 = 2L /1, λ 2 = 2L /2, etc.
Using Equation 5 in Equation 38 gives us as the
Frequency of Guitar String Waves formula for the frequency of the traveling wave
What notes do you hear when you pluck a guitar string?
That depends very much upon how you pluck it. 1 T
Usually when you pluck the string, you create a number f = (39)
λ µ
of the standing wave patterns at one time, and the note
you hear is a rich mixture of the frequencies of the The same Equation 39 must also apply to the standing
individual waves. waves on the guitar string. We get for the frequency fn
of the nth harmonic, which has a wavelength fn ,
With care, however, you can pluck the string so that
most of the vibration is in one of the harmonics. A
gentle pluck at the center of the string will excite mostly 1 T frequency of the
the fundamental or first harmonic. Pluck the string 1 4 fn = (40)
λn µ nth harmonic
of the way from one end and briefly place your finger
at the center of the string to create a node there. This
way you can excite mostly the second harmonic. You For anyone who has tuned a guitar, Equation 40 makes
will then notice that the sound of the note is one octave a lot of sense. First note that when you go from the first
above the sound of the fundamental. Accomplished harmonic λ 1 = 2L, to the second harmonic λ 2 = L, the
guitar players can selectively excite still higher har- wavelength is cut in half and the frequency doubles. A
monics. doubling of the frequency of a note corresponds to
going up one octave.
What are the frequencies of oscillations of these vari-
ous standing wave patterns? We can answer this When you are tuning the guitar, you raise the frequency
question because of our knowledge that standing waves of a string by tightening it and increasing the tension.
That is also predicted by Equation 40.
15-22 One Dimensional Wave Motion

On a guitar or most stringed instruments the low notes Exercise 9

are played on fat wires that have a greater mass per unit One end of a wire is attached to a post as shown in
length µ than the skinny wires used for the high notes. Figure (16). The wire is then run over a pulley where a
The reason for using the fat wire is that you can increase mass m is hung on the other end. The distance d from
the tension and still keep the frequency down. The the post to the pulley is 1 meter and the mass of one
more tension in the wire and the more mass in the wire, meter length of the wire is 5 grams.
the more energy you can store in the wire and the louder
(a) How big a mass m must be hung on the wire in order
the sound you can produce. It is hard to get as much
to get the wire to vibrate in its fundamental mode at a
sound out of the low frequency strings and you need all
frequency of 440 cycles/second, which is middle A?
the help you can get. (Answer: 395.10 kg).

Exercise 8 (b) Describe four distinct ways one could double the
frequency of oscillation of the wire.
You have a guitar string of length L, with a tension T and
a mass per unit length µ. (L is the distance from the nut (c) How much mass would you have hung on the wire
to the bridge.) in Figure (16) to get the wire to oscillate in its fundamen-
(a) What is the f frequency of the fundamental mode of tal mode at a frequency two octaves above middle A?
vibration? Express your answer in terms of L, T, and µ. (Answer 6321.63 kg.)

(b) Show that the nth harmonic has a frequency n times

as great as the fundamental. Sound Produced by a Guitar String
When you pluck a guitar string, the standing wave on
the string produces a traveling sound wave in the air.
This is analogous to plucking the end of a rope to
produce a traveling wave along the rope as illustrated
in Figure (8b). The wavelength of the sound wave is
determined by the frequency of oscillation of the string
and the speed of sound. (It is not the same as the
d wavelength of standing waves on the string.)

Exercise 10
A guitar string is tuned to oscillate at a frequency of 440
Figure 16 cycles/second in its fundamental mode. What are the
An easy way to adjust the wavelengths of the sound waves produced by the first
tension in a vibrating string three harmonics
m
(a) in air at 20° C

(b) in helium at 20° C

Chapter 16
Fourier Analysis,
Normal Modes and Sound

CHAPTER 16 FOURIER ANALYSIS, NOR-

MAL MODES AND SOUND

In Chapter 15 we discussed the principle of superposi- This example demonstrates how the ability to separate
tion—the idea that waves add, producing a composite a complex waveform (in this case white sunlight) into
wave that is the sum of the component waves. As a its component wavelengths or frequencies, can be a
result, quite complex wave structures can be built from powerful research tool. The sounds we hear, like those
relatively simple wave forms. In this chapter our focus produced by an orchestra, are also a complex mixture
will be on the analysis of complex wave forms, finding of waves. Even individual instruments produce com-
ways to determine what simple waves went into con- plex wave forms. Our ears are very sensitive to these
structing a complex wave. wave forms. We can distinguish between a note played
on a Stradivarius violin and the same note played by the
As an example of this process, consider what happens same person on an inexpensive violin. The only
when white sunlight passes through a prism. White difference between the two sounds is a slight difference
light is a mixture of all the colors, all the wavelengths in the mixture of the component waves, the harmonics
of the visible spectrum. When white light passes through present in the sound. You could not tell which was the
a prism, a rainbow of colors appears on the other side. better violin by looking at the waveform on an oscillo-
The prism separates the individual wavelengths so that scope, but your ear can easily tell.
we can study the composition of the white light. If you
look carefully at the spectrum of sunlight, you will You can hear these subtle differences because the ear
observe certain dark lines; some very specific wave- is designed in such a way that it separates the complex
lengths of light are missing in light from the sun. These incoming sound wave into its component frequencies.
wavelengths were absorbed by elements in the outer The information your brain receives is not what the
atmosphere of the sun. By noticing what wavelengths shape of the complex sound wave is, but how much of
are missing, one can determine what chemical ele- each component wave is present. In effect, your ear is
ments are in the sun’s atmosphere. This is how the acting like a prism for sound waves.
element Helium (named after Helios, Greek for sun)
was discovered.
16-2 Fourier Analysis, Normal Modes and Sound

When we study sound in the laboratory, the usual under a curve, a process known in calculus as integra-
technique is to record the sound wave amplitude using tion. While the idea of measuring the area under a
a microphone, and display the resulting waveform on curve is not a difficult concept to grasp, the actual
an oscilloscope or computer screen. If you want to, you process of doing this, particularly for complex wave
can generate more or less pure tones that look like sine shapes, can be difficult. Everyone who takes a calculus
waves on the screen. Whistling is one of the best ways course knows that integration can be hard. The
to do this. But if you record the sound of almost any integrals involved in Fourier analysis, particularly the
instrument, you will not get a sine wave shape. The analysis of experimental data are much too hard to do
sound from virtually all instruments is some mixture of by hand or by analytical means.
different frequency waves. To understand the subtle
differences in the quality of sound of different instru- People find integration hard to do, but computers
ments, and to begin to understand why these differ- don’t. With a computer one can integrate any experi-
ences occur, you need to be able to decompose the mental wave shape accurately and rapidly. As a result,
complex waves you see on the oscilloscope screen into Fourier analysis using a computer is very easy to do. A
the individual component waves. You need something particularly fast way of doing Fourier analysis on the
like an ear or a prism for these waves. computer was discovered by Cooley and Tukey in the
1950s. Their computer technique or algorithm is
A way to analyze complex waveforms was discovered known as the Fast Fourier Transform or FFT for
by the French mathematician and physicist Jean short. This algorithm is so commonly used that one
Baptiste Fourier, who lived from 1786 to 1830. Fou- often refers to a Fourier transform as an FFT.
rier was studying the way heat was transmitted through
solids and in the process discovered a remarkable The ability to analyze data using a computer, to do
mathematical result. He discovered that any continu- things like Fourier analysis, has become such an
ous, repetitive wave shape could be built up out of important part of experimental work that older tech-
harmonic sine waves. His discovery included a math- niques of acquiring data with devices like strip chart
ematical technique for determining how much of each recorders and stand alone oscilloscopes have become
harmonic was present in any given repetitive wave. obsolete. With modern computer interfacing tech-
This decomposition of an arbitrary repetitive wave niques, important data is best recorded in a computer
shape into its component harmonics is known as for display and analysis. We have developed the
Fourier analysis. We can think of Fourier analysis MacScope™ program, which will be used often in this
effectively serving as a mathematical prism. text, for recording and displaying experimental data.
The main reason for writing the program was to make
The techniques of Fourier analysis are not difficult to it a simple and intuitive process to apply Fourier
understand. Appendix A of this chapter is a lecture on analysis. In this chapter you will be shown how to use
Fourier analysis developed for high school students this program. With the computer doing all the work of
with no calculus background (explicitly for my the analysis, it is not necessary to know the mathemati-
daughter’s high school physics class). To apply Fou- cal processes behind the analysis, the steps are dis-
rier analysis you have to be able to determine the area cussed in the appendix. But a quick reading of the
appendix should give you a feeling for how the process
works.
 16-3

HARMONIC SERIES To begin our discussion of Fourier analysis and the

We begin our discussion with a review of the standing building up of waveforms from a harmonic series, we will
waves on a guitar string, shown in Figure (15-15) first study the motion of two air carts connected by
reproduced here. We saw that the wavelengths λ n of springs and riding on an air track. We will see that these
the allowed standing waves are given by the formula coupled air carts have several distinct modes of motion.
Two of the modes of motion are purely sinusoidal, with
λ n = 2L
n (15-37) precise frequencies. But any other kind of motion
appears quite complex.
Where L is the length of the string, and n takes on
integer values n = 1, 2, 3, ..... However, when we record the complex motion, we
discover that the velocity of either cart is repetitive. A
Each of these standing waves has a definite frequency graph of the velocity as a function of time gives us a
of oscillation that was given in Equation 15-38 as continuous repetitive wave. According to Fourier’s
theorem, this waveform can be built up from sinusoidal
cycles v meters sec
fn sec = waves of the harmonic series whose fundamental fre-
λ n meters cycle
quency is equal to the repetition frequency of the wave.
(15-38)
v cycles When we use Fourier analysis to see what harmonics are
= involved in the motion, we will see that the apparent
λ n sec
complex motion of the carts is not so complex after all.
If we substitute the value of λ n from Equation 15-37 bridge string nut
into Equation 15-38, we get as the formula for the
corresponding frequency of vibration
v v v
f n = wave = wave = n wave (1)
λn 2L n 2L first
harmonic or λ 1 = 2L
For n = 1, we get fundamental
L
1

v
f1 = wave (2)
2L second λ 2 = 2L
All the other frequencies are given by harmonic 2

fn = n f 1 harmonic series (3)

third λ 3 = 2L
harmonic 3
This set of frequencies is called a harmonic series. The
fundamental frequency or first harmonic is the fre-
quency f1 . The second harmonic f2 has twice the fourth λ 4 = 2L
harmonic 4
frequency of the first. The third harmonic f3 has a
frequency three times that of the first, etc. Note also that
the fundamental has the longest wavelength, the second fifth λ 5 = 2L
harmonic has half the wavelength of the fundamental, harmonic 5

the third harmonic one third the wavelength, etc.

n th λ n = 2L
It was Fourier’s discovery that any continuous repetitive harmonic n
wave could be built up by adding together waves from a
harmonic series. The correct harmonic series is the one
where the fundamental wavelength λ 1 is equal to the Figure 15-15
Standing waves on a guitar string.
period over which the waveform repeats.
16-4 Fourier Analysis, Normal Modes and Sound

NORMAL MODES OF OSCILLATION Things become more interesting if we place two carts
The reason musical instruments generally produce on the air track, connected by springs as shown in
complex sound waves containing various frequency Figure (2). We will call this a system of two coupled
components is that the instrument has various ways to air carts.
undergo a resonant oscillation. Which resonant oscil- When we analyzed the one cart system, we found that
lations are excited with what amplitudes depends upon the force on the cart was simply F = –kx, where x is the
how the instrument is played. displacement of the cart from its equilibrium position.
In Chapter 14 we studied the resonant oscillation of a With two carts, the force on one cart depends not only
mass suspended from a spring, or equivalently, of a cart on the position of that cart, but also on how far away the
on an air track with springs attached to the end of the other cart is. A full analysis of this coupled cart system,
cart, as shown in Figure (1). This turns out to be a very using Newton’s second law, leads to a pair of coupled
simple system—there is only one resonant frequency, differential equations whose solution involves matri-
given by ω = k m . The only natural motion of the ces and eigenvalues. In this text we do not want to get
mass is purely sinusoidal at the resonant frequency. into that particular branch of mathematics. Instead we
This system does not have the complexity found in will study the motion of the carts experimentally, and
most musical instruments. find that the motion, which at first appears complex,
can be explained in simple terms.
ω
In order to record the motion of the aircarts, we have
air cart mounted the velocity detector apparatus shown in
Figure (3). The apparatus consists of a 10 turn wire coil
Figure 1 mounted on top of the cart, that moves through the
Cart and springs on an air track. magnetic field of the iron bars suspended above the
coil. The operation of the velocity detector apparatus
depends upon Faraday’s law of induction which will be
discussed in detail in Chapter 30 on Faraday’s law. For
Figure2 now all we need to know is that a voltage is induced in
System of coupled air carts. the wire coil, a voltage whose magnitude is propor-
tional to the velocity of the cart. This voltage signal

Figure 3
Recording the velocity
of one of the aircarts. A
10 turn coil is mounted
on top of one of the
carts. The coil moves
through the magnetic
field between the angle
irons, and produces a
voltage proportional to
the velocity of the cart.
This voltage is then
recorded by the
Macintosh oscilloscope.
 16-5

from the wire coil is carried by a cable to the Macintosh carts come together and go apart as if one cart were the
oscilloscope where it is displayed on the computer mirror image of the other. This motion of the carts is
screen. illustrated in Figure (4a). We will call this the vibra-
tional mode of motion.
When you first start observing the motion of the
coupled aircarts, it appears chaotic. One cart will stop In Figure (4b) we have used the velocity detector to
and reverse direction while the other is moving toward record the motion of one of the coupled air carts when
or away from it, and there is no obvious pattern. But the carts are moving in the vibrational mode. You can
after a while, you may discover a simple pattern. If you see that the curve closely resembles a sine wave.
pull both carts apart and let go in just the right way, the

air cart air cart air cart air cart

air cart air cart air cart air cart

air cart air cart air cart air cart

air cart air cart air cart air cart

air cart air cart air cart air cart

Figure4a Figure 5a
Vibrational motion of the coupled air carts. Sloshing mode of motion of the coupled aircarts.

Figure 4b Figure 5b
Vibrational mode of oscillation of the coupled A pure sloshing mode is harder to get. Here we
aircarts. The voltage signal is proportional to came close, but it is not quite a pure sine wave.
the velocity of the cart that has the coil on top.
16-6 Fourier Analysis, Normal Modes and Sound

If you play around with the carts for a while longer, you FOURIER ANALYSIS
will discover another way to get a simple sinusoidal As we mentioned in the introduction, Fourier analysis
motion. If you pull both carts to one side, get the is essentially a mathematical prism that allows us to
positions just right, and let go, the carts will move back decompose a complex waveform into its constituent
and forth together as illustrated in Figure (5a). We will pure frequencies, much as a prism separates sunlight
call this the sloshing mode of motion of the coupled air into beams of pure color or wavelength. We have just
carts. In Figure (5b) we have recorded the velocity of studied the motion of coupled air carts, which gave us
one of the carts in the sloshing mode, and see that the an explicit example of a relatively complex waveform
curve is almost sinusoidal. to analyze. While the two carts can oscillate with
In general, the motion of the two carts is not sinusoidal. simple sinusoidal motion in the vibrational and slosh-
For example, if you pull one cart back and let go, you ing modes of Figures (4) and (5), in general we get
get a velocity curve like that shown in Figure (6). If you complex patterns like those in Figures (6) and (7).
start the carts moving in slightly different ways you get What we will see is that, by using Fourier analysis, the
differently shaped curves like the one seen in Figure waveforms in Figures (6) and (7) are not so complex
(7). Only the vibrational and sloshing modes result in after all.
sinusoidal motion, all other motions are more complex. The MacScope program was designed to make it easy
To study the complex motion of the carts, we will use to perform Fourier analysis on experimental data. The
the techniques of Fourier analysis. MacScope tutorial gives you considerable practice
using MacScope for Fourier analysis. What we will do
here is discuss a few examples to see how the program,
and how Fourier analysis works. We will then apply
Fourier analysis to the curves of Figures (6) and (7) to
see what we can learn. But first we will see how
MacScope handles the analysis of more standard curves
like a sine wave or square wave.

Figure 6 Figure 7
Complex motion of the coupled air carts. Another example of the complex
motion of the coupled air carts.
 16-7

Analysis of a Sine Wave The selection rectangle is obtained by holding down

In Figure (8a) we attached MacScope to a sine wave the mouse bottom and dragging across the desired
generator and recorded the resulting waveform. The section of the curve. The starting point of the selection
sine wave generator, which is usually called a signal rectangle can be moved by holding down the shift key
generator, is an electronic device that we will use while moving the mouse. The data box shows the
extensively in laboratory work on the electricity part of period T of the selected rectangle, and the correspond-
the course. Typically the device has a dial that allows ing frequency f. If you wish the data box to remain after
you to select the frequency of the wave, a knob that the selection is made, hold down the option key when
allows you to adjust the wave amplitude, and some you release the mouse button. This immediately gives
buttons by which you can select the shape of the wave. you the ImageGrabber™, which allows you to select
Typically you can choose between a sine wave shape, any section of the screen to save as a PICT file for use
, a square wave shape in a report or publication.
and a triangular wave .
In Figure (9), we went up to the Analyze menu of
In Figure (8a) we have selected one cycle of the wave MacScope and selected Fourier Analysis. As a result
and see that the frequency of the wave is 1515 Hz we get the window shown in Figure (10). At the top we
(1.515 KHz), which is about where we set the fre- see the selected one cycle of a sine wave. Beneath, we
quency dial at on the signal generator. To get Figure see a rectangle with one vertical bar and a scale labeled
(8b) we pressed the Expand button that appears once Harmonics. The vertical bar is in the first position,
a section of curve has been selected. This causes the indicating that the section of the wave which we
selected section of the curve to fill the whole display selected has only a first harmonic. Beneath the ex-
rectangle.

Figure 8a Figure 8b
Sine wave from a signal generator. Selecting one cycle Once we have selected a section of curve, a new control
of the wave, we see that the frequency of the wave is labeled Expand appears. When we press the Expand
about 1.5 kilocycles (1.515 KHz). (Here we also see button, the selected section of the curve fills the entire
some of the controls that allow you to study the display rectangle as seen above. (The control then
experimental data. The scrollbar labeled “S” lets you becomes a Reset button which takes us back to the full
move the curve sideways. The “T” scrollbar changes curve.)
the time scale, and the “O” scrollbar moves the curve
up and down. In the text, we will usually not show
controls unless they are important to the discussion.)
16-8 Fourier Analysis, Normal Modes and Sound

panded curve, there is a printout of the period and The basic function of the Fourier analysis program in
frequency of the selected section. Here we see again MacScope is to determine how you can construct the
that the frequency is 1.515 KHz. selected section of the wave from harmonic sine
waves. The first harmonic has the frequency of the
For Figure (11), we pressed Reset to see the full curve, selected section. For example, suppose that you had a
selected 4 cycles of the sine wave, and expanded that. 10 cycle per second sine wave and selected one cycle.
In the lower rectangle we now see one vertical bar over That selection would have a frequency of 10 Hz and a
the 4th position, indicating that for the selected wave period of 0.1 seconds. The second harmonic would
we have a pure fourth harmonic. In the MacScope have a frequency of 20 Hz, and the third harmonic 30
tutorial we give you a MacScope data file for a section Hz, etc. The nth harmonic frequency is n times greater
of a sine wave. Working with this data file, you should than the first harmonic. Fourier’s discovery was that
find that if you select one cycle of the wave, anywhere any continuous repeating wave form can be con-
along the wave, you get an indication of a first har- structed from the harmonic sine waves. So far we have
monic. Select two cycles, and you get an indication of considered only the obvious examples of sections of a
a pure second harmonic, etc. pure sine wave. We will now go on to more complex
examples to see how a waveform can be constructed by
adding up the various harmonics.

Figure 9
Choosing Fourier Analysis.

Figure 10 Figure 11
In the MacScope program, Fourier Analysis acts only When 4 cycles are selected, the wave form consists
on the selected section of the curve. Since we selected of a pure 4th harmonic. We see the period of the
only one cycle, we see only a first harmonic. selected section of wave, which is 4 times longer
than one cycle. This makes the frequency 1/4 as
high. (The difference between 1515 Hz /4 and 390.6
Hz indicates the accuracy of graphical selection.)
 16-9

Analysis of a Square Wave shows in the upper rectangle. In the lower rectangle,
The so called square wave , whose which we will now call the FFT window, we see that
shape is shown in Figure (12), is commonly used in this section of the square wave is made up from various
electronics labs to study the response of various elec- harmonics. The MacScope program calculates 128
tronic circuits. The waveform regularly jumps back harmonics, but we have clicked three times on the
and forth between two levels, giving it a repeated Scale button to expand the harmonics scale so that we
rectangular shape. The ideal mathematical square can study the first 16 harmonics in more detail.
wave jumps instantaneously from one level to another.
The square waves we study in the lab are not ideal; In Figure (14) we have clicked on the first bar in the FFT
some time is always required for the transition. window, the bar that represents the amplitude of the
first harmonic. In the upper window you see one cycle
It is traditional to use the square wave as the first of a sine wave superimposed upon the square wave.
example to show students how a complex wave form This is a picture of the first harmonic. It represents the
can be constructed from harmonic sine waves. This is best possible fit of the square wave by a single sine
a bit ironic, because the ideal square wave has discon- wave. If you want a better fit, you have to add in more
tinuous jumps from one level to another, and therefore sine waves.
does not satisfy Fourier’s theorem that any continuous
wave shape can be made from harmonic sine waves. In Figure (15) we clicked on the bar in the 3rd harmonic
The result is that if you try to construct an ideal square position, the bar representing the amplitude of the 3rd
wave from sine waves, you end up with a small blip at harmonic in the square wave. In the upper window you
the discontinuity (called the Gibb’s effect). Since our see a sine wave with a smaller amplitude and three
focus is experimental data where there is no true times the frequency of the first harmonic. If you select
discontinuity, we will not encounter this problem. a single harmonic, as we have just done in Figure (15),
MacScope prints the frequency of both the first har-
In Figure (12) we have selected one cycle of the square monic and the selected harmonic above the FFT win-
wave. Selecting Fourier Analysis, we get the result dow. Here you can see that the frequency of the first
shown in Figure (13). We have clicked on the Expand harmonic is 201.6 Hz and the selected harmonic fre-
button so that only the selected section of the wave quency is 604.9 Hz as expected.

Figure 12 Figure 13
The square wave. The wave shape goes back Expanding the one cycle selected, and choosing
and forth periodically between two levels. Here Fourier Analysis, we see that this wave form has
we have selected one cycle of the square wave. a number of harmonics. The computer program
calculates the first 128 harmonics. We used the
Scale button to display only the first 16.
16-10 Fourier Analysis, Normal Modes and Sound

If we select both the first and third harmonics together

(either by dragging a rectangle over both bars, or by
holding down the shift key while selecting them indi-
vidually), the upper window displays the sum of the
first and third harmonic, as shown in Figure (16). You
can see that the sum of these two harmonics gives us a
waveform that is closer to the shape of the square wave
than either harmonic alone. We are beginning to build
up the square wave from sine waves.

In Figures (17, 18 and 19), we have added in the 5th, 7th

and 9th harmonics. You can see that the more harmon-
ics we add, the closer we get to the square wave.

One of the special features of a square wave is that it

Figure 14
Select the first harmonic by clicking on the first bar. contains only odd harmonics—all the even harmonics
are absent. Another is that the amplitude of the nth
harmonic is 1/n times as large as that of the first
harmonic. For example, the third harmonic has an
amplitude only 1/3 as great as the first. This is
represented in the FFT window by drawing a bar only
1/3 as high as that of the first bar. In the MacScope
program, the harmonic with the greatest amplitude is
represented by a bar of height 1. All other harmonics
are represented by proportionally shorter bars.

Figure 17
Sum of the
harmonics
Figure 15 1, 3, and 5.
Select the second harmonic by clicking on the second bar.

Figure 18
Sum of the
harmonics
1, 3, 5, 7.

Figure 19
Sum of the
harmonics
1, 3, 5, 7, 9.
Figure 16
Sum of the first and second harmonic
is obtained by selecting both.
 16-11

Repeated Wave Forms your data is not repeating then you have to deal one way
Before we apply the Fourier transform capability of or another with this problem. One technique often used
MacScope to the analysis of experimental data, there is by engineers is to select a long section of data and
one more feature of the analysis we need to discuss. smoothly force the ends of the data to zero so that the
What we are doing with the program is reconstructing selected data can be treated as repeating data. Hope-
a selected section of a waveform from harmonic sine fully, forcing the ends of the data to zero does not
waves. Anything we build from harmonic sine waves destroy the information you are interested in. You can
exactly repeats at the period of the first harmonic. often accomplish the same thing by throwing away
Thus our reconstructed wave will always be a repeating higher harmonics, assuming that the lower harmonics
wave, beginning again at the same height as the begin- contain the interesting features of the data. You can see
ning of the previous cycle. that neither of these techniques would work well for our
one and a half cycles of a sine wave selected in Figure
You can most easily see what we mean if you select a (21).
non repeating section of a wave. In Figure (20) we have
gone back to a sine wave, but selected one and a half In this text, our use of Fourier analysis will essentially
cycles. In the FFT window you see a slew of harmon- be limited to the analysis of repeating waveforms. As
ics. To see why these extra harmonics are present, we long as we select a section that repeats, we do not have
have in Figure (21) selected the first 9 of them and in the to worry about the spurious harmonics.
upper window see what they add up to. It is immedi-
ately clear what has gone wrong. The selected harmon- (Most programs for the acquisition and analysis of
ics are trying to reconstruct a repeating version of our experimental data have an option for doing Fourier
1.5 cycle of the sine wave. The extra spurious harmon- analysis. Unfortunately, few of them allow you to
ics are there to force the reconstructed wave to start and select a precise section of the experimental data for
stop at the same height as required by a repeating wave. analysis. As a result the analysis is usually done on a
non repeating section of the data, which distorts the
If you are analyzing a repeating wave form and select resulting plot of the harmonic amplitudes. For a
a section that repeats, then your harmonic reconstruc- careful analysis of data, the ability to precisely select
tion will be accurate, with no spurious harmonics. If the data to be analyzed is an essential capability.)

Figure 20 Figure 21
When we select one and a half cycles of a sine wave, The spurious harmonics result from the fact that
we get a whole bunch of spurious harmonics. the reconstructed wave must be repeating—must
start and stop at the same height.
16-12 Fourier Analysis, Normal Modes and Sound

ANALYSIS OF THE When the carts were released in an arbitrary way, we

COUPLED AIR CART SYSTEM generally get the complex motion seen in Figures (6)
and (7). What we wish to do now is apply Fourier
We are now ready to apply Fourier analysis to our
analysis to these waveforms to see if any simple
system of coupled air carts. Recall that there were two
features underlie this complex motion.
modes of motion that resulted in a sinusoidal oscilla-
tion of the carts, the vibrational motions shown in In Figure (24), which is the waveform of Figure (6), we
Figure (4) and the sloshing mode shown in Figure (5). see that there is a repeating pattern. The fact that the
In Figures (22) and (23) we expanded the time scales so pattern repeats means that it can be reconstructed from
that we could accurately measure the period and fre- harmonic sine waves, and we can use our Fourier
quency of these oscillations. From the data rectangles, analysis program to find out what the component sine
we see that the frequencies were 1.11Hz and 0.336 Hz waves are.
for the vibrational and sloshing modes respectively.
In Figure (24) we have selected precisely one cycle of
the repeating pattern. This is the crucial step in this
experiment—finding the repeating pattern and select-
ing one cycle of it. How far you have to look for the
pattern to repeat depends upon the mass of the carts and
the strength of the springs.

Figure 22
Vibrational mode of Figure (4). We have expanded
the time scale so that we could accurately measure
the period and frequency of the oscillation.

Figure 24
Complex mode of the coupled aircarts. We
see that the waveform repeats, and have
selected one cycle of the repeating wave.

Figure 23
Sloshing mode of Figure (5). Less of an expansion of
the time scale was needed to measure the period here.
 16-13

Expanding the repeating section of the complex wave- drawn in the display window of Figure (26), and we
form, and choosing Fourier Analysis gives us the find that the frequency of this harmonic is 0.336 Hz.
results shown in Figure (25). What we observe from the This is the frequency of the sloshing mode of Figure
Fourier analysis is that the complex waveform is a (23). Clicking on the bar above the tenth harmonic, we
mixture of two harmonics, in this case the third and get the harmonic drawn in the display window of
tenth harmonic. If we click on the bar showing the Figure (27), and see that the frequency of this mode is
amplitude of the third harmonic, we see that harmonic 1.12 Hz, within a fraction of a percent of the 1.11 Hz
frequency of the vibrational mode of Figure (22).

Figure 25 Figure 27
Fourier analysis of one cycle of the complex Selecting the tenth harmonic, we see that its
waveform. The FFT rectangle shows us that the frequency is essentially equal to the frequency
wave consists of only two harmonics. of the vibrational mode of motion.

Figure 26 Figure 28
When we click on the third harmonic bar, we see that Selecting both modes shows us that the complex motion
the frequency of the third harmonic is precisely the is simply the sum of the two sinusoidal modes of motion.
frequency of the sloshing mode of oscillation.
16-14 Fourier Analysis, Normal Modes and Sound

If we select both the third and tenth harmonics, the sum or in some combination of the two modes. No other
of these two harmonics is shown in Figure (28). These kinds of motion are allowed! The various complex
two harmonics together so closely match the experi- motions are just different combinations of the two
mental data that we had to move the experimental curve modes.
down in order to see both curves. What we have
learned from this experiment is that the complex Adding another aircart so that we have three coupled
motion of Figure (6) is a mixture of the two simple, aircarts, the motion becomes still more complex. How-
sinusoidal modes of motion. ever if we look carefully, we find that the waveform
eventually repeats. Selecting one repeating cycle and
Back in Figures (6) and (7), we displayed two wave- choosing Fourier Analysis, we got the results shown in
forms, representing different complex motions of the Figure (30). We observe that this complex motion is
same two carts. Starting with the second waveform of made up of three harmonics.
Figure (7), we selected one cycle of the motion, ex-
panded the selected section, and chose Fourier Analy- The sinusoidal modes of motion of the coupled air carts
sis. The result is shown in Figure (29). What we see is are called normal modes. The general rule is that if you
that the second complex waveform is also a mixture of have n coupled objects, like n carts on an air track
the third and tenth harmonics. The first waveform in connected by springs, and they are confined to move in
Figure (25) had more of the third harmonic, more of the 1 dimension, there will be n normal modes. (With 2
sloshing mode, while the second waveform of Figure carts, we saw 2 normal modes. With 3 carts, 3 normal
(29) has more of the tenth harmonic, the vibrational modes, etc.) This result, which will play an important
mode. Both complex waveforms are simply mixtures role in our discussion of the specific heat of molecules,
of the vibrational and sloshing modes. They have can be extended to motion in 2 and 3 dimensions. For
different shapes because they are different mixtures. example, if you have n coupled particles that can move
in 3 dimensions, as in the case of a molecule with n
This experiment is beginning to demonstrate that for atoms, then the system should have 3n normal modes
the two coupled aircarts, there is a strict limitation to the of motion. Such a molecule should have 3n indepen-
kind of motion the carts can have. They can either dent ways to vibrate or move. We will have more to say
move in the vibrational mode, or in the sloshing mode, about this subject in the next chapter.

Figure 29 Figure 30
The second complex mode of motion, from One cycle of the waveform for three coupled aircarts.
Figure (7), is simply a different mixture of the With three carts, we get three normal modes of motion.
same vibrational and sloshing modes of motion.
 16-15

THE HUMAN EAR the air. This vibrational motion is then transferred via a lever
The human ear performs a frequency analysis of sound system of three bones (the malleus, incus, and stapes) to a
waves that is not unlike the Fourier analysis of wave motion small membrane covering the oval window of the snail
which we just studied. In the ear, the initial analysis is done shaped cochlea.
mechanically, and then improved and sharpened by a
The cochlea, shown unwound in Figure (32), is a fluid filled
sophisticated data analysis network of nerves. We will
cavity surrounded by bone, that contains two main channels
focus our attention on the mechanical aspects of the ear’s
separated by a membrane called the basilar membrane.
frequency analysis.
The upper channel (scala vestibuli) which starts at the oval
Figure (31) is a sketch of the outer and inner parts of the window, is connected at the far end to the lower channel
human ear. Sound waves, which consist of pressure varia- (scala tympani) through a hole called the helicotrema. The
tions in the air, are funneled into the auditory canal by the lower channel returns to the round window which is also
external ear and impinge on the eardrum, a large membrane covered by a membrane. If the stapes pushes in on the
at the end of the auditory canal. The eardrum (tympanic) membrane at the oval window, fluid flows around the
membrane vibrates in response to the pressure variations in helicotrema and causes a bulge at the round window.

Semicircular
canals

Malleus
Oval
Incus window Cochlea
Auditory nerve
Stapes
Eardrum
Auditory canal

Round Window

Cochlea unwound
Figure 31 Malleus Helicotrema
The human ear. Sound, Incus
entering the auditory Stapes at oval window
canal, causes vibrations
Scala vestibuli
of the eardrum. The
vibrations are transferred Tympanic
membrane
by a bone lever system to (ear drum)
the membrane covering
the oval window. Scala tympani
Basilar Membrane
Vibrations of the oval Round Window
window membrane then
cause wave motion in the Figure 32
fluid in the cochlea. Lever system of the inner ear and an unwound view of the cochlea. The
(Adapted from Lindsey and
basilar membrane separates the two main fluid channels in the cochlea.
Norman, Human
Vibrations of the basilar membrane are detected by hair cells. (Adapted
Information Processing.)
from Principles of Neural Science Edited by E. R. Kandel and J. H. Schwartz,
Elsevier/North-Holland, p260.)
16-16 Fourier Analysis, Normal Modes and Sound

The purpose of the lever system between the eardrum membrane is narrow and stiff, while at the far end it is
and the cochlea is to efficiently transfer sound energy about 5 times as wide and much more floppy. To see
to the cochlea. The eardrum membrane is about 25 why the basilar membrane has this structure, we have
times larger in area than the membrane across the oval in Figure (34) sketched a mechanical model that has a
window. The lever system transfers the total force on similar function as the membrane. In this model we
the eardrum to an almost equal force on the oval have a series of masses mounted on a flexible steel band
window membrane. Since force equals pressure times and attached by springs to fixed rods as shown. The
area, a small pressure variation acting on the large area masses are small and the springs stiff at the front end.
of the eardrum membrane results in a large pressure If we shake these small masses, they resonate at a high
variation at the small area at the oval window. The frequency ω = k m . Down the membrane model,
higher pressures are needed to drive a sound wave the masses get larger and the springs weaken with the
through the fluid filled cochlea. result that the resonant frequency becomes lower.

If the oval window membrane is struck by a pulse, a If you gently shake the steel band at some frequency ω 0
pressure wave travels down the cochlea. The basilar a small amplitude wave will travel down the band and
membrane, which separates the two main fluid chan- soon build up a standing wave of that frequency. If ω 0
nels, moves in response to the pressure wave, and a is near the resonant frequency of one of the masses, that
series of hair cells along the basilar membrane detect mass will oscillate with a greater amplitude than the
the motion. It is the way in which the basilar membrane others. Because the masses are connected by the steel
responds to the pressure wave that allows for the band, the neighboring masses will be carried into a
frequency analysis of the wave. slightly larger amplitude of motion, and we end up with
a peak in the amplitude of oscillation centered around
Figure (33) is an idealized sketch of a straightened out the mass whose frequency ω is equal to ω 0 . (In the
cochlea. (See Appendix B for more realistic sketches.) sketch, we are shaking the band at the resonant fre-
At the front end, by the oval window, the basilar quency ω 7 of the seventh mass.)

scala vestibuli
helicotrema
scala tympani
Figure 33
The basilar membrane in
the cochlea. (Adapted from oval
window bone
Green, An Introduction to
basilar membrane
Hearing, John Wiley &
500 µm

Sons, p66.) stapes 100µm basilar membrane

round window 33 mm

Figure 34
rod
Spring model of the basilar
masses membrane. As we go down the
steel band, the masses become
larger, the springs weaker, and
the resonant frequency drops.
If we vibrate the end of the
ω7 steel band at some frequency, the
1 2 3 4 band mass which resonates at that
5 6 7 8 frequency will have the biggest
9 10 11 amplitude of oscillation.
 16-17

Because the masses m get larger and the spring con- 3

stants k get smaller toward the far end, there is a 25 Hz

continuous decrease in the resonant frequency as we go 0

down the band. If we start shaking at a high frequency,
the resonant peak will occur up near the front end. As 3
we lower the frequency, the peak of the oscillation will 50 Hz
move down the band, until it finally gets down to the
0
lowest resonant frequency mass at the far end. We can
thus measure the frequency of the wave by observing
3
where along the band the maximum amplitude of

relative amplitude of movement (µm)

100 Hz
oscillation occurs.
0
The basilar membrane functions similarly. The stiff
narrow membrane at the front end resonates at a high 3
frequency around 20,000 Hz, while the wide floppy 200 Hz

back end has a resonant frequency in the range of 20 to

0
30 Hz. Figure (35) shows the amplitude of the oscilla-
tion of the membrane in response to driving the fluid in
3
the cochlea at different frequencies. We can see that as 400 Hz
the frequency increases, the location of the maximum
amplitude moves toward the front of the membrane, 0
near the stapes and oval window. Figure (36) depicts
the shape of the membrane at an instant of maximum 3
8 00 Hz
amplitude when driven at a frequency of a few hundred
Hz. The amplitude is greatly exaggerated; the basilar 0
membrane is about 33 millimeters long and the ampli-
tude of oscillation is less than .003 mm. 3
1600 Hz
Although the amplitude of oscillation is small, it is
accurately detected by a system of about 30,000 hair 0
0 10 20 30 mm
cells. How the hair cells transform the oscillation of the distance from stapes (mm)
membrane into nerve impulse signals is discussed in Figure 35
Appendix B at the end of this chapter. Amplitude of the motion of the basilar membrane at
different frequencies. (Adapted from Principles of
The human ear is capable of detecting tiny changes in Neural Science Edited by E. R. Kandel and J. H.
frequency and very subtle mixtures of harmonics in a Schwartz, Elsevier/North-Holland, p 263.)
sound. Looking at the curves in Figure (35) (which
were determined from a cadaver and may not be quite
as sharp as the response curves from a live membrane),
it is clear that it would not be possible to make the ear’s
fine frequency measurements simply by looking for
the peak in the amplitude of the oscillation of the
membrane. But the ear does not do that. Instead,
measurements are continuously made all along the
membrane, and these results are fed into a sophisticated
data analysis network before the results are sent to the Figure 36
brain. The active area of current research is to figure out Response of the basilar membrane to a moderately low
how this data analysis network operates. frequency driving force. (From Vander,A; Sherman,J;
and Luciano,D. Human Physiology, 4th edition, 1985,
P662. McGraw Hill Publishing Co., NY.)
16-18 Fourier Analysis, Normal Modes and Sound

STRINGED INSTRUMENTS To demonstrate the difference between electric and

The stringed instruments provide the clearest example acoustic stringed instruments, and to illustrate how
of how musical instruments function. The only pos- Fourier analysis can be used to study these differences,
sible modes of oscillation of the string are those with my daughter played the same note, using the same
nodes at the ends, and we have seen that the frequencies bowing technique, on the open E string of both her
of these modes form a harmonic series fn = n f 1. This electric and her acoustic violins. Using the same
suggests that if we record the sound produced by a microphone in the same setting to record both, we
stringed instrument and take a Fourier transform to see obtained the results shown in Figure (37).
what harmonics are present in the sound, we can tell From Figure (37) we immediately see why acoustic
from that what modes of oscillation were present in the stringed instruments sound differently from their elec-
vibrating string. tric counterparts. With the acoustic instruments you
This is essentially correct for the electric stringed get a far richer mixture of harmonics. In the first trial,
instruments like the electric guitar and electric violin. labeled E(1) Electric Violin, the string was bowed so
Both of these instruments have a magnetic pickup that that it produced a nearly pure 4th harmonic. The sound
detects the velocity of the string at the pickup, using the you hear corresponds to a pure tone of frequency
same principle as the velocity detector we used in the 657.8/4 = 264 Hz. The corresponding sound produced
air cart experiments discussed earlier in this chapter. by the acoustic violin has predominately the same 4th
The voltage signal from the magnetic pickup is then harmonic, but a lot of the sound is spread through the
amplified electronically and sent to a loudspeaker. first 8 harmonics.
Thus the sound we hear is a fairly accurate representa- A number of recordings were made, so that we could
tion of the motion of the string, and an analysis of that see how the sounds varied from one playing to the next.
sound should give us a good idea of which modes of The examples shown in Figure (37) are typical. It is
oscillation of the strings were excited. clear in all cases that for this careful bowing a single
The situation is different for the acoustic stringed harmonic predominated in the electric violin while the
instruments, like the acoustic guitar used by folk sing- acoustic violin produced a mixture of the first eight. It
ers, and the violin, viola, cello and base, found in is rather surprising that the ear hears all of these sounds
symphony orchestras. In these instruments the vibra- as representing the same note, but with a different
tion of the string does not produce that much sound quality of sound.
itself. Instead, the vibrating string excites resonances We chose the violin for this comparison because by
in the sound box of the instrument, and it is the sound using a bow, one can come much closer to exciting a
produced by the resonating sound box that we hear. As pure mode of vibration of the string. We see this
a result the quality of the sound from an acoustic string explicitly in the E(1) Electric Violin example. When
instrument depends upon how the sound box was you pluck or strum a guitar, even if you pluck only one
constructed. Subtle differences in the shape of the string, you get a far more complex sound than you do
sound box and the stiffness of the wood used in its for the violin. If you pluck a chord on an acoustic guitar,
construction can lead to subtle differences in the har- you get a very complex sound. It is the complexity of
monics excited by the vibrating string. The human ear the sound that gives the acoustic guitar a richness that
is so sensitive to these subtle differences that it can makes it so effective for accompanying the human
easily tell the difference between a great instrument voice.
like a 280 year old Stadivarius violin, and even the best
of the good instruments being made today. (It may be Figure 37
that it takes a couple of hundred years of aging for a very Comparison of the sound of an electric and an
good violin to become a great one.) acoustic violin. In each case the open E string was
bowed as similarly as possible. The electric violin
produces relatively pure tones. The interaction
between the string and the sound box of the acoustic
violin gives a much richer mixture of harmonics.
 16-19

E(1) Electric Violin E(1) Acoustic Violin

E(2) Electric Violin E(2) Acoustic Violin

E(3) Electric Violin E(3) Acoustic Violin

16-20 Fourier Analysis, Normal Modes and Sound

Recording the sound of an instrument and using Fou- WIND INSTRUMENTS

rier analysis can be an effective tool for studying While the string instruments are all based on the
musical instruments, but care is required. For example, oscillation of a string, the wind instruments, like the
in comparing two instruments, start by choosing a organ, flute, trumpet, clarinet, saxophone, and glass
single note, and try to play the note the same way on bottle, are all based on the oscillations of an air column.
both instruments. Make several recordings so that you Of these, the bottle is the most available for studying the
can tell whether any differences seen are due to the way nature of the oscillations of an air column.
the note was played or due to the differences in the
instruments themselves. With careful work, you can When you blow carefully across the top of a bottle, you
learn a lot about the nature of the instrument. hear a sound with a very definite frequency. Add a little
water to the bottle and the frequency of the note rises.
In the 1970s, before we had personal computers, stu- When you shortened the length of a string, the pitch
dents doing project work would analyze the sound went up, thus it is not surprising that the pitch also goes
produced by instruments or the voice, using a time up when you shorten the length of the air column.
sharing mainframe computer system to do the Fourier
analysis. Hours of work were required to analyze a You might guess that the mode of vibration you set up
single sound, but the results were so interesting that by blowing across the top of the bottle has a node at the
they served as the incentive to develop the MacScope™ bottom of the bottle and an anti node at the top where
program when the Macintosh computer became avail- you are blowing. For a sine wave the distance from a
able node to the next anti node is 1/4 of a wavelength, thus
you might predict that the sound has a wavelength 4
I particularly remember an early project in which two times the height of the air column, and a frequency
students compared the spinet piano, the upright piano
and the grand piano. They recorded middle C played vsound v
f = = sound for λ = 4d (4)
on each of these pianos. Middle C on the spinet λ 4d
consisted of a wide band of harmonics. From the This prediction is not quite right as you can quickly find
upright there were still a lot of harmonics, but the first, out by experimenting with various shaped bottles. Add
third, and fifth began to predominate. The grand piano water to different shaped bottles, adjusting the levels of
was very clean with essentially only the first, third, and the water so that all the bottles have the same height air
fifth present. You could clearly see the effect of the columns. When you blow across the top of the different
increase in the size of the musical instrument. bottles, you will hear distinctly different notes, the
fatter bottles generally having lower frequencies than
The same year, another student, Kelly White, took a the skinny ones. Unlike the vibrating string, it is not just
whale sound from the Judy Collins record Sound of the the length of the column and the speed of the wave that
Humpback Whales. Listening to the record, the whale determine the frequency of oscillation, the shape of the
sounds are kind of squeaky. But when the sound was container also has a noticeable effect.
analyzed, the results were strikingly similar to those of
the grand piano. The analysis suggested that the whale
sounds were by an instrument as large as, or larger, than
a grand piano. (The whale’s blowhole acts as an organ Figure 38
λ
pipe when the whale makes the sound.) Blowing across the top of
a bottle produces a note 4
whose wavelength is
approximately 4 times the
height of the air column.
 16-21

Despite this additional complexity, there is one com- The various modes of oscillation of an air column in an
mon feature to the air columns encountered in musical organ pipe are shown in Figure (39). All have an anti
instruments. They all have unique frequencies of node at the end with sharp edges where the turbulence
oscillation. Even an organ pipe that is open at both excites the oscillation. The open ended pipes shown in
ends—a situation where you might think that the length Figure (39a) also have anti nodes at the open end, while
of the column is not well defined—the air column has the closed pipes of Figure (39b) have a node at the far
a precise set of frequencies. The fussiness in defining end. The pictures in Figure (39a) are a bit idealized, but
the length of the column does not fuzz out the sharpness give a reasonably accurate picture of the shape of the
of the resonance of the column. standing wave. We can compensate for a lack of
accuracy of these pictures by saying, for example, that
Organ pipes, with their straight sides, come closest to the anti node of the open ended pipes lies somewhat
the simple standing waves we have seen for a stretched beyond the end of the pipe.
instrument string. In all cases, the wave is excited at one
end by air passing over a sharp edge creating a turbulent
Exercise 1
flow behind the edge. This turbulence excites the air
column in much the same way that dragging a sticky (a) Find the formula for the wavelength λn of the nth
bow across a violin string excites the oscillation of the harmonic of the open ended pipes of Figure (39).
string. (b) Assuming that the frequencies of vibration are given
by f(cycle/sec) = v(meter /sec) λ (meter /cycle) where v
is the speed of sound, what is the formula for the allowed
frequencies of the open organ pipe of length L?

(c) What should be the length L of an open ended organ

pipe to produce a fundamental frequency of 440 cycles/
second, middle A?

(d) Repeat the calculations of parts a, b, and c for the

closed organ pipes of Figure (39).

(e) If you have the opportunity, find a real organ and

check the predictions you have made (or try the experi-
ment with bottles).
a)
If you start with an organ pipe, and drill holes in the side,
essentially converting it into a flute or one of the other
wind instruments like a clarinet, you considerably alter
the shape and frequency of the modes of oscillation of
the air column. As a first approximation you could say
that you create an anti node at the first open hole. But
then when you play these instruments you can make
more subtle changes in the pitch by opening some holes
and closing others. The actual patterns of oscillation
can become quite complex when there are open holes,
but the simple fact remains that, no matter how com-
plex the wave pattern, there is a precise set of resonant
frequencies of oscillation. It is up to the maker of the
Figure 39 b)
Modes of oscillation of an air column instrument to locate the holes in such a way that the
in open and closed organ pipes. resonances have the desired frequencies.
16-22 Fourier Analysis, Normal Modes and Sound

PERCUSSION INSTRUMENTS a drumhead at a precise frequency than it is to find the

We all know that the string and wind instruments normal modes of the coupled air cart system.
produce sound whose frequency we adjust to produce The experiment is illustrated in Figure (40). The
melodies and chords. But what about drums? They apparatus consists of a hollow cardboard cylinder with
seem to just make noise. Surprisingly, drumheads have a rubber sheet stretched across one end to act as a
specific modes of oscillation with definite frequencies, drumhead. At the other end is a loudspeaker attached
just as do vibrating strings and air columns. But one to a signal generator. When the frequency of the signal
does not usually adjust the fundamental frequency of generator is adjusted to the resonant frequency of one
oscillation of the drumhead, and the frequencies of the of the normal modes of the drumhead, the drumhead
higher modes of oscillation do not follow the harmonic will start to vibrate in that mode of oscillation.
patterns of string and wind instruments.
To observe the shape and motion of the drumhead in
To observe the standing wave patterns corresponding one of its vibrational modes, we place a strobe light to
to modes of vibration of a drumhead, we can drive the one side of the drumhead as shown. If you adjust the
drumhead at the resonant frequency of the oscillation strobe to the same frequency as the normal mode
we wish to study. It turns out to be a lot easier to drive vibration, you can stop the motion and see the pattern.

plywood frame
rubber
sheet

speaker

strobe
light

Figure 40
Studying the modes of oscillation of a drumhead.

a) b) c)

d) e) f)
Figure 41
Modes of oscillation of a drumhead. (Adapted from Vibration and Sound
by Phillip M. Morse, 2nd ed., McGraw-Hill, New York, 1948.)
 16-23

Turn the frequency a bit off resonance, and you get a When you strike a drumhead you excite a number of
slow motion moving picture of the motion of that modes at once and get a complex mixture of frequen-
mode. cies. However, you do have some control over the
modes you excite. Bongo drum players, for example,
Some of the low frequency normal mode or standing get different sounds depending upon where the drum is
wave patterns of the drumhead are illustrated in Figure struck. Hitting the drum in the center tends to excite the
(41). In the lowest frequency mode, Figure (41a), the lowest mode of vibration and produces a lower fre-
entire center part of the drumhead moves up and down, quency sound. Striking the drum near the edges excites
much like a guitar string in its lowest frequency mode. the higher harmonics, giving the drum a higher fre-
In this pattern there are no nodes except at the rim of the quency sound.
drumhead.
Even more complex than the modes of vibration of a
In the next lowest frequency mode, shown in Figure drumhead are those of the components of a violin. To
(41b), one half the drumhead goes up while the other construct a successful violin, the front and back plates
half goes down, again much like the second harmonic of a violin must be tuned before assembly. Figure (42)
mode of the guitar string. The full two dimensional shows a violin backplate under construction, while
nature of the drumhead standing waves begins to Figure (43) shows the first 6 modes of oscillation of a
appear in the next mode of Figure (41c) where the completed backplate. Note again that the resonant
center goes up while the outside goes down. Now we frequencies do not form a harmonic series.
have a circular node about half way out on the radius of
the drumhead.

As we go up in frequency, we observe more complex

patterns for the higher modes. In Figure (41d) we see
a pattern that has a straight node like (41b) and a circular Figure 42
node like (41c). This divides the drumhead into 4 Back plate of a violin
separate regions which oscillate opposite to each other. under construction. The
resonant frequencies are
Finer division of the drumhead into smaller regions can tuned by carving away
be seen in Figures (41e) and (41f). The frequencies of wood from different
the various modes are listed with each diagram. You sections of the plate.
can see that there is no obvious progression of frequen-
cies like the harmonic progression for the modes of a
stretched string.

116Hz 167Hz 222Hz 230Hz 349Hz 403Hz

Figure 43
Modes of oscillation of the backplate made visible by holographic techniques. Quality violins
are made by tuning the frequencies of the various modes. (Figures 42 and 43 from “The
Acoustics of Violin Plates,” by Carleen Maley Hutchins, Scientific American, October 1981.)
16-24 Fourier Analysis, Normal Modes and Sound

SOUND INTENSITY Taking the logarithm to the base 10 of these ratios gives
One of the amazing features of the human eardrum is
its ability to handle an extreme range of intensities of Log10 I0 /I0 = Log10 1 = 0
sound waves. We define the intensity of a sound wave Log10 I1 /I0 = Log10 10 = 1
as the amount of energy per second being carried by
Log10 I2 /I0 = Log10 100 = 2
a sound wave through a unit area. In the MKS system
of units, this would be the number of joules per second ........
passing through an area of one square meter. Since one Log10 I12/I0 = Log10 10 12 = 12 (6)
joule per second is a unit of power called a watt, the
MKS unit for sound intensity is watts per square meter.
The human ear is capable of detecting sound intensities Bells and Decibels
as faint as 10 – 12 watts /m2, but can also handle inten- The scale of loudness defined as Log10 I/I0 with
sities as great as 1 watt /m2 for a short time. This is an I0 = 10 –12watts/m 2 is measured in bells, named after
astounding range, a factor of 10 12 in relative intensity. Alexander Graham Bell, the inventor of the telephone.
The ear and brain handle this large range of intensities
loudness of a sound I
by essentially using a logarithmic scale. Imagine, for ≡ Log10
example, you are to sit in front of a hi fi set playing a measured in bells I0
pure tone, and you are told to mark off the volume (7)
control in equal steps of loudness. The first mark is I0 = 10 –12watts/m2
where you just barely hear the sound, and the final mark
is where the sound just begins to get painful. Suppose From this equation, we see that the faintest sound we
you are asked to divide this range of loudness into what can hear, at I = I0 , has a loudness of zero bells. The
you perceive as 12 equal steps. If you then measured most intense one we can stand for a short while has a
the intensity of the sound you would find that the loudness of 12 bells. All other audible sounds fall in the
intensity of the sound increased by approximately a range from 0 to 12 bells.
factor of 10 after each step. Using the faintest sound
you can hear as a standard, you would measure that the It turns out that the bell is too large a unit to be
sound was 10 times as intense at the end of the first step, convenient for engineering applications. Instead one
100 times as intense after the second step, 1000 times usually uses a unit called the decibel (db) which is
at the third, and 10 12 times as intense at the final step. 1/10 of a bell. Since there are 10 decibels in a bell, the
formula for the loudness β , in decibels, is
The idea that the intensity increases by a factor of 10 for
each equal step in loudness is what we mean by the I
β decibels = 10 db Log10 (8)
statement that the loudness is based on a logarithmic I0
scale. We take the faintest sound we can hear, an
intensity I0 = 10 – 12watts/m 2 as a basis. At the first On this scale, the loudness of sounds range from 0
setting I = I0 , at the second setting I1 = 10 I0 , at the decibels for the faintest sound we can hear, up to 120
3rd setting I2 = 100 I0 , etc. The factor I/I0 by which decibels ( 10 × 12 bells) for the loudest sounds we can
the intensity has increased is thus
tolerate.
I0 /I0 = 1 (I0 = 10 – 12 watts/m 2 )
I1 /I0 = 10
I2 /I0 = 100
........
I12/I0 = 10 12 (5)
 16-25

The average loudness of some of the common or well Our sensitivity to sound depends not only to the inten-
known sounds is given in Table 1. sity of the sound, but also to the frequency. About the
lowest frequency note one can hear, and still perceive
Table 1 Various Sound Levels in db
as being sound, is about 20 cycles/second. As you get
threshold of hearing 0 older, the highest frequencies you can hear decreases
rustling leaves 10 from around 20,000 cycles/sec for children, to 15,000
whisper at 1 meter 20 Hz for young adults to under 10,000 Hz for older
city street, no traffic 30 people. If you listen to too much, too loud rock music,
quiet office 40 you can also decrease your ability to hear high fre-
office, classroom 50 quency sounds.
normal conversation at 1 meter 60
busy traffic 70 Figure (44) is a graph of the average range of sound
average factory 80 levels for the human ear. The faintest sounds we can
jack hammer at 1 meter 90 detect are in the vicinity of 4000 Hz, while any sound
old subway train 100 over 120 db is almost uniformly painful. The fre-
rock band 120 quency ranges and sound levels usually encountered in
jet engine at 50 meters 130 music are also shown.
Saturn rocket at 50 meters 200

An increase in loudness of 10 db corresponds to an

increase of 1 bell, or an increase of intensity by a factor
of 10. A rock band at 110 db is some 100 times as
intense (2 factors of 10) as a jack hammer at 90 db. A
Saturn rocket is about 10 20 times as intense as the
faintest sound we can hear.

120
threshold of pain

100
sound level (db)

Figure 44 80
Average range of sound
levels for the human ear.
60 music
Only the very young can
hear sound frequencies up
to 20,000 Hz. (Adapted 40
from Fundamental
Physics by Halliday and
Resnick, John Wiley & 20
Sons.)
threshold of hearing
0

20 100 1000 10,000 20,000

frequency (Hz)
16-26 Fourier Analysis, Normal Modes and Sound

Sound Meters Perhaps the easiest way to begin to calibrate a micro-

Laboratory experiments involving the intensity or loud- phone for measuring sound intensities is to use the fact
ness of sound are far more difficult to carry out than that very little sound energy is lost as sound travels out
those involving frequencies like the Fourier analysis through space. Suppose you had a speaker radiating
experiments already discussed. From the output of any 100 watts of sound energy, and for simplicity let us
reasonably good microphone, you can obtain a rela- assume that the speaker radiates uniformly in all direc-
tively good picture of the frequencies involved in a tions and that there are no nearby walls.
sound wave. But how would you go about determining
the intensity of a sound from the microphone output? If we are 1 meter from this speaker, all the sound energy
(There are commercial sound meters which have a is passing out through a 1 meter radius sphere centered
scale that shows the ambient sound intensity in deci- on the speaker. Since the area of a sphere is 4πr 2 , this
bels. Such devices are often owned by zoning boards 1 meter radius sphere has an area of 4πmeters2 , and the
for checking that some factory or other noise source average intensity of sound at this 1 meter distance must
does not exceed the level set by the local zoning be
ordinance, often around 45 db. The point of our average intensity of
question is, how would you calibrate such a device if sound 1 meter from = 100 watts2
you were to build one?) a 100 watt speaker 4π meters
(9)
The energy in a wave is generally proportional to the = 8.0 watts 2
square of the amplitude of the wave. A sound wave can meters
be viewed as oscillating pressure variations in the air,
and the energy in a sound wave turns out to be propor- If we wish to convert this number to decibels, we get
tional to the square of the amplitude of the pressure
sound intensity 1 meter
variations. The output of a microphone is more or less β
proportional to the amplitude of the pressure varia- from a 100 watt speaker
tions, thus we expect that the intensity of a sound wave I
should be more or less proportional to the square of the = 10 db Log
I0
voltage output of the microphone. However, there is a
great variation in the sensitivity of different micro- 8.0 watts / m2
= 10 db Log10
phones, and in the amplifier circuits used to produce 10 –12 watts / m2
reasonable signals. Thus any microphone that you
wish to use for measuring sound intensities has to be
calibrated in some way. = 10 db × Log10 8 × 10 12
= 10 db × 12.9
= 129 db (10)

From our earlier discussion we see that this exceeds the

threshold of pain. One meter from a 100 watt speaker
is too close for our ears. But we could place a
microphone there and measure the amplitude of the
signal output for our first calibration point.
 16-27

Move the speaker back to a distance of 10 meters and Exercise 3

the area that the sound energy has to pass through
You are playing a monophonic record on your stereo
increases by a factor of 100 since the area of a sphere
system when one of your speakers cuts out. How many
is proportional to r 2 . Thus as the same 100 watts passes db did the loudness drop? (Assume that the intensity
through this 100 times larger area, the intensity drops dropped in half when the speaker died. Surprisingly
to 1/100 of its value at 1 meter. At a distance of 10 you can answer this question without knowing how loud
meters the intensity is thus 8/100 = .08 watts/m 2 and the stereo was in the first place. The answer is that the
the loudness level is loudness dropped by 3 db).

10 meters from a
β Speaker Curves
100 watt speaker
When you buy a hi fi loudspeaker, you may be given a
frequency response curve like that in Figure (45), for
.08 watts / m2 your new speaker. What the curve measures is the
= 10 db Log10
10 –12 watts / m2 intensity of sound, at a standard distance, for a standard
amount of power input at different frequencies. It is a
= 109 db (11) fairly common industry standard to say that the fre-
quency response is “flat” over the frequency range
We see that when the intensity drops by a factor of 100, where the intensity does not fall more than 3 db from its
it drops by 20 db or 2 bells. average high value. In Figure (45), the response of that
To calibrate your sound meter, record the amplitude of speaker, with the woofer turned on, is more or less
the signal on your microphone at this 10 meter distance, “flat” from 62 Hz up to 30,000 Hz.
then set the microphone back to a distance of 1 meter, Why the 3 db cutoff was chosen, can be seen in the
and cut the power to the speaker until the microphone result of Exercise 3. There you saw that if you reduce
reads the same value as it did when you recorded 100 the intensity of the sound by half, the loudness drops
watts at 10 meters. Now you know that the speaker is by 3 db. This is only 3/120 (or 1/36) of our total hearing
emitting only 1/100th as much power, or 1 watt. range, not too disturbing a variation in what is supposed
Repeating this process, you should be able to calibrate to be a flat response of the speaker.
a fair range of intensities for the microphone signal. If
you get down to the point where you can just hear the 10

sound, you could take that as your value of I0 , which +3 db

should presumably be close to I0 = 10 – 12watts/m 2 . 0

–3 db
Then calibrate everything in db and you have built a
Amplitude in db

loudness meter. (The zoning board, however, might –10

not accept your meter as a standard for legal purposes.)

–20

Exercise 2
What is the loudness, in db, 5 meters from a 20 watt –30

speaker? (Assume that the sound is radiated uniformly

in all directions). –40
10 100 1000 10000
Frequency in Hz

Figure 45
Speaker response curve from a recent audio magazine.
The dashed line shows the response when the woofer is
turned off. (We added the dotted lines at + and – 3 db.)
16-28 Fourier Analysis, Normal Modes and Sound

APPENDIX A Calculating Fourier Coefficients

In general we cannot construct an arbitrary wave out of
FOURIER ANALYSIS LECTURE
just sine waves, because sine waves, sin(t), sin(2t), etc.,
In our discussion of Fourier Analysis, we saw that all have a value 0 at t = 0 and at t = 2π. If our wave is
any wave form can be constructed by adding to- not zero at the beginning (t = 0) of our selected period,
gether a series of sine and cosine waves. You can or not zero at the end (t = 2π), then we must also include
think of the Fourier transform as a mathematical cosine waves which have a value 1 at those points.
prism which breaks up a sound wave into its various Thus the general formula for breaking an arbitrary
wavelengths or frequencies, just as a light prism repetitive wave into sine and cosine waves is:
breaks up a beam of white light into its various colors
or wavelengths. In MacScope, the computer does F(t) = A0 + A1 cos(1t) + A2 cos(2t)
the calculations for us, figuring out how much of + A3 cos(3t) + ...
each component sine wave is contained in the sound
wave. The point of this lecture is to give you a + B1 sin (1t) + B2 sin (2t)
feeling for how these calculations are done. The + B3 sin (3t) + ... (A-1)
basic ideas are easy, only the detailed calculations
that the computer does would be hard for us to do. The question is: How do we find the coefficients
A0, A1, A2, B1, B2 etc. in Equation (A1)? (These are
the Fourier coefficients.)
Square Wave
In Figure A-1 we show a MacScope window for a To see how we can determine the Fourier coefficients,
square wave produced by a Hewlett Packard oscilla- let us take an explicit example. Suppose we wish to find
tor. the coefficient B3, representing the amount of sin(3t)
present in the wave. We can find B3 by first multiplying
We have selected precisely one cycle of the wave, and Equation (1) through by sin(3t) to get:
see that the even harmonics are missing. A careful
investigation shows that the amplitude of the Nth odd
harmonic is 1/N as big as the first (e.g., the 3rd
harmonic is 1/3 as big as the first, etc.). Thus the
mathematical formula for a square wave F(t) can be
written:
F(t) = (1)sin(t) + (1/3)sin(3t) + (1/5)sin(5t)
+ (1/7)sin(7t) ...
where, for now, we are assuming that the period of the
wave is precisely 2π seconds. The coefficients (1),
(1/3), (1/5), (1/7), which tell us how much of each sine
wave is present, are called the Fourier coefficients.
Our goal is to calculate these coefficients.

Figure A-1
The square wave has only odd harmonics.
 16-29

F(t) sin(3 t) = A 0 sin(3t) Area under

+ A 1 cos (1t) sin(3t) sin(3t)sin(3t) 2π
= sin(3t)sin(3t)dt (A-3)
+ A 2 cos (2t) sin(3t) from t = 0 0
+ A 3 cos (3t) sin(3t) to t = 2π
+ ...
(Those who have had calculus say we are taking the
+ B 1 sin (1t) sin(3t) integral of the term sin(3t)sin(3t).)
+ B 2 sin (2t) sin(3t)
A basic rule learned in algebra is that if we do the
+ B 3 sin (3t) sin(3t)
same thing to both sides of an equation, the sides will
+ ... (A-2) still be equal. This is also true if we do something as
peculiar as evaluating the net area under the curves
At first it looks like we have created a real mess. We on both sides of an equation.
have a lot of products like cos(t)sin(3t), sin(t)sin(3t),
sin(3t)sin(3t), etc. To see what these products look like, If we take the net area under the curves on the right side
we plotted them using True BASIC™ and obtained of Equation A-2, only the sin 2 (3t) term survives and
the results shown in Figure (A2) (on the next page). we get:

Notice that in all of the plots involving sin(3t), the 2π 2π

F(t)sin(3t)dt = B3 sin 2 (3t)dt (A-4)
product sin(3t)sin(3t) = sin 2(3t) is special; it is the 0 0
only one that is always positive. (It has to be since it is
a square.) A careful investigation shows that, in all the In Figure (A4), we have replotted the curve sin 2 (3t) ,
other non square terms, there is as much negative area and drawn a line at height y = .5. We see that the peaks
as positive area, as indicated by the two different above the y = .5 line could be flipped over to fill in the
shadings in Figure (A3). If we define the “net area” valleys below the y = .5 line. Thus sin 2 (3t) in the
under a curve as the positive area minus the negative interval t = 0 to t = 2π has a net area equal to that of a
area, then only the sin(3t)sin(3t) term on the right side rectangle of height .5 and length 2π. I.e., the net area
of the Equation A-2 has a net area. is π:
2π
2
The mathematical symbol for finding the net area sin (3t)dt = π (A-5)
under a curve (in the interval t = 0 to t = 2π ) is: 0

sin(1t) sin(3t) sin(3t) sin(3t)

1 1
positive area
.5

0 0
rectangle area = .5 2π = π
negative
area (Crests above .5 just fill
in the troughs below .5.)
–1 –1
0 2π 0 2π
Figure A-3 Figure A-4
Only the square terms such as sin(3t)sin(3t) The area under the curve sin2 (3t) is equal to
have a net area. This curve has no net area. the area of a rectangle .5 high by 2π long.
16-30 Fourier Analysis, Normal Modes and Sound

sin(1t) sin(1t) sin(3t) sin(3t)

1 1

0 0

–1 –1
0 2π 0 2π

sin(1t) sin(2t) cos(1t) sin(1t)

1 1

0 0

–1 –1
0 2π 0 2π

sin(1t) sin(3t) cos(1t) sin(3t)

1 1

0 0

–1 –1
0 2π 0 2π

sin(1t) sin(4t) cos(2t) sin(3t)

1 1

0 0

–1 –1
0 2π 0 2π

sin(1t) sin(1t) cos(2t) sin(3t)

1 1

0 0

–1 –1
0 2π 0 2π
Figure A-2
Product wave patterns.
 16-31

Substituting Equation A-5 in Equation A-4 and solving The function cos(t – φ ) is illustrated in Figure (A-5).
for the Fourier coefficient B3 gives: We see that C is the amplitude of the wave, and the
phase angle φ is the amount the wave has been moved
B3 =
1 2π
F(t)sin(3t)dt (A-6) to the right. (When t = 0, cos(t – φ ) = cos(– φ ).) With
π 0 Equation A-9, we can rewrite equation A-1 in the form:
Similar arguments show that the general formulas for
F(t) = C0 + C1 cos( t – φ1)
the Fourier coefficients An and Bn are:
+ C2 cos(2t – φ2)
1 2π + C3 cos(3t – φ3)
An = F(t)cos(nt)dt (A-7)
π 0 + ... (A-10)
The advantage of Equation A-10 is that the coeffi-
1 2π cients C represent how much of each wave is present,
Bn = F(t)sin(nt)dt (A-8) and sometimes we do not care about the phase angle
π 0
φ . For example our ears are not particularly sensi-
tive to the phase of the harmonics in a musical note,
These integrals, which were nearly impossible to do
thus the tonal quality of a musical instrument is
before computers, are now easily performed even on
determined almost entirely by the amplitudes C of
small personal computers. Thus the computer has
the harmonics the instrument produces.
made Fourier analysis a practical experimental tool.
cos(t – φ)
1
Amplitude and Phase
Instead of writing the Fourier series as a sum of separate
sine and cosine waves, it is often more convenient to φ
use amplitudes and phases. The basic formula we use 0
is **
Acos(t) + Bsin(t) = Ccos(t – φ ) (A-9)
where –1
0 2π
when t = φ,
C = amplitude cos(t – φ) = cos(φ – φ) = cos(0) = 1
C
C2 = A 2 + B2 B Figure A-5
φ The function cos(t– φ) . When φ = 90°
φ = phase you get a sine wave.
A
tan(φ ) = B/A

** Start with
cos (x–y) = cos (x)×cos (y) + sin(x)×sin(y)
Let x = t, y = φ, and multiply through by C to get
C cos(t–φ) = {C cos(φ)} cos(t) + {C sin(φ)} sin(t)
This is Equation A–9 if we set
A = {C cos(φ)}; B = {C sin(φ)}
Thus tanφ = sinφ/cos φ = B/A
16-32 Fourier Analysis, Normal Modes and Sound

In the Fourier transform plots we have shown so far, the

graph of the harmonics has been representing the
amplitudes C. If you wish to see a plot of the phases φ ,
then press the button labeled ø as shown, and you get
the result seen in Figure(A-6). In that figure we are
looking at the phases of the odd harmonic sine waves
that make up a square wave. Since
sin(t) = cos(t - 90°)
all the sine waves should have a phase shift of 90°.

If for any reason, you need accurate values of the

Fourier coefficients, they become available if you
press the FFT Data button to get the results shown
in Figure (A-7). When you do this, the Editor Figure A-6
window is filled with a text file containing the A, B When you press the ø button, the Fourier
coefficients accurate to 3 or 4 significant figures. Transform display shifts from amplitudes to
phases. Since the square wave is made up of
pure odd harmonic sine waves, each odd
harmonic should have a 90 degree or
π /2 phase shift.

Figure A-7
For greater numerical accuracy, you can press
the FFT Data button. This gives you a text file
with the A, B coefficients given to four places.
 16-33

Amplitude and Intensity

An experiment that has become possible with Mac-
Scope, is to have students compare the Fourier
transform of a multiple slit grating with the diffrac-
tion pattern produced by a laser beam passing through
that grating. For example, in Figure (A-8) we have
taken the transform of a 3-slit grating. In this case,
the 3-slit “pattern” was made simply by turning a 2
volt power supply on and off. We are now working
on ways for students to record the slit pattern di-
rectly.

The problem with Figure (A-8) is that the Fourier

transform of the slits gives the amplitude of the
Figure A-8 diffraction pattern, while in the lab one measures the
Amplitude. The Fourier transform of a 3-slit pattern intensity of the diffraction pattern. The intensity of
gives the amplitude of the diffraction pattern that a light wave is proportional to the square of its
would be produced by a laser beam passing through amplitude.
these slits. (Selecting the data to give wider slits would
correspond to using a different wavelength laser beam.) In order that students can compare a slit Fourier
transform with an experimental diffraction pattern,
we have designed MacScope so that one more press
on the ø button takes us from a display of phases to
intensities. Explicitly, the ø button cycles from
amplitudes to phases to intensities. In Figure (A-9)
we have clicked over to intensities, and this pattern
may be directly compared with the intensity of the
3-slit diffraction pattern seen in lab.

(The Fourier transform of the diffraction pattern

amplitude should give the slit pattern. Unfortu-
nately, if you take the Fourier transform of the
experimental diffraction pattern, you are taking the
transform of the intensity, or square, of the ampli-
tude. What you get, as Chris Levey of our depart-
ment demonstrated, is the convolution of the slit
pattern with itself.)
Figure A-9
Intensity. In the lab, you see the intensity of
the diffraction pattern. MacScope will display
the intensity of the Fourier transform if you
click one more time on the ø button.
16-34 Fourier Analysis, Normal Modes and Sound

APPENDIX B As seen in Figure (35), the lower the frequency, the

farther down the membrane the maximum amplitude
INSIDE THE COCHLEA occurs. Along the top of the basilar membrane is a
In Figure (32), our simplified unwound view of the system of hair cells that detects the motion of the
cochlea, we show only the basilar membrane separated membrane and sends the needed information to the
by two fluid channels (the scala vestibuli which starts brain.
at the oval window, and the scala tympani which ends
at the round window). That there is much more Figure (B-2a) is a close up view of the hair cells that sit
structure in the cochlea is seen in the cochlea cross atop of the basilar membrane. (There are about 30,000
section of Figure (B-1). The purpose of this additional hair cells in the human ear.) Above the hair cells is
structure is to detect the motion of the basilar mem- another membrane called the tectorial membrane which
brane in a way that is sensitive to the harmonic content is hinged on the left hand side of that figure. Fine hairs
of the incoming sound wave. go from the top of each hair cell up to the tectorial
membrane as shown. When the basilar membrane is
Recall that when the basilar membrane is excited by a deflected by an incoming sound wave the hairs are bent
sinusoidal oscillation, the maximum amplitude of the as shown in Figure (B-2b). It is the bending of the hairs
response of the basilar membrane is located at a posi- that triggers an electrical impulse in the hair cell.
tion that depends upon the frequency of the oscillation.
Figure (B-3) is a mechanical model of how the bending
of the hairs creates the electrical impulse. The fluid in
Stapes at oval window the cochlea duct surrounding the hair cells has a high
Cochlea unwound concentration of positive potassium ions ( k + ). The
Scala vestibuli
3
25 Hz

ear drum 0
cross cross section Scala tympani
3
Basilar Membrane 50 Hz
Round Window
0
Figure 32 (repeated)
The cochlea unwound. 3
relative amplitude of movement (µm)

100 Hz

0
scala vestibular membrane cochlea
vestibuli duct
tectorial 3
membrane 200 Hz

0
ne
bo

3
hair cells 400 Hz

3
8 00 Hz

3
1600 Hz

0
scala 0 10 20 30 mm
auditory nerve basilar membrane tympani distance from stapes (mm)

Figure 35 (repeated)
Figure B-1
Amplitude of the motion of the basilar
Cross section of the cochlea. (From Vander,A; Sherman,J; membrane at different frequencies, as
and Luciano,D. Human Physiology, 4th edition, 1985, P662.
we go down the basilar membrane.
McGraw Hill Publishing Co., NY.)
 16-35

bending of the hair cell opens small channels allowing the amplitude, in db, of the sound wave required to
potassium ions to flow into the hair cell. This flow of excite a nerve fiber connected to that particular region
positive charge into the cell changes the electrical of hair cells. You can see that the nerve is most sensitive
potential of the cell, triggering reactions that will to a 2 killocycle (2kHz) frequency. At 2 kHz, that nerve
eventually result in an electrical impulse in the nerve fires when excited by a 15 db sound wave. It is not
fiber that is connected to the hair cell. excited by a 4 kHz wave until the sound intensity rises
to 80 db.
After the channel at the top of the hair cell closes, the
excess potassium is pumped out of the hair cell, and the Ultimately the exquisite sensitivity of the human ear to
cell returns to its normal resting voltage, ready to fire different frequency components in a sound wave re-
again. sults from the fact that there are about 30,000 hair cells
continuously monitoring the motion of the basilar
There are various ways that a hair cell can transmit membrane. Effective processing of this vast amount of
frequency information to the nervous system. One is information leads to the needed sensitivity. Much of
by its location down the basilar membrane. The lower this processing of information occurs in the nervous
the frequency of the sound wave, the farther down the system in the ear, before the information is sent to the
membrane an oscillation of the membrane takes place. brain.
Thus high frequency waves excite cells at the front of
the basilar membrane, while low frequency oscilla-
tions excite cells at the back end.

Secondly, hair calls in a given area show special

sensitivities to different frequencies. Figure (B-4) shows

Figure B-4 100

Frequency dependence.
A much lower amplitude
Figure B-3 75
Amplitude (db)

sound will excite this

Model of the valves at the nerve fiber at 2kHz than
base of the hair cells. any other frequency. 50
(From Shepard,G.M.,
Different nerve fibers
Neurobiology, 3rd Edition,
connected to the hair 25
1994, P316. Oxford Univ.
cells have different
Press.)
frequency dependence.
0.5 1.0 1.5 2.0 5.0
Log frequency (kHz)

tectorial hairs bent

membrane

lls
hair ce

hair cells

bone
angle of deflection
basilar membrane

Figure B-2 a,b

When the basilar membrane is deflected by a sound wave, the hair cells are bent. This opens a
valve at the base of the hair call. (Figures 2 & 4 adapted from Kandel, E; Schwartz, J; and Jessell, T;
Principles of Neural Science, 3rd Edition, 1991; pages 486 and 489.)
Chapter 17
Atoms, Molecules
and Atomic Processes

CHAPTER 17 ATOMS, MOLECULES AND ATOMIC

PROCESSES

To extract the basic laws of mechanics from the variety Our knowledge of atoms comes not from looking
and confusion of the world around us required looking through microscopes, but instead from the study of
at matters on a large scale, looking out at the moon and chemical reactions, the measurement of the physical
planets whose motion is regular, periodic, and easier properties of substances, and the bombardment of
to understand. In this and the next two chapters we take materials with x-rays and other particles. This study
a similarly large leap to the small scale of distance essentially began with John Dalton’s construction of
where simplicity and periodic behavior again allow us the first periodic table in 1808. Other milestones were
to gain insight into the working of nature. Here we find Thomson’s discovery of the electron in 1895,
the world of atoms and their constituent particles, a Rutherford’s discovery of the atomic nucleus in 1912,
world in which we observe the basic forces and par- Neils Bohr’s model of the hydrogen atom in 1913, and
ticles ultimately responsible for the variety about us. the discovery of the rules of quantum mechanics in the
mid 1920s.
The jump down to the small scale of atoms is compa-
rable to the jump out from the study of projectile motion
in the lab to the analysis of satellite orbits. Imagine, for
example, that we could enlarge the golf balls used in
our strobe labs to the size of the earth. The same
enlargement of a hydrogen atom would give us an
object about the size of a golf ball.

Only with the development of the new generation of

microscopes in the late 1980s has it become possible to
see and work with individual atoms. Figure (1) is the
first atomic sized logo consisting of xenon atoms on a
background of nickel, made by scientists at the IBM
Research Laboratories in 1990. But despite great Figure 1
improvements in seeing and working with individual Thirty five xenon atoms were dragged across a nickel
atoms, the images we now get are still fuzzy and we are surface to form the letters IBM. (D. M. Eigler & E. K.
Schweizer, Nature, 5 April 1990.)
restricted to looking at atoms in solid structures, atoms
that do not move around.
17-2 Atoms, Molecules and Atomic Processes

MOLECULES
Atoms attract each other to form molecules, like the
water molecule H20 sketched in Figure (2). It was from
x-ray studies of ice, the crystalline form of water, that
we know the distance from the center of the oxygen
atom to the center of a hydrogen atom is .958 × 10 – 8 cm
oxygen
and that the hydrogen atoms are spread out at an angle atom 0.958 X 10
–8
cm
of 104.5 degrees as shown.

X-ray studies of large biological molecules began in hydrogen

atom
the late 1950s. For example, myoglobin is a substance
found in muscle tissue. The myoglobin molecule
contains over 2500 atoms, mostly carbon, hydrogen,
oxygen, nitrogen, and one iron atom. For determining
the precise structure of the myoglobin molecule from
x-rays of crystals of myoglobin, John Kendrew and
Max Peritz received the 1963 Nobel Prize in chemistry. 104.5°
Their model of the molecule is shown in Figure (3). Figure 2
The water molecule H2 O . We know the precise
Recent advances in computer modeling now provide location of the centers of the three atoms.
detailed views of numerous kinds of molecules. An
example is Figure (4) showing the cholera toxin B-
subunit.

Figure 4
Computer model of the cholera
toxin B-subunit. (Courtesy of
Argonne National Laboratory.)
 17-3

Figure 3
Model of the myoglobin molecule. (Photograph courtesy of J.C. Kendrew and H.C. Watson.)
17-4 Atoms, Molecules and Atomic Processes

ATOMIC PROCESSES
The myoglobin molecule provides a hint of the com-
plex structures that can be formed from atoms, a
complexity we wish to avoid in this chapter by concen-
trating on basic, simpler atomic processes. To help
illustrate these processes, we have illustrated in Figures
(5) through (11) a set of sketches drawn on the black-
board by Richard Feynman and copied to his book of
introductory physics lectures. Such simple sketches,
full of information, were characteristic of Feynman’s
style.

In the first three Figures (5, 6, and 7) we have views of

three forms of matter made from water molecules,
namely ice, water and steam. In the form of ice, the
water molecules fit into a hexagonal structure with a
hole in the center, as seen in Figures (5a,b). When
water freezes to form snowflakes, the hexagonal struc- Figure 5a
ture repeats and we get a six sided symmetry seen in all Ball and stick model of an ice crystal.
snowflakes, examples of which is shown in Figure
(5c).

When ice melts to form water, shown in Figure (6), the

rigid structure of ice disappears and the water mol-
ecules can now slide past each other. In addition the
holes in the hexagonal structure fill in with the result
that water is more dense than ice. That is why ice floats
in water.

The third form of water is steam, shown in Figure (7). hydrogen

oxygen
In the gaseous state the water molecules move freely Figure 5b
about, interacting only when they collide with each Sketch of the arrangement of the
other. The picture of steam is more or less what we water molecules in an ice crystal.
would see if we could look at the steam emerging from
a teakettle on an atomic scale. The separation of the
molecules is on the average about 10 times the diameter
of a water molecule. As a result, the steam is about 1000
times less dense than liquid water. The transition from
water to steam, either by evaporation or by boiling,
involves a competition between molecular forces and
thermal forces. We will discuss that competition shortly.

Figure 5c
Snowflakes reflect the 6-sided structure
of the ice crystal.
 17-5

Some atomic processes are illustrated in Figures (8) other to form solids like diamond, graphite, soot, or
through (10). Figure (8) is what you might see in a Buckeyballs (soccerball shaped structures of carbon).
snapshot of the surface of a glass of water. On our scale But there is a greater attraction between a carbon and an
of distance, such a surface looks quiet, but on an atomic oxygen atom than between two carbon or two oxygen
scale it is an active place with molecules continually atoms. If the carbon and oxygen are heated, the various
entering and leaving the water. Evaporation occurs if atoms bounce into each other at high speeds, the old
more water molecules leave the water than return. If structures break apart and molecules of carbon monox-
you put a cover over the glass, the concentration of ide and carbon dioxide are formed. Energy
water molecules in the air above the water builds up to is released in the process in the form of heat and light,
the point where just as many water molecules return as and we say that the carbon is burning.
leave, and the level of the water stops dropping. We
would then say that the evaporation has ceased.

In Figure (9) we see what happens to a block of carbon

when it is heated in an atmosphere of oxygen. By
themselves oxygen atoms combine in pairs to form O2
or oxygen molecules, and carbon atoms attract each

oxygen hydrogen nitrogen

Figure 8
Water evaporating in air.

Figure 6
Water magnified a billion times.

Figure 7
Steam.

oxygen carbon
Figure 9
Carbon burning in oxygen.
17-6 Atoms, Molecules and Atomic Processes

The process of salt dissolving in water is illustrated in THERMAL MOTION

Figure (10). Table salt is a stable crystal structure made What we have not been able to display in the sketches
from sodium and chlorine atoms. Strong electric forces of atomic processes is the constant juggling of the
hold these atoms together in the crystal. But let water atoms and molecules. This juggling, which we will call
molecules come into contact with the salt crystal, and thermal motion, becomes more intense as a substance
the water molecules work their way in between the becomes warmer and can cause major changes in the
sodium and chlorine atoms, allowing these atoms to structure of matter. When ice is warmed to above the
move freely and independently throughout the water. melting point, the juggling or thermal motion breaks up
One of our favorite sketches is Figure (11), the odor of the rigid structure of the ice crystal, allowing the water
violets. Many of our common experiences have a molecules to slide past each other to form liquid water.
simple origin on an atomic scale. With more heating, the thermal motion can increase to
the point that the water molecules fly apart.

Surprisingly this thermal motion can be seen on a

considerably larger scale than the scale of atoms. In
1827, the botanist Robert Brown observed that tiny
pollen particles in water, when seen through a micro-
scope, moved around in a juggling, random fashion.
Wondering whether these particles were alive and
swimming, Brown studied these and other small par-
ticles in circumstances where nothing should be alive,
and concluded that this random motion, now called
Brownian motion, had nothing to do with life, but was
related to the motion of the molecules.

With a laser, microscope and TV camera, it is easy to

Figure 10 chlorine sodium set up a demonstration of the Brownian motion of
Salt dissolving in water. cigarette smoke particles in air. The apparatus, shown
in Figure (12a), consists of a small cavity between two
microscopic slides, into which we inject smoke through
a small tube.
microscope
microscope
slides
laser
beam

smoke-filled
cavity
Figure 12a
Brownian Motion. A small cavity, made from
microscope slides is filled with cigarette smoke, is
illuminated from the side by a laser beam and viewed
oxygen hydrogen from above by a microscope and TV camera.
nitrogen water

carbon
Figure 11
Odor of violets.
 17-7

Once inside the cavity the smoke is illuminated from The air molecules themselves are not all the same; air
the side by a laser beam. Through the microscope, is mostly nitrogen and oxygen, some carbon dioxide
whose image can be displayed using a TV camera, we and water, and smaller amounts of other gases and
clearly see the illuminated smoke particles moving pollutants. It turns out that each species of molecule in
around in a relatively slow jagged motion. Figure (12b) the air has precisely the same average kinetic energy
is one minute movie showing the motion of the smoke due to thermal motion. As a result, the oxygen mol-
particles. ecules, for example, having a slightly greater mass than
the nitrogen molecules, must have a slightly smaller
In Figure (12c), a student recorded the motion of a average speed in order that the average kinetic energy
single smoke particle for 38 TV frames. Although we
1/2 mv 2 be the same. The smoke particles, with their
may think of smoke particles as being small, they are
huge masses, have a much slower average speed than
huge compared to the air molecules in which they are
the air molecules. The air molecules move at roughly
immersed. Smoke particles have a mass many orders
the speed of sound in air, while the smoke particles
of magnitude larger than that of the air molecules. Yet
move slowly enough for us to see and follow them on
the motion we see is caused by the constant bombard-
the TV screen.
ment of these huge particles by the air molecules.
16 Brownian motion
15
At first you might believe that if a large particle were of a smoke particle
constantly bombarded on all sides by billions of tiny 7
11
particles the effect of the collisions would cancel out 10 17
and the big particle would just sit there. But it turns out 12/14 18
8
that if the collisions are random, then fluctuations in the 9
13 19
6
collisions will cause the large particle to move around 5 20
with the jerky motion we see in the Brownian motion 25

demonstration. What is more remarkable, the average 4

21
38
kinetic energy of the smoke particles, as they wander 22 23
24
about, is the same as the average kinetic energy of the 3 37
air molecules bombarding them. 2 27

26
31/33 34
28
29 32 36
1
30

∆t = 1/6 second 35

0 .01 .02 .03 mm

Figure 12c
Brownian motion of a smoke particle. This is the result
of a student project by Lisa Stigler, where the motion of
a smoke particle was recorded by a TV camera using
the apparatus of Figure (12a). (We cannot use this plot
to estimate the average speed of the smoke particles,
Figure 12b
because the smoke particle undergoes many collisions
First frame of a one minute movie showing the
between TV frames. However this plot does illustrate
brownian motion of smoke particles. The frames
the random walk nature of the motion of the particle.
are 1/15 of a second apart in this movie. (Click
One feature of a random walk, that may be observable
on the image to see the movie. Press the “esc”
from plots like this, is that the average distance from
button to close the movie. )
the starting point should be proportional to the square
root of the elapsed time.)
17-8 Atoms, Molecules and Atomic Processes

Exercise 1 THERMAL EQUILIBRIUM

If we know the relative masses of the molecules in air, If you place a hot cup of coffee and a cold glass of milk
and know the average speed of one of the species, we on the table and leave the room for several hours, when
can use the fact that all species of particles have the you return the coffee and the milk are both at room
same average kinetic energy, in order to calculate the temperature. We say that the coffee, the milk, and the
average speed of the other particles. It turns out that if air in the room are in thermal equilibrium. A faster way
we have a sample of air at room temperature (27° C), the to reach thermal equilibrium, at least between the
average speed of the nitrogen molecule is 483 meters/ coffee and the milk, is to pour the milk into the coffee
sec. (We will calculate this number shortly.)
and stir.
Using the mass of a hydrogen atom as a standard mass
On an atomic scale, what does it mean to say that the
of 1, the mass of a nitrogen molecule is 28, and an
oxygen molecule 32. (These are often called the
molecules of the coffee and those of the milk are in
molecular weights of the molecules). In Table 1 we thermal equilibrium? We obtain a hint from our
have listed various constituents of the cigarette smoke, discussion of Brownian motion. The cigarette smoke
the relative mass of the particles, and the average represents a well stirred mixture of smoke particles and
thermal speed of two of the species. Use the fact that air molecules. These should therefore be in thermal
all the species have the same average kinetic energy to equilibrium, just as the well stirred molecules in the
fill in the table of average speeds. (We gave you the coffee and milk. In the case of Brownian motion, the
speed of helium so that you could check your calcula- smoke particles and the air molecules had the same
tions.) average thermal kinetic energy. We expect that the
same may be true for the molecules in the mixture of
coffee and milk.
Particle Symbol Mass* Speed** It is an almost general rule that when two objects are in
Hydrogen molecule H2 2.0 ... thermal equilibrium, the molecules that make up these
Helium atom He 4.0 1370 m/sec objects have the same average thermal kinetic energy.
Water molecule H20 18.0 ... When we first placed a hot cup of coffee and a cold cup
Nitrogen molecule N2 28.0 518 m/sec of milk on the table, the molecules in the coffee had a
Oxygen molecule 02 32.0 ...
greater average kinetic energy, and the molecules in the
Carbon Dioxide CO2 44.0 ...
milk a lesser average kinetic energy, than the mol-
Smoke particle 1010
ecules in the air. But after a few hours, the coffee
* Mass relative to Hydrogen atom
molecules slowed down and the milk molecules speeded
** Average speed at room temperature
up until all three sets of molecules, coffee, milk, and air
Table 1 attained the same average kinetic energy.
Particles involved in the Brownian motion
demonstration. Fill in the column for average The process of reaching thermal equilibrium is usually
speed, using the fact that these particles have a result of random collisions between molecules. If a
the same average kinetic energy. fast molecule collides with a slow one, chances are that
the slow one will speed up and the fast one will slow
down. It requires a detailed analysis, which we will not
attempt, to show that if the collisions are random, they
tend to equalize the kinetic energy of the particles.
 17-9

TEMPERATURE Absolute Zero

If there is any scientific concept familiar to everybody, An immediate consequence of temperature being re-
it is the concept of temperature. All of our lives we have lated to thermal kinetic energy, is that there must be a
been poked with thermometers and listened to weather lowest possible temperature. When the thermal kinetic
forecasts about tomorrow’s temperature. How is that energy is gone, you cannot go any lower in tempera-
quantity, measured by various kinds of thermometers, ture. It thus seems reasonable to define an absolute zero
related to the atomic processes we have been discuss- of temperature as the state where the molecules have no
ing? thermal kinetic energy, and choose a temperature scale
that starts at this absolute zero and goes up proportion-
Most thermometers are a black art which depends upon ally to the thermal kinetic energy.
such properties as the thermal expansion of mercury or
alcohol, the stiffness of a spring, or the color changes However, as you approach absolute zero, as you try to
of a material, etc. There is, however, one kind of a remove the last vestiges of thermal kinetic energy,
thermometer whose function can be understood from nature has a surprise in store. No matter what you do,
a simple molecular picture. That is the ideal gas there is some unremovable kinetic energy left. One of
thermometer which we will discuss shortly. We will the basic predictions of quantum mechanics is that a
see that for an ideal gas thermometer, the temperature confined particle cannot have zero kinetic energy, and
reading is proportional to the average thermal kinetic the closer the confinement the more kinetic energy it
energy of the gas molecules in the thermometer. has to have. A molecule in a liquid or a solid is confined
to the small volume bounded by its neighbors, and
When you measure the temperature of an object, you therefore cannot have a kinetic energy less than that
have to wait until the thermometer and the object are in required for that volume.
thermal equilibrium. (You wait until the reading on the
thermometer in your mouth stops changing.) When in The unremovable kinetic energy is called zero point
thermal equilibrium, the molecules of the object and energy. This energy is so small that for most sub-
those of the thermometer have the same average ther- stances it is not noticeable unless you carry out spe-
mal kinetic energy. If we are using an ideal gas cially designed experiments to detect it. However, zero
thermometer, the reading is proportional to this aver- point energy shows up clearly in the case of liquid
age kinetic energy. Thus if we use an ideal gas helium. All substances except helium freeze when
thermometer as an experimental definition of tempera- cooled to a sufficiently low temperature. We can
ture, we are effectively defining temperature as being remove enough kinetic energy from the molecules so
proportional to the average thermal kinetic energy of that they settle into a solid structure. But the molecular
the molecules. force between helium atoms is so weak that the zero
point kinetic energy alone is enough to keep helium a
liquid. You cannot freeze helium by cooling alone, you
must also subject it to high pressures.

The existence of zero point energy suggests that we will

encounter problems with the definition of temperature
as we approach absolute zero. Suppose, for example,
we have two substances with different zero point
energies in thermal equilibrium. If the temperature is
so low that any thermal kinetic energy is much less than
the zero point energies, then we have a situation in
 17-10 Atoms, Molecules and Atomic Processes

which molecules with different vibrational kinetic Temperature Scales

energies are in thermal equilibrium. If we insist that For the rest of this chapter, we will put aside any worries
two substances in thermal equilibrium are at the same about zero point energy, and simply assume that the
temperature, then we can no longer say that tempera- temperature of an object is proportional to the average
ture is proportional to the vibrational kinetic energy of thermal kinetic energy of the molecules in the object,
the molecules. and that absolute zero is where no thermal kinetic
energy remains.
The ideal gas thermometer does not get us out of this
problem because it does not work at very low tempera- From this point of view, the simplest way to define a
tures. Before the zero point energies become impor- temperature scale is to equate the temperature with the
tant, any gas we use in an ideal gas thermometer average thermal kinetic energy, and measure tempera-
becomes liquid or solid and we no longer have an ideal ture in energy units such as ergs as shown in Figure
gas as a working substance. (13). But you probably have not heard anyone describe
temperature in ergs, and for good reason. Telling your
In the next chapter on entropy and the second law of doctor that you are running a fever of 6.4423 × 10 – 14,
thermodynamics, we will discuss the consequences of an increase of 23 × 10 – 18 over normal could be a bit
the basic idea that order does not naturally arise from hard to explain when you are sick. It is much easier to
disorder. In that discussion we will describe a method say that you have a temperature of 100° F or about 38°
of defining temperature that applies to all temperature C. Ergs are too awkward a unit for most purposes.
ranges. This thermodynamic definition of temperature
is consistent with the ideal gas thermometer over the Historically, thermometers were invented and tem-
range that ideal gas thermometers operate, but also perature scales established long before the relation
correctly describes temperatures near absolute zero between temperature and the average kinetic energy of
where we have to deal with zero point energy. molecules became known. Throughout the world the
most widely used temperature scale is the Centigrade
scale, where the temperature of melting ice is arbi-
trarily set at 0° C (zero degrees Centigrade), and the
ergs absolute Centigrade Fahrenheit
average kinetic energy (Kelvin)
of gas molecules

– 14
7.72 x 10 ergs water boils
boiling point 7.72 x 373° K 100° C 212° F
normal temperature (98.6° F)
– 14 10 – 14ergs
5.65 x 10 ergs ice melts

freezing point 5.65 x 273° K 0° C 32° F

10 – 14ergs
dry ice

nitrogen becomes liquid

(liquid air)
helium becomes liquid
0 ergs absolute zero absolute zero 0 ergs 0° K – 273° C – 459° F

Figure 13a Figure 13b

Temperature scale in ergs. Comparison of various temperature scales.
 17-11

boiling of water at 100° C. Commonly, changes in

Boltzman's ergs
temperature are measured with a mercury thermom- = 1.38 × 10 – 16 (2)
eter. This device registers temperature changes when constant k kelvin
the mercury in a thin glass column expands or contacts.
On the Centigrade scale, the distance between 0° C and If you are using MKS units, the value of k is
100° C is marked into 100 equally spaced smaller 1.38 × 10 – 23 joules/kelvin .
intervals which we call degrees. The important feature of Equation 1 is that the average
A less arbitrary scale is the Kelvin or absolute scale, kinetic energy of the molecules is proportional to the
which measures temperature in Centigrade size de- absolute temperature measured in kelvins. We have
grees beginning at absolute zero. Using the absolute written the proportionality constant as 3/2 k, putting in
scale, we find that helium boils at 4 degrees Kelvin, ice the numerical factor of 3/2 to get rid of a factor 2/3, as
melts at 273 degrees Kelvin and water boils at 373 you will see shortly. Basically, think of Boltzman’s
degrees Kelvin. A comparison of various temperature constant as the conversion factor to go from tempera-
scales (ergs, degrees Kelvin, degrees Centigrade, and ture units to energy units or vice versa.
degrees Fahrenheit) is shown in Figure (13b).
Exercise 2
Those who define standard nomenclature for physical
Use Equation 1 to calculate the temperature of melting
quantities have decided, in their great wisdom, that the ice in ergs. Compare your answer with the result in
word “degrees” shall be omitted when talking about Figure (13).
temperature in degrees Kelvin. Thus we should say
that helium boils at 4 kelvins or 4K, ice melts at 273 Exercise 3
kelvins or 273K, and the temperature difference be- What would be the temperature in Kelvins of a gas if the
tween melting ice and boiling water is 100 kelvins or particles in the gas had an average kinetic energy of 1
100K. At least this nomenclature is easy to say and erg?
should not be confusing when you get used to it. We
do not feel the same way about all recent changes in Exercise 4
nomenclature. In Exercise 1 we said that the average speed of nitrogen
molecules at room temperature was 518 meters/sec,
The conversion from one temperature scale to another and asked you to use that result to calculate the average
is a relatively straightforward process. If you went to speed of the other molecules and particles in the
an American school, somewhere along the way you cigarette smoke. Now you are to calculate the speed of
were taught how to convert from Fahrenheit to Centi- the nitrogen molecules using the fact that their average
grade degrees. You do not need to worry about that kinetic energy is 3/2 kT.
because we will not be using the obsolete Fahrenheit
It is traditional to take room temperature as 300K = 27°
scale. But we will often want to convert from the
C. Assume that a nitrogen molecule is 28 times as
absolute scale to the energy units, ergs or joules. The
massive as a hydrogen atom, whose mass is essentially
conversion is written in the somewhat peculiar form the same as a proton, or 1.67 × 10– 24 grams. See if you
get the answer of 518 meters/sec.
average kinetic energy
3
of gas molecules = kT (1)
2
in ergs or joules

where T is the temperature in kelvins, and the conver-

sion factor k, known as Boltzman’s constant has the
numerical value
17-12 Atoms, Molecules and Atomic Processes

MOLECULAR FORCES pressure, elasticity of rubber, and the behavior of an

Much of the behavior of matter we see as we look ideal gas. At the end of Chapter 19 we will discuss the
around us is the result of a competition between mo- so-called Leonard Jones potential as a model for
lecular forces holding atoms together and thermal molecular forces. That model is far more detailed than
motion tending to pull them apart. Molecular forces we need for our current discussion. All we need to
can be subtle enough to form objects as complex as the know now is reviewed in Figure (14).
myoglobin molecule. Yet knowing just some of the In Figure (14a), where the atoms are about an atomic
basic features of molecular forces is enough to provide diameter apart, the attractive molecular force between
an insight into processes like evaporation, osmotic the two atoms is less than one percent of its maximum
value. The point is that unless the atoms are very close
together, within an atomic diameter of each other,
r molecular forces are negligible. This is why atoms in
less than 1% of
a gas often act as independent free particles. In the air
a)
the maximum force we breath, the average spacing of atoms is about ten
molecular diameters, so that molecular forces play no
role except when molecules collide.
b) half the maximum
attractive force When atoms get closer than an atomic diameter, the
attractive molecular force increases rapidly, reaching a
maximum at a separation at about one tenth of an
c) maximum atomic diameter as shown in Figure (14c). Then the
attractive force
force rapidly drops to zero at the spacing shown in
Figure (14d). When the force is zero, this is the
equilibrium distance which determines the size of the
d) equilibrium
(no force) atom in a chunk of matter. Effectively we can say that
when the atoms are at their equilibrium separation, they
are just touching, as we drew them in Figure (14d).
e) strong
repulsion Try to shove the atoms closer together than the equilib-
rium position, and you encounter a repulsive force that
potential
builds very rapidly, much faster than the attractive
energy of repulsive force increases as you pull the atoms apart. This
the two core
particles repulsion makes atoms behave as hard, nearly incom-
f) r pressible spherical objects. This repulsive force is
r=0 (separation between often referred to as the repulsive core of the atom.
molecular centers)
maximum
attractive force In Figure (14f) we have sketched the potential energy
no force corresponding to the molecular force. As you can see
Figure 14 the potential energy forms a well with the bottom at the
Interaction of two atoms via a Leonard Jones equilibrium position. When two atoms form a mol-
potential (f). When the atoms have an equilibrium ecule, like hydrogen ( H 2 ), oxygen ( O 2 ) or nitrogen
separation (d), their potential energy of interaction is ( N 2 ), you can picture one of the atoms as sitting in the
a minimum, and we can visualize one of the atoms as
potential well created by the other, and vice versa. We
sitting at the bottom of the potential energy well. If
the separation either increases or decreases, there is a only have to think about one of the atoms, for the same
force back toward equilibrium. The repulsion quickly thing is happening to the other.
builds up if you try to shove the atoms together, and
the attraction dies rapidly after the atoms become
separated by about one atomic diameter.
 17-13

If the atom is sitting at rest at the bottom of its potential The advantage of the potential energy diagram is that
well, it is at its equilibrium position shown in Figure it allows us to think of atomic processes in terms of the
(14d). If the atom moves either way, in or out, it is energy involved. The distance from the zero of poten-
subject to a restoring force pushing it back to the tial energy down to the bottom of the well is the binding
equilibrium position. If it does not move too far from energy of the molecule as indicated in Figure (16). This
its equilibrium position, the restoring force is very is the energy required to pull a molecule apart, starting
similar to the restoring force of a spring at equilibrium, with the atoms in their equilibrium position.
as indicated in Figure (15). That is why one can often
quite accurately picture molecular forces as spring We have seen that the average thermal kinetic energy
forces between atoms. of an atom or molecule is 3/2 kT where k is Boltzman’s
constant and T the temperature in Kelvins. If the
thermal kinetic energy is much less than the binding
Leonard Jones
potential
energy of the molecule, as we have shown in Figure
(16), then the atom can move back and forth around the
bottom of the potential well but not climb out.
parabolic
approximation From the depth and shape of the potential well one can
of spring force deduce general features of the behavior of matter, such
as why solids and liquids expand when heated. But
equilibrium before we look at such fine details, there is much to
separation
understand about atomic processes just from the fact
a) Comparison of the Leonard Jones potential and the
that there is an attractive molecular force with a repul-
parabolic potential of a spring force. sive core. We will look at these more general features
first and then return to the details we see in Figure (16).

equilibrium potential
separation energy of
molecular force
binding
energy of
parabolic energy
well of spring force molecule
bottom of Leonard Jones
potential energy well 3/2 kT
b) Modeling the molecular force as a spring force. As long
as the atom stays near the equilibrium position, the Figure 16
Leonard Jones force and the spring force are equivalent.
Binding energy. If an atom is in its equilibrium
position, we can think of it as sitting at the bottom
Figure 15 of its potential energy well. To remove the atom
Physics and chemistry texts often picture molecules from the molecule, we have to lift it out of the well.
with spring forces between the atoms. At first this Thus the binding energy is the depth of the well. If
may seem to be a highly unrealistic picture. But the atom has a thermal kinetic energy 3/2 kT, and
when you carefully compare the spring potential this thermal energy is less than the binding energy,
energy and the Leonard Jones potential energy right the molecule should stay together.
near the equilibrium position, the two curves have
the same shape. Thus the spring force is a good
model as long as the atoms stay near their
equilibrium positions.
17-14 Atoms, Molecules and Atomic Processes

Evaporation surface requires that you do work against these attrac-

Simple features of molecular forces lead to a reason- tive forces. The amount of work required to extract a
able understanding of the transition from a liquid to a molecule from the surface depends upon the type of
gaseous state, the process of evaporation. We start with liquid and the temperature of the liquid, but some
a picture of a liquid as a collection of molecules that all energy is required as long as the surface exists.
attract each other, can move around past each other, but
are nearly incompressible because the repulsive core in Example 2
the molecular force prevents atoms from being squeezed To estimate the amount of energy required to extract a
into each other. The incompressibility of water can be water molecule from the surface of water, we note that
seen from the fact that water in the deepest parts of the to boil 1 gram of water requires 2.25 × 1010 ergs of
ocean, where the pressures are some 800 times atmo- energy. Since there are 3.3 × 1022 molecules in 1 gram
spheric pressure, is only about 3% denser than the of water, this represents an energy of 7 × 10–13 ergs per
water at the surface. molecule. Some of the energy you supply goes into
displacing the air above the water to make room for the
A molecule in a liquid is free to move around because steam, but most of it goes into supplying the energy
of its thermal kinetic energy and because there is each molecule needs to escape water at 100° C (373K).
essentially no net force on it. Although attracted to all
of its neighbors, the neighbors surround the molecule Exercise 5
as shown in Figure (17), and the net force is zero. (a) What is the average kinetic energy of a molecule at
a temperature of 373K?
The situation is different for a molecule on the surface
(b) Is this enough energy for an average water molecule
as shown in Figure (18). Such a molecule has neigh-
to escape through the surface of the water?
bors only to the sides and below. If we try to lift such
a molecule out of the surface, there will be a net force (c) At what temperature does the thermal kinetic energy
exerted by all of the molecules beneath it, pulling the equal to the 7 × 10–13 ergs needed to escape?
molecule back in. To extract a molecule from the

Figure 17 Figure 18
A molecule in the interior of a liquid is attracted by A molecule on the surface(b)is attracted to its neighbors
all its neighbors which surround it. As a result the beneath it. To pull a molecule out of the surface, you
net force is zero and the molecule is free to move have to overcome these forces. As a result it takes
about through the liquid. energy to remove a molecule from the liquid. This
surface force is often referred to as surface tension.
 17-15

If you worked Exercise 6, you realize that the aver- Whether you get evaporation or condensation de-
age molecule, even in boiling water, does not have pends upon the number of water molecules in the air
nearly enough thermal kinetic energy to escape above the water. If you cover a glass of water with
through the surface. Yet even at room temperature a dish, soon the number of water molecules in the air
water evaporates; even at these lower temperatures in the glass builds up to the point that there is a
some molecules escape through the surface. The balance between molecules leaving and molecules
reason is that, while 3/2 kT is the average thermal entering the liquid surface. When this balance is
kinetic energy of the molecules, some molecules achieved, evaporation ceases and we say that the air
have more kinetic energy than average, some less. above the water is at 100% relative humidity. In
Some have so much more kinetic energy than aver- order to get cooling from evaporation, the relative
age that they can escape. humidity of the air must be less than 100%.
The rate of evaporation depends very much on the The human body uses evaporation for cooling which
distribution of thermal kinetic energies. At a given is effective on a hot, dry day but not on a humid one.
temperature T what fraction of the molecules have a When the relative humidity approaches 100% there
kinetic energy sufficiently far above average to be is no net loss of water molecules and no cooling.
able to escape? It turns out that for a substance in Incidentally, you blow on soup to cool it, not neces-
thermal equilibrium, there is a precise formula for sarily because your breath is cooler than the soup,
the distribution of thermal kinetic energies, a for- but because you are replacing the moist air over the
mula known as the Boltzman distribution which we soup with drier air so that more evaporation can take
discuss in Chapter 22. For now we will not go into place.
that much detail. Instead, we will simply recognize
that some molecules are hotter than average, some
colder than average, and that it is the very hottest
ones that have enough energy to escape.

If it is the hot molecules that escape during evapora-

tion, then the cooler ones must be left behind and
evaporation must be a cooling process. There must,
however, be a net loss of molecules from the surface
for cooling to occur. As we noted at the beginning
of the chapter, the surface of water is a dynamic
place where water molecules are continually leaving
and returning. A returning water molecule, even if
relatively cool when in the air above the water, gains
as much kinetic energy when it reenters the water as
the hot molecule lost when escaping. Thus reenter-
ing molecules become hot when they get back in the
water, and thus the returning or condensation of
water molecules is a warming process.
17-16 Atoms, Molecules and Atomic Processes

PRESSURE One of the simplest demonstrations of the pressure

When you try to compress a liquid, you are trying to exerted by a gas is provided by a rubber balloon. When
shove the atoms into each other which is very difficult you blow up a balloon, the rubber of the balloon is
to do because of the repulsive core of the molecular trying to compress the gas, force it down to a smaller
force. It is also hard to compress a gas—try blowing air volume. The molecules of the gas exert an outward
into a soda bottle. But when you compress air you are force on the rubber, preventing it from collapsing. This
not squeezing air molecules together, you are not trying outward force is caused by the collisions of the gas
to overcome a molecular force at all. The air that you molecules with the rubber, as illustrated in Figure (19).
breathe is mostly empty space, the separation of air Whenever an air molecule strikes and bounces off of
molecules being about 10 molecular diameters. There the rubber, there is a net transfer of outward directed
is a completely different explanation for why it is linear momentum to the rubber. Since the collisions are
difficult to compress a gas, why a gas exerts a pressure occurring continually, there is a continual transfer of
that you must overcome to compress it. momentum to the rubber, which, according to Newton’s
second law F = dp/dt, means that the molecules exert
a force on the rubber. On the average, the direction of
momentum transfer is outward, perpendicular to the
surface of the rubber. Thus the force exerted by the gas
molecules is also perpendicular to the surface.
The total force exerted on some part of the balloon
depends upon the area of the balloon we are talking
about. We can simplify the discussion by talking about
the force on a unit area of the balloon surface, and give
that force the special name pressure. We know that
force is a vector quantity, but the force that a gas exerts
on a surface is always directed perpendicular to the
surface, no matter what orientation the surface has.

Figure 19a Figure 19b

The air molecules bouncing off the inside surface of Balloon placed on liquid nitrogen. If you cool the air
the balloon, transfer an outward directed momentum to molecules inside the balloon, they do not strike the
the rubber. The average momentum per second rubber as hard, exert less of a force, and the balloon
transferred by these collisions is the average force the collapses. In the final picture, there is only a puddle
molecules exert on a section of the balloon surface. of liquid air inside.
 17-17

Thus we can let the word “pressure” stand for the If you take the balloon out of the liquid nitrogen, the air
magnitude of the force on a unit area, and determine the inside warms up and the balloon expands again to its
direction of the force from the orientation of the sur- original volume. Remarkably, a typical balloon can
face. undergo this cycle a number of times without breaking.

magnitude of the force Stellar Evolution

pressure The balloon has features in common with our sun. The
= exerted by the gas on a (3)
of a gas sun is not the solid object it appears. Abetter model is
unit area of surface
that the sun is a bag of gas, somewhat like a balloon, but
with the constraining force of the rubber replaced by
In the case of a gas inside a balloon, it is clear that the gravity. The sun looks like it has a distinct surface, but
behavior of the gas molecules is more or less the same that is an illusion created by the fact that the gas
throughout the balloon. As a result, the pressure should molecules in the sun, which are mostly hydrogen,
be essentially the same on all surface areas of the become ionized and opaque at temperatures in excess
balloon. Instead of using the word “pressure” to merely of 3000 Kelvins. As you go down into the sun, the
describe the force on surfaces of the balloon, we can say temperature of the gas increases. The distance at which
that the pressure is in the gas itself. Once you know the it reaches 3000 Kelvins, the gas changes from transpar-
pressure of the gas, you can then calculate the force the ent to opaque and that is what we see as the surface of
gas exerts on some particular surface by multiplying the sun.
the pressure times the area of the surface, and noting
that the force is directed perpendicular to the surface. For the past five billion years, the sun has gotten energy
from the conversion of hydrogen nuclei to helium
The pressure of the gas, the force the gas molecules nuclei. And there is another five billion years worth of
exert on a surface, depends upon how fast the mol- hydrogen left before the sun runs out of fuel. At that
ecules are moving when they hit the surface. The faster point the sun will do something spectacular. It will
their average speed, the greater the force and pressure. expand so that the earth will be orbiting near the sun’s
Since the motion of the molecules is thermal motion, surface. (We discuss this process in more detail in
whose average kinetic energy is 3/2 kT, the average Chapter 20.) But eventually the sun will settle down
speed and pressure must increase with temperature. If and begin to cool off.
you heat the balloon, the gas pressure increases and the
balloon expands. If you cool it, it contracts. As the sun cools, it will contract very much like the
balloon in our demonstration. And like the balloon, the
An excellent demonstration of the dependence of air collapse will be halted when the atoms are so close
pressure on temperature is to place a balloon in a bucket together that the repulsive core of the molecular forces
of liquid nitrogen, as shown in Figure (19b). A prevents further contraction. At that point the sun will
common Styrofoam ice bucket makes an excellent have become what is called a white dwarf, an object
container for liquid nitrogen for this demonstration. about the size of the earth slowly cooling until it
When you place the balloon on the surface of the liquid becomes a dark ember.
nitrogen, the balloon sits there for a while, and then
begins to shrink as the air molecules cool down. The If the sun were just a bit bigger, about 1.4 times its
shrinking continues until the balloon collapses and all current mass, the gravitational collapse would not be
you have inside is a puddle of liquid air. Now any halted by the molecular forces between atoms. That is
further contraction of the balloon would require squeez- the mass at which gravity is strong enough to crush the
ing the air molecules themselves together which is atoms together, to overcome the atomic repulsive
opposed by the repulsive core of the molecular forces. cores, with the result that you end up with a neutron star.
In a sense, the collapse of the cooling gas in the balloon That is also a topic we discuss in Chapter 20.
was halted by molecular forces.
17-18 Atoms, Molecules and Atomic Processes

THE IDEAL GAS LAW the cylinder from expanding. We are assuming that
We have discussed the picture of how the repeated there is no gas behind the piston, so that only the force
collisions of the gas molecules inside a balloon exert an F keeps the gas from expanding.
outward force on the rubber, and how, if we raise the We know that the force F exerted by the gas increases
temperature, the molecules travel faster, strike harder, with temperature because the balloon expanded when
and exert a greater force. We will now, with a fairly we heated it. We also know that the average thermal
simple derivation, obtain an explicit relation between kinetic energy of the gas molecules increases as 3/2 kT
the temperature of the molecules and the force they as the temperature rises. What we wish to do now is
exert, a relationship known as the ideal gas law. relate these observations to obtain a formula for how
There are many ways to derive the ideal gas law, the force F depends upon the temperature T.
depending upon what assumptions you are willing to To relate F to T we start with the simplest possible
make about averaging over molecular speeds. The less
model of the gas in the cylinder, namely a gas consist-
you are willing to assume, the harder the derivation is.
ing of one molecule, bouncing back and forth at a speed
But there is one rather surprising feature of all the
v, as shown in Figure (21). Each time the molecule
derivations. They all give the same correct answer.
strikes and bounces off the piston, its momentum
The usual procedure in textbooks is to make the deri-
changes by 2mv. Since linear momentum is conserved
vation as complex as students will tolerate, apologize
for or hide the approximations, and announce that the during the collision, the piston picks up an outward, x-
answer is correct. What we will do is present the directed, linear momentum of magnitude 2mv as a
simplest derivation we can find that gives the right result of the collision. (Remember problem 7-5, where
answer. When an argument looks too simple to be true, two skaters on frictionless ice were tossing a ball back
but gives the right answer, that means that you may and forth. When one of the skaters caught the ball, she
have extracted an important basic feature from a com- picked up the ball’s momentum mv. When she threw
plex situation. the ball back, she recoiled, picking up an additional mv
of momentum. As a result she picked up 2mv of
The Ideal Piston momentum with each catch and toss.)
While a balloon is a very practical container for gas, the
If we designate by ∆p the magnitude of the x-directed
curved surfaces make it a bit difficult to use for theoreti-
linear momentum that the piston gains from each
cal analysis. Instead we will, more or less as a thought
collision we have
experiment, use an idealized device called the friction-
less piston. Figure (20) is a diagram of a frictionless momentum
piston in a cylinder of cross-sectional area A. In the transferred ∆p = 2mv (4)
per collision
cylinder is a gas -- like air at room temperature. The gas
molecules are bouncing around, colliding with the The momentum ∆p = 2mv is transferred to the pis-
walls of the cylinder and the face of the piston. Because ton each time the molecule comes back and strikes the
of the collisions, the gas molecules exert a force on the piston. If ∆ t is the time between collisions, then the
piston, and because the piston is frictionless, we must amount of momentum transferred per second is ∆p/∆t .
exert an oppositely directed force F, as shown, to keep
z
v
area A F x
m
y
x
Figure 20
Figure 21
Frictionless cylinder in a piston. Picture the force that
Analysis of a one molecule gas.
the gas molecules exert on the piston as being
counterbalanced by an external force F as shown.
 17-19

Using Newton’s second law, in the form F = dp/dt , we Since, in this approximation, all the molecules have the
see that this rate of transfer of linear momentum ∆p/∆t same speed v, then the factor 1/2 mv 2 in Equation 8
is just the average force F that the molecule is exerting must be the average thermal kinetic energy of the
on the piston molecules. Replacing 1/2 mv 2 by 3/2 kT gives
average force average rate
F N molecule = 2 N × 3 kT
F = ∆
exerted by p of momentum gas (9)
3 2
molecule ∆t transferred (5)
on piston to the piston Now you see why the factor of 3/2 was inserted into the
To calculate the time ∆ t between collisions, note that formula 3/2 kT for the average thermal kinetic energy.
if the distance from the end of the cylinder to the piston The 3/2 cancels the 2/3 that appeared in Equation 9, and
is , and the molecule is traveling at a speed v, it covers we get
the distance down and back, 2 , in a time
F N molecule
gas = N kT (10)
∆t = 2 cm = 2 sec (6)
v cm/sec v
We are almost finished. Equation 10, despite our
With Equation (4) for ∆p and (6) for ∆t in (5), we get, approximations, is the correct formula for the force
for the average force F exerted by this one molecule of exerted by a gas of N molecules at a temperature T. The
gas only problem with Equation 10 is the explicit depen-
∆p 2 dence on the length of the cylinder. We can remove
F = = 2mv = mv (7) this explicit dependence by expressing the force on the
∆t 2 /v
cylinder in terms of the pressure P of the gas.
Note the appearance of mv 2 in Equation 7. We are
already beginning to see a connection between the In our earlier discussion, we said that the pressure of a
molecule’s kinetic energy and the force it exerts. gas inside a balloon was equal to the force per unit area
exerted by the gas on the surface of the balloon (Equa-
If you could actually set up a one molecule, one tion 3). If we have a gas at pressure P in a cylinder of
dimensional, gas like that shown in Figure (21), Equa- cross-sectional area A, as shown in Figures (20) and
tion 7 would accurately describe the average force of (21), then the force exerted on the piston, whose area is
that gas on the piston. No approximations have been A, must be
made yet. The approximations enter when we go to a F = PA
pressure
(11)
gas of N molecules, moving in three dimensions, at times area
various speeds. The simplest, most outrageous ap- Substituting Equation 11 for F in Equation 10, and
proximation we can make is that all of the molecules multiplying through by the cylinder length , gives
have the same speed v, and that 1/3 of them are
bouncing back and forth in the x-direction, 1/3 in and PA = N kT (12)
out in the y-direction, and 1/3 up and down in the z- The final step is to note that A , the area of the cylinder
direction. Such a gas with N/3 molecules bouncing times its length, is the volume V of the cylinder. Thus
back and forth, would exert a force N/3 times as great we get
as our one molecule gas in Figure (21)
PV = NkT ideal gas law (13)
F N molecule
gas = N F 1 molecule
gas
3
Equation 13 is known as the ideal gas law. Despite the
= N mv 2 approximations we used to derive it, it is accurate as
3 long as the particles in the gas are separated enough that
one can neglect the molecular forces between particles.
= 2 × N × 1 mv 2 (8) To express this another way, any gas that obeys Equa-
3 2
tion 13 is known as an ideal gas. To a very high degree
of accuracy, the air around us behaves as an ideal gas.
17-20 Atoms, Molecules and Atomic Processes

Ideal Gas Thermometer plug will slide down until it sits on the remaining air.
The ideal gas law, PV = NkT, incorporates such laws There will be a vacuum above the plug. How high the
as Boyle’s law and Charle’s law which you may have plug rides depends upon the length of the mercury
encountered in an introductory chemistry course. One plug and how much gas you left in the tube before
can construct numerous examples and homework prob- sealing it.
lems applying this law. What we will do for our first
application is to describe the ideal gas thermometer We can use the ideal gas law to predict how the height
which we will use, at least for now, as our experimental of the plug varies with the temperature of the gas in the
definition of temperature. tube. When we do this, we obtain a messy looking
formula with factors like the density ρ of the mercury,
An example of an ideal gas thermometer is shown in
the number N of air molecules in the tube, the area A of
Figure (22). The glass tube and the plug of mercury
the tube, the acceleration g due to gravity and Boltzman’s
come about as close as you can get to a cylinder with a
constant k. But when we take another look at the result
frictionless piston. You can make one of these devices
we see that most of the factors are constants, and the
by sealing a glass tube at the bottom, pouring some
height h of the air column turns out to be strictly
mercury in, evacuating most of the air above the
proportional to the temperature T of the gas in the tube.
mercury, and sealing the top of the tube. If the tube is
Let us see how this all works out.
fairly small, when you turn the tube over, the mercury
Rewriting the ideal gas law as an equation for the
temperature T of the gas molecules, we get

T = PV = PhA (14)
Nk Nk
vacuum where V = hA is the volume occupied by the air which
is in a column of height h and area A. The mercury plug
of length riding on top of the gas, exerts a gravitational
mercury force mg on the gas, where the mass m of the mercury
plug
is equal to the mercury’s density ρ times its volume A.
Thus
glass tube
weight of
mercury = mg = ρ Ag (15)
air column
h
The force mg is the total force exerted by the mercury
column on the air. The force per unit area, which must
equal the pressure of the gas if the plug is balanced on
the gas is
Figure 22 pressure of gas
Ideal gas thermometer. If we heat the gas beneath beneath a plug mg ρ Ag
the mercury plug, the gas expands raising the plug. of mercury
P= = =ρ g (16)
A A
If we have an ideal gas, then the height of the plug of length
depends only on the temperature of the gas and not
the kind of gas we used. Equation 16 will turn out to be useful in other experi-
ments, for it tells us how to measure the pressure of a gas
in terms of the height of a mercury plug that the gas
can support.
 17-21

Using Equation 16 in 14, we get the desired result you have only 2 experimental points, the height at 0° C
ρ gA and at 100° C. Connect these two points by a straight
T = PhA = h = χh (17) line (that is what the formula T = χh says you should
Nk Nk
do), and you find that h goes to zero at a temperature of
where χ = ρ gA/Nk is a collection of constants. The – 273° C. That is all there is to it!
basic result is that the gas temperature T is strictly
From our discussion of molecular forces, you can see
proportional to the height h of the air column.
that any ideal gas thermometer you actually build
To use an ideal gas thermometer, we do not need to has to fail before you get to absolute zero. At some
evaluate the constants in the formula for χ . Instead point as you cool the air in the thermometer, you end
immerse the thermometer in ice water and mark the up with a puddle of liquid air as we did in the balloon
bottom of the plug 0°C. Then put the thermometer in demonstration. Even before the air becomes liquid,
boiling water and mark that 100°C. Mark off the the spacing between the air molecules is reduced to
distance between 0°C and 100°C in 100 equally spaced the point where the molecular forces between air
intervals and you have a centigrade thermometer. molecules becomes important. The attractive mo-
lecular forces reduce the pressure of the gas, the gas
The fascinating feature of an ideal gas thermometer is no longer obeys the ideal gas law, and we cannot
that you can quickly determine the temperature at
believe the readings of the thermometer. This prob-
which the gas volume should go to zero, the tempera-
lem can be put off by using helium gas that remains
ture we have called absolute zero. On a sheet of graph
a gas down to a temperature of 4 kelvins, but that’s
paper, mark off a temperature scale on the bottom that
the limit. To work at temperatures closer to absolute
runs backwards from 100° C to 0° C and goes on out
zero you need a different experimental definition of
quite away into negative temperatures. On the vertical
axis plot the height h of the air column. For this plot, temperature, like the thermodynamic definition we
discuss in Chapter 18.

Height of
air column Figure 23
Absolute zero. If you plot the height of the air
100°C • column in an ideal gas thermometer as a function
of temperature, drawing a straight line between the
Height two known data points at 100° C (boiling water) and
At 0°C • 0° C (melting ice), and continue the line down to
zero height, the intersection is at – 273° C. This
represents an absolute low value for temperature, as
defined by the ideal gas thermometer.

0cm
100°C 0°C -100°C -200°C -273°C
Temperature
17-22 Atoms, Molecules and Atomic Processes

The Mercury Barometer in Figure (25), the molecules of the air are colliding
and Pressure Measurements with the surface of the mercury, exerting a force in the
Columns of mercury are useful not only for making same way that the air molecules in a balloon push out
thermometers, but also for making devices to measure on the rubber. If the air molecules are at a pressure Pa
pressure. We will begin with a discussion of the (pressure of the atmosphere) and the glass tube has an
mercury barometer whose construction is shown in area A, the force exerted by the air is the pressure times
Figure (24). the area.

As shown in Figure (24a), start with a u shaped glass force exerted

by atmosphere = Pa A (18)
tube about a meter long, sealed at one end, and work on air column
mercury into it until the sealed section is nearly full.
Then invert the tube as shown in Figure (24b). The The weight mg of the mercury column pushing down
mercury in the sealed section will slide down, leaving at point (2) is
a vacuum behind it, until the difference in the heights mg = ρ Ah g (19)
of the mercury columns is about 76 cm as shown.
where ρ is the density of mercury and Ah is the volume
The height difference of the two columns tells us the of mercury in the column of area A and height h.
pressure of the atmosphere. To see why, conceptually
weight of pressure
break the mercury column up into two parts as shown mercury of the
in Figure (24c). The bottom part is the loop of mercury column atmosphere
that goes from point 1 (at the open end of the mercury) mg Pa A
to point 2 (at the equal height in the closed section). The
upper part, goes from point (2) up to the vacuum, a 2 1
section whose height we designate by the letter h. This
column sits over the bottom loop and exerts a down-
ward force equal to the weight mg of a column of
mercury of height h.

The mercury in the bottom section between points (1)

and (2), is completely free to move up one side or the
other. Since it does not move, the weight mg of the Figure 25
mercury column pushing down on the left side at point The weight of the mercury column above
(2) must be balanced by the force of the atmosphere point (2) must be balanced by the force
pushing down at the open end, point (1). As indicated exerted by the atmosphere at point (1).

mercury area A

Figure 24 vacuum
Construction of a mercury
barometer. When you turn the
about 100 cm

mercury
76 cm

tube over, going from (a) to (b),

h
the mercury slides down the
sealed leg, leaving a vacuum
behind. The difference in
heights h of the two columns (c) 2 1
is a measure of atmospheric
pressure.
mg Pa A

(a) (b) (c)

 17-23

Equating the forces on the two sides of the mercury in What is much easier is to simply express the pressure
the bottom section gives in terms of the height of the mercury column that the
atmospheric pressure will support. “A low pressure
Pa A = mg = ρhAg
system moved over the area today, and the barometer
reading dropped to 752 millimeters of mercury” sounds
Pa A = ρgh
(20) much better than saying that the pressure dropped to
1.0023 x 10 5 pascals.
The result is that the atmospheric pressure is propor-
tional to the height difference h in the two columns of The millimeter of mercury, as a unit of pressure, has
mercury. As we have mentioned, the dimensions of been given the name torr, one more name inflicted
pressure in CGS units is dynes per square centimeter, upon students by those who decide what the standard
while in MKS units it is Newton’s per square meter, a names shall be. For comparison’s sake, we can express
set of dimensions given the name pascal. Neither set atmospheric pressure as
of units is particularly convenient. Using
ρ = 13.6 gm/cm 3 for the density of mercury, and Pa = 1.01 × 10 6 dynes/cm 2
using the value h = 76 cm for an average value for the
= 1.01 × 10 5 pascals
height of the mercury column, we get
= 101 kilo pascals
gm
Pa A = 13.6 × 980 cm2 × 76 cm = 76 cm Hg (22)
cm 3 sec = 760 mm Hg
2 = 760 torr
gm cm/sec
= 1.01 × 10 6 = 14.7 lbs/in 2
cm 2
At different times you may encounter any of these
Pa A = 1.01 × 10 6 dynes /cm 2 (21a) units.
Converting to MKS units, where one newton When you are working with vacuum pumps and
equals 10 5 dynes and 1 m 2 = 10 4 cm 2 we have vacuum gauges, even the torr, 1 millimeter of mercury,
is too large a unit to be convenient. Many gauges are
dynes 10 4 cm 2 m 2 calibrated in microns, which is the pressure exerted by
Pa A = 1.01 × 10 6 ×
cm 2 10 5 dynes/newton one micron or one millionth of a meter of mercury.
1 "micron" = 10 – 6 meters Hg
= 1.01 × 10 newtons
5
= 10 – 4 cm Hg
m2
= 10 – 3 mm Hg (23)
= 1.01 × 10 5 pascals (21b) = 10 – 3 torr
In the electron gun experiments we discuss in Chapter
Just as it was inconvenient to measure temperature in 28, the glass tube containing the electron beam is
ergs, it is rather inconvenient for the weatherman to evacuated to a pressure of around one micron. Current
announce today’s barometric pressure in either dynes technology allows you to work with much better
per square centimeter or pascals. Numbers in the range vacuums in the range of 10 – 6 to 10 – 7 microns. Such
of 10 6 or 10 5 do not go over well with the listening vacuums are needed to maintain clean surfaces when
audience. studying the atomic structure of surfaces or creating
complex electronic chips.
17-24 Atoms, Molecules and Atomic Processes

Exercise 6 The number 6 × 10 23 (more accurately 6.02 × 10 23 ),

What is the pressure P in dynes/cm2 and pascals,
which is known as Avogadro’s number or constant, is
inside an apparatus where the pressure gauge reads essentially the number of hydrogen atoms in one gram
of hydrogen. Since hydrogen atoms and protons have
a) 1 cm Hg essentially the same mass, a mole of protons also has a
b) 1 micron
mass of 1 gram.

c) 1 kilo pascal When you have a bottle of hydrogen gas, the hydrogen
atoms combine in pairs to form hydrogen molecules.
d) 1 torr Thus a mole of hydrogen molecules has a mass of 2
e) 10–9 torr grams. An oxygen atom is 16 times as massive as a
hydrogen atom, thus a mole of oxygen molecules, with
f) 50 microns 2 atoms in each molecule, has a mass of
2 × 16 = 32 grams . The mass of a mole of a given
g) 30 lbs/in2 (U.S. tire pressure gauge)
kind of molecule is usually called the molecular weight,
h) 200 kilo pascals (European tire pressure gauge) but more properly the molecular mass, of that kind of
molecule.

(The integer numbers that appear in the mass of atoms

AVOGADRO’S LAW arises from the fact that protons and neutrons which
Speaking of inconvenient units like measuring tem- make up the atomic nucleus have about the same mass,
perature in ergs or pressure in dynes per cm2, we have and the mass of the electrons is much much smaller.
something particularly inconvenient in the form of the Most oxygen atoms for example have a nucleus with 8
ideal gas law PV = NkT. To use this equation , we have protons and 8 neutrons, and that is why an oxygen atom
to know the number N of the molecules in the gas. To is 16 times as massive as a hydrogen atom.)
actually count the molecules is essentially an impos-
sible requirement. We will use the symbol N A to designate Avogadro's
number
Instead of counting individual molecules, we can lump
them in large units, and count the number of units. The particles Avogadro's
standard unit for counting molecules is the mole. If you NA = 6 × 10 23 number
(24)
mole
have a mole of molecules or any other object, you have
6 × 10 23 of them. We can now rewrite the ideal gas law in the form

1 mole of = 6 × 10 23 objects PV = NkT = N × kNA T (25)

objects NA
(23)
We do this because N/N A is the number of moles of the
The idea of a mole is that it is a convenient counting
device for handling large numbers. You might, for substance rather than the number of molecules. Desig-
example, hear an astronomer say that there is about a nating this by the symbol n, we have
mole of stars in the visible universe. By that the
n ≡ N molecules = N moles (26)
astronomer would mean that he thinks that the visible NA molecules/mole NA
universe contains about 6 × 10 23 stars. (That estimate
may not be too many orders of magnitude off.) As
another example, it would take about a mole of base-
balls to fill the volume of the earth with baseballs.
 17-25

In addition the product of Boltzman’s constant k times The numerical value is

Avogadro’s number N A is called the gas constant R
V = 8.31 × 2735 m 3
R ≡ kN A gas constant 1.01 × 10

= 22.4 × 10 –3 m 3
joules 6.02 × 10 23
= 1.38 × 10 – 23
× (27)
kelvin mole Noting that 10 –3 m 3 is one liter, we get
joules volume of 1 mole
R = 8.31 MKS units of any gas at 0° C
mole kelvin V = 22.4 liters (29)
and atmospheric
pressure
In this case the MKS units are much more convenient.
The gas constant R in the CGS system is 10 – 7 times The important point is that a mole of any kind of gas has
smaller. a volume of 22.4 liters at the standard conditions of 0°
C and atmospheric pressure. (One often uses the
Expressing the ideal gas law in terms of the number of
notation STP for this standard temperature and pres-
moles n and the gas constant R gives
sure.) At STP, 22.4 liters of hydrogen have a mass of
PV = nRT (28) 2 grams, nitrogen 28 grams, and oxygen 32 grams.
which is the alternate form of the ideal gas law. If your
Exercise 7
perspective is from an atomic point of view, you would
use the form PV = NkT. But if you were a chemist and A helium nucleus contains 2 protons and 2 neutrons.
had to actually measure the quantities involved, you The mass of 22.4 liters of helium gas at STP is 4 grams.
would use the form PV = nRT. What does that say about the molecular force between
helium atoms?
Our first example of the use of Equation 28, will be to
determine the volume of one mole of molecules at 0° The calculation of the volume of a mole of a gas
C (273 K) and atmospheric pressure (1.01 × 10 5 pas- emphasizes an important point about the behavior of an
cals) . We have ideal gas. Namely, if we have equal volumes of gas at
the same temperature and pressure, the volumes will
PV = nRT contain the same number of molecules. This was first
suggested by the Italian scientist Amedeo Avogadro
1.01 * 10 5 newtons ×V (1776--1856), and is known as Avogadro’s law.
meter 2
joules
= 1 mole × 8.31 × 273 K
mole K
First let us check the dimensions. The moles and the
kelvins cancel on the right side, and we get

m 2 = kg m 2 ×
V ~ joules × newton m2
sec 2 kg m/sec 2

= meter 3
17-26 Atoms, Molecules and Atomic Processes

HEAT CAPACITY Specific Heat

From our discussion of temperature and other pro- The amount of heat energy required to raise the tem-
cesses from an atomic and molecular point of view, it perature of a unit mass of a substance one degree is
is obvious that the way to raise the temperature of an called the specific heat capacity or specific heat of the
object is to add energy. At higher temperatures the substance. For example, since it requires one calorie to
average thermal kinetic energy of the molecules in- heat one gram of water one degree centigrade, we can
creases, and that energy must come from somewhere. say that the specific heat of water is 1 calorie/gm °C, or
Historically the relationship between heat and energy 4.186 joules/gm °C.
was not so clear. As late as 1798, 71 years after
Newton’s death, there were accepted theories that Molar Heat Capacity
treated heat as a substance called caloric that flowed If, instead of measuring the heat capacity of a unit mass,
from hot substances to cooler ones. we measure the heat capacity of a mole of a substance,
It was Benjamin Thomson, later known as Count we call the result the molar heat capacity. For example
Rumford, who proposed that heat was, in fact, a form a water molecule H2O with 2 hydrogen and 1 oxygen
of energy. Rumford was boring cannons for Prince atom is 18 times as massive as a hydrogen atom. Its
Maximilian of Bavaria, and was quite aware that when molecular weight is 18, and thus a mole of water has a
the drills were dull, the cannons became hot. Thomson mass of 18 grams. As a result it takes 18 calories to raise
proposed that the mechanical work he put into turning the temperature of a mole of water 1 degree centigrade,
the drills was converted to heat energy that raised the and thus the molar heat capacity of water is 18 calories/
temperature of the cannons. Forty years later, Joule mole °C or 18 × 4.186 = 75.3 joules/mole °C. For the
accurately measured the amount of work required to units, instead of degrees centigrade, we can use kelvins,
raise the temperature of various substances. which are the same size. Thus we can write

Traditionally heat energy was defined as the amount of molar specific joules
= 75.3 (31)
heat required to raise the temperature of one gram of heat of water mole K
water one degree centigrade. This unit of heat is called as an example of a molar specific heat.
the calorie. In terms of mechanical energy, the conver-
sion factor is Predicting the specific heat of a substance, even with an
1 calorie = 4.186 joules (30) understanding of the atomic and molecular processes
involved, turned out to be a much more difficult subject
This is the relationship between mechanical work and than expected. The first time a failure of Newtonian
heat that Joule studied. mechanics was detected was during the efforts to
predict the specific heats of various gases. This failure
Exercise 8 was due to quantum mechanics being necessary to fully
A 1 kilogram mass is dropped into a bucket containing understand what happened to the added heat energy.
1 liter ( 103 cm3 ) of water. Assume that all of the kinetic
energy of the mass ends up as heat energy, raising the There is one example, however, where the simple
temperature of the water. picture of atoms we have been discussing gives the
correct answer. That is for the specific heat of helium
(a) From what height would you have to drop the mass gas. We will discuss that example here, and leave all
to raise the temperature one degree centigrade? other discussions of specific heat to Chapter 20, an
(b) (More realistic question.) How much would the entire chapter devoted to the subject.
temperature rise if you dropped the mass from a height
of one meter?
 17-27

Molar Specific Heat of Helium Gas Other Gases

A gas of helium atoms is about the simplest substance It took almost no effort to correctly predict the specific
you can picture. Since helium does not form mol- heat of helium gas. What complications do we face
ecules, the gas simply consists of individual atoms when we try to predict the specific heat of other gases?
moving around and bouncing off of each other. If the
temperature of the gas is T, then the average thermal The problem is that other gases form molecules. From
kinetic energy of the atoms is 3/2 kT. one point of view the molecules themselves are the gas
particles, so that their average thermal kinetic energy
If you have a mole of helium gas at a temperature T, must be 3/2 kT just like the helium atoms. So far so
then the thermal energy of the atoms should be the good, but molecules have an internal structure. An
average energy of 1 atom, 3/2 kT, times the number N A oxygen molecule, for example, consists of two oxygen
atoms in a mole. Thus we easily estimate that the atoms held together by the molecular force we dis-
thermal energy E He of a mole of helium atoms is cussed back in Figures (14, 15, and 16). As we saw in
Figure (15), we can fairly accurately picture the mol-
3 ecule as two atoms held together by a spring force as
EHe = N A × kT
2 shown here in Figure (26).
3 If an oxygen molecule collides with another molecule
= N kT
2 A in the gas, one would expect that the molecule would
start to vibrate, and perhaps rotate. This vibration and
Using the fact that N A k = R, the gas constant, we get
rotation represent forms of internal motion of the
3 thermal energy molecule that are quite distinct from the motion of the
EHe = RT of a mole of (32) molecule as a whole, distinct from what we would call
2 helium atoms the center of mass motion.
If we raise the temperature one degree, from T to (T +
If the center of mass motion has an average thermal
1), the thermal energy goes from 3/2 RT to
kinetic energy 3/2 kT just like helium atoms, but the
3/2 R(T + 1), an increase of 3/2 R. Thus the molar
molecules can have internal motions and internal en-
specific heat, which we will call C V, is
ergy, you would expect that it would require more
joules energy to heat a mole of oxygen than a mole of helium.
CV = 3 R = 3 × 8.31 For with the oxygen you not only have to supply the
2 2 mole K
kinetic energy of the center of mass motion, but also the
joules internal energy of the molecules. And that is correct.
CV = 12.5 = 3R (33) The molar specific heat of oxygen is 20.8
(helium) mole K 2
joules/(mole K) , as compared to 12.5 joules/(mole K)
for helium.
As we mentioned, we get the right answer. Equation 33
is in agreement with experiment. However it is when we try to calculate how much the
internal energy of the molecules contribute to the
The subscript V on the symbol C V is there to remind us
specific heat, we run into trouble. As far back as 1858,
to measure the specific heat at constant volume. If you
James Clerk Maxwell, who was working on these
add heat to a gas, and at the same time allow the gas to
calculations, repeatedly failed, and suspected that the
expand, some of the energy goes into the work required
failure was due to a problem with Newtonian mechan-
to expand the volume, pushing the surrounding gas
aside. This is a complication that we will discuss in ics.
Chapter 18. For now we will leave the subscript V on
oxygen oxygen
C V to remind us not to let the volume increase.

Figure 26
Model of an oxygen molecule.
17-28 Atoms, Molecules and Atomic Processes

EQUIPARTITION OF ENERGY motion, the so-called normal modes. The carts were
The success of our calculation of the specific heat of not free to move in an arbitrary way, their motion had
helium gas lies in the fact that in our discussion of to either be all of the vibrational mode, or all of the
thermal processes, the helium atom can be treated as a sloshing mode, or some combination of the two. The
rigid, undeformable sphere. Striking a helium atom is only thing that was arbitrary about the motion of the
more analogous to striking a golf ball than hitting carts was how much of each of the two normal modes
whiffle ball. When you hit a golf ball, most of the was present.
energy of the impact goes into the kinetic energy of the In our earlier discussion of center of mass motion in
ball, and the motion of the ball is quite predictable. Hit Chapter 11, we considered the example of two air carts,
a whiffle ball and most of the energy goes into mushing joined to each other by a spring, but free to move down
the ball; predicting where the whiffle ball will go is the track as shown in Figure (11-9). We saw that when
difficult. we gave one of the carts a shove, the center of mass of
All gas atoms except helium form molecules. While the two carts moved at a uniform speed down the track,
the individual atoms can usually be treated as hard while the carts themselves oscillated about the center of
spheres, for thermal calculations, the molecule as a mass. In this example we again have two normal
whole is generally not rigid. Strike a molecule and modes of motion. One is the motion of the center of
some of the energy goes into center of mass motion of mass, and the other is the oscillation about the center of
the molecule, but some goes into internal motions of mass. If we shove the carts just right we can have pure
the individual atoms. It seems that it would be rather center of mass motion. Or we can have the carts
hard to say much about the energy of an object that is oscillate with no center of mass motion. Or we can have
vibrating, rotating, and flying through space. some combination of the two kinds of motion.

However, if you have a gas of molecules in thermal These examples begin to show a pattern. If you have
equilibrium, the laws of Newtonian mechanics com- two masses connected by springs, that are constrained
bined with the mathematical laws of probability, make to move on a one dimensional track, the objects will
a surprisingly simple prediction of where the energy have precisely two normal modes of motion. What the
goes. This prediction is called the equipartition of actual modes are depends upon the way the springs are
energy theorem which we will now describe. connected. When the springs were connected to the
ends of the air track, there was no center of mass
As a background for the concepts involved in the motion, but we had two vibrational modes. When the
equipartition of energy theorem, let us go back to the carts were free to move down the track, we had the
normal modes experiment of Chapter 16 where we had mode representing center of mass motion, but only one
two air carts on an air track, connected by springs as vibrational mode.
shown in Figure (16-3). We found that the air carts had
two distinct kinds of motion. There was the high Another term often used to describe the way these carts
frequency mode of motion where the two air carts are moving is the expression degrees of freedom . Two
oscillated against each other, moving together and then carts moving in one dimension have 2 degrees of
apart in a sinusoidal motion. Then there was the low freedom of motion. The degrees of freedom are the
frequency sloshing mode where the carts went back center of mass motion and the vibrational mode shown
and forth along the track more or less together. in Figure (11-9), or the two vibrational modes that we
get from the setup in Figure (16-3).
When we started the carts moving in a random way,
recorded the motion, and did a Fourier analysis, we
found that the apparently complex motion was merely Figure 11-9
a combination of the two simple sinusoidal modes of Oscillating carts
 17-29

If one connects three air carts with springs and starts the picture we are developing, we will view the atoms
them moving, the resulting motion can be quite com- themselves as perfectly smooth spheres, so that you
plex. But if you record the motion, select one cycle of cannot tell whether the atom itself is rotating or not.
the repeating pattern and do a Fourier analysis, you get From this point of view, if the separation of two atoms
the simple result that there are three normal modes, as in a diatomic molecule is along the z axis as shown in
seen in the results from a student project shown in Figure (27), then rotation about the z axis cannot be
Figure (16-30). What is not so easy is to find out what detected and does not count as one of the degrees of
the individual normal modes are. And for our discus- freedom. Only rotations about the x and y axis contrib-
sion now, it is not important to find them. The ute. Thus for a diatomic molecule moving in 3 dimen-
important result is that if you have three carts connected sions, the 6 available degrees of freedom are 3 for
in some way by springs, and they are constrained to center of mass motion, 2 for rotation, and one for
move in one dimension there will be three normal vibration. x
modes or degrees of freedom. (If you have no springs,
you still have 3 degrees of freedom, namely the center
of mass motion of each of the 3 carts. z

The counting of normal modes or degrees of freedom y

generalizes to more than one dimension. If you have
one particle moving in three dimensions, it has three rotation
a) about x axis
degrees of freedom, one for center of mass motion in
the x direction, one for center of mass motion in the y
direction and one for center of mass motion in the z
direction.
If we have two particles in 3 dimensions, there are 6
degrees of freedom. If they are independent particles, rotation
b) about y axis
then each has three degrees of freedom of motion of the
center of mass. If they are connected by a spring, which
is a good model of a diatomic molecule, there are still
6 degrees of freedom but we count them in a different
way. There are the 3 degrees of freedom of the center
of mass motion, and one degree of freedom for the kind
of vibrational motion we saw in Figure (11-9) where rotation
we had two air carts connected by a spring. That c) about z axis
accounts for 4 degrees of freedom; what are the other
two?
Figure 27
When two connected particles move in three dimen- The three independent rotations of a molecule. For a
sions, what they can do that they could not do in one diatomic molecule, the rotation about the z axis does
not count. (Picture the atoms as perfectly smooth
dimension is rotate about the center of mass. One can
spheres. Then a collision could not start the z axis
envision independent rotations about the x, the y, and rotation, and you could not tell that it was rotating
the z axis, but one of these rotations does not count. In this way.)
17-30 Atoms, Molecules and Atomic Processes

If we have a system of n small spherical particles The theorem is particularly easy to understand for the
connected by spring-like forces, a reasonable model case of a gas of monatomic particles. If we have a single
for many molecules, there should be 3n degrees of particle moving in 3 dimensions, we can write its
freedom. For example, the ammonia molecule with kinetic energy 1/2 mv 2 in the form
2 2 2
1/2 m(vx + vy + vz ) where we used the
one nitrogen and three hydrogen atoms shown in
Figure (28) should have 12 degrees of freedom. Three Pythagorean theorem to express v 2 in terms of its
are center of mass motion, and now all 3 degrees of components. Thus the kinetic energy breaks up into 3
rotation should be counted. The remaining 6 must be distinct terms
vibrational normal modes, one for each spring like
force. If you could kick an ammonia (or, let us say, a kinetic = 1 mv 2 + 1 mv 2 + 1 mv 2
energy 2 x 2 y 2 z
large scale model of one), record the motion of one of
the molecules, and Fourier analyze a repeated pattern which we can call the kinetic energies of x motion, y
of the motion, you should be able to detect up to 6 motion, and z motion respectively.
normal mode frequencies of vibration.
If the particle is in thermal equilibrium, then on the
We are now ready to state the equipartition of energy average vx 2 should be the same as vy 2 and vz2 . Thus
theorem that was first derived by James Clark Maxwell the kinetic energies associated with each of the degrees
in 1858. Using Newtonian mechanics and the math- of freedom of the molecule (x motion, y motion, and z
ematical laws of probability, Maxwell showed that if a motion) should be the same, and the sum should be the
gas of molecules is in thermal equilibrium, then the total average kinetic energy 3/2 kT. If 3/2 kT is shared
average thermal energy of each molecule is 1/2 kT 3 ways, each degree of freedom should get 1/2 kT
times the number of degrees of freedom possessed by kinetic energy on the average, as required by the
the molecule. The theorem implies that, as you add equipartition of energy theorem.
thermal energy to a system of molecules, the energy is
shared equally, on the average, between the available
degrees of freedom. Real Molecules
Applying Maxwell's equipartition of energy theorem,
N we can make definite predictions about the specific
heat of various kinds of molecules. We will begin with
a brief review of the calculation of the specific heat of
a monatomic gas like helium. A monatomic gas atom
has 3 degrees of freedom, thus on the average its kinetic
energy is

E 1 molecule = 3 × 1 kT = 3 kT (34)
2 2
H
Since the specific heat C V deals with one mole of a
substance, we multiply Equation 34 through by
H H Avagadro's number N A (number of particles in a
mole) to get
Figure 28
The ammonia molecule is a tetrahedral structure with E 1 mole = NA 3 kT = 3 NAk T
one nitrogen atom and three hydrogen atoms. Here we 2 2
are modeling the forces between atoms as spring forces.
= 3 RT (35)
2
where N Ak = R is the gas constant.
 17-31

Finally C V is defined as the change in energy ∆E for FAILURE OF CLASSICAL PHYSICS

a small change in temperature ∆T . Differentiating If you worked out a complex theory that made detailed
Equation (35) gives ∆E = (33 2)R∆T
2 , so that predictions, and when you compared the predictions
∆E ≡ C = 3 R (36)
with experiment, you got the results shown in Table 1,
∆T V 2 you should be disappointed. The agreement is simply
terrible. The predictions work only for the monatomic
where the subscript V reminds us to keep the volume V
gases (gases that remain individula atoms and do not
of the gas constant so that none of the gas energy goes
form molecules). There is an increase in specific heat
into doing the work of expanding the gas.
when we go to larger molecules, but not the predicted
In the above derivation, we immediately see that the increase. In going from carbon dioxide to methane,
factor of 3 in the formula C V = 3/2 R came from our where there is a considerable increase in the number of
assumption that the molecule has 3 degrees of freedom. degrees of freedom, there is actually a decrease in the
If we had a molecule with n degrees of freedom, then specific heat.
the equipartition of energy theorem predicts that the
Maxwell worked on this problem for a number of
specific heat should be
years, carefully checking that he had correctly applied
for a molecule prediction of the the mathematical laws of statistics to Newtonian me-
with n degrees C V = n R equipartition of (37) chanics, but he could find no error in his work. By 1879
of freedom 2 energy theorem
he became convinced that Newtonian mechanics was
To see how good the predictions of the equipartition of flawed, and that eventually some new theory would
energy theorem are, we have in Table 2 listed the have to be developed to replace it. The new theory, of
specific heats of some common gasses, and compared course, was quantum mechanics, discovered nearly 50
the results with the predicted values. years later. Maxwell's work in the 1860s and 1870s
provided the first real evidence that Newtonian me-
chanics was not correct in all applications.

Molecule Number of Expected number Expected CV Experimental

particles degrees of freedom (joules/mole) CV

helium 1 3 3/2 R = 12.5 12.5

argon 1 3 3/2 R = 12.5 12.6

nitrogen N2 2 6 6/2 R = 25 20.7

oxygen O2 2 6 6/2 R = 25 20.8

carbon dioxide CO2 3 9 9/2 R = 37.5 29.7

methane NH4 5 15 15/2R = 62.5 29.0

Table 2
Specific heats of various molecules. Theory and
experiment agree only for the monatomic gases.
17-32 Atoms, Molecules and Atomic Processes

Freezing Out of Degrees of Freedom degrees of freedom appear to freeze out. By room
While a straight forward application of the equipartition temperature we have lost two degrees of freedom, and
of energy theorem fails miserably, the ideas involved down at 100 K, only the 3 translational degrees of
are not completely useless. A look at the specific heat freedom are left. Why the degrees of freedom freeze
of hydrogen gas as a function of temperature, Figure out is what is not explained by Newtonian mechanics.
(29) gives us a clue as to what is happening at low This is purely a quantum mechanical effect.
temperatures below 100 K. The specific heat is 3/2 R,
which is what we would expect for a monatomic gas In Maxwell's time, the idea that matter consisted of
with only 3 degrees of freedom. At these temperatures atoms was a hypothesis rather than an experimentally
none of the thermal energy is going into exciting the proved fact. What atoms consisted of, whether they
internal motion of the atoms. At these low tempera- were indivisible hard spheres or had an internal struc-
tures, the hydrogen molecules are acting like incom- ture was unknown. By applying Newtonian mechanics
pressible hard spheres. to models of atoms and molecules, he was trying to
learn about the nature of these objects. The fact that
Up at room temperature, the specific heat of hydrogen monatomic gases have a specific heat C V = 3/2R
has jumped to 5/2 R. It appears that two additional was evidence that the atoms were in fact acting like
degrees of freedom have appeared, and some of the hard, indivisible objects. The failure to predict molecu-
thermal energy is now going into internal motions of lar specific heats turned out to be evidence that
the molecule. At still higher temperatures, just as the Newtonian mechanics was failing.
molecules are being torn apart, their specific heat
reaches 7/2 R, indicating 7 degrees of freedom, one We know that atoms themselves consist of many
more than we expected. We can explain the 7 degrees particles—a nucleus surrounded by electrons. If we
of freedom by assuming that 1/2 kT of thermal energy applied Newtonian mechanics to this structure, we
goes, on the average, into the spring potential energy. would assume that each atom should have many de-
grees of freedom and that the nucleus and electrons
Going back down in temperature, we have the follow- should individually pick up thermal energy as the atom
ing picture. At very high temperature all the degrees of is heated. This simply does not happen. Applying the
freedom are active and energy is shared equally among language we have used above, we can say that tempera-
them as required by the equipartition of energy theo- tures at which we ordinarily study atoms, the internal
rem. As we go down in temperature some of the degrees of freedom of the atom are frozen out.

4
7
2
Figure 29 3
vibration
Specific heat of the hydrogen
5
molecule. If each degree of
2
CV/R

freedom contributes 1/2R to the

specific heat, then as the 2
rotation
temperature drops, we see that 3
various degrees of freedom 2
freeze out. (Diagram adapted
1
from Halliday and Resnick.)
translation

0
20 50 100 200 500 1000 2000 5000 10,000
temperature(K)
 17-33

THERMAL EXPANSION the potential energy well, due to the fact that the
When you heat a substance, in most cases the substance repulsive core rises faster than the attractive side, the
expands. For example, the mercury (or alcohol) in the average separation 2r T of the atoms at a temperature T
bulb at the bottom of a thermometer expands when is greater than the average separation 2r 0 at low
heated, forcing more mercury up the small tube above temperatures.
the bulb. If the column is marked off in degrees Although our discussion of molecular forces focused
centigrade, you have a typical mercury thermometer as on two atom molecules, the general shape of the
illustrated in Figure (30). As another example, you are molecular force potential well is the same when you
aware of the cracks left between sections of cement in have many atoms forming a liquid or a solid. Thus
sidewalks and cement highways. These cracks are when you heat a liquid or a solid, the atoms gain an
there so that on a hot day when the cement expands, the average thermal kinetic energy 3/2 kT, effectively rise
sidewalk or road will not buckle. up in the molecular force potential well, and due to the
The reason for thermal expansion can be understood at shape of the well, have a slightly greater average
an atomic level in terms of the shape of the molecular separation. The substance expands.
force potential well. Figure (31) is a redrawing of the You can see that the amount of expansion depends
molecular force potential well of Figure (16), with upon the detailed shape of the potential well which
some added information. Let 2r be the separation of the varies from one substance to another. Thus a thermom-
two atoms as indicated at the top of the diagram. If the eter based on the expansion properties of mercury does
molecule is at a very low temperature, the atom will not have to give precisely the same reading as a
essentially sit at the bottom of the potential well and the thermometer using alcohol, except at the calibration
separation will be 2r 0 as shown. points 0° C and 100° C. And neither of these thermom-
If we raise the temperature of the molecule, the atoms eters has to agree with the ideal gas thermometer. Since
gain thermal kinetic energy whose average value is the ideal gas thermometer is based on the universal
3/2 kT. As a result they will move back and forth at a ideal gas law, one should use the ideal gas thermometer
higher level in the potential well, a height we have as a standard against which you calibrate mercury,
indicated as level 1 in the diagram. Due to the shape of alcohol, and other thermometers.
r r

boiling
water 100° C

Average separation
rT at a temperature T
100
divisions
Repulsive co

ide
Binding es
energy of ctiv
a
tr

molecule
At

leval 1
e r

3/2 kT
melting leval 0
0° C
ice r0
bulb full Equilibrium separation
of mercury at very low temperatures
Figure 31
Figure 30
As you raise the temperature of a substance, the
The typical mercury thermometer is based on
average thermal kinetic energy 3/2 kT rises and the
the thermal expansion of mercury. There is no
molecule sits higher in its potential well. Because the
guarantee that a mercury thermometer and an
well is lopsided, the average separation of the atom
ideal gas thermometer will agree at any
becomes greater and the substance expands.
temperatures except 0° C and 100 ° C.
17-34 Atoms, Molecules and Atomic Processes

OSMOTIC PRESSURE Assume that some definite fraction, for instance 50%,
We would like to conclude this chapter, this brief view of all water molecules that strike the membrane pass
of atoms, molecules and thermal processes, with a through it. Initially, more water molecules strike the
discussion of two familiar phenomena that can be membrane from side 2 than from side 1 simply because
understood qualitatively from a molecular point of there are more water molecules on side 2. If more water
view. One is the elasticity of rubber, and the other is the molecules strike from side 2 than side 1, and if 50% of
process of osmosis, which is essential for biological all the water molecules striking the membrane pass
systems. through it, there must be a net flow of water from side
2 to side 1.
Osmosis is a rather peculiar but important effect that is
easily explained with an atomic model. Ordinarily, As the flow continues, the level of the liquid on side 1
when a liquid can flow between two vessels at the same rises and the solution becomes diluted. As the solution
height, the liquid will tend to seek the same height in becomes further diluted, the number of water mol-
both vessels. But this does not always happen. Sup- ecules on side 1 facing each cm2 of the membrane
pose we have a tank separated by a membrane, as increases, the more flow back to side 2, so that the net
shown in Figure (32). On the right side (side 2) of the flow into side 1 decreases. From this description alone,
membrane we place pure water, indicated by the small however, we would not expect the flow to stop, since
molecules. On the left (side 1) we place a solution of we never get pure water on side 1.
water and some other substance consisting of large The flow does stop eventually though, because the
molecules. The membrane has a special characteristic: level in side 1 rises to such an extent that the pressure
the small water molecules can pass through it easily, at the bottom of side 1 becomes considerably greater
whereas the big molecules are prevented from passing than the pressure at the bottom of side 2 (a result of the
through because the holes in the membrane are too increased weight of the column of water). This addi-
small. Initially, the two compartments are filled to the tional pressure, known as osmotic pressure, finally
same level on each side of the membrane. stops the flow of water from side 2 to side 1. The flow
of the small molecules through the membrane is called
osmosis; thus osmotic pressure is the pressure that
finally stops osmosis.

mixture
mixture pure water 1
1 2
pure water
2

(a) initially (b) finally

Figure 32
Osmosis. The two sides of the container are separated by a membrane that allows the water
(small molecules) but not the large molecules to pass through. If the liquid levels are the
same initially, as in (a), some of the pure water will flow through the membrane, raising the
level on the side with the large molecules as shown in (b). This process is called osmosis.
 17-35

Osmosis and osmotic pressure are crucial in biological ELASTICITY OF RUBBER

processes. Osmosis is involved in the separation of A model for the elasticity of rubber was presented by
nutrients and wastes in our own cells, the flow of fluids Richard Feynman in a lecture to freshmen at Caltech in
in our bodies, the flow of sap in plant life, and a number 1960. We select this model, not so much for its
of other important processes. accuracy in describing the detailed behavior of mol-
The function and composition of blood is critically ecules in rubber, but for developing an intuition for
dependent on osmosis and osmotic pressure. Blood thermal processes. The mechanisms underlying the
consists of red cells, white cells, and a fluid called model and the behavior of rubber are fundamentally
plasma. The red cells are membrane sacs containing the same. The beauty of the model is that it is so
about 60% water and 40% hemoglobin molecules, outrageous that you are forced to think differently
molecules closely related to, but about four times as about thermal processes.
large as, the giant myoglobin molecule described in at
the beginning of this chapter. We may think of the red
blood cell as representing side 1 in Figure (32) where
the big molecules are the hemoglobin molecules. If red
blood cells are removed from blood and placed in pure
water, they absorb so much water by osmosis that they
burst. The red hemoglobin flows away, leaving an
empty, pale misshapen sac.

The function of the red blood cell and its hemoglobin

is to carry oxygen to the other cells in the body. The
plasma, which consists of 90% water, 9% protein
molecules, and 1% salts, serves as a fluid in which to
dissolve needed proteins and salts to be carried to the
cells, and to make the blood fluid enough to flow
through the minute capillaries.

The capillary walls through which blood flows are

porous membranes that permit water and salts to pass
freely through, but that restrict the passage of proteins.
Pure water could not be pumped through the blood-
stream because it would leak out through the capillary
walls. You may wonder how blood plasma, which is
90% water, can be pumped through the porous capillar-
ies. The reason is that the 9% protein molecules in the
plasma is sufficient to draw just enough water back into
the capillary by osmosis to replace the water molecules
that do leak out. Just as many water molecules are
drawn back in as leak out, even though the pressure of
the plasma inside the capillary is greater than the
pressure of the fluids outside the leaky walls. Thus,
side 1 in Figure (32) behaves in the same way as the
capillary with the blood plasma inside it.
17-36 Atoms, Molecules and Atomic Processes

A Model of Rubber the colliding cannonballs, and suddenly pull the sides
Imagine that you enter a large room where there are a of the room apart so that the chains are straight and
number of heavy chains loosely suspended from one tight. When we suddenly straighten out the kinked
end of the room to the other, as shown in Figure (33). chains, the chains will slap against the cannonballs
These are massive chains, like the anchor chains used transforming the work we do pulling the chains straight
on old sailing ships, but they are hanging loosely, so into increased thermal kinetic energy of the cannon-
that except for their weight, they are not exerting any balls. As a result by suddenly stretching the chains we
force pulling the walls together. raise the temperature of the cannonballs.

On the floor are hundreds of cannonballs, lying there a If we let the walls go back suddenly, the chains initially
couple of layers deep. This is our room at “absolute go slack, and it takes some of the thermal kinetic
zero”. energy of the cannonballs to kink the chains up again.
As a result of unstretching the chains, the temperature
Now turn up the temperature in the room. The cannon- of the cannonballs drops.
balls start to jiggle and vibrate with an average thermal
kinetic energy 3/2 kT. In this model, nothing melts. One’s lips are a good detector of small temperature
Instead, as we turn up the temperature the jiggling changes. Place a loose rubber band between your lips
becomes stronger and stronger. When the average and suddenly stretch it. You will notice that the rubber
thermal kinetic energy 3/2 kT becomes as large as the band becomes distinctly warmer. Now quickly re-
gravitational potential energy mgh of a cannonball near lease the rubber band by bringing your hands together.
the ceiling, then we will have cannonballs flying all The rubber band becomes distinctly cool. Rubber
around the room. We will have a gas of cannonballs. consists of a long chain of molecules that are kinked up
by thermal motion. When you stretch the rubber band,
As the cannonballs fly around, they strike the chains, you increase the thermal kinetic energy of the mol-
kinking them up as indicated in Figure (34). The ecules and raise their temperature. Releasing the band
kinked-up chains are no longer hanging loose, instead reduces the thermal motion and drops the temperature.
they are taut and pulling the side walls of the room in. The elastic restoring force you felt when you stretched
the band is caused by thermal motions kinking the long
If we raise the temperature of the gas of cannonballs,
chain molecules.
the cannonballs strike the chains harder and the chains
pull harder on the walls.

Here is an experiment that stretches the imagination

even more. Suppose we start with the room with a gas
of cannonballs at a temperature T, and chains kinked by

Figure 33 Figure 34
Room with suspended chains and cannonballs. Top view of chains being struck by cannonballs.
Chapter 18
Entropy

CHAPTER 18 ENTROPY
Movie Undive

The focus of the last chapter was the thermal energy of 3600 seconds, or 1000 × 3600 joules. Thus at a rate of
the atoms and molecules around us. While the thermal 10 cents per kilowatt hour, the thermal energy of a mole
energy of an individual molecule is not large, the of helium gas is worth only .01 cents.
thermal energy in a reasonable collection of mol-
ecules, like a mole, is a noticeable amount. Suppose, The value .01 cents does not sound like much, but that
for example, you could extract all the thermal energy was the value of the energy in only one mole of helium.
in a mole of helium atoms at room temperature. How Most substances have a greater molar heat capacity
much would that energy be worth at a rate of 10 cents than helium due to the fact that energy is stored in
per kilowatt hour? internal motions of the molecule. Water at room
temperature, for example, has a molar heat capacity
This is an easy calculation to do. The atoms in helium six times greater than that of helium. Thus we would
gas at room temperature have an average kinetic expect that the thermal energy in a mole of water
energy of 3/2 kT per molecule, thus the energy of a mole should be of the order of 6 times greater than that of
is N A 3 2KT
2 = 3 2RT,
2 where N A is Avagadro’s helium, or worth about .06 cents.
number and R = N A K. Since ice melts at 273 K and
water boils at 373 K, a reasonable value for room A mole of water is only 18 grams. A kilogram of water,
temperature is 300 K, about one quarter of the way up 1000 grams of it, is 55 moles, thus the thermal energy
from freezing to boiling. Thus the total thermal energy in a kilogram or liter of water should have a value in the
in a mole of room temperature helium gas is neighborhood of .06 cents/mole × 55 moles = 3.3 cents.
Now think about the amount of water in a swimming
thermal energy pool that is 25 meters long, 10 meters wide, and 2
in a mole of
room temperature = 3 RT meters deep. This is 500 m 3 or 500 ×10 3 liters, their
2 being 1000 liters/ m 3. Thus the commercial value of
helium gas
the heat energy in a swimming pool of water at room
joules
= 3×8 × 300K temperature is 5 × 10 5 × 3.3cents or over $16,000.
2 mole K
= 3600 joules The point of this discussion is that there is a lot of
thermal energy in the matter around us, energy that
This is enough energy to lift 1 kilogram to a height of would have enormous value if we could get at it. The
360 feet! question is why don’t we use this thermal energy rather
To calculate the monetary value of this energy, we note than getting energy by burning oil and polluting the
that a kilowatt hour is 1000 watts of electric power for atmosphere in the process?
18-2 Entropy

INTRODUCTION Let us look at this process from the point of view of the
A simple lecture demonstration helps provide insight energy involved. Before the drop hits, it has kinetic
into why we cannot easily get at, and use, the $16,000 energy due to falling. When it hits, this kinetic energy
of thermal energy in the swimming pool. In this goes into the kinetic energy of wave motion. The
demonstration, illustrated in Figure (1), water is pumped waves then flow into the bucket, eventually dissipate,
up through a hose and squirted down onto a flat plate and all the kinetic energy becomes thermal energy of
placed in a small bucket as shown. If you use a vibrator the water molecules in the bucket. This causes a slight,
pump, then the water comes out as a series of droplets almost undetectable, increase in the temperature of the
rather than a continuous stream. water in the bucket (if the water drop had the same
temperature as the water in the bucket, and we neglect
To make the individual water droplets visible, and to cooling from evaporation).
slow down the apparent motion, the water is illumi-
nated by a strobe. If the time between strobe flashes is We have selected this demonstration for discussion,
just a bit longer than the time interval which the drops because with a slight twist of the knob on the strobe, we
are ejected, the drops will appear to move very slowly. can make the process appear to run backwards. If the
This allows you to follow what appears to be an time interval between flashes is just a bit shorter than
individual drop as it moves down toward the flat plate. the pulse interval of the pump, the drops appear to rise
from the plate and go back into the hose. The situation
For our discussion , we want to focus on what happens looks funny, but it makes a good ending to the demon-
to the drop as it strikes the plate. As seen in the series stration. Everyone knows that what they see couldn’t
of pictures in Figure (2), when the drop hits it flattens possibly happen—or could it?
out, creating a wave that spreads out from where the
drop hits. The wave then moves down the plate into the Does this reverse flow violate any laws of physics?
pool of water in the bucket. [In Figure (1) we see several Once a drop has left the plate it moves like a ball thrown
waves flowing down the plate. Each was produced by up in the air. From the point of view of the laws of
a separate drop.] physics, nothing is peculiar about the motion of the
drop from the time it leaves the plate until it enters the
hose.

Figure 1
Water droplets are created water
by a vibrator pump. If you droplets
illuminate the drops with a
strobe light, you can make
them appear to fall or rise.

vibrator
pump
 18-3

Where the situation looks funny is in the launching of It is not impossible to convert thermal energy into
the drop from the plate. But it turns out that none of the organized kinetic energy or what we call useful work.
laws of physics we have discussed so far is violated Steam engines do it all the time. In a modern electric
there either. Let us look at this launching from the point power plant, steam is heated to a high temperature by
of view of the energy involved. Initially the water in the burning some kind of fuel, and the steam is sent through
bucket is a bit warm. This excess thermal energy turbines to produce electricity. A certain fraction of the
becomes organized into a wave that flows up the plate. thermal energy obtained from burning the fuel ends up
As seen in Figure (3) the wave coalesces into a drop that as electrical energy produced by the electric generators
is launched up into the air. No violation of the law of attached to the turbines. This electric energy can then
conservation of energy is needed to describe this be used to do useful work running motors.
process.
The important point is that power stations cannot
While the launching of the drop in this reversed picture simply suck thermal energy out of a reservoir like the
may not violate the laws of physics we have studied, it ocean and turn it into electrical energy. That would
still looks funny, and we do not see such things happen correspond to our water drop being launched by the
in the real world. There has to be some reason why we thermal energy of the water in the bucket.
don’t. The answer lies in the fact that in the reversed
process we have converted thermal energy, the disor-
ganized kinetic energy of individual molecules, into
the organized energy of the waves, and finally into the
more concentrated kinetic energy of the upward trav-
elling drop. We have converted a disorganized form of
energy into an organized form in a way that nature does
not seem to allow.

Figure 2
Falling water drop creating wave.

Figure 3
Rising wave launching water drop.
18-4 Entropy

Even more discouraging is the fact that power plants do his studies Carnot found that there was a theoretically
not even use all the energy they get from burning fuel. maximum efficient engine whose efficiency depended
A typical high efficiency power plant ends up discard- upon the temperature of the boiler relative to the
ing, into the atmosphere or the ocean, over 2/3 of the temperature of the boiler’s surroundings.
energy it gets from burning fuel. Less than 1/3 of the
energy from the fuel is converted into useful electrical To make his analysis, Carnot had to introduce a new
energy. Car engines are even worse. Less than assumption not contained in Newton’s law of mechan-
1/5 of the energy from the gasoline burned goes into ics. Carnot’s assumption is equivalent to the idea that
powering the car; most of the rest comes out the exhaust you cannot convert thermal energy into useful work
pipe. in a process involving only one temperature. This is
why you cannot sell the $16,000 worth of thermal
Why do we tolerate these low efficiency power plants energy in the swimming pool—you cannot get it out.
and even lower efficiency car engines? The answer lies
in the problem of converting a disorganized form of This law is known as the Second Law of Thermody-
energy into an organized one. Or to state the problem namics. (The first law is the law of conservation of
more generally, of trying to create order from chaos. energy itself.) The second law can also be expressed in
terms of entropy which we now know represents the
The basic idea is that a disorganized situation does not disorder of a system. The second law states that in any
naturally organize itself—in nature, things go the other process, the total entropy (disorder) of a system either
way. For example, if you have a box of gas, and initially stays the same or increases. Put another way, it states
the atoms are all nicely localized on one side of the box, that in any process, the total order of a system cannot
a short time later they will be flying around throughout increase; it can only stay the same, or the system can
the whole volume of the box. On their own, there is become more disordered.
almost no chance that they will all move over to that one
side again. If you want them over on one side, you have To develop his formulas for the maximum efficiency of
to do some work, like pushing on a piston, to get them engines, Carnot invented the concept known as a
over there. It takes work to create order from disorder. Carnot engine, based on what is called the Carnot
cycle. The Carnot engine is not a real engine, no one has
At first, it seems that the concepts of order and disorder, ever built one. Instead, you should think of it as a
and the related problems of converting thermal energy thought experiment, like the ones we used in Chapter
into useful work should be a difficult subject to deal 1 to figure out what happened to moving clocks if the
with. If you wished to formulate a physical law, how principle of relativity is correct.
do you go about even defining the concepts. What, for
example, should you use as an experimental definition The question we wish to answer is, how efficient can
of disorder? It turns out, surprisingly, that there is a you make an engine or a power plant, if the second law
precise definition of a quantity called entropy which of thermodynamics is correct? If you cannot get useful
represents the amount of disorder contained in a sys- work from thermal energy at one temperature, how
tem. Even more surprising, the concept of entropy was much work can you get if you have more than one
discovered before the true nature of heat was under- temperature? It turns out that there is a surprisingly
stood. simple answer, but we are going to have to do quite a
bit of analysis of Carnot’s thought experiment before
The basic ideas related to entropy were discovered in we get the answer. During the discussion of the Carnot
1824 by the engineer Sadi Carnot who was trying to engine, one should keep in mind that we are making this
figure out how to improve the efficiency of steam effort to answer one basic question—what are the
engines. Carnot was aware that heat was wasted in the consequences of the second law of thermodynamics—
operation of a steam engine, and was studying the what are the consequences of the idea that order does
problem in an attempt to reduce the waste of heat. In not naturally arise from disorder.
 18-5

WORK DONE BY After this expansion, the volume of the gas has in-
AN EXPANDING GAS creased by an amount ∆V = A∆x . Thus Equation 2
can be written in the form ∆W = pA∆x or
The Carnot thought experiment is based on an analysis
of several processes involving the ideal piston and
cylinder we discussed in the last chapter. We will ∆W = p∆V (3)
discuss each of these processes separately, and then put
them together to complete the thought experiment. Equation 3 is more general than our derivation indi-
cates. Any time a gas expands its volume by an amount
The ideal piston and cylinder is shown in Figure (4). A ∆V, the work done by the gas is p∆V no matter what
gas, at a pressure p, is contained in the cylinder by a the shape of the container. For example, if you heat the
frictionless piston of cross-sectional area A. (Since no gas in a balloon and the balloon expands a bit, the work
one has yet built a piston that can seal the gas inside the done by the gas is p∆V where ∆V is the increase in
cylinder and still move frictionlessly, we are now the volume of the balloon.
already into the realm of a thought experiment.) A
force F is applied to the outside of the piston as shown Exercise 1
to keep the piston from moving. The gas, at a pressure In our introduction to the concept of pressure, we
p, exerts an outward force dipped a balloon in liquid nitrogen until the air inside
became a puddle of liquid air (see Figure 17-19). When
p newtons × A meter 2 = pA newtons we took the balloon out of the liquid nitrogen, the air
meter2 slowly expanded until the balloon returned to its original
size. During the expansion, the rubber of the balloon
on the cylinder, thus F must be given by was relatively loose, which means that the air inside the
balloon remained at or very near to atmospheric pres-
F = pA (1) sure during the entire time the balloon was expanding.
to keep the cylinder from moving. (a) If the final radius of the balloon is 30 cm, how much
work did the gas inside the balloon do as the balloon
If we decrease the force F just a bit to allow the gas in
expanded? (You may neglect the volume of the liquid
the cylinder to expand, the expanding gas will do work air present when the expansion started.) (Answer:
on the piston. This is because the gas is exerting a force 1.1 × 104 joules.)
pA on the cylinder, while the cylinder is moving in the
direction of the force exerted by the gas. If the piston (b) Where did the gas inside the balloon get the energy
moves out a distance ∆x as shown in Figure (5), the required to do this work?
work ∆W done by the gas is the force pA it exerts times
the distance ∆x ∆V = A∆x
∆W = pA ∆x (2)
pA
piston of area A
gas at a Figure 5
∆x
F
pressure p The work done by an expanding gas is
equal to the force pA it exerts, times
the distance ∆x the piston moves.
Figure 4
In the ideal piston and cylinder, the piston
confines the gas and moves frictionlessly.
18-6 Entropy

SPECIFIC HEATS CV AND Cp energy required increase in thermal

to heat 1 mole energy of the gas
In our earlier discussion of specific heat, we dealt of a gas 1K at
Cp = when the temperature
exclusively with the “molar” specific heat at constant constant pressure increases 1K
volume C V . We always assumed that we kept the gas
at constant volume so that all the energy we added work done
by the
would go into the internal energy of the gas. If we had + expanding (4)
allowed the gas to expand, then some of the energy gas
would have gone into the work the gas did to expand its
Noting that since C V is equal to the increase in thermal
volume, and we would not have had an accurate
energy of the gas, and that the work done is p∆V, we
measure of the amount of energy that went into the gas
get
itself.
C p = C V + p∆V (5)
Sometimes it is convenient to heat a gas while keeping
the gas pressure, rather than volume, constant. This is In the special case of an ideal gas, we can use the ideal
more or less the case when we heat the gas in a balloon. gas law pV = nRT, setting n = 1 for 1 mole
The balloon expands, but the pressure does not change
very much if the expansion is small. pV=RT (1moleofgas) (6)
If we let the gas expand a bit at constant pressure, we get
Earlier we defined the molar heat capacity C V as the differentiating Equation 6, keeping p constant**
amount of energy required to heat one mole of a
substance one kelvin, if the volume of the substance is p∆V = R∆T (if p is constant) (7)
kept constant. Let us now define the molar heat
capacity C p as the amount of energy required to heat
one mole of a substance one kelvin if the pressure is ** By differentiating the equation (pV = RT), we
kept constant. For gases C p is always larger than C V . mean that we wish to equate the change in (pV) to the
This is because, when we heat the gas at constant change in (RT). To determine the change in (pV), for
pressure, the energy goes both into heating and expand- example, we let (p) go to (p + ∆p) and (V) go to
ing the gas. When we heat the gas at constant volume, (V + ∆V), so that the product (pV) becomes
the energy goes only into heating the gas. We can write
this out as an equation as follows pV → (p + ∆p)(V +∆V)
= pV + (∆p)V + p∆V + (∆p)∆V

If we neglect the second order term (∆p)∆V then

pV → pV + (∆p)V + p∆V
Then if we hold the pressure constant (∆p = 0), we see
that the change in (pV) is simply (p∆V).Since R is
constant, the change in RT is simply R∆T.
 18-7

If the temperature increase is ∆T = 1 kelvin, then Exercise 2

Equation 7 becomes (a) Back in Table 2 on page 31 of Chapter 17, we listed
p∆V = R (1 mole, p constant, ∆T = 1K) (8) the values of the molar specific heats for a number of
gases. While the experimental values of did not agree
Using Equation 8 in Equation 5 we get the simple result in most cases with the values predicted by the
equipartition of energy, you can use the experimental
Cp = CV + R (9) values of CV to accurately predict the values of Cp for
these gases. Do that now.

The derivation of Equation 9 illustrates the kind of (b) Later in this chapter, in our discussion of what is
steps we have to carry out to calculate what happens to called the adiabatic expansion of a gas (an expansion
the heat we add to substances. For example, in going that allows no heat to flow in), we will see that the ratio
from Equation 6 to Equation 7, we looked at the change of Cp CV plays an important role in the theory. It is
in volume when the temperature but not the pressure common practice to designate this ratio by the Greek
was varied. When we make infinitesimal changes of letter γ
some quantities in an equation while holding the quan-
γ ≡ Cp CV (10)
tities constant, the process is called partial differentia-
tion. In this text we will not go into a formal discussion
of the ideas of partial differentiation. When we encoun- (i) Explain why, for an ideal gas, γ is always greater than
ter the process, the steps should be fairly obvious as 1.
they were in Equation 7. (ii) Calculate the value of γ for the gases, listed in Table
2 of Chapter 17.
(The general subject that deals with changes produced
by adding or removing heat from substances is called Answers:
thermodynamics. The full theory of thermodynamics
relies heavily on the mathematics of partial deriva- gas γ
tives. For our discussion of Carnot’s thought experi- helium 1.66
ments, we need only a small part of thermodynamics argon 1.66
theory.) nitrogen 1.40
oxygen 1.40
CO2 1.28
NH4 1.29
18-8 Entropy

ISOTHERMAL EXPANSION However, we wish to study the process in which the gas
AND PV DIAGRAMS expands at constant temperature. To keep the tempera-
ture from dropping, we have to let heat flow into the
In the introduction, we pointed out that while a swim-
gas. Since the temperature of the gas is constant, there
ming pool of water may contain $16,000 of thermal
is no change in the thermal energy of the gas. Thus all
energy, we could not extract this energy to do useful
the heat that flows in goes directly into the work done
work. To get useful energy, we have to burn fuel to get
heat, and convert the heat to useful work. What seemed by the gas.
like an insult is that even the most efficient power plants To calculate the amount of work done , we have to add
turn only about 1/3 of the heat from the fuel into useful up all the p∆V ’s as the gas goes from a volume V1 to
work, the rest being thrown away, expelled either into a volume V2 . If we graph pressure as a function of
the atmosphere or the ocean. volume, in what is called a pV diagram, we can easily
visualize these increments of work p∆V as shown in
We are now going to discuss a process in which heat is
Figure (7).
converted to useful work with 100% efficiency. This
involves letting the gas in a piston expand at constant Suppose the pressure of the gas is initially p 1 when the
temperature, in a process called an isothermal expan- volume of the cylinder is V1 . As the cylinder moves
sion. (The prefix “iso” is from the Greek meaning out and the gas expands, its pressure will drop as shown
“equal,” thus, isothermal means equal or constant in Figure (7), reaching the lower value p 2 when the
temperature.) This process cannot be used by power cylinder volume reaches V2 . At each step ∆V i , when
plants to make them 100% efficient, because the pro- the pressure is p i , the amount of work done by the gas
cess is not repetitive. Some work is required to get the is p i ∆V i . The total work, the sum of all the p i ∆V i , is
piston back so that the expansion can be done over just the total area under the pressure curve, as seen in
again. Figure (7).
Suppose we start with a gas in a cylinder of volume V1 The nice feature of a graph of pressure versus volume
and let the gas slowly expand to a volume V2 as shown like that shown in Figure (7), is the work done by the
in Figure (6). We control the expansion by adjusting gas is always the area under the pressure curve, no
the force F exerted on the back side of the piston. matter what the conditions of the expansion are. If we
had allowed the temperature to change, the shape of the
While the gas is expanding, it is doing work on the
pressure curve would have been different, but the work
piston. For each ∆V by which the volume of the gas
done by the gas would still be the area under the
increases, the amount of work done by the gas is p∆V .
pressure curve.
The energy required to do this work must come from
somewhere. If we did not let any heat into the cylinder, pressure
the energy would have to come from thermal energy, p1 p1V1
and the temperature would drop. (This is one way to get
work out of thermal energy.)

V1 F
area = pi ∆Vi
pi
p2V2
V2 F p2

Figure 6
Isothermal expansion of the gas in a cylinder. The V1 ∆Vi volume V2
force F on the cylinder is continually adjusted so Figure 7
that the gas expands slowly at constant temperature. The work done by an expanding gas is equal to the sum
of all p ∆V 's , which is the area under the pressure curve.
 18-9

Isothermal Compression ADIABATIC EXPANSION

If we shoved the piston back in, from a volume V2 to We have seen that we can get useful work from heat
a volume V1 in Figure (7), we would have to do work during an isothermal expansion of a gas in a cylinder.
on the gas. If we kept the temperature constant, then the As the gas expands, it does work, getting energy for the
pressure would increase along the curve shown in work from heat that flows into the cylinder. This
Figure (7) and the work we did would be precisely represents the conversion of heat energy at one tem-
equal to the area under the curve. In this case work is perature into useful work. The problem is that there is
done on the gas (we could say that during the compres- a limited amount of work we can get this way. If we
sion the gas does negative work). When work is done shove the piston back in so that we can repeat the
on the gas, the temperature of the gas will rise unless we process and get more work, it takes just as much work
let heat flow out of the cylinder. Thus if we have an to shove the piston back as the amount of work we got
isothermal compression, where there is no increase in out during the expansion. The end result is that we have
the thermal energy of the gas, then we have the pure gotten nowhere. We need something besides isother-
conversion of useful work into the heat expelled by the mal expansions and compressions if we are to end up
piston. This is the opposite of what we want for a power with a net conversion of heat into work.
plant.
Another kind of expansion is to let the cylinder expand
without letting any heat in. This is called an adiabatic
Isothermal Expansion of an Ideal Gas
expansion, where adiabatic is from the Greek (a-not +
If we have one mole of an ideal gas in our cylinder, and dia-through + bainein-to go). If the gas does work
keep the temperature constant at a temperature T1 , then during the expansion, and we let no heat energy in, then
the gas will obey the ideal gas equation. all the work must come from the thermal energy of the
pV = RT1 = constant (11) gas. The result is that the gas will cool during the
expansion. In an adiabatic expansion, we are convert-
Thus the equation for the pressure of an ideal gas during ing the heat energy contained in the gas into useful
an isothermal expansion is work. If we could keep this expansion going we could
suck all the thermal energy out of the gas and turn it into
p = constant (11a) useful work. The problem, of course, getting the piston
V
back to start the process over again.
and we see that the pressure decreases as 1/V. This
decrease is shown in the pV diagram of Figure (8).
pressure
p1 p1V1

pV = constant
(p = const / V)
is o
th e
rma
l e xp
area under a n sio
n p2V2
curve = work
p2
done by gas

V1 volume V2
Figure 8
In the isothermal expansion of an ideal gas, we have
pV = constant. Thus the pressure decreases as 1/V.
18-10 Entropy

It is instructive to compare an isothermal expansion to Exercise 3

an adiabatic expansion of a gas. In either case the In the next section, we will discuss a way of connecting
pressure drops. But in the adiabatic expansion, the adiabatic and isothermal expansions and compres-
pressure drops faster because the gas cools. In Figure sions in such a way that we form a complete cycle (get
(9), we compare the isothermal and adiabatic expan- back to the starting point), and get a net amount of work
sion curves for an ideal gas. Because the adiabatic out of the process. Before reading the next section, it is
curve drops faster in the pV diagram, there is less area a good exercise to see if you can do this on your own.
under the adiabatic curve, and the gas does less work.
In order to see whether or not you are getting work out
This is not too surprising, because less energy was
or putting it in, it is useful to graph the process in a pV
available for the adiabatic expansion since no heat
diagram, where the work is simply the area under the
flowed in. curve.
For an ideal gas, the equation for an adiabatic expan- To get you started in this exercise, suppose you begin
sion is with an ideal gas at a pressure p1 , volume V1 , and
Cp temperature T1 , and expand it isothermally to p2, V2, T1
pV γ = constant; γ = (12) as shown in Figure (10a). The work you get out is the
CV
area under the curve.
a result we derive in the appendix. (You calculated the
value of γ for various gases in Exercise 2.) The If you then compressed the gas isothermally back to
p1, V1, T1 , this would complete the cycle (get you back to
important point now is not so much this formula, as the
where you started), but it would take just as much work
fact that the adiabatic curve drops faster than the
to compress the gas as you got from the expansion.
isothermal curve. Thus there is no net work gained from this cycle. A more
complex cycle is needed to get work out.
If we compress a gas adiabatically, all the work we do
goes into the thermal energy of the gas, and the If we add an adiabatic expansion to the isothermal
temperature rises. Thus with an adiabatic expansion expansion as shown in Figure (10b) we have the start of
we can lower the temperature of the gas, and with an something more complex. See if you can complete this
adiabatic compression raise it. cycle, i.e., get back to p1, V1, T1 , using adiabatic and
pressure
isothermal expansions or compressions, and get some
p1 p1V1 net work in the process. See if you can get the answer
before we give it to you in the next section. Also show
graphically, on Figure (10b) how much work you do get
γ out.
pV = constant (for γ = 1.667)

pV = constant pressure
p1V1T1
is o
the
rma
l exp
a n sio isothermal
a di n expansion
aba
ti c e x p2
p a n sio n
p3

volume p2V2T1
V1 V2

Figure 9
Comparison of isothermal and adiabatic
expansions. In an adiabatic expansion the gas
cools, and thus the pressure drops faster. volume
V1 V2
Figure 10a
pV diagram for an isothermal expansion
from volume V1 to volume V2 .
 18-11

THE CARNOT CYCLE cycle by heating the cool gas back up to its original
With the isothermal and adiabatic expansion and com- temperature. In this way you get back to the original
pression of an ideal gas in a frictionless cylinder, we volume and temperature (and therefore pressure) of the
now have the pieces necessary to construct a Carnot cylinder; you have a complete cycle, and hopefully you
cycle, the key part of our thought experiment to study have gotten some useful work from the cycle.
the second law of thermodynamics.
To cool the gas, and then later heat it up again, Carnot
The goal is to construct a device that continually used an adiabatic expansion and then an adiabatic
converts heat energy into work. Such a device is called compression. We can follow the steps of the Carnot
an engine. Both the isothermal and adiabatic expan- cycle on the pV diagram shown in Figure (11). The gas
sions of the gas converted heat energy into work, but starts out at the upper left hand corner at a high
the expansions alone could not be used as an engine temperature T1 , volume V1 , pressure p1 . It then goes
because the piston was left expanded. Carnot’s re- through an isothermal expansion from a volume V1 to
quirement for an engine was that after a complete cycle a volume V2 , remaining at the initial temperature T1 .
all the working parts had to be back in their original The hot gas is then cooled down to a low temperature
condition ready for another cycle. Somehow the gas in T3 by an adiabatic expansion to a volume V3 . The cool,
the cylinder has to be compressed again to get the piston low pressure gas is then compressed isothermally to a
back to its original position. And the compression volume V4 , where it is then heated back to a higher
cannot use up all the work we got from the expansion, temperature T1 by an adiabatic compression. The
in order that we get some net useful work from the volume V4 is chosen just so that the adiabatic compres-
cycle. sion will bring the temperature back to T1 when the
volume gets back to V1 .
The idea for Carnot’s cycle that does give a net amount
of useful work is the following. Start off with the gas pressure
in the cylinder at a high temperature and let the gas p1V1T1
expand isothermally. We will get a certain amount of p1
work from the gas. Then rather than trying to compress
i s ot h

the hot gas, which would use up all the work we got,
er m

cool the gas to reduce its pressure. Then isothermally

al e
adiabati

compress the cool gas. It should take less work to

xpa

compress the low pressure cool gas than the work we

ns
io n
c co m

got from the high pressure hot gas. Then finish the
p2V2T1
p2
pre s

pressure
a di
ab
si o

p1V1T1
ati
n

ce

p4 is o
xp

p4V4T3 c o th er m nsio
a

mp n
isothermal r e s s al
p3 io n p3V3T3
expansion
volume
V1 V4 V2 V3
p2V2T1
Figure 11
adiabatic The Carnot Cycle. The gas first expands at a high
expansion temperature T1. It is then cooled to a lower
temperature T3 by an adiabatic expansion. Then it is
p3V3T3
compressed at this lower temperature, and finally
heated back to the original temperature T1 by an
volume adiabatic compression. We get a net amount of
V1 V2 V3
Figure 10b work from the process because it takes less work to
pV diagram for an isothermal expansion compress the cool low pressure gas than we got
followed by an adiabatic expansion. from the expansion of the hot high pressure gas.
18-12 Entropy

In this set of 4 processes, we get work out of the two Thermal Efficiency
expansions, but put work back in during the two of the Carnot Cycle
compressions. Did we really get some net work out? The net effect of the Carnot cycle is the following.
We can get the answer immediately from the pV During the isothermal expansion while the cylinder is
diagram. In Figure (12a), we see the amount of work at the high temperature TH , a certain amount of thermal
we got out of the two expansions. It is the total area or heat energy, call it Q H , flows into the cylinder. Q H
under the expansion curves. In Figure (12b) we see must be equal to the work the gas is doing during the
how much work went back in during the two compres- isothermal expansion since the gas’ own thermal en-
sions. It is the total area under the two compression ergy does not change at the constant temperature.
curves. Since there is more area under the expansion (Here, all the heat in becomes useful work.)
curves than the compression curves, we got a net
amount of work out. The net work out is, in fact, just During the isothermal compression, while the cylinder
equal to the 4 sided area between the curves, seen in is at the lower temperature TL , (the gas having been
Figure (13). cooled by the adiabatic expansion), an amount of heat
Q L is expelled from the cylinder. Heat must be ex-
pressure
pelled because we are doing work on the gas by
p1V1 T1 compressing it, and none of the energy we supply can
go into the thermal energy of the gas because its
temperature is constant. (Here all the work done be-
comes expelled heat.)

p2V2T1
Since no heat enters or leaves the cylinder during the
adiabatic expansion or compression, all flows of heat
have to take place during the isothermal processes.
Thus the net effect of the process is that an amount of
p3V3T3 thermal energy or heat Q H flows into the cylinder at the
high temperature TH , and an amount of heat Q L flows
V1 V3 volume out at the low temperature TL , and we get a net amount
Figure 12a
of useful work W out equal to the 4-sided area seen in
The work we get out of the two expansions is
equal to the area under the expansion curves. Figure (13). By the law of conservation of energy, the
pressure
pressure

p1V1 T1

heat in at high
TH temperature TH

net work done

during cycle

p4V4T3

TL
p3V3T3
heat out at low
volume
temperature TL
V1 V3 volume
Figure 12b Figure 13
The work required to compress the gas The net work we get out of one complete cycle is equal
back to its original volume is equal to to the area bounded by the four sided shape that lies
the area under the compression curves. between the expansion and compression curves.
 18-13

work W must be equal to the difference between Q H Reversible Engines

in and Q L out In our discussion of the principle of relativity, it was
immediately clear why we developed the light pulse
W = QH – QL (13)
clock thought experiment. You could immediately see
We see that the Carnot engine suffers from the same that moving clocks should run slow, and why that was
problem experienced by power plants and automobile a consequence of the principle of relativity.
engines. They take in heat Q H at a high temperature
(produced by burning fuel) and do some useful work We now have a new thought experiment, the Carnot
W, but they expel heat Q L out into the environment. To engine, which is about as idealized as our light pulse
be 100% efficient, the engine should use all of Q H to clock. We have been able to calculate the efficiency of
produce work, and not expel any heat Q L . But the a Carnot engine, but it is not yet obvious what that has
Carnot cycle does not appear to work that way. to do either with real engines, or more importantly with
the second law of thermodynamics which we are
One of the advantages of the Carnot cycle is that we can studying. It is not obvious because we have not yet
calculate Q H and Q L , and see just how efficient the discussed one crucial feature of the Carnot engine.
cycle is. It takes a couple of pages of calculations,
which we do in the appendix, but we obtain a remark- The Carnot engine is explicitly designed to be revers-
ably simple result. The ratio of the heat in, Q H , to the ible. As shown in Figure (14) , we could start at point
heat out, Q L , is simply equal to the ratio of the high 1 and go to point 4 by an adiabatic expansion of the gas.
temperature TH to the low temperature TL . During this expansion the gas would do work but no
heat is allowed to flow in. Thus the work energy would
for a carnot come from thermal energy and the gas would cool from
QH TH
= cycle based on (14) TH to TL .
QL TL an ideal gas
pressure
One suspects that if you do a lot of calculation involv-
ing integration, logarithms, and quantities like the 1
specific heat ratio, and almost everything cancels to
i s o th

leave such a simple result as Equation 14, then there

er m

might be a deeper significance to the result than ex-

TH
al c
adiabati

pected. Equation 14 was derived for a Carnot cycle

om

operating with an ideal gas. It turns out that the result

pre
ss
c exp

is far more general and has broad applications.

io n

2
a n si

a di
on

Exercise 4
ab
a
tic

A particular Carnot engine has an efficiency of 26.8%.

co

4 is o pr
m

That means that only 26.8% of QH comes out as useful e x therm es

pan al sio
work W and the rest, 73.2% is expelled at the low sio n n
temperature TL . The difference between the high and TL 3
low temperature is 100 K ( TH – TL = 100 K). What are volume
the values of TH and TL ? First express your answer in
kelvins, then in degrees centigrade. (The answer Figure 14
should be familiar temperatures.) The Carnot cycle run backward.

Exercise 5
If you have a 100% efficient Carnot engine, what can
you say about TH and TL ?
18-14 Entropy

The next step of the reverse Carnot cycle is an isother- In going from point 3 to point 2, we adiabatically
mal expansion from a volume V4 to a volume V3 . compress the gas to heat it from the lower temperature
During this expansion, the gas does an amount of work TL to the higher temperature TH . Since the compres-
equal to the area under the curve as shown in Figure sion is adiabatic, no heat flows in or out.
(15a). Since there is no change in the internal energy of
a gas when the temperature of the gas remains constant, In the final step from point 2 to point 1, we isothermally
the heat flowing in equals the work done by the gas. compress the gas back to its original volume V1 . Since
This is the same amount of heat Q L that flowed out the gas temperature remains constant at TH , there is no
when the engine ran forward. change in thermal energy and all the work we do,
shown as the area under the curve in Figure (15b), must
pressure be expelled in the form of heat flowing out of the
1 cylinder. The amount of heat expelled is just Q H , the
amount that previously flowed in when the engine was
run forward.
TH
We have gone through the reverse cycle in detail to
heat in equals emphasize the fact that the engine should run equally
work done by well both ways. In the forward direction the engine
2 the gas takes in a larger amount of heat Q H at the high
temperature TH , expels a smaller amount of heat Q L
TL
4 at the lower temperature TL , and produces an amount
of useful work W equal to the difference Q H – Q L .
3
In the reverse process, the engine takes in a smaller
V4 V3 volume
amount of heat Q L at the low temperature TL , and
a) During the isothermal expansion, some expels a larger amount Q H at the higher temperature
heat flows into the gas to supply the energy
needed for the work done by the gas.
TH . Since more heat energy is expelled than taken in,
an amount of work W = Q H – Q L must now be sup-
pressure plied to run the engine. When we have to supply work
1 to pump out heat, we do not usually call the device an
engine. The common name is a refrigerator. In a
TH heat expelled refrigerator, the refrigerator motor supplies the work
equals work done W, a heat Q L is sucked out of the freezer box, and a total
on the gas amount of energy Q L + W = Q H is expelled into the
higher room temperature of the kitchen. If we have a
2 Carnot refrigerator running on an ideal gas, then the
heats Q L and Q H are still given by Equation 14
4 TL Q L TL
= (14repeated)
3 Q H TH

V1 V2 volume where TL and TH are the temperatures on a scale

starting from absolute zero such as in the kelvin scale.
b) A lot more work is required, and a lot
more heat is expelled when we compress
the hot gas isothermally. Exercise 6
How much work must a Carnot refrigerator do to remove
Figure 15
1000 joules of energy from its ice chest at 0° C and expel
Heat flow when the Carnot cycle runs backward.
Since more heat flows out than in, some work W is the heat into a kitchen at 27° C?
required for the cycle. The net effect is that the work
W pumps heat out of the gas, giving us a refrigerator.
 18-15

ENERGY FLOW DIAGRAMS Figure (17) is the energy flow diagram for a Carnot
Because of energy conservation, we can view the flow refrigerator. A heat Q L is sucked out of the low
of energy in much the same way as the flow of some temperature reservoir, an amount of work W is sup-
kind of a fluid. In particular we can construct flow plied (by some motor), and the total energy
diagrams for energy that look much like plumbing Q H = Q L + W is expelled into the high temperature
diagrams for water. Figure (16) is the energy flow reservoir.
diagram for a Carnot engine running forward. At the
top and the bottom are what are called thermal reser- Maximally Efficient Engines
voirs—large sources of heat at constant temperature We are now ready to relate our discussion of the Carnot
(like swimming pools full of water). At the top is a cycles to the second law of thermodynamics. The
thermal reservoir at the high temperature TH (it could statement of the second law we will use is that you
be kept at the high temperature by burning fuel) and at cannot extract useful work from thermal energy at one
the bottom is a thermal reservoir at the low temperature temperature. (The colloquial statement of the first law
TL . For power plants, the low temperature reservoir is of thermodynamics — conservation of energy—is that
often the ocean or the cool water in a river. Or it may you can’t get something for nothing. The second law
be the cooling towers like the ones pictured in photo- says that you can’t break even.)
graphs of the nuclear power plants at Three Mile Island.
Up until now we have had to point out that our formula
In the energy flow diagram for the forward running for the efficiency of a Carnot engine was based on the
Carnot engine, an amount of heat Q H flows out of the assumption that we had an ideal gas in the cylinder. If
high temperature reservoir, a smaller amount Q L is we use the second law of thermodynamics, we can
expelled into the low temperature reservoir, and the show that it is impossible to construct any engine, by
difference comes out as useful work W. If the Carnot any means, that is more efficient than the Carnot engine
engine is run on an ideal gas, Q H and Q L are always we have been discussing. This will be the main result
related by Q H/Q L = TH/TL . of our thought experiment.

high temperature reservoir high temperature reservoir

TH TH

QH Carnot QH Carnot
W engine W refrigerator
QL QL

low temperature reservoir low temperature reservoir

TL TL

Figure 16 Figure 17
Energy flow diagram for a Carnot engine. Since energy The Carnot refrigerator is a Carnot engine run
is conserved, we can construct a flow diagram for backwards. The work W plus the heat QL equals the heat
energy that resembles a plumbing diagram for water. In QH pumped up into the high temperature reservoir.
a Carnot cycle, QH flows out of a “thermal reservoir”
at a temperature TH . Some of this energy goes out as
useful work W and the rest, QL , flows into the low
temperature thermal reservoir at a temperature TL .
18-16 Entropy

Let us suppose that you have constructed a Super The next thing you notice is that you do not need the low
engine that takes in more heat Q H * from the high temperature reservoir. All the heat expelled by your
temperature reservoir, and does more work W * , while Super engine is taken in by the Carnot refrigerator. The
rejecting the same amount of heat Q L as a Carnot low temperature reservoir can be replaced by a pipe and
engine. In the comparison of the two engines in Figure the new plumbing diagram for the combined Super
(18) you can immediately see that your Super engine is engine and Carnot refrigerator is shown in Figure (20).
more efficient than the Carnot engine because you get The overall result of combining the super engine and
more work out for the same amount of heat lost to the Carnot refrigerator is that a net amount of work
low temperature reservoir. Wnet = W *–W is extracted from the high temperature
reservoir. The net effect of this combination is to
Now let us run the Carnot engine backwards as a produce useful work from thermal energy at a single
refrigerator as shown in Figure (19). The Carnot temperature, which is a violation of the second law of
refrigerator requires an amount of work W to suck the thermodynamics.
heat Q L out of the low temperature reservoir and expel
the total energy Q H = W + Q L into the high tempera-
Exercise 7
ture reservoir.
Suppose you build a Super engine that takes the same
You do not have to look at Figure (19) too long before amount of heat from the high temperature reservoir as
you see that you can use some of the work W * that your a Carnot engine, but rejects less heat Q*L< QL than a
Super engine produces to run the Carnot refrigerator. Carnot engine into the low temperature. Using energy
Since your engine is more efficient than the Carnot flow diagrams show what would happen if this Super
cycle, W * > W and you have some work left over. engine were connected to a Carnot refrigerator. (You
would still be getting useful work from thermal energy at
high temperature reservoir TH
some temperature. From what temperature reservoir
Super engine Carnot engine would you be getting this work?)

Q*H QH high temperature reservoir TH

W* W Super engine Carnot refrigerator
QL QL

Wnet
QH
low temperature reservoir TL
QL QL

Figure 18
Comparison of the Super engine QL
with the Carnot engine.
Figure 20
high temperature reservoir TH If you connect the Super engine to the Carnot
Super engine Carnot refrigerator refrigerator, you can eliminate the low temperature
reservoir and still get some work Wnet out. This
machine extracts work from a single temperature, in
Q*H QH
W* W violation of the second law of thermodynamics.
QL QL

low temperature reservoir TL

Figure 19
Now run the Carnot engine
backward as a refrigerator.
 18-17

Reversibility APPLICATIONS OF
We have just derived the rather sweeping result that if THE SECOND LAW
the second law of thermodynamics is correct, you During the oil embargo in the middle 1970s, there was
cannot construct an engine that is more efficient than a a sudden appreciation of the consequences of the
Carnot engine based on an ideal gas. You may wonder second law of thermodynamics, for it finally became
why the cycles based on an ideal gas are so special. It clear that we had to use energy efficiently. Since that
turns out that they are not special. What was special time there has been a growing awareness that there is
about the Carnot engine is that it was reversible, that it a cost to producing energy that considerably exceeds
could be run backwards as a refrigerator. You can use what we pay for it. Burning oil and coal depletes
precisely the same kind of arguments we just used to natural limited resources and adds carbon dioxide to
show that all reversible engines must have precisely the atmosphere which may contribute to global climate
the same efficiency as a Carnot engine. It is a changes. Nuclear reactors, which were so promising in
requirement of the second law of thermodynamics. the 1950s, pose unexpected safety problems, both now
There were two reasons we went through the detailed as in the example of Chernobyl, and in the very distant
steps of constructing a Carnot engine using an ideal gas future when we try to deal with the storage of spent
in a frictionless piston. The first was to provide one reactor parts. Hydroelectric power floods land that
example of how an engine can be constructed. It is not may have other important uses, and can damage the
a very practical example, commercial engines are agricultural resources of an area as in the case of the
based on different kinds of cycles. But the Carnot Aswan Dam on the Nile River. More efficient use of
engine illustrates the basic features of all engines. In all energy from the sun is a promising idea, but technology
engines the process must be repetitive, at least two has not evolved to the point where solar energy can
temperatures must be involved, and only some of the supply much of our needs. What we have learned is
heat extracted from the high temperature reservoir can that, for now, the first step is to use energy as efficiently
be converted to useful work. Some heat must be as possible, and in doing this, the second law of
expelled at a lower temperature. thermodynamics has to be our guide.

While all reversible engines have the same efficiency, During the 1950s and 60s, one of the buzz words for
we have to work out at least one example to find out modern living was the all electric house. These houses
what that efficiency is. You might as well choose the were heated electrically, electric heaters being easy and
simplest possible example, and the Carnot cycle using inexpensive to install and convenient to use. And it also
an ideal gas is about as simple as they get. Because of represents one of the most stupid ways possible to use
the second law of thermodynamics, you know that energy. In terms of a heat cycle, it represents the 100%
even though you are working out a very special ex- conversion of work energy into thermal energy, what
ample, the answer Q H/Q L = TH/TL applies to all re- we would have called in the last section, a 0% efficient
versible engines operating between two temperature engine. There are better ways of using electric power
reservoirs. This is quite a powerful result from the few than converting it all into heat.
pages of calculations in the appendix. You can see where the waste of energy comes in when
you think of the processes involved in producing
electric power. In an electric power plant, the first step
is to heat some liquid or gas to a high temperature by
burning fuel. In a common type of coal or oil fired
power plant, mercury vapor is heated to temperatures
of 600 to 700 degrees centigrade. The mercury vapor
is then used to run a mercury vapor turbine which cools
the mercury vapor to around 200° C. This cooler
mercury vapor then heats steam which goes through a
steam turbine to a steam condenser at temperatures
18-18 Entropy

around 100° C. In a nuclear reactor, the first step is temperature at absolute zero, which is not only impos-
often to heat liquid sodium by having it flow through sible to achieve but even difficult to approach. You can
pipes that pass through the reactor. The hot sodium can see from this equation why many power plants are
then be used to heat mercury vapor which runs turbines located on the shore of an ocean or on the bank of a large
similar to those in a coal fired plant. The turbines are river. These bodies are capable of soaking up large
attached to generators which produce the electric power. quantities of heat at relatively low temperatures. If an
ocean or river is not available, the power plant will have
Even though there are many stages, and dangerous and large cooling towers to condense steam. Condensing
exotic materials used in power stations, we can esti- steam at atmospheric pressure provides a low tempera-
mate the maximum possible efficiency of a power plant ture of TL = 100° C or 373 K.
simply by knowing the highest temperature TH of the
boiler, and the lowest temperature TL of the condenser. Equation 15 also tells you why power plants run their
If the power plant were a reversible cycle running boilers as hot as possible, using exotic substances like
between these two temperatures, it would take in an mercury vapor or liquid sodium. Here one of the
amount of heat Q H at the high temperature and reject limiting factors is how high a temperature turbine
an amount of heat Q L at the low temperature, where blades can handle without weakening. Temperatures
Q H and Q L are related by Q H /Q L = TH /TL (Eq. 14). as high as 450° C or around 720 K are about the limit
The work we got out would be of current technology. Thus we can estimate the
maximum efficiency of power plants simply by know-
amount of work from
W = QH – QL a reversible cycle (15) ing how high a temperature turbine blades can with-
stand, and that the plant uses water for cooling. You do
We would naturally define the efficiency of the cycle not have to know the details of what kind of fuel is used,
as the ratio of the work out to the heat energy in what kind of exotic materials are involved, or how
Q – QL turbines and electric generators work, as long as they
efficiency = W = H (16) are efficient. Using the numbers TH = 720 k ,
QH QH
TL = 373 k we find that the maximum efficiency is
If we solve Equation 14 for Q L about
T
QL = QH L maximum = TH – TL = 720 – 373
TH efficiency TH 720
and use this in Equation 16, we get (18)
= .48
QH – QL Q H 1 – TL/TH
efficiency = = Thus about 50% represents a theoretical upper limit to
QH QH the efficiency of power plants using current technol-
ogy. In practice, well designed power plants reach only
TH – TL efficiency of a about 33% efficiency due to small inefficiencies in the
efficiency = reversible cycle
(17)
TH many steps involved.

Since by the second law of thermodynamics no process You can now see why the all electric house was such a
can be more efficient than a reversible cycle, Equation bad idea. An electric power plant consumes three times
17 represents the maximum possible efficiency of a as much fuel energy as it produces electric energy.
power plant. Then the electric heater in the all electric house turns
this electric energy back into thermal energy. If the
The important thing to remember about Equation 17 is house had a modern oil furnace, somewhere in the
that the temperatures TH and TL start from absolute order of 85% of the full energy can go into heating the
zero. The only way we could get a completely efficient house and hot water. This is far better than the 33%
engine or power plant would be to have the low efficiency from heating directly by electricity.
 18-19

Electric Cars The Heat Pump

One of the hot items in the news these days is the There is an intelligent way to heat a house electrically,
electric car. It is often touted as the pollution free and that is by using a heat pump. The idea is to use the
solution to our transportation problems. There are electric energy to pump heat from the colder outside
advantages to electric cars, but not as great an advan- temperature to the warmer inside temperature. Pump-
tage as some new stories indicate. ing heat from a cooler temperature to a warmer tem-
perature is precisely what a refrigerator does, while
When you plug in your electric car to charge batteries, taking heat from the freezer chest and exhausting it into
you are not eliminating the pollution associated with the kitchen. The heat pump takes heat from the cooler
producing useful energy. You are just moving the outside and exhausts it into the house.
pollution from the car to the power plant, which may,
however, be a good thing to do. A gasoline car may As we saw in our discussion, it takes work to pump heat
produce more harmful pollutants than a power station, from a cooler to a higher temperature. The ratio of the
and car pollutants tend to concentrate in places where heat Q L taken in at the low temperature, to the heat Q H
people live creating smog in most major cities on the expelled at the higher temperature, is Q H/Q L = TH/TL
earth. (Some power plants also create obnoxious for a maximally efficient refrigerator. The amount of
smog, like the coal fired plants near the Grand Canyon work W required is W = Q H – Q L . The efficiency of
that are harming some of the most beautiful scenery in this process is the ratio of the amount of heat delivered
the country.) to the work required.

In addition to moving and perhaps improving the efficiency of Q QH

nature of pollution, power plants have an additional = H =
heat pump W QH – QL
advantage over car engines—they are more efficient. (19)
Car engines cannot handle as high a temperature as a TH
=
power plant, and the temperature of the exhaust from a TH – TL
car is not as low as condensing steam or ocean water.
Car engines seldom have an efficiency as high as 20%; where the last step in Equation 19 used Q L = Q HTL/TH .
in general, they are less than half as efficient as a power
plant. Thus there will be a gain in efficiency in the use Exercise 8
of fuel when electric cars come into more common use. Derive the last formula in Equation 19.

(One way electric cars have for increasing their effi- When the temperature difference TH – TL is small, we
ciency is to replace brakes with generators. When can get very high efficiencies, i.e., we can pump a lot
going down a hill, instead of breaking and dissipating of heat using little work. In the worst case, where
energy by heating the brake shoes, the gravitational TL = 0 and we are trying to suck heat from absolute
potential energy being released is turned into electric zero, the efficiency of a heat pump is 1—heat delivered
energy by the generators attached to the wheels. This equals the work put in—and the heat pump is acting
energy is then stored as chemical energy in the batteries like a resistance heater.
as the batteries are recharged.)
18-20 Entropy

To illustrate the use of a heat pump let us assume that Exercise 9

it is freezing outside ( TL = 0° C = 273 K ) and you This so-called “heat of fusion” of water is 333kJ/kg.
want the inside temperature to be 27°C = 300 K . Then What that means is that when a kilogram (1 liter) of water
a heat pump could have an efficiency of freezes (going from 0° C water to 0° ice), 333 kilojoules
of heat are released. Thus to freeze a liter of 0° C water
in your refrigerator, the refrigerator motor has to pump
efficiency of heat TH 333 × 103 joules of heat energy out of the refrigerator
pump running from = into the kitchen. The point of the problem is to estimate
TH – TL
0°C to 27°C how powerful a refrigerator motor is required if you want
to be able to freeze a liter of water in 10 minutes.
300 k
=
300 k – 273 k Assume that the heat is being removed at a temperature
of 0°C and being expelled into a kitchen whose tem-
= 11.1 perature is 30°C, and that the refrigerator equipment is
100% efficient. (We will account for a lack of efficiency
In other words, as far as the second law of thermody- at the end of this problem.)
namics is concerned, we should be able to pump eleven In the United Sates, the power of motors is generally
times as much heat into a house, when it is just freezing given in “horsepower”, a familiar but archaic unit. The
outside, as the amount of electrical energy required to conversion factor is 1 horsepower = 746 watts, and a
pump the heat. Even if the electrical energy is produced power of 1 watt is 1 joule per second.
at only 30% efficiency, we should still get
.30 × 11.1 = 3.3 times as much heat into the house as Calculate the horsepower required, then double the
by burning the fuel in the house at 100% conversion of answer to account for lack of efficiency. (Answer:
0.16 horsepower.)
fuel energy into heat.
Exercise 10
Here is a problem that should give you some practice
with the concepts of efficiency. You have the choice of
buying a furnace that converts heat energy of oil into
heat in the house with 85% efficiency. I.e., 85% of the
heat energy of the oil goes into the house, and 15% goes
up the chimney. Or you can buy a heat pump which is
half as efficient as a Carnot refrigerator. (This is a more
realistic estimate of the current technology of refrigera-
tion equipment.) At very low temperatures outside, heat
pumps are not as efficient, and burning oil in your own
furnace is more efficient. But if it does not get too cold
outside, heat pumps are more efficient. At what outside
temperature will the heat pump and the oil furnace have
the same efficiency? Assume that the electric energy
you use is produced by a power plant that is 30%
efficient. (Answer: - 26° C.)
 18-21

The Internal Combustion Engine While the piston goes back down, the valves are set so
We finish this section on practical applications with a that a mixture of air and fuel are sucked into the piston.
brief discussion of the internal combustion engine. The When the cylinder is at the bottom of the piston, we
main point is to give an example of an engine that runs have a cool, low pressure fuel air mixture filling the full
on a cycle that is different from a Carnot cycle. It is volume V4 . We are now at the position labeled (4) in
more difficult to apply the second law of thermody- Figure (21). It took two strokes (up and down) of the
namics to an internal combustion engine because it piston to go from position 3 to position 4.
does not take heat in or expel heat at constant tempera-
tures like the Carnot engine, but we can still analyze the In the final stroke, the valves are shut and the rising
work we get out using a pV diagram. piston adiabatically compresses the gas back to the
starting point p 1 , V1 , T1 . While the increase in tem-
The pV diagram for an internal combustion engine is perature during this compression is what is needed to
shown in Figure (21). At position 1, a fuel and air ignite the diesel fuel, you do not want the temperature
mixture have been compressed to a small volume V1 to rise enough to ignite the air gasoline mixture in a
by the piston which is at the top of the cylinder. If it is gasoline engine. This can sometimes happen in a
a gasoline engine, the fuel air mixture is ignited by a gasoline engine, causing a knock in the engine, or
spark from a sparkplug. If it is a diesel engine, the sometimes allowing the engine to run for a while after
mixture of diesel fuel and air have been heated to the you have shut off the ignition key and stopped the spark
point of combustion by the adiabatic compression from plug from functioning.
point 4 to point 1 that has just taken place. One of the
pressure
advantages of a diesel engine is that an electrical
2
system to produce the spark is not needed. This is power stroke
particularly important for boat engines where electric
systems give all sorts of problems. (We said this was
a section on practical applications.) ad ab
i a ti
spark 1 c
After ignition, the pressure and temperature of the gas
rise rapidly to p 2 , T2 before the piston has had a chance ad 3
ia b exhaust
to move. Thus the volume remains at V1 and the pV a ti c
curve goes straight up to point 2. The heated gas then 4
expands adiabatically, and cools some, driving the compression
piston down to the bottom of the cylinder. This is the volume
V1 V4
stroke from which we get work from the engine.

We now have a cylinder full of hot burned exhaust V1

gases. In a 4 cycle engine, a valve at the top of the V4
cylinder is opened, and a piston is allowed to rise,
pushing the hot exhaust gases out into the exhaust
pipes. Not much work is required to do this. This is the
part of the cycle where (relatively) low temperature
thermal energy is exhausted to the environment. Figure 21
PV diagram for an internal combustion
engine. When the piston is all the way up in
the cylinder the volume is V1 . When it is all
the way down, the volume has increased to V2 .
18-22 Entropy

ENTROPY We see that the inefficient engine expelled more Q/T

The second law of thermodynamics provided us with than it took in. The inefficient, non reversible, engine
the remarkable result that the efficiency of all revers- creates Q/T while reversible engines do not.
ible engines is the same. Detailed calculation of this As we have done throughout the course, whenever we
efficiency using a Carnot engine based on an ideal gas encounter a quantity that is conserved, or sometimes
gave us a surprisingly simple formula for this effi- conserved, we give it a name. We did this for linear
ciency, namely Q H/Q L = TH/TL . Our preceding ex- momentum, angular momentum, and energy. Now we
amples involving car engines, power plants, refrigera- have a quantity Q/T that is unchanged by reversible
tors and heat pumps illustrate how important this engines, but created or increased by irreversible ineffi-
simple relationship is to mankind. cient ones. We are on the verge of defining the quantity
When you do a calculation and a lot of stuff cancels out, physicists call entropy. We say on the verge of defining
it suggests that your result may have a simpler interpre- entropy, because Q/T is not entropy itself; it represents
tation than you originally expected. This turns out to be the change in entropy. We can say that when the gas
true for our calculations of the heat flow in a Carnot expanded, the entropy of the gas increased by Q H /TH .
engine. To get a new perspective on our equation for And when the gas was compressed, the entropy de-
heat flow, let us write the equation in the form creased by Q L/TL .
QH Q For a reversible engine there is no net change in entropy
= L (20) as we go around the cycle. But for an irreversible,
TH TL
inefficient engine, more entropy comes out than goes in
In this form the equation for heat flow is beginning to during each cycle. The net effect of an inefficient
look like a conservation law for the quantity Q/T. engine is to create entropy.
During the isothermal expansion, an amount Q H /TH
of this quantity flowed into the piston. During the What is this thing called entropy that is created by
isothermal compression, Q L /TL flowed out. We find inefficient irreversible engine? Consider the most
that if the engine is reversible, the amount of Q/T that inefficient process we can imagine—the electric heater
flowed in is equal to the amount of Q/T that flowed out. which converts useful work in the form of electrical
The net effect is that there was no change in Q/T during energy into heat. From one point of view, the device
the cycle. does nothing but create entropy. If the heater is at a
temperature T, and the electric power into the heater is
To get a better insight into what this quantity Q/T may W watts, all this energy is converted to heat and entropy
be, consider a nonreversible engine operating between is produced at a rate of W/T in units of entropy per
TH and TL , an engine that would be less efficient than second. (Surprisingly, there is no standard name for a
the Carnot engine. Assume that the less efficient unit of entropy. The units of Q/T are of course joules/
engine and the Carnot engine both take in the same kelvin, the same as Boltzman’s constant.)
amount of heat Q H at the high temperature TH . Then
the less efficient engine must do less work and expel The process of converting energy in the form of useful
more heat at the low temperature TL . Thus Q L for the work into the random thermal energy of molecules can
less efficient engine is bigger than Q L for the Carnot be viewed as the process of turning order into disorder.
engine. Since Q L/TL = Q H/TH for the Carnot engine, Creating entropy seems to be related to creating disor-
Q L/TL must be greater than Q H/TH for the less effi- der. But the surprising thing is that we have an explicit
cient engine formula Q/T for changes in entropy. How could it be
for an engine that
possible to measure disorder, to have an explicit for-
QL Q mula for changes in disorder? This question baffled
> H is less efficient than (21)
TL TH a Carnot engine physicists for many generations.
 18-23

Ludwig Boltzman proposed that entropy was related to When working with large numbers, it is easier to deal
the number of ways that a system could be arranged. with the logarithm of the number than the number
Suppose, for example, you go into a woodworking itself. There are approximately 10 51 protons in the
shop and there are a lot of nails on the wall with tools earth. The log to the base 10 of this number is 51 and
hanging on them. In one particular woodworking shop the natural logarithm, ln 10 51 , is 2.3 times bigger or
you find that the carpenter has drawn an outline of the 117. In discussing the number of protons in the earth,
tool on the wall behind the nail. You enter her shop you the number 117 is easier to work with than 10 51 ,
find that the tool hanging from each nail exactly particularly if you have to write out all the zeros.
matches the outline behind it. Here we have perfect
order, every tool is in its place and there is one and only If we describe the disorder of our tool hanging system
one way the tools can be arranged. We would say that, in terms of the logarithm of the number of ways the
as far as locating tools is concerned the shop is in perfect tools can be hung, we get a much more reasonable set
order, it has no disorder or entropy. of numbers, as shown in Table 1.

On closer inspection, we find that the carpenter has two Setup Number of ways Logarithm of the
to arrange tools number of ways
saws with identical outlines, a crosscut saw and a rip
saw. We also see that the nails are numbered, and see all tools have
the cross cut saw on nail 23 and the rip saw on nail 24. unique positions 1 0
A week later when we come back, the cross cut is on two saws can
nail 24 and the rip saw on 23. Thus we find that her be interchanged 2 .7
system is not completely orderly, for there are two 25 tools on
different ways the saws can be placed. This way of unmarked nails 1.5 × 1025 58
organizing tools has some entropy. 50 tools on
unmarked nails 3.0 × 1065 148
A month later we come back to the shop and find that
another carpenter has taken over and painted the walls.
TABLE 1
We find that there are still 25 nails and 25 tools, but now
there is no way to tell which tool belongs on which nail.
The table starts off well. If there is a unique arrange-
Now there are many, many ways to hang up the tools
ment of the tools, only one way to arrange them, the
and the system is quite disordered. We have the feeling
logarithm of the number of ways is 0. This is consistent
that this organization, or lack of organization, of the
with our idea that there is no disorder. As the number
tools has quite high entropy.
of tools on unmarked nails increases, the number of
To put a numerical value on how disorganized the ways they can be arranged increases at an enormous
carpenter shop is, we go to a mathematician who tells pace, but the logarithm increases at a reasonable rate,
us that there are N! ways to hang N tools on N nails. approximately as fast as the number of tools and nails.
Thus there are 25! = 1.55 × 10 25 different ways the 25 This logarithm provides a reasonable measure of the
tools can be hung on the 25 nails. We could use this disorder of the system.
number as a measure of the disorder of the system, but
the number is very large and increases very rapidly with
the number of tools. If, for example, there were 50 tools
hung on 50 nails, there would be 3.04 × 10 64 different
ways of hanging them. Such large numbers are not
convenient to work with.
18-24 Entropy

We could define the entropy of the tool hanging system Thus Boltzman’s equation gives us an explicit formula
as the logarithm of the number of ways the tools could for the fractional increase in the number of ways the
be hung. One problem, however, is that this definition atoms in the gas atoms in the cylinder can be arranged.
of entropy would have different dimensions than the
definition introduced earlier in our discussions of en- Boltzman committed suicide in 1906, despondent over
gines. There changes in entropy, for example Q H /TH , the lack of acceptance of his work on the statistical
had dimensions of joules/kelvin, while our logarithm is theory of matter, of which Equation 22 is the corner-
dimensionless. However, this problem could be fixed stone. And in 1906 it is not too surprising that physi-
by multiplying our dimensionless logarithm by some cists would have difficulty dealing with Boltzman’s
fundamental constant that has the dimensions of joules/ equation. What is the meaning of the number of ways
kelvin. That constant, of course, is Boltzman’s con- you can arrange gas atoms in a cylinder? From a
stant k, where k = 1.38 × 10 –23 joules/kelvin. We Newtonian perspective, there are an infinite number of
could therefore take as the formula for the entropy (call ways to place just one atom in a cylinder. You can
it S) of our tool hanging system as count them by moving the atoms an infinitesimal
distance in any direction. So how could it be that 10 24
entropy of our tool atoms in a cylinder have only a finite way in which they
S = k ln n (22)
hanging system can be arranged?

where n is the number of ways the tools can be hung. This question could not be satisfactorily answered in
Multiplying our logarithm by k gives us the correct 1906, the answer did not come until 1925 with the
dimensions, but very small values when applied to as discovery of quantum mechanics. In a quantum pic-
few items as 25 or 50 tools. ture, an atom in a cylinder has only certain energy
levels, an idea we will discuss later in Chapter 35. Even
Equation 22 appears on Boltzman’s tombstone as a when you have 10 24 atoms in the cylinder, the whole
memorial to his main accomplishment in life. Boltzman system has only certain allowed energy levels. At low
believed that Equation 22 should be true in general. temperatures the gas does not have enough thermal
That, for example, it should apply to the atoms of the energy to occupy very many of the levels. As a result
gas inside the cylinder of our heat engine. When heat the number of ways the atoms can be arranged is
flows into the cylinder and the entropy increases by an limited and the entropy is low. As the temperature is
amount Q H /TH , the number of ways that the atoms increased, the gas atoms can occupy more levels, can
could be arranged should also increase, by an amount be arranged in a greater number of ways, and therefore
we can easily calculate using Equation 22. Explicitly, have a greater entropy.
if before the heat flowed in there were n old ways the
atoms could be arranged, and after the heat flowed in The concept of entropy provides a new definition of
n new ways, then Equation 22 gives absolute zero. A system of particles is at absolute zero
when it has zero entropy, when it has one uniquely
change in QH defined state. We mentioned earlier that quantum
entropy = k ln n new – k ln n old = T mechanics requires that a confined particle has some
H
kinetic energy. All the kinetic energy cannot be
Since ln n new – ln n old = ln n new /n old , we get removed by cooling. This gives rise to the so-called
n new QH zero point energy that keeps helium a liquid even at
ln = (23) absolute zero. However, a bucket of liquid helium can
n old kTH
be at absolute zero as long as it is in a single unique
Taking the exponent of both sides of Equation 16, using quantum state, even though the atoms have zero point
the fact that e ln x = x , we get kinetic energy.
n new QH /kTH
n old = e (24)
 18-25

In our discussion of temperature in the last chapter, we For a while there was a debate among physicists as to
used the ideal gas thermometer for our experimental whether the second law of thermodynamics was the
definition of temperature. We pointed out, however, only law in nature that could be used to distinguish
that this definition would begin to fail as we ap- between time running forward and time running back-
proached very low temperatures near absolute zero. At ward.
these temperatures we need a new definition which
agrees with the ideal gas thermometer definition at When you study processes like the decay of one kind of
higher temperatures. The new definition which is used elementary particle into another, the situation is so
by the physics community, is based on the efficiency of simple that the concepts of entropy and the second law
a reversible engine or heat cycle. You can measure the of thermodynamics do not enter into the analysis. In
ratio of two temperatures TH and TC , by measuring the these cases you can truly study whether nature is
heats Q H and Q C that enter and leave the cycle, and use symmetric with the respect to the reversal of time. If
the formula TH/TC = Q H/Q C . Since this formula is you take a moving picture of a particle decay, and run
based on the idea that a reversible cycle creates no the movie backwards, will you see a process that can
entropy ( Q H/TH = Q C/TC ), we can see that the con- actually happen? For example if a muon decays into an
cept of entropy forms the basis for the definition of electron and a neutrino, as happened in our muon
temperature. lifetime experiment, running that moving picture back-
ward would have neutrinos coming in, colliding with
an electron, creating a muon. Thus, if the basic laws of
The Direction of Time physics are truly symmetric to the reversal of time, it
We began the chapter with a discussion of a demonstra- should be possible for a neutrino and an electron to
tion that looked funny. We set the strobe so that the collide and create a muon. This process is observed.
water drops appeared to rise from the plate in the bucket
and enter the hose. Before our discussion of the second In 1964 Val Fitch and James Cronin discovered an
law of thermodynamics, we could not find any law of elementary particle process which indicated that nature
physics that this backward process appeared to violate. was not symmetric in time. Fitch found a violation of
Now we can see that the launching of the drops from the this symmetry in the decay of a particle called the
plate is a direct contradiction of the second law. In that neutral k meson. For this discovery, Fitch was awarded
process, heat energy in the bucket converts itself at one the Nobel prize in 1980. Since the so-called “weak
temperature into pure useful work that launches the interaction” is responsible for the decay of k mesons,
drop. the weak interaction is not fully symmetric to the
reversal of time. The second law of thermodynamics
When you run a moving picture of some action back- is not the only law of physics that knows which way
wards, effectively reversing the direction of time, in time goes.
most cases the only law of physics that is violated is the
second law of thermodynamics. The only thing that
appears to go wrong is that disordered systems appear
to organize themselves on their own. Scrambled eggs
turn into an egg with a whole yoke just by the flick of
a fork. Divers pop out of swimming pools propelled,
like the drops in our demonstration, by the heat energy
in the pool (see movie). All these funny looking things
require remarkable coincidences which in real life do Movie
not happen. Time reversed motion picture of dive
18-26 Entropy

APPENDIX: CALCULATION OF THE Using Equation 1, we get

EFFICIENCY OF A CARNOT CYCLE V2 V2
The second law of thermodynamics tells us that the W = pdV = RT dV
efficiency of all reversible heat engines is the same. V
V1 V1
Thus if we can calculate the efficiency of any one V2
engine, we have the results for all. Since we have based V2
so much of our discussion on the Carnot engine running = RT dV = RTlnV
V V1
on an ideal gas, we will calculate the efficiency of that V1
engine.
= RT ln V2 – ln V1 (A-2)
To calculate the efficiency of the ideal gas Carnot
engine, we need to calculate the amount of work we get V2
= RT ln
out of (or put into) isothermal and adiabatic expan- V1
sions. With these results, we can then calculate the net
amount of work we get out of one cycle and then the Thus the work we get out is RT times the logarithm of
efficiency of the engine. To simplify the formulas, we the ratio of the volumes.
will assume that our engine is running on one mole of
an ideal gas. Adiabatic Expansion
It is a bit trickier to calculate the amount of work we get
Isothermal Expansion out of an adiabatic expansion. If we start with a mole
Suppose we have a gas at an initial volume V1 , pressure of ideal gas at a volume V1 , pressure p 1 , and tempera-
p 1 , temperature T, and expand it isothermally to a ture TH , the gas will cool as it expands because the gas
volume V2 , pressure p 2 , and of course the same does work and we are not letting any heat in. Thus
temperature T. when the gas gets to the volume V2 , at a pressure p 2 ,
its temperature TC will be cooler than its initial tem-
The P-V diagram for the process is shown in Figure perature TH .
(A-1). The curve is determined by the ideal gas law,
which for 1 mole of an ideal gas is The pV diagram for the adiabatic expansion is shown
in Figure (A-2). To get an equation for the adiabatic
pV=RT (A-1) expansion curve shown, let us assume that we change
The work we get out of the expansion is the shaded area the volume of the gas by an infinitesimal amount ∆V .
under the curve, which is the integral of the pressure With this volume change, there will be an infinitesimal
curve from V1 to V2 . pressure drop ∆p , and an infinitesimal temperature
pressure pressure
p1 p1 TH

p2

p2 TL
isothermal adiabatic
V1 V2 volume V1 V2 volume
Figure A-1 Figure A-2
Isothermal expansion. Adiabatic expansion.
 18-27

drop ∆T . We can find the relationship between these It is standard notation to define the ratio of specific
small changes by differentiating the ideal gas equa- heats by the constant γ
tions. Starting with Cp
pv=RT ≡ γ (A-7)
CV
and differentiating we get thus Equation A-6 can be written in the more compact
form
p∆V + ∆p V = R∆T (A-3)
γp∆V + ∆pV = 0 (A-8)
We now have to introduce the idea that the expansion
is taking place adiabatically, i.e., that no heat is enter- The next few steps will look like they were extracted
ing. That means that the work p∆V done by the gas from a calculus text. They may or may not be too
during the infinitesimal expansion ∆V must all have familiar, but you should be able to follow them step-by-
come from thermal energy. But the decrease in thermal step.
energy is C V∆T . Thus we have from conservation of
First we will replace ∆V and ∆p by dV and dp to
energy
indicate that we are working with calculus differen-
p∆V + C V∆T = 0 tials. Then dividing through by the product pV gives
dp
or γ dV + p = 0 (A-9)
V
– p∆V
∆T = (A-4) Doing an indefinite integration of this equation gives
CV
The – (minus) sign tells us that the temperature drops γ ln V + ln p = const (A-10)
as work energy is removed. The γ can be taken inside the logarithm to give
Using Equation A-4 for ∆T in Equation A-3 gives ln V γ + ln p = const (A-11)
– p∆V Next exponentiate both sides of Equation A-11 to get
p∆V + ∆pV = R
CV
eln Vγ + ln p = econst = another const (A-12)
Combining the p∆V terms gives
where e const is itself a constant. Now use the fact that
p∆V 1 + R + ∆pV = 0 e a + b = e a e b to get
CV
eln Vγ eln p = const (A-13)
CV + R
p∆V + ∆pV = 0 (A-5) Finally use e ln(x) = x to get the final result
CV

pV γ = const
adiabatic
Earlier in the chapter, in Equation 9, we found that for expansion
(A-14)
an ideal gas, C V and C p were related by C p = C V + R .
Thus Equation A-5 simplifies to
Cp
p∆V + ∆pV = 0 (A-6)
CV
18-28 Entropy

Equation A-14 is the formula for the adiabatic curve The heat Q L expelled at the low temperature TL is
seen in Figure (A-2). During an isothermal expansion, equal to the work we do compressing the gas isother-
we have pV= RT where T is a constant. Thus if we mally in going from point 3 to 4. This work is
compare the formulas for isothermal and adiabatic
expansions, we have for any ideal gas W3 4 = Q L = RTC ln V4 /V3 (A-17)

Taking the ratio of Equations A-16 to A-17 we get

pV = const isothermal expansion

pV γ = const adiabatic expansion QH TH ln V2 /V1

= (A-18)
γ = C p /C V ratio of specific heats (A-15) QL TC ln V4 /V3
Cp = CV + R The next step is to calculate the ratio of the logarithms
of the volumes using the adiabatic expansion formula
The Carnot Cycle pV γ = constant .
We now have the pieces in place to calculate the In going adiabatically from 2 to 3 we have
efficiency of a Carnot cycle running on one mole of an
ideal gas. The cycle is shown in Figure (11) repeated p 2V2 γ = p 3V3 γ (A-19)
here as Figure (A-3).
and in going from 4 to 1 adiabatically we have
During the isothermal expansion from point 1 to 2, the
p 4V4 γ = p 1V1 γ (A-20)
amount of heat that flows into our mole of gas is equal
to the work one by the gas. By Equation A-2, this work Finally, use the ideal gas law pV = RT to express the
is pressure p in terms of volume and temperature in
W12 = Q H = RTH ln V2 /V1 (A-16) Equations A-19 and A-20. Explicitly use

p 1 = RTH / V1 ; p 2 = RTH / V2
pressure (A-21)
p 3 = RTH / V3 ; p 4 = RTH / V4
p1V1T1
p1 to get for Equation (19)
is o th

RTH γ RTC γ
er m

V2 = V
V2 V3 3
al e
adiabati
xpa

or
ns

THV2 γ – 1 = TCV3 γ – 1
io n
c co m

p2V2T1
(A-22)
p2
pre s

ad

and similarly for Equation (20)

iab
si o

ati
n

THV1 γ – 1 = TCV4 γ – 1
ce

p4 is o (A-23)
xp

p4V4T3 c o th e r m n sio
a

mp n
r e s s al
p3 io n p3V3T3 as you can check for yourself.
volume
V1 V4 V2 V3
Figure A-3
The Carnot cycle.
 18-29

If we divide Equation A-22 by A-23 the temperatures

TH and TC cancel, and we get
V2 γ – 1 V3 γ – 1
=
V1 γ – 1 V4 γ – 1
or
γ–1 γ–1
V2 V3
= (A-24)
V1 V4

Taking the ( γ – 1 )th root of both sides of Equation A-

24 gives simply
V2 V3
= (A-25)
V1 V4
Since V2 /V1 = V3 /V4 , the logarithms in Equation A-
18 cancel, and we are left with the surprisingly simple
result
QH T
= H (14repeated)
QL TL
which is our Equation 14 for the efficiency of a Carnot
cycle. As we mentioned, when you are doing a
calculation and a lot of stuff cancels to give a simple
result, there is a chance that your result is more general,
or has more significance than you expected. In this
case, Equation 14 is the formula for the efficiency of
any reversible engine, no matter how it is constructed.
We happened to get at this formula by calculating the
efficiency of a Carnot engine running on one mole of
an ideal gas.
Chapter 19
The Electric Interaction
Atomic & Molecular Forces
CHAPTER 19 THE ELECTRIC IN-
TERACTION

THE FOUR BASIC INTERACTIONS

The world around us is a complex place with enor-
mous variety of a myriad of interactions. But if you
look in the right places, from the right point of view,
you may find great simplicity. Planetary motion is
one example. If you look at the sun and planets alone,
ignoring things on a larger scale like other stars and
galaxies, and on a smaller scale like the makeup of
the planets and the atmosphere of the sun, you have
a system of 10 objects whose behavior is accurately
determined by a single force law. The system is
simple enough that mankind learned about physics
by studying it.

In this and the next chapter we take a first look at

several of the basic patterns and laws of physics.
Perhaps the most important discovery in the twenti-
eth century, actually more of a gradual realization
made during the first half of the twentieth century,
was that all of the phenomena of nature, everything
we see around us, can be explained in terms of four
basic forces or interactions. In some circumstances,
as in the case of planetary motion, a single force
dominates and the structures it creates are obvious.
In most cases however, the structures are complex
and it is no easy task to uncover the underlying
forces.

What we will do in these two chapters is to describe

the four basic interactions, focusing our attention on
examples where the action of the force is most clearly
seen. For the gravitational force, a planetary system
19-2 The Electric Interaction

provides the most clear and detailed example of a When you bring complete atoms close to each other,
structure created by gravity. Using this example as there are weak residual electric forces which result
a guide we can see that larger structures like globu- from small distortions of the atomic structure. These
lar clusters and galaxies are an extension of the weak residual forces are the molecular forces that
gravitational interaction to more complex situa- hold atoms together to form molecules, crystals and
tions. Although we do not attempt to calculate in most of the variety we see in the world about us. In
detail the motion of the millions or billions of stars this chapter we will consider only the simplest
in such objects, we gain an intuitive feeling for the example of a molecular force to see how such
kind of structures the gravitational interaction cre- residual forces arise when atoms interact.
ates.
In the next chapter we leave the scale of atoms and
The electric interaction, the second of the four basic molecules and look down inside the atomic nucleus,
interactions to be discovered, is most clearly seen at where we find two more forces at work. There is an
an atomic level. To introduce the electric interac- attractive force, called the nuclear interaction, that is
tion, we will take the simple point of view that atoms even stronger than electricity. It holds the nucleus
consist of a tiny nucleus made from protons and together against the electric repulsion between the
neutrons, surrounded by electrons held to the nucleus proton in the nucleus. There is also another force,
by the electric force. The model should be familiar, called the weak interaction, which allows neutrons to
it is essentially a scaled down version of the solar decay into protons and protons into neutrons (the β
system, with an electric force replacing the gravita- decay reaction discussed in Chapter 6). The nature of
tional force. After a brief discussion of the kinds of nuclear reactions, the stability of atomic nuclei, and the
nuclei that can be made from protons and neutrons, abundance of the elements depend upon a delicate
we will look at the properties of the electric interac- interplay of the nuclear, the weak, and the electric
tion that give rise to complete atoms. This involves interactions.
the concept of electric charge, the fact that electric-
ity, like gravity, is a 1 r 2 force law, and the fact that, In the past 30 years we have been able to sharpen
on an atomic scale, the electric force is much, much our view of nuclear and subnuclear matter. We have
stronger than gravity. been able to look inside the proton and neutron and
discover that they are not elementary particles, but
instead, composite objects made from quarks. We
Jupiter electron
have also found that the basic nuclear force is the
one between quarks. The force that holds protons
together in the nucleus is a residual of the quark
sun proton
force, just as molecular forces that hold atoms
together are a residual of the electric force. Because
of a mathematical analogy to the theory of color, the
planetary hydrogen force between quarks is called the color force. The
system atom color force now replaces the residual nuclear force
Figure 0 as one of the four basic interactions.
The classical model of the hydrogen atom
closely resembles a planetary system Once we see how nature can be explained in terms
of just four basic forces, we cannot help wondering
why the number is four. Are there more basic forces,
some of them yet undetected? Or are there fewer
than four basic forces, some of the four being equiva-
lent on a more fundamental level?
 19-3

Einstein spent the latter half of his life trying to find a ATOMIC STRUCTURE
unified way of viewing the gravitational and electric To set the stage for our discussion of the electric
interactions. He tried to find a single more fundamen- interaction, we will first construct a brief overview
tal theory from which both electricity and gravity of atomic and nuclear structure. The components of
emerged. He did not succeed, and no one else has yet atoms and nuclei that are of interest are the proton
done so either. However from the knowledge gained by and neutron found in the nucleus, and the electron
looking inside the proton and neutron, the knowledge which orbits outside. Protons and neutrons are each
that led to the discovery of quarks, Steven Weinberg about 1836 times as massive as the electron, thus
and Sheldon Glashow were able to construct a theory most of the mass of an atom is located in the nucleus.
that unified the electric and weak interactions. These
two forces which appeared so different in their effects The simplest of all atoms is hydrogen with one proton
on matter, turn out to be two components of a more for a nucleus and one electron outside. The electron is
fundamental interaction. Thus we are now down to attracted to the proton by a 1/r 2 electric force, just as
three basic interactions. Why three? Can these be the earth is attracted to the sun by the 1/r 2 gravitational
unified? We do not know yet. force. And because the proton is much more massive
than the electron, the proton sits nearly at rest at the
We end this discussion with another look at the center of the atom while the electron orbits outside,
gravitational interaction. On an atomic scale, grav- much as the earth orbits the sun.
ity is so weak compared to electricity that only
recently has it been experimentally determined that Also like the solar system, the atom is mostly empty
electrons fall down rather than up in the earth’s space. A proton has a diameter of about 10 –13 cm,
gravitational field. The only reason we human while a hydrogen atom is one hundred thousand
beings personally know about gravity is the fact that times bigger. If the hydrogen atom were enlarged to
we are standing on a huge chunk of matter, the earth. the point where the proton nucleus were the size of
It takes a lot of matter to create a big enough the sun, the electron would be orbiting out at a
gravitational force for us to notice. distance over 10 times the radius of the pluto’s orbit.
In this sense there is more empty space in an atom
But there is a lot of matter in the universe. Some- than in the solar system.
times, as in the case of a neutron star, so much
matter is packed in such a small space that the In Newton’s law of gravity, the gravitational force
gravitational force becomes stronger than the elec- on an object is proportional to an object’s mass. The
tric force. In a neutron star, gravity has forced the fact that all gravitational forces are attractive can be
electrons back into the nucleus, to form pure nuclear viewed as a consequence of the fact that there is only
matter. If the neutron star gets too big, if gravity gets positive mass. In the electric interaction, there are
a bit stronger, it can overwhelm the nuclear force both attractive and repulsive electric forces. We will
and crush the star to form a black hole. Gravity, a see that for the electric interaction, the concept of
force so weak that we could barely detect it using the electric charge plays a role similar to that of mass for
Cavandish experiment, can become the strongest of the gravitational interaction. The existence of both
all forces. attractive and repulsive electric forces leads to hav-
ing both positive and negative charge.
19-4 The Electric Interaction

One new feature of having both attractive and repul- The chemistry of an atom depends upon the behav-
sive electric forces is that the net electric force ior of the electrons in an atom, and the electron
between two objects can be zero, due to the cancel- behavior depends significantly on the number of
lation of attractive and repulsive components. This electrons. Since an electrically neutral atom has the
cancellation of electric force can be represented by same number of electrons as protons, different ele-
a cancellation of electric charge, giving us an object ments with different numbers of protons have differ-
which we say is electrically neutral. Because there ent numbers of electrons and thus different chemical
are no repulsive gravitational forces and no negative properties. For example, hydrogen with one elec-
mass, there is no such thing as a gravitationally tron is an excellent fuel, helium with 2 electrons is
neutral object. chemically inert, and lithium with 3 electrons is a
highly reactive alkali metal.
When an atom has the same number of electrons in
orbit as the number of protons in its nucleus, the The periodic table is not merely a list of atoms
atom is electrically neutral. If you have two electri- according to the number of protons in the nucleus.
cally neutral atoms separated by a reasonable dis- The table exhibits many striking patterns or regu-
tance, like the atoms in a gas, the electric forces larities in the chemical behavior of the elements. For
between the atoms cancel. Thus neutral atoms in a example, helium with 2 electrons, neon with 10,
gas move by each other with almost no interaction. argon with 18, krypton with 36, xenon with 54 and
There is an interaction only when the atoms get too radon with 86 electrons are all chemically inert
close in a collision and the electric forces no longer gases. These so-called noble gases enter into few if
cancel. any chemical reactions. Now add one electron to
each of these atoms (and one proton to the nucleus),
Atoms are classified by the number of protons in the and you get lithium (3 electrons), sodium (11 elec-
nucleus. If the nucleus has one proton, the atom trons), potassium (19 electrons), etc., all reactive
belongs to the element hydrogen. If there are 2 alkali metals.
protons in the nucleus, it is a helium atom. Three
protons gives us lithium, on up through the periodic The patterns in the chemical properties of the ele-
table. The largest naturally occurring atom, on the ments exhibited by the periodic table are a conse-
earth at least, is uranium with 92 protons. Atoms quence of the electric interaction, but they cannot be
with over 100 protons in the nucleus have been explained using Newtonian mechanics. Scientists
created artificially. had to wait until the discovery of quantum mechan-
ics before a detailed explanation of the periodic table
The periodic table, and the classification of the unfolded. After we have discussed some of the basic
elements, were developed by chemists studying the ideas of quantum mechanics in later chapters, we
chemical properties of matter. However chemical will see that there are fairly simple explanations for
reactions, with the possible exception of cold fusion, the main features of the periodic table, like the
have virtually no effect on the atomic nucleus. If you difference between noble gasses and alkali metals
have a lead nucleus with its 82 protons, there is no set mentioned above. For now, however, where we
of chemical reactions that can change it to a gold have only developed a background of Newtonian
nucleus with 79 protons. This is what doomed the mechanics, we will go no further than treating the
alchemists of the middle ages to failure. periodic table as a list of the elements according to
the number of protons in the nucleus.
 19-5

Element Chemical No. of No. of Element Chemical No. of No. of

symbol protons neutrons symbol protons neutrons
Hydrogen H 1 0 Iodine I 53 74
Helium He 2 2 Xenon Xe 54 78
Lithium Li 3 4 Cesium Cs 55 78
Beryllium Be 4 5 Barium Ba 56 82
Boron B 5 6 Lanthanum La 57 82
Carbon C 6 6 Cerium Ce 58 82
Nitrogen N 7 7 Praseodymium Pr 59 82
Oxygen O 8 8 Neodymium Nd 60 82
Fluorine F 9 10 Promethium Pm 61 86
Neon Ne 10 10 Samarium Sm 62 90
Sodium Na 11 12 Europium Eu 63 90
Magnesium Mg 12 12 Gadolinium Gd 64 94
Aluminum Al 13 14 Terbium Tb 65 94
Silicon Si 14 14 Dysprosium Dy 66 98
Phosphorus P 15 16 Holmium Ho 67 98
Sulfur S 16 16 Erbium Er 68 98
Chlorine Cl 17 18 Thulium Tm 69 100
Argon A 18 22 Ytterbium Yb 70 104
Potassium K 19 20 Lutetium Lu 71 104
Calcium Ca 20 20 Hafnium Hf 72 108
Scandium Sc 21 24 Tantalum Ta 73 108
Titanium Ti 22 26 Tungsten W 74 110
Vanadium V 23 28 Rhenium Re 75 112
Chromium Cr 24 28 Osmium Os 76 116
Manganese Mn 25 30 Iridium Ir 77 116
Iron Fe 26 30 Platinum Pt 78 117
Cobalt Co 27 32 Gold Au 79 122
Nickel Ni 28 30 Mercury Hg 80 122
Copper Cu 29 34 Thallium Tl 81 124
Zinc Zn 30 34 Lead Pb 82 126
Gallium Ga 31 38 Bismuth Bi 83 126
Germanium Ge 32 42 Polonium Po 84 124 (3 yr)
Arsenic As 33 42 Astatine At 85 125 (8 hr)
Selenium Se 34 46 Radon Rn 86 136 (3 days)
Bromide Br 35 44 Francium Fr 87 136 (21 min)
Krypton Kr 36 48 Radium Ra 88 138 (1622 yr)
Rubidium Rb 37 48 Actinium Ac 89 138 (22 hr)
Strontium Sr 38 50 Thorium Th 90 140 (80,000 yr)
Yttrium Y 39 50 Protactinium Pa 91 140 (34,000 yr)
Zirconium Zr 40 50 Uranium U 92 146 (4.5 billion yr)
Niobium Nb 41 52 Neptunium Np 93 144 (2.2 million yr)
Molybdenum Mo 42 56 Plutonium Pu 94 145 (24,000 yr)
Technetium Tc 43 54 (> 100 yr) Americium Am 95 144 (490 yr)
Ruthenium Ru 44 58 Curium Cm 96 146 (150 day)
Rhodium Rh 45 58 Berkelium Bk 97 150 (1000 yr)
Palladium Pd 46 60 Californium Cf 98 153 (800 yr)
Silver Ag 47 60 Einsteinium Es 99 155 (480 days)
Cadmium Cd 48 66 Fermium Fm 100 153 (23 hr)
Indium In 49 66 Mendelevium Md 101 155 (1.5 hr)
Tin Sn 50 70 Nobelium No 102 152 (3 sec)
Antimony Sb 51 70 Lawrencium Lw 103 154 (8 sec)
Tellurium Te 52 78

Table 1
The most commonly found (most abundant in nature) isotope of each element is listed. In
cases where an element has no stable isotopes, the isotope with the longest lifetime is listed.
19-6 The Electric Interaction

Isotopes nuclei, which have 2 protons usually also have 2

The atomic nucleus contains not only the protons we neutrons, but can be found with only one neutron.
have been discussing, but also neutrons. The nuclear Generally the light elements have roughly equal
force, unlike the electric force, is the same between numbers of protons and neutrons while the heavy
protons and neutrons, and ignores electrons. The elements like uranium have a considerable excess of
nuclear force between nucleons (protons or neu- neutrons.
trons), is attractive if the nucleons are close but not
too close, and repulsive if you try to shove nucleons Atoms of the same element with different numbers
into each other. As a result, due to the nuclear force, of neutrons in the nucleus are called different iso-
protons and neutrons in a nucleus stick to each other topes of the element. We distinguish different
forming a ball of nuclear matter as indicated in isotopes of an element by appending to the name of
Figure (1). The nuclear force is strong enough to the element a number equal to the total number of
hold the nucleus together despite the electrical re- protons and neutrons in the nucleus. The hydrogen
pulsion between the protons. In the nucleus of a atom with just a single proton for a nucleus is
given element, there is no precisely fixed number of hydrogen-1. If there is one neutron in addition to the
neutrons. The hydrogen nucleus which has one proton, the atom is called hydrogen-2, and with 2
proton, can have zero, one, or two neutrons. Helium neutrons and one proton, we have hydrogen-3. These
isotopes and the isotopes of helium are indicated
schematically in Figure (2).
Figure 1a
Picture the nucleus
The naturally occurring isotope of uranium, an un-
as a spherical ball stable element, but one with such a long half life (4.5
of protons and billion years) that it has survived since the formation
neutrons. of the earth, is uranium-238 or U-238. Since ura-
nium has 92 protons, the number of neutrons in U–
238 must be 238 - 92 = 144. Another uranium
isotope, U–235 with 3 fewer neutrons, is a more
highly radioactive material from which atomic bombs
can be constructed. (U–238 is so long–lived, so
stable that it is quite safe to handle. There was some
discussion of using U–238 for the keels of America’s

p
Hydrogen-1

p p
n n p
Hydrogen-2 Helium-3
(Deuterium)

p p n
n n n p
Hydrogen-3 Helium-4
Figure 1b
(Tritium)
Model of the uranium nucleus constructed from
Figure 2
styrofoam balls. The dark balls represent protons.
Isotopes of hydrogen and helium.
 19-7

cup yachts because of its very high density, but this THE ELECTRIC FORCE LAW
was disallowed as being too high tech.) A list of the Since electrons, protons, and neutrons make up
most common or longest lived isotopes for each almost everything we see around us (except for
element is given in Table 1. photons or light itself), a description of the electric
For historical reasons, the isotopes of hydrogen are force between these three particles provides a fairly
given special names. Hydrogen-2, with one proton complete picture of the electric interaction, insofar
and one neutron, is known as deuterium. Just over as it affects our lives. For electrons, protons, and
one in ten thousand hydrogen atoms in naturally neutrons at rest, this interaction is completely sum-
occurring hydrogen are the deuterium isotope. Water marized in Figure (3).
molecules (H20), in which one of the hydrogen As we see, protons repel each other, electrons repel
atoms is the deuterium isotope, are called heavy each other, and a proton and an electron attract each
water. Heavy water played an important role in the other. There is no electric force on a neutron. The
unsuccessful German effort to build a nuclear bomb strength of the electric force between these particles
during World War II. Hydrogen–3, with one proton drops off as 1/r 2 and has a magnitude shown in
and 2 neutrons, is called tritium. Tritium is unstable Equation 1. We know, to extremely high precision,
with a half life of 12.5 years. Along with deuterium, that the attractive force between an electron and
tritium plays an important role in mankind’s attempt proton has the same strength as the repulsive force
to build a nuclear fusion reactor. between two protons or two electrons, when the
Why are some isotopes stable while others are not? particles have the same separation r.
Why are there roughly equal numbers of neutrons as It is surprising how complete a summary of the
protons in the small stable isotopes and an excess of electric interaction Figure (3) represents. We have
neutrons in the large ones? Why are some elements only shown the forces between the particles at rest.
more abundant than others—for example, why does But if you combine these results with the special
the earth have an iron core? These are questions theory of relativity, you can deduce the existence of
whose answers depend upon an interplay inside the magnetism and derive the formulas for magnetic
nucleus between the nuclear, the electric, and the forces. We will do this in Chapter 28.
weak interaction. We reserve a discussion of these
questions for the next chapter where we discuss the
nuclear and weak interactions in more detail. r
Fe Fe
proton–proton
p p force

Fe Fe
electron–electron
e e force

Fe Fe
proton–electron
p e force

–19
Fe (dynes) = 2.3×10 (CGS units) (1)
r 2(cm2)

Figure 3
The electric interaction between protons and electrons
at rest. There is no electric force on a neutron.
19-8 The Electric Interaction

Strength of the Electric Interaction ELECTRIC CHARGE

If two electrons or two protons are separated by a From a historical perspective, the electric interac-
distance of one centimeter, then according to Equa- tion was carefully studied and the electric force law
tion 1 there is a repulsive force between them— well known long before the discovery of electrons
whose strength is 2.3×10 –19 dynes. Since a dyne is and nuclei, even before there was much evidence for
the weight of one milligram of mass, 2.3×10 –19 the existence of atoms. The simple summary of the
dynes is a very small force. But, of course, protons electric force law given by Equation 1 could only be
and electrons are very small particles. written after the 1930s, when we finally began to
understand what was going on inside an atom. Prior
To get a better idea of how strong the electric force is,
to that, the electric force law was expressed in terms
let us compare it with the gravitational force. If we have
of electric charge, a concept invented by Benjamin
2 protons any distance r apart, then the ratio of the
Franklin. What we want to do in this section is to
electric repulsion to the gravitational attraction is
show how the concept of electric charge evolves
from the forces pictured in Figure (3), and why
electric repulsion Fe
= electric charge is such a useful concept.
gravitational attraction
Fg
To convert Equation 1 into the more standard form of
(2) the electric force law, we will begin by writing
– 19 2
= 2.3 * 10 /r the numerical constant 2.3×10 –19 dyne cm2 , or
Gm pm p /r 2 2.3×10 –20 newton meter2 , in the form Ke2 to give
Since both electricity and gravity are 1/r 2 forces, the 2 electric force
r 2 cancel out in Equation 2 and we are left with Fe electron = Ke2 between two (4)
r electrons

where e is called the charge on an electron and k is

= 2.3 × 102
Fe – 19
a numerical constant whose size depends on the
Gm p system of units we are using.
Fg

The form of Equation 4 is chosen to make the electric

= 2.3 × 10 – 19
2 force law look like the gravitational force law. To see
6.67 × 10 – 8 × 1.67 × 10 – 24 this explicitly, compare the formulas for the magnitude
of the electric and the gravitational forces between two
= 1240000000000000000000000000000000000 electrons
(3) Gm em e
Fgravitational =
r2
The electrical force is some 10 36 times stronger than
gravity. This is true no matter how far apart the protons
are. The only reason that electric forces do not com- Felectric = K e2 e (5)
r
pletely swamp gravitational forces is that there are both
attractive and repulsive electric forces which tend to
cancel on a large scale, when many electrons and
protons are involved.
 19-9

In words, we said that the gravitational force be- however, to be far more convenient when working
tween two electrons was proportional to the product with practical or engineering applications of electri-
of the masses m e , and inversely proportional to the cal theory. As a result we will use the MKS system
square of the separation 1/r 2. Now we say that the throughout the chapters on electric fields and their
electric force is proportional to the product of the application, and restrict our use of the CGS system
charges (e), and inversely proportional to the square to discussions of atomic phenomena.
of the distance 1/r 2. By introducing the constant (e)
as the charge on the electron, we have electric You will notice that the dimensions of e, displayed in
charge playing nearly the same role for the electric Equation 4 are fairly messy. To avoid writing
force law as mass does for the gravitational force dynes cm 2 all the time, this set of units is given the
law. name esu which stands for electric charge as measured
in the electrostatic system of units. Thus we can
To get the numerical value for the charge (e) on an rewrite Equation 7 as
electron, we note that in the CGS system of units it is
traditional to set the proportionality constant (K) equal e = 4.8 × 10 – 10 esu (7a)
to one, giving
as the formula for the amount of charge on an
2 electric force electron. (In the MKS system, electric charge is
Fe CGS = e2 law in CGS measured in coulombs rather than esu. The differ-
r units (K = 1) ence between a coulomb and an esu arises not only
(6)
from the different set of units (newtons vs dynes) but
Comparing Equations 1 and 6 we get also from the different choice of K in the MKS
system. Any further discussion of the MKS system
e 2 = 2.3×10 – 19 dynes cm 2 will be reserved for later chapters.
charge on
e = 4.8×10 – 10 dynes cm 2 electron in Exercise 1
CGS units What would be the value of the electric force constant
K in a system of units where distance was measured
(7) in centimeters and the charge e on the electron was
In the MKS system, the proportionality constant K set equal to 1?
is not 1, therefore Equation 6 does not apply to that
system. Calculations involving electric forces on an
atomic scale are simpler in the CGS system because
of the choice K = 1. The MKS system turns out,
19-10 The Electric Interaction

Positive and Negative Charge Addition of Charge

It was Ben Franklin who introduced the concept of The concept of charge is particularly useful when we
two kinds of electric charge. Franklin noticed that have to deal with complex structures involving
you get opposite electrical effects when you rub a many particles. To see why, let us start with the
glass rod with silk, or rub a rubber rod with cat fur. simplest electrical structure, the hydrogen atom, and
He decided to call the charge left on the glass rod gradually add more particles. We will quickly see
positive charge, and the charge left on the rubber rod that the electric force law, in the form Fe = e 2 /r 2
negative charge. What we will see in this section is becomes difficult to use.
how this choice of positive and negative charge
leads to the electron having a negative charge – e, In Figure (4) we have a model of the hydrogen atom
and a proton a positive charge + e. consisting of a proton at the center and an electron
orbiting about it. The proton sits nearly at rest at the
charge on an electron = – e (8a) center because it is 1836 times as massive as the
electron, much as our sun sits at the center of our
charge on a proton = +e (8b)
solar system because it is so much more massive
Despite the fact that the electron’s charge turns out than the planets.
to be negative, e is still called the charge on an
electron. The proton and electron attract each other with a
force of magnitude (in CGS units) of Fe = e 2 /r 2 .
The basis for saying we have two kinds of charge is Since this is similar in form to the gravitational force
the fact that with the electric interaction we have between the earth and the moon, we expect the
both attractive and repulsive forces. With the choice electron to travel in elliptical orbits around the
that electrons are negative and protons are positive, proton obeying Kepler’s laws. This would be ex-
then the rules shown in Figure 3 can be summarized actly true if Newton’s laws worked on the small
as follows: like charges (2 protons or 2 electrons) scale of the hydrogen atom as they do on the larger
repel, opposite charges (proton and electron) at- scale of the earth-moon system.
tract. We can explain the lack of any force on the
neutron by saying that the neutron has no charge—
that it is neutral. e
Fe

Choosing one charge as positive and one as negative

automatically gives us a reversal in the direction of
the force when we switch from like to opposite Fe
p
charges
+e +e e2
Fe = =
proton proton r2 r2
Figure 4
–e –e 2 The hydrogen atom consists of a proton at the center with
Fe electronelectron
= 2
= e2 an electron moving about it. The particles are held
r r together by the attractive electric force between them.
– e +e 2
Fe = = – e2
electronproton r2 r
(9)
 19-11

Exercise 2 (Do this now) If the separation r of the atoms is large compared to
the diameter of each hydrogen atom, then all these
Hydrogen atoms are approximately 10–8 cm in diam-
eight forces have essentially the same magnitude
eter. Assume that the electron in the hydrogen atom
in Figure (4) is traveling in a circular orbit about the
e 2/r2. Since half are attractive and half are repulsive,
proton. Use Newton’s law F = ma and your knowl- they cancel and we are left with no net force between
edge about the acceleration a of a particle moving in the atoms.
a circular orbit to predict the speed, in cm/sec, of the
electron in its orbit. How does the electron’s speed Exercise 3
compare with the speed of light? (It had better be A complete carbon atom has a nucleus with 6 pro-
less.) tons, surrounded by 6 orbiting electrons. If you have
two complete separate carbon atoms, how many
An analysis of the hydrogen atom is easy and straight forces are there between the particles in the two
forward using the force law Fe = e 2 /r 2 . But the different atoms?
analysis gets more difficult as the complexity of the
problem increases. Suppose, for example, we have With just two simple hydrogen atoms we have to
two hydrogen atoms separated by a distance r. Let deal with 8 forces in order to calculate the total force
r be quite a bit larger than the diameter of a hydrogen between the atoms. If we have to deal with some-
atom, as shown in Figure (5a). thing as complex as calculating the force between
two carbon atoms, we have, as you found by doing
Even though r is much larger than the size of the
Exercise 3, to deal with 72 forces. Yet the answer is
individual hydrogen atoms, there are still electric
still zero net force. There must be an easier way to
forces between the protons and electrons in the two
get this simple result.
atoms. The two protons repel each other with a force
Fpp, the electrons repel each other with a force Fee, The easier way is to use the concept of net charge Q
the proton in the left atom attracts the electron in the which is the sum of the charges in the object. A
right atom with a force Fep. Sa you can see, eight hydrogen atom has a net charge
separate forces are involved, as shown in Figure
(5b). Q hydrogen = + e proton
+ –e electron
r
= 0 (10)
The net force between two objects with net charges
H atom H atom Q 1 and Q 2 is simply
(a) Two hydrogen atoms separated by a distance r
KQ 1Q 2 Coulomb's law, where
p p
Fnet = 2 K = 1 for CGS units
(11)
Fpp Fpe Fpe Fpp r
Fee Fpe Fpe Fee
e e Equation 1, which looks very much like Newton’s
(b) Forces between the particles in the two atoms law of gravity, except that charge replaces mass, is
Fpp = Fee = Fpe = e 2 /r 2 known as Coulomb’s law. The proportionality
constant K is 1 for CGS units.
Figure 5
When we have two hydrogen atoms fairly far apart,
then there is essentially no net force between the atoms.
The reason is that the repulsive forces between like
particles in the two atoms are cancelled by the
attractive force between oppositely charged particles.
19-12 The Electric Interaction

Applying Coulomb’s law to the force between two Exercise 4

complete carbon atoms, we see immediately that the This exercise is designed to give you a more intuitive
complete atoms have zero net charge, and therefore feeling for the enormous magnitude of the electric
by Coulomb’s law there is no net force between force, and how complete the cancellation between
them. The cancellation of individual forces seen in attractive and repulsive forces is in ordinary matter.
Figure (4) is accounted for by the cancellation of
charge in Coulomb’s law. Imagine that you could strip all the electrons from two
garden peas, leaving behind two small balls of pure
To see that the addition of charge and Coulomb’s positive charge. Assume that there is about one mole
law work in situations where charge does not cancel, (6 x 1023) protons in each ball.
suppose we had two helium nuclei separated by a (a) What is the total charge Q on each of these two
distance r. (A bare helium nucleus, which is a balls of positive charge? Give the answer in esu.
helium atom missing both its electrons, would be
called a doubly ionized helium atom.) In Figure (6) (b) The two positively charged peas are placed one
we have sketched the forces between the two pro- meter (100 cm) apart as shown. Use Coulomb’s law
tons in each nucleus. Both protons in nucleus Fe Fe
number 1 are repelled by both protons in nucleus
number 2, giving rise to a net repulsive force four
times as strong as the force between individual r = 1 meter
protons, or a force of magnitude 4 e 2/r2.
to calculate the magnitude of the repulsive force
Applying Coulomb’s law to these two nuclei, we see between them. Give your answer in dynes and in
metric tons. (1 metric ton = 109 dynes ≈ 1 english ton.)
that the charge on each nucleus is 2e, giving for the
charges Q 1 and Q 2
If you worked Exercise 4 correctly, you found that
Q 1 = 2e = total charge on nucleus #1 two garden peas, stripped of all electrons and placed
one meter apart, would repel each other with a force
Q 2 = 2e = total charge on nucleus #2 of nearly 10 16 tons! Yet when you actually place
Thus Coulomb’s law gives (with K = 1 for CGS two garden peas a meter apart, or only a centimeter
units) apart, there is no observable force between them.
The 10 16 ton repulsive forces are so precisely can-
Q 1Q 2 2e 2e 2 celled by 10 16 ton attractive forces between elec-
Fe = 2 = 2 = 4e2 (12)
r r r trons and protons that not even a dyne force remains.
which is in agreement with Figure (6).
Exercise 5
1 2 What would be the repulsive force between the peas
p p
if only one in a billion (one in 109) electrons were
p p removed from each pea, and the peas were placed
r one meter apart?
e2
Ftot = 4 ( )
Figure 6 r2
We see that the repulsive force between 2 helium
nuclei is 4 times as great as the repulsion e2 / r2
between 2 protons. Using Coulomb’s law
F = Q1 Q2 / r2 with Q1 = Q2 = 2 e for the helium
nuclei gives the same result.
 19-13

CONSERVATION OF CHARGE with a particle or an antiparticle. We denote antipar-

Up to this point, we have used the concept of electric ticles by putting a bar over the symbol, thus ν represents
charge to simplify the calculation of the electric force a neutrino, and ν an antineutrino. Since a particle and
between two objects containing many electrons and an antiparticle can annihilate each other, a particle and
protons. But the fact that electrons and protons have an antiparticle must have opposite electric charges if
precisely opposite charges suggests that in nature elec- they carry charge at all, so that charge will not be lost
tric charge has a deeper significance. That deeper in the annihilation. As a result the antiparticle of the
significance is the conservation of electric charge. Like electron e – is the positively charged positron which we
the conservation of energy, linear momentum, and designate e +.
angular momentum, the conservation of electric charge Using these conventions, we have the following nota-
appears to be a basic law with no known exceptions. tion for the particles under discussion (photons and
When we look beyond the familiar electrons and neutrinos are uncharged):
protons, into the world of subnuclear particles, we find Notation Particle Particle
a bewildering array of hundreds of different kinds of p+ proton ν0 antineutrino
particles. In the chaos of such an array of particles, two n0 neutron µ– muon
features stand out. Almost all of the particles are e– electron π+ pi plus
unstable, and when the unstable particles decay, elec- e+ positron π0 pi naught
tric charge is conserved. In looking at the particle γ0 photon π– pi minus
decays, it becomes clear that there really is something ν0 neutrino
we call electric charge that is passed from one particle
to another, and not lost when a particle decays. The particle decays we just described can now be
written as the following reactions.
We will illustrate this with a few examples. We have
already discussed several unstable particles, the muon µ– → e– + ν0 muon decay (a)
introduced in the muon lifetime experiment, the π n → p+ + e– + ν0 neutron decay (b)
mesons, created for cancer research, and the neutron π– → e+ + ν0 pi minus decay (c)
which, by itself outside a nucleus, has a half life of nine
minutes. The muon decays into an electron and a π0 → γ 0 + γ 0 pi naught decay (d)
neutrino, and the neutron decays into a proton, electron π– → e– + ν0 pi minus decay (e) (14)
and an antineutrino (the antiparticle of the neutrino).
There are three separate π mesons. The negative Note that in all of these decays, the particles change
charged one decays into an electron and an antineutrino, but the charge does not. If we start with a negative
the positive one into a positron (antielectron) and charge, like the negative muon, we end up with a
neutrino, and the neutral one into two photons. negative particle, the electron. If we start with a
neutral particle like the π 0, we end up with no net
We can shorten our description of these decays by charge, in this case two photons.
introducing shorthand notation for the particles and
their properties. We will use the Greek letter µ (mu) Among the hundreds of elementary particle decays
for the muon, π for the π mesons, ν (nu) for the neutrino that have been studied, no one has found an example
and γ (gamma) for protons. We designate the charge where the total charge changed during the process.
of the particle by the superscript + for a positive charge, It is rather impressive that the concept of positive
– for a negative, and 0 for uncharged. Thus the three π and negative charge, introduced by Ben Franklin to
mesons are designated π +, π 0, and π – for the positive, explain experiments involving rubber rods and cat
neutral and negative ones respectively. In later discus- fur, would gain even deeper significance at the
sions, it will be useful to know whether we are dealing subnuclear level.
19-14 The Electric Interaction

Stability of Matter experiment. A mole of protons (hydrogen) weighs one

The conservation of electric charge may be related to gram, a billion moles is a million kilograms or a
the stability of matter. The decay of elementary thousand metric tons. You get that much mass in a cube
particles is not an exceptional occurrence, it is the of water 10 meters on a side, or in a large swimming
general rule. Of the hundreds of particles that have pool. For this reason, experiments designed to detect
been observed, only four are stable, the proton, the the decay of the proton had to be able to distinguish a
electron, the photon and the neutrino. (Neutrons are few proton decays per year in a swimming pool sized
also stable if buried inside a nucleus, for reasons we will container of water.
discuss in the next chapter.) All the other particles
eventually, and often very quickly, decay into these So far none of these detectors has yet succeeded in
four. detecting a proton decay (but they did detect the
neutrinos from the 1987 supernova explosion). We
The question we should ask is not why particles decay, now know that the proton half-life is in excess of 10 32
but instead why these four particles do not. We know years, and as a result the Grand Unified Theories are in
the answer in the case of two of them. Photons, and trouble. We still do not know whether the proton is
perhaps, neutrinos, have zero rest mass. As a result they stable, or just very long lived.
travel at the speed of light, and time does not pass for
them. If a photon had a half life, that half life would Quantization of Electric Charge
become infinite due to time dilation.
Every elementary particle that has been detected indi-
Why is the electron stable? It appears that the stability vidually by particle detectors has an electric charge that
of the electron is due to the conservation of energy and is an integer multiple of the charge on the electron.
electric charge. The electron is the least massive Almost all of the particles have a charge
charged particle. There is nothing for it to decay into + e or – e, but since the 1960s, a few particles with
and still conserve charge and energy. charge 2e have been observed. Until the early 1960s
it was firmly believed that this quantization of charge
That leaves the proton. Why is it stable? We do not in units of + e or – e was a basic property of electric
know for sure. There are a couple of possibilities which charge.
are currently under study. One is that perhaps the
proton has some property beyond electric charge that is In 1961, Murray Gell-Mann, who for many years had
conserved, and that the proton is the least massive been trying to understand the bewildering array of
particle with this property. This was the firm belief elementary particles, discovered a symmetry in the
back in the 1950s. masses of many of the particles. This symmetry, based
on the rather abstract mathematical group called SU2,
In the 1960s, with the discovery of quarks and the
combining of the electric and weak interaction theo- predicted that particles with certain properties, could
ries, it was no longer obvious that the proton was stable. be grouped into categories of 8 or 10 particles. This
Several theories were proposed, theories that attempted grouping was not unlike Mendeleev’s earlier grouping
to unify the electric, weak, and nuclear force. These so– of the elements in the periodic table.
called Grand Unified Theories or GUT for short,
When the periodic table was first constructed, there
predicted that protons should eventually decay, with a
were gaps that indicated missing, as yet undiscovered
half life of about 10 31 years. Since the universe is only
10 10 years old, that is an incredibly long time.
elements. In Gell-Mann’s SU3 symmetry there were
also gaps, indicating missing or as yet undetected
It is not impossible to measure a half life of 10 31 years. elementary particles. In one particular case, Gell-
You do not have to wait that long. Instead you look at Mann accurately predicted the existence and the prop-
10 31 or 10 32 particles, and see if a few decay in one erties of a particle that was later discovered and named
year. Since a mole of particles is 6×10 23 particles, you the ω – (omega minus). The discovery of the ω – verified
need about a billion moles of protons for such an the importance of Gell-Mann’s symmetry scheme.
 19-15

In 1964 Gell-Mann, and independently George Zweig MOLECULAR FORCES

also from Caltech, found an exceedingly simple A naïve application of Coulomb’s law would say
model that would explain the SU3 symmetry. They that complete atoms do not interact. A complete
found that if there existed three different kinds of atom has as many electrons outside as protons in the
particles which Gell-Mann called quarks, then you nucleus, and thus zero total charge. Thus by
could make up all the known heavy elementary Coulomb’s law, which says that the electric force
particles out of these three quarks, and the particles between two objects is proportional to the product of
you make up would have just the right SU3 symme- the charges on them, one would predict that there is
try properties. It was an enormous simplification to no electric force between two complete atoms. Tell
explain hundreds of “elementary particles” in terms that to the two hydrogen atoms that bind to form
of 3 kinds of quarks. hydrogen molecules, oxygen atoms that bind to
Our discussion of quarks will be mainly reserved for form to O 2 molecules we breathe, or the hydrogen
the next chapter. But there is one property of quarks and oxygen atoms that combine to form the water
that fits into our current discussion of electric charge. molecules we drink. These are all complete atoms
The charge on a quark or anti quark can be + 1/3 e, that have combined together through electric forces
– 1/3 e, + 2/3 e or – 2/3 e. The charge e on the electron to form molecular structures.
turns out not to be the fundamental unit of charge. The reason that neutral atoms attract electrically to
An even stranger property of quarks is the fact that form molecules is the fact that the negative charge in
an atom is contained in the electrons which are
they exist only inside elementary particles. For
moving about the nucleus, and their motion can be
example, a proton or neutron is made up of three
affected by the presence of other atoms.
quarks and a π meson of two. All particles made
from quarks have just the right number of quarks, in When trying to understand molecular forces, the
just the right combination, so that the total charge of planetary picture of an atom, with electrons in orbits
the particle is an integer multiple of the electron like the planets moving around the sun, is not a
charge e. Although the quarks themselves have a particularly useful or accurate model. A more useful
fractional charge ± 1/3 e or ± 2/3 e, they are always picture, which has its origin both in quantum me-
found in combinations that have an integer net chanics and Newtonian mechanics is to picture the
charge. electrons as forming a cloud of negative charge
surrounding the nucleus. You can imagine the
You might ask, why not just tear a proton apart and electrons as moving around so fast that, as far as
look at the individual quarks? Then you would see neighboring atoms are concerned, the electrons in an
a particle with a fractional charge. It now appears atom simply fill up a region around the nucleus with
that, due to an unusual property of the so-called negative electric charge. When doing accurate cal-
color force between quarks, you cannot simply pull culations with quantum mechanics, one finds that
quarks out of a proton. The reason is that the color the electron clouds have definite shapes, shapes
force, unlike gravity and electricity, becomes stron- which the chemists call orbitals.
ger, not weaker, as the separation of the particles
increases. We will see later how this bizarre feature One does not need quantum mechanics in order to
of the color force makes it impossible to extract an get a rough understanding of the origin of molecular
individual quark from a proton. forces. Simple arguments about the behavior of the
electron clouds gives a fairly good picture of what
the chemists call covalent bonding. We will illus-
trate this with a discussion of the hydrogen mol-
ecule.
19-16 The Electric Interaction

Hydrogen Molecule Since the two protons are identical, the electron will
To construct a hydrogen molecule, imagine that we have no reason to prefer one proton over the other,
start with a single proton and a complete hydrogen and will form a symmetric electron cloud about both
atom as indicated in Figure (7a). Here we are protons as shown in Figure (7c). The result is a
representing the hydrogen atom by a proton with the complete and stable object called a hydrogen mol-
electron moving around to more or less fill a spheri- ecule ion. electron
cal region around the proton. In this case the
external proton is attracted to the sphere of negative
charge by a force that is as strong as the repulsion protron
from the hydrogen nucleus. As a result there is very hydrogen atom
little force between the external proton and the
(a) A proton far from a complete hydrogen atom.
neutral atom. Here, Coulomb’s law works.

Now bring the external proton closer to the hydro-

gen atom, as shown in Figure (7b). Picture the Fe Fe
hydrogen nucleus as fixed, nailed down, and look at
protron
what happens to the hydrogen electron cloud. The
hydrogen atom
electron is now beginning to feel the attraction of the (b) The external proton is brought closer
distorting the electron cloud
external proton as well as its own proton. The result
is that the electron cloud is distorted, sucked over a
electron cloud formed
bit toward the external proton. Now the center of the by one electron
electron cloud is a bit closer to the external proton
than the hydrogen nucleus, and the attractive force
between the electron cloud and the proton is slightly
stronger than the repulsive force between the pro-
tons. The external proton now feels a net attractive (c) Electron orbits both protons Hydrogen molecule ion
force to the neutral hydrogen atom because of the
Figure 7a
distortion of the electron cloud. A naïve application Formation of a hydrogen molecule ion. To visualize
of Coulomb’s law ignores the distortion of the how a hydrogen molecule ion can be formed,
electron cloud and therefore fails to predict this imagine that you bring a proton up to a neutral
attractive force. hydrogen.
Figure 7b
Since the external proton in Figure (7b) is attracted When the proton gets close, it distorts the hydrogen
to the hydrogen atom, if we let go of the proton, it electron cloud. Since the distorted cloud is closer to
will be sucked into the hydrogen atom. Soon the the external proton, there is a net attractive force
electron will start orbiting about both protons, and between the proton and the distorted hydrogen atom.
the external proton will be sucked in until the repul- Figure 7c
sion between the protons just balances the electrical If the protons get too close, they repel each other. As
attraction. a result there must be some separation where there is
neither attraction or repulsion. This equilibrium
separation for the protons in a hydrogen molecule ion
is 1.07 Angstroms. (1 Angstrom = 10– 8 cm.)
 19-17

The final step in forming a hydrogen molecule is to 2 electrons moving

about both protons
note that the hydrogen molecule ion of Figure (7c)
has a net charge +e, and therefore will attract another
electron. If we drop in another electron, the two
electrons form a new symmetric cloud about both
protons and we end up with a stable H 2 molecule,
shown in Figure (8). Hydrogen molecule
Figure 8
Although the discussion related to Figures (7, 8) is The hydrogen molecule ion of Figure (7c) has a net
qualitative in nature, it is sufficient to give a good positive charge +e, and therefore can attract and hold
picture of the difference in character between atomic one more electron. In that case both electrons orbit both
protons and we have a complete hydrogen molecule.
and molecular forces. Atomic forces, the pure e 2/r2 The equilibrium separation expands to 1.48 Angstroms.
Coulomb force that binds electrons to the nucleus, is
very strong and fairly simple to understand. Mo-
lecular forces, which are also electrical in origin but
which depend on subtle distortion of the electron
clouds, are weaker and more complex. Molecular
forces are so subtle that you can make very complex
objects from them, for example, objects that can
read and understand this page. The sciences of
chemistry and biology are devoted primarily to
understanding this complexity.
19-18 The Electric Interaction

Molecular Forces—A More

Quantitative Look
electron
It is commonly believed that quantum mechanics, F1
which can be used to predict the detailed shape of
electron clouds, is needed for any quantitative un- F2
derstanding of molecular forces. This is only partly
true. We can get a fair understanding of molecular
forces from Newtonian mechanics, as was demon-
strated by the student Bob Piela in a project for an protons
introductory physics course. This section will closely a) Electric force acting on the electron.
follow the approach presented in Piela’s project.

In this section we will discuss only the simplest of all

molecules, the hydrogen molecule ion consisting of
two protons and one electron and depicted in Figure
(7c). We will use Newtonian mechanics to get a
better picture of how the electron holds the molecule
together, and to see why the lower, the more nega-
tive the energy of the electron, the more tightly the
protons are bound together.

If you do a straightforward Newtonian mechanics

calculation of the hydrogen molecule ion, letting all b) Line drawing plot of the orbit of
three particles move under the influence of the the electron about the two protons
Coulomb forces between them, the system eventu-
ally flies apart. As a number of student projects
using computer calculations have shown, eventu-
ally the electron gets captured by one of the protons
and the other proton gets kicked out of the system.
With Newtonian mechanics we cannot explain the
stability of the hydrogen molecule ion, quantum
mechanics is required for that.

Piela avoided the stability problem by assuming that

the two protons were fixed at their experimentally
known separation of 1.07 × 10 – 8 cm (1.07 ang-
stroms) as shown in Figure (9a), and let the com- c) Dots showing the position of the
puter calculate the orbit of the electron about the two electron at equal time intervals
fixed protons, as seen in Figure (9b). By letting the (effectively a strobe photograph).
calculation run for a long time and plotting the
position of the electron at equal intervals of time, as Figure 9
in a strobe photograph of the electron’s motion, one Orbit of an electron about two fixed protons.
obtains the dot pattern seen in Figure (9c). This dot
pattern can be thought of as the classical electron
cloud pattern for the electron.
 19-19

The Bonding Region equal time above and below the protons. Thus on the
Of the dots we have drawn in Figure (9c) some are average the y components cancel, and the net effect
more effective than others in holding the molecule of the electron is described by the x components F1x
together. When the electron is between the protons, and F2x alone. We can see from Figure (10a) that
it pulls in on both protons providing a net bonding F1x and F2x are pulling the protons together. This
force. But when the electron is outside to the left or electron is clearly in the bonding region.
right, it tends to pull the protons apart. We can call
the region where the electron gives rise to a net It is a little bit harder to see the anti-bonding forces.
attractive force the bonding region, while the rest of In Figure (10b), we show the electron first to the left
space, where the electron tends to pull the protons of the protons, then to the right. Again we concen-
apart, can be called the anti-bonding region. trate on the x components F1x and F2x because the y
components will, on the average, cancel. However,
To see how we can distinguish the bonding from the for orbits like that shown in Figure (9b), the electron
anti-bonding region, consider Figure (10) where we spends the same amount of time on the left as the
show the forces the electron exerts on the protons for right, and thus we should average F1x and F2x for
several positions of the electron. In (10a), the these two cases. When we do this, we see that the
electron is between but above the protons, giving average F1x points left, the average F2x points right,
rise to the forces F1 and F2 shown. We also show and these two average forces are pulling the protons
the x components which are F1x and F2x . These apart.
components are of more interest to us than F1 and
F2 because the electron, while in orbit, will spend an Average x component of force
average F2x average F1x
p1 p2

Electron on left Electron on right

electron
electron electron

F1 F2 F2
F1 F2 F1
anti- anti- anti-
bonding p p bonding F1x F2x bonding F1x F2x
F1x F2x p1 p2 region p1 p2
region region

bonding bonding bonding

region region region

a) When the electron is in the bonding region, b) When the electron is in the anti-bonding region, the average
the x component of its electric forces pulls x component of its electric forces pulls the protons apart.
the protons together. (The y components
average out since the electron spends as
much time above as below the protons.)

Figure 10
Bonding and anti bonding regions. When the electron is in the bonding region,
the electric force exerted by the electron on the protons pulls the protons together.
When in the anti-bonding region, the electric force pulls the protons apart.
19-20 The Electric Interaction

If you are calculating an orbit and want to test Electron Binding Energy
whether the electron is in the bonding or anti- One of the features of Figure (12) is that there are
bonding region, simply compare F1x and F2x . If the many dots out in the anti-bonding region. It looks as
electron is to the right of the protons and F1x is if there are more dots out there pulling the protons
bigger than F2x , the protons are being pulled apart. apart than inside holding them together. In this case
If you are to the left of the protons, and F2x is bigger is the electron actually helping to hold the molecule
than F1x , then again the protons are being pulled together?
apart. Otherwise the protons are being pulled to-
gether and the electron is in the bonding region. One way to tell whether or not a system will stay
together or fall apart is to look at the total energy of
In Figure (11), we replotted the electron dot pattern the system. If it costs energy to pull a system apart,
of Figure (9c), but before plotting each point, checked it will stay together. But if energy is released when
whether the electron was in the bonding or anti- a system comes apart, it will fall apart.
bonding region. If it were in the bonding region, we
plotted a dot, and if it were in the anti-bonding region In our earlier discussion in Chapter 8 of the motion
we drew a cross. After the program ran for a while, of a satellite around the earth, we saw that if the total
it became very clear where to draw the lines separat- energy of the satellite were negative, the satellite
ing the bonding from the anti-bonding regions. would be bound to the earth and could not escape.
On the other hand, if the total energy of the satellite
were positive, the satellite would eventually escape
no matter what direction it was heading (assuming it
did not crash). These predictions about total energy
applied if we did not include rest energy, and as-
sumed that the gravitational potential energy was
zero when the satellite and earth were infinitely far
apart. This gave us as the formula for gravitational
potential energy
planet
Etot and = 1 m v 2 – Gm s Me (8-29)
satellite 2 s r
a) Electron cloud for a –10eV electron. Points in
the bonding region are plotted as dots, outside
in the anti-bonding region as crosses.
where m s and Me are the masses of the satellite and
earth, and v the speed of the satellite.

In describing the motion of an electron in an atom or

molecule, we can use the convention that the
electron's electric potential energy is zero when it is
anti- bonding anti-
infinitely far away from the protons. With this
bonding region bonding convention, the formula for the electric potential
region region
energy between an electron and a proton a distance
r apart is – Ke2 / r , which is analogous to the
gravitational potential energy – Gm sMe / r . (Sim-
ply replace Gm sMe by K e2 to go from a discussion
of gravitational forces in satellite motion to electric
b) The area of dots in a) show us where the
bonding region is. forces in atoms.)
Figure 12
Determining the bonding region from the computer plot.
 19-21

Thus the formula for the total energy of an electron The value –13.6 eV means that, in order to pull the
in orbit about a single stationary proton should be (in electron out of the hydrogen atom, we would have to
analogy to Equation 8-29) supply 13.6 eV of energy. In other words, the
binding energy of the electron in a cold hydrogen
electron 1 m v 2 – Ke2
Etot and = (15) atom is 13.6 eV. The number is 13.6 eV is much
proton 2 e r
easier to discuss and remember than
2.16 × 10 – 18 joules .
where m e is the mass of the electron, v its speed, and
r the separation between the electron and proton. Another unit that is convenient for discussing atoms
is the angstrom (abbreviated A° ) which is 10 – 8cm
Just as in the case of satellite motion, the total energy
or 10 – 10 meters.
tells you whether the electron is bound or will
eventually escape. If the total energy is negative, the 1 Angstrom A ° = 10 – 8cm
electron cannot escape, while if the total energy is = 10 – 10m (17)
positive, it must escape.
A hydrogen atom has a diameter of 1 A ° and all
Another way of describing the electron's behavior is atoms are approximately the same size. Even the
to say that if the electron's total energy is negative, largest atom, Uranium, has a diameter of only a few
it is down in some kind of a well and needs outside angstroms. In the hydrogen molecule ion, the sepa-
help, outside energy, in order to escape. The more ° .
ration of the protons is 1.07 A
negative the electron's total energy, the deeper it is
in the well, and the more tightly bound it is. We can
call the amount by which the electron's total energy Electron Energy
is negative the binding energy of the electron. The in the Hydrogen Molecule Ion
binding energy is the amount of energy that must be We have seen that the strength of the binding of an
supplied to free the electron. electron in an atom is related to the total energy of
the electron. The more negative the energy of the
electron, the more tightly it is bound.
Electron Volt as a Unit of Energy
In discussing the motion of an electron in an atom, Let us now return to our discussion of the hydrogen
quantities like meters and kilograms and joules are molecule ion to see if the total energy of the electron
awkwardly large. There is, however, a unit of in its orbit about the two protons is in any way related
energy that is particularly convenient for discussing to the effectiveness which the electron binds the
many applications, including the motion of elec- proton together.
trons in atoms. This unit of energy, called the
electron volt (abbreviated eV), is the amount of When the electron is orbiting about two protons,
energy an electron would gain if it hopped from the there are two electric potential energy terms, one for
negative to the positive terminal of a 1 volt battery. each proton. Thus the formula for the electron's total
The numerical value is energy is
electron
1 m v 2 – Ke2 – Ke2
1 electron volt eV = 1.6 × 10 – 19 joules Etot in H2+ =
2 e r1 r2 (18)
molecule
= 1.6 × 10 – 12ergs (16)
which is the same as Equation 15 except for the
As an example, an electron in a cold (unexcited) additional potential energy term. In this equation, r1
hydrogen atom has a total energy of –13.6 eV. The is the distance from the electron to proton #1, and r2
fact that the electron's energy is negative means that to proton #2.
the electron is bound to the proton–cannot escape.
19-22 The Electric Interaction

When we wrote computer programs for satellite around in its orbit, we can name the orbit by E tot .
motion, we found that it was much easier to work in For example, the orbit shown back in Figures (9) and
a system of units where the earth mass, earth radius (11), had a total energy of –10 eV. We can say that
and hour were set to 1. In these units the gravita- this was a "–10 eV orbit".
tional constant G was simply 20, and we never had
to work with awkwardly large numbers. When Bob Piela did his project on the hydrogen
molecule ion, his main contribution was to show
Similarly, we can simplify electron orbit calcula- how the energy of the electron in orbit was related to
tions by choosing a set of units that are convenient the bonding force exerted by the electron on the
for these calculations. In what we will call atomic protons. Piela's results are easily seen in Figure (13).
units, we will set the mass m e of the electron, the In (13a), we show the –10 eV orbit superimposed
electric charge e, the angstrom, and the electron volt upon a sketch of the bonding region. In (13b), the
all to 1. When we do this, the electric force constant same orbit is shown as a strobe photograph. As we
K has the simple value of 14.40. These choices are mentioned earlier, there appear to be a lot more dots
summarized in Table 2. outside the bonding region than inside, and it does
not look like a –10 eV electron does a very good job
Using atomic units, the formula for the total energy of binding the protons in the H+2 molecule.
of the electron in the hydrogen molecule ion (Equa-
tion 18) reduces to In Figure (13c) we have plotted a –20 eV orbit. The
striking feature is that as the electron energy is
2 reduced, made more negative, the electron spends
E tot H+2 = v – 14.40 14.40
r1 – r2 (18a)
2 more time in the bonding region doing a better job of
holding the molecule together. In Figure (13d) the
since m e = e = 1 . The real advantage of this for-
electron energy is dropped to –30 eV and the major-
mula is that it directly gives the electron's total
ity of the electron cloud is now in the bonding
energy in electron volts, no conversion is required.
region. It is easy to see that at –30 eV the electron
Because energy is conserved, because the electron's
does a good job of binding the protons. With Piela's
total energy does not change as the electron goes
diagrams it is easy to see how the electron bonds
more strongly when its energy is lowered.
ATOMIC UNITS
Constant Symbol Atomic Units MKS Units
electron volt eV 1 1.6 × 10 – 19 joules
angstrom °
A 1 10 – 10m
electron mass me 1 9.1 × 10 – 31 kg
electric charge e 1 1.6 × 10 – 19coulombs
electric force constant K 14.40 9 × 10 9 m/ farad
Bohr radius rb °
.51 A
separation of protons in H+2 °
1.07 A
Table 2
 19-23

a) Orbit of a -10eV electron superimposed c) Electron cloud for a –20eV electron. More
on the bonding region of the dots are inside the bonding region,
with the result the protons are more
tightly bound.

b) Electron cloud for a –10eV electron. d) Electron cloud for a –30eV electron. You
can see that the lower the electron energy,
the stronger the bonding.
Figure 13
As we decrease the electron’s total energy, the electron spends more time in the bonding region, with the result that
the protons are more tightly bound. Thus we see that the lower the electron energy, the stronger the binding.
19-24 The Electric Interaction

!Calculate force
! HYDROGEN MOLECULE ION LET F1 = K*Qp*Qe/R^2 - Kr*Qe*Qp/R^3 !Force by proton 1
LET F2 = K*Qp*Qe/S^2 - Kr*Qe*Qp/S^3 !Force by proton 2
! --------- Plotting window
! (x axis = 1.5 times y axis) LET F1x = -(Rx/R)*F1 !points in -R direction
SET WINDOW -3,6,-3,3 LET F1y = -(Ry/R)*F1

LET F2x = -(Sx/S)*F2 !points in -S direction

! --------- Experimental constants in Atomic Units LET F2y = -(Sy/S)*F2
LET K = 14.40 !Electric force constant
LET Kr = K/3 !Ficticious repulsive force
LET Qe = 1 !Charge on electron (magnitude) LET Fx = F1x + F2x !Vector sum of forces
LET Qp = 1 !Charge on Proton LET Fy = F1y + F2y
LET Me = 1 !Electron mass
LET Mp = 1836.1 !Proton mass !Newton's Second law
LET Rbohr = .5292 !Bohr radius LET Ax = Fx/Me
LET Dion = 1.07 !Proton separation LET Ay = Fy/Me

! --------- Position of proton #2 LET Vold = SQR(Vx*Vx + Vy*Vy)

LET Zx = Dion LET Vx = Vx + Ax*dt
LET Zy = 0 LET VY = Vy + Ay*dt
LET Z = SQR(Zx*Zx + Zy*Zy) LET Vnew = SQR(Vx*Vx + Vy*Vy)
LET V = (Vold + Vnew)/2
! --------- Plot crosses at protons
LET Rx = 0 LET T = T + dt
LET Ry = 0 LET i = i + 1
CALL BigCROSS
LET Rx = Zx IF MOD(i,50) = 0 THEN
LET Ry = Zy
CALL BigCROSS PLOT Rx,Ry

! --------- Initial conditions IF Rx > Zx THEN

LET Rx = 1.5 IF -F2x > -F1x THEN CALL CROSS
LET Ry = 1.6 END IF
LET R = SQR(Rx*Rx + Ry*Ry)
IF Rx < 0 THEN
LET Sx = Rx - Zx !Vector equation (S = R - Z) IF -F1x < -F2x THEN CALL CROSS
LET Sy = Ry - Zy END IF
LET S = SQR(Sx*Sx + Sy*Sy)
END IF
LET Vx = -1
LET Vy = 0 LOOP UNTIL T > 100
LET V = SQR(Vx*Vx + Vy*Vy)
! --------- Subroutine ENERGY prints out total energy.
LET T = 0 SUB ENERGY
LET i = 0 LET Etot = Me*V*V/2 - K*Qe*Qp/R - K*Qe*Qp/S
!Add potential energy of repulsive core
! --------- Print total energy LET Etot = Etot + (1/2)*Kr*Qe*Qp/R^2 + (1/2)*Kr*Qe*Qp/S^2
CALL ENERGY PRINT T,Etot
END SUB
! --------- Computer time step
LET dt = .001 ! --------- Subroutine CROSS draws a cross at Rx,Ry.
SUB CROSS
! --------- Calculational loop PLOT LINES: Rx-.01,Ry; Rx+.01,Ry
DO PLOT LINES: Rx,Ry-.01; Rx,Ry+.01
LET Rx = Rx + Vx*dt END SUB
LET RY = Ry + Vy*dt
LET R = SQR(Rx*Rx + Ry*Ry) ! --------- Subroutine BigCROSS draws a cross at Rx,Ry.
SUB BigCROSS
LET Sx = Rx - Zx PLOT LINES: Rx-.04,Ry; Rx+.04,Ry
LET Sy = Ry - Zy PLOT LINES: Rx,Ry-.04; Rx,Ry+.04
LET S = SQR(Sx*Sx + Sy*Sy) END SUB

END

electron electron

F1

R S F2

proton Z proton proton proton

Figure 14
Computer program for the hydrogen molecule ion.
Chapter 20
Nuclear Matter

In the last chapter our focus was on what one might call
electronic matter—the structures that result from the
interaction of the electrons in atoms. Now we look at
CHAPTER 20 NUCLEAR MATTER nuclear matter, found both in the nuclei of atoms and
in neutron stars. The structures we see result from an
interplay of the basic forces of nature. In the atomic
nucleus, the nuclear, electric, and weak interactions
are involved. In neutron stars and black holes, gravity
also plays a major role.
20-2 Nuclear Matter

NUCLEAR FORCE we can use the word nucleon to describe either a proton
In 1912 Ernest Rutherford discovered that all the or neutron, and talk about the nuclear force between
positive charge of an atom was located in a tiny dense nucleons. Another feature of the nuclear force is that it
object at the center of the atom. By the 1930s, it was ignores electrons. We could say that electrons have no
known that this object was a ball of positively charged nuclear charge.
protons and electrically neutral neutrons packed closely The properties of the nuclear force can be deduced
together as illustrated in Figure (19-1) reproduced here. from the properties of the structures it creates—namely
Protons and neutrons are each about 1.4 × 10 – 13 cm atomic nuclei. The fact that protons and neutrons
in diameter, and the size of a nucleus is essentially the maintain their size while inside a nucleus means that
size of a ball of these particles. For example, iron 56, the nuclear force is both attractive and repulsive. Try
with its 26 protons and 30 neutrons, has a diameter of to pull two nucleons apart and the attractive nuclear
about 4 proton diameters. Uranium 235 is just over 6 force holds them together, next to each other. But try
proton diameters across. (One can check, for example, to squeeze two nucleons into each other and you
that a bag containing 235 similar marbles is about six encounter a very strong repulsion, giving the nucleons
marble diameters across.) essentially a solid core.
That the nucleus exists means that there is some force We have seen this kind of behavior before in the case
other than electricity or gravity which holds it together. of molecular forces. Molecular forces are attractive,
The protons are all repelling each other electrically, the holding atoms together to form molecules, liquids and
neutrons are electrically neutral, and the attractive crystals. But if you try to push atoms into each other,
gravitational force between protons is some 10 – 38 try to compress solid matter, the molecular force be-
times weaker than the electric repulsive force. The comes repulsive. It is the repulsive part of the molecu-
force that holds the nucleus together must be attractive lar force that makes solid matter hard to compress, and
and even stronger than the electric repulsion. This the repulsive part of the nuclear force that makes
attractive force is called the nuclear force. nuclear matter nearly incompressible.
The nuclear force treats protons and neutrons equally.
In a real sense, the nuclear force cannot tell the differ-
ence between a proton and a neutron. For this reason,

Figure 19-1a Figure 19-1b

Sketch of an atomic nucleus, showing Styrofoam model of a Uranium nucleus.
it as a ball of protons and neutrons. (The dark balls represent protons.)
 20-3

Range of the Nuclear Force NUCLEAR FISSION

While the attractive nuclear force must be stronger than One way the instability of large nu-
the electric force to hold the protons together in the clei shows up is in the process of
nucleus, it is not a long range 1/r 2 force like electricity nuclear fission, a process that is ex-
and gravity. It drops off much more rapidly than 1/r 2 , plained by the liquid drop model of
with the result that if two protons are separated by more the nucleus developed by Neils Bohr
than a few proton diameters, the electric repulsion and John Wheeler in 1939.
becomes stronger than the nuclear attraction. The
separation R 0 at which the electric repulsion becomes In this model, we picture nuclear
stronger than the nuclear attraction, is about 4 proton matter as being essentially an incom-
diameters. This distance R 0 , which we will call the pressible liquid. The nucleons can-
range of the nuclear force, can be determined by not be pressed into each other, or
looking at the stability of atomic nuclei. pulled apart, but they are free to slide
around each other like the water mol-
If we start with a small nucleus, and keep adding ecules in a drop of water. As a result
nucleons, for a while the nucleus becomes more stable of the liquid nature of nuclear matter,
if you add the right mix of protons and neutrons. By we can learn something about the
more stable, we mean more tightly bound. To be behavior of nuclei by studying the
explicit, the more stable, the more tightly bound a behavior of drops of water.
nucleus, the more energy that is required, per nucleon,
to pull the nucleus apart. This stability, this tight In our discussion of entropy at the
binding, is caused by the attractive nuclear force be- beginning of Chapter 18, we dis-
tween nucleons. cussed a demonstration in which a
stream of water is broken into a series
Iron 56 is the most stable nucleus. It takes more energy of droplets by vibrating the hose lead-
per nucleon to take an Iron 56 nucleus apart than any ing to the stream. If you put a strobe
other nucleus. If the nucleus gets bigger than Iron 56, light on the stream, you can stop the
it becomes less stable, less tightly bound. If a nucleus apparent motion of the individual
gets too big, bigger than a Lead 208 or Bismuth 209 droplets. The result is a strobe pho-
nucleus, it becomes unstable and decays by itself. tograph of the projectile motion of
the droplets.
The stability of Iron 56 results from the fact that an Iron
56 nucleus has a diameter about equal to the range of the If you use a closely focused televi-
nuclear force. In an Iron 56 nucleus every nucleon is sion camera, you can follow the mo-
attracting every other nucleon. If we go to a nucleus tion of individual drops. Adjust the
larger than Iron 56, then neighboring nucleons still strobe so that the drop appears to fall
attract each other, but protons on opposite sides of the slowly, and you can watch an indi-
nucleus now repel each other. This repulsion between vidual drop oscillate as it falls. As
distant protons leads to less binding energy per particle, shown in Figure (1), the oscillation
and instability. is from a rounded pancake shape
(images 3 & 4) to a vertical jelly bean
shape (images 6 & 7) . Bohr and
Wheeler proposed that similar oscil-
lations should take place in a large
nucleus like Uranium, particularly if
the nucleus were
struck by some out- Figure 1
Oscillations of
side particle, like an a liquid drop.
errant neutron.
20-4 Nuclear Matter

Suppose we have an oscillating Uranium nucleus, and less neutron rich nuclei, some free neutrons are also
at the present time it has the dumbbell shape shown in emitted as indicated in Figure (3). These free neutrons
Figure (2a,b). In this shape we have two nascent may go out and strike other Uranium nuclei, causing
spheres (shown by the dotted circles) connected by a further fission reactions.
neck of nuclear matter. The nascent spheres are far
enough apart that they are beyond the range R 0 of the If you have a small block of Uranium, and one of the
nuclear force, so that the electrical repulsion is stronger Uranium nuclei fissions spontaneously (it happens
than the nuclear attraction. The only thing that holds once in a while), the extra free neutrons are likely to
this nucleus together is the neck of nuclear matter pass out through the edges of the block and nothing
between the spheres. happens. If, however, the block is big enough, (if it
exceeds a critical mass of about 13 pounds for a
If the Uranium nucleus is struck too vigorously, if the sphere), then neutrons from one fissioning nucleus are
neck is stretched too far, the electric force will cause the more likely to strike other Uranium nuclei than to
two ends to fly apart, releasing a huge quantity of escape. The result is that several other nuclei fission,
electrical potential energy. This process, shown in and each of these cause several others to fission.
Figure (3) is called nuclear fission. Quickly you have a large number of fissioning nuclei
in a process called a chain reaction. This is the process
In the fission of Uranium 235, the large Uranium that occurs in an uncontrolled way in an atomic bomb
nucleus breaks up into two moderate sized nuclei, for and in a controlled way in a nuclear reactor.
example, Cesium 140 and Zirconium 94. Because
larger nuclei have a higher percentage of neutrons than The energy we get from nuclear fission, the energy
smaller ones, when Uranium breaks up into smaller, from all commercial nuclear reactors, is electrical
potential energy released when the two nuclear frag-
R0 ments fly apart. The fragments shown in Figure (3) are
+ + + +
+ +
at that point well beyond the range R 0 of the attractive
+ + + ++++ + + + + +
+ + + + nuclear force, and essentially feel only the repulsive
+ + ++ ++ + ++ + + electric force between the protons. These two balls of
+ + ++ + ++ + + +
+ + ++ + + + positive charge have a large positive electric potential
energy which is converted to kinetic energy as the
Figure 2a
fragments fly apart.
Uranium nucleus in a dumbbell shape.

+ + + + +
+ ++ +
+ + + + + + +
+ + + + + + + + + +
+ + + + + + + +
+ + + + + + + + + +
+ + + + + +

Figure 3
When the nucleus flies apart, an
enormous amount of electric
potential energy is released.

Figure 2b
Styrofoam model of a Uranium
nucleus in a dumbbell shape.
 20-5

To get a feeling for the amount of energy released in a To compare the strength of nuclear fission reactions to
fission reaction, let us calculate the electric potential chemical reactions, we can compare the electric poten-
energy of two fragments, say a Cesium and a Zirco- tial energies in Equations 2 and 3. If we take the range
nium nucleus when separated by a distance 2 R 0 , twice R 0 of the nuclear force to be 4 proton diameters then
the range of the nuclear force.
R 0 = 4 × 1.4 × 10 – 13 cm = 5.6 × 10 – 13 cm
In CGS units, the formula for the electric potential
energy of 2 particles with charges Q 1 and Q 2 sepa- Since the Bohr radius is 5 × 10 – 9 cm , we see that R 0
rated by a distance R is is essentially 10 – 4 R b or ten thousand times smaller
electric Q 1Q 2 than the Bohr radius.
potential U electric = r CGS units (1)
energy R 0 = 10 – 4 r b (4)
For our problem, let Q 1 be the charge on a Cesium Substituting Equation 4 into 2 gives
nucleus (55 protons) and Q 2 the charge on a Zirconium
e2
nucleus (40 protons). U electric = 1.1 × 10 3
R0
Q Cesium = 55e
e2
Q Zirconium = 40e = 1.1 × 10 3
r = 2R 0 10 – 4 rb
and we get
e2
=1.1 × 10 7
55 e × 40 e rb
U electric =
2R 0
Using the fact that e 2/r b has a magnitude of 13.6
3 e2 electron volts, we get
= 1.1 × 10 (2)
R0
U electric = 1.1 × 10 7 × 13.6 eV
We would like to compare the energy released in
nuclear fission reactions with the energies typically = 150 × 10 6 eV (5)
involved in chemical reactions. It takes a fairly violent
= 150 MeV
chemical reaction to rip the electron completely out of
a hydrogen atom. The amount of energy to do that, to where 1MeV is one million electron volts. From Equa-
ionize a hydrogen atom is e 2 /r b where r b is the Bohr tion 5, we see that, per particle, some ten million times
radius of 5 × 10 – 9 cm . more electric potential energy is released in a nuclear
2 fission reaction than in a violent chemical reaction.
= er = 13.6 eV
energy to ionize
a hydrogen atom
(3) Many millions of electron volts are involved in nuclear
b
reactions as compared to the few electron volts in
We evaluated the number e 2 /r b earlier and found it to chemical reactions. You can also see that a major
have a numerical value of 13.6 electron volts. This is reason for the huge amounts of energy in a nuclear
a large amount of energy for a chemical reaction, more reaction is the small size of the nucleus (the fact that
typical chemical reactions, arising from molecular R 0 << r b ).
forces, have involved energies in the 1 to 2 electron volt
range.
20-6 Nuclear Matter

NEUTRONS AND If you have an isolated free neutron, the β decay of

THE WEAK INTERACTION Equation 5 occurs with a half life of 15 minutes. Such
a reaction can occur only if energy can be conserved in
The stability of the iron nucleus and the instability of
the process. But the neutron is sufficiently massive to
nuclei larger than Uranium results primarily from the
decay into a proton and an electron and still have some
fact that the attractive part of the nuclear force has a
energy left over. Expressing rest mass or rest energy in
short range R 0 over which it dominates the repulsive
units of millions of electron volts (MeV), we have for
electric force between protons. The range R 0 is about
the particles in the neutron decay reaction (5),
4 proton diameters, the diameter of an iron nucleus. In
larger nuclei, not all nucleons attract each other, and neutron rest mass mn = 939.6 MeV
this leads to the kind of instability we see in a fissioning proton rest mass mp = 938.3 MeV
Uranium nucleus. (6)
electron rest mass me = 0.511 MeV
The range of the nuclear force is not the only important neutrino rest mass mν = 0
factor in determining the stability of nuclei. There is no
electric repulsion between neutrons, neutrons are at- where
tracted equally to neutrons and protons. Adding neu- 1 MeV = 10 6 eV = 1.6 × 10 – 6 ergs
trons to a nucleus increases the attractive nuclear force
without enhancing the electric repulsion. This is why You can see that the neutron rest mass is 1.3 MeV
the most stable large nuclei have an excess of neutrons greater than that of a proton, and .8 MeV greater than
over protons. The neutron excess acts as a nuclear glue, the combined rest masses of the proton, electron and
diluting the repulsion of the protons. neutrino. Thus when a neutron β decays, there is an
excess of .8 MeV of energy that is released in the form
If adding a few extra neutrons increases the stability of of kinetic energy of the reaction products.
a nucleus, why doesn’t adding more neutrons give even
more stable nuclei? Why is it that nuclei with too many As we have mentioned, when the β decay process was
excess neutrons are in fact unstable? Why can’t we first studied in the 1920s, the neutrino was unknown.
make nuclei out of pure neutrons and avoid the proton What was observed was that in β decays, the proton and
repulsion altogether? the electron carried out different amounts of energy,
sometimes all of the available energy, but usually just
The answer lies in the fact that, because of the weak part of it. To explain the missing energy, Wolfgang
interaction, and because of a small excess mass of a Pauli proposed the existence of an almost undetectable,
neutron, a neutron can decay into a proton and release uncharged, zero rest mass particle which Fermi named
energy. This is the beta decay reaction we discussed the little neutral one or neutrino. As bizarre as Pauli’s
earlier, and is described by the equation hypothesis seemed at the time, it turned out to be
correct. When a neutron decays, it decays into 3
n 0 → p + + e– + ν0 (5)
particles, and the .8 MeV available for kinetic energy
The neutral neutron ( n 0 ) decays into a positive proton can be shared in various ways among the 3 particles.
( p + ), a negative electron ( e– ), and a neutral Because the neutrino has no rest mass, it is possible for
antineutrino ( ν0 ), thus electric charge is conserved in the proton and electron to get all .8 MeV of kinetic
the process. It is called a beta β decay reaction energy. At other times the neutrino gets much of the
because the electrons that come out were originally kinetic energy, so that if you did not know about the
called beta rays before their identity as electrons was neutrino, you would think energy was lost.
determined.
 20-7

NUCLEAR STRUCTURE Our simple calculation shows that for the neutron to
Free neutrons decay in fifteen minutes, but neutrons decay into a proton in deuterium, the neutron would
inside a nucleus seem to live forever. Why don’t they have to create nearly as much electrical potential
decay? The answer to this question is an energy energy (.72MeV) as the available .8 MeV neutron mass
balance. Like a rock dropped into a well, an atomic energy. A more complete analysis shows that the
nucleus will fall down to the lowest energy state available .8 Mev is not adequate for the neutron β
available. The neutron will decay if the result is a lower decay, with the result the neutron in deuterium is stable.
energy, less massive nucleus. Otherwise the neutron We can now see the competing processes involved in
will be stable. the formation of nuclei. To construct stable nuclei you
We have seen that an isolated neutron can decay into want to add neutrons to give more attractive nuclear
the less massive proton and release energy. Now forces and dilute the repulsive electric forces between
consider a neutron in a nucleus. Take the simplest protons. However, the rest mass of a neutron is greater
nucleus with a neutron in it, namely Deuterium . If that than the rest mass of a proton and an electron, and the
neutron decayed we would end up with a helium weak interaction allows the neutron to decay into these
nucleus consisting of two protons only, plus an elec- particles. Thus neutrons can shed mass, and therefore
tron and a neutrino. We can write this reaction as energy, by decaying.

deuterium decay But when a neutron inside of a nucleus decays into a

2
H → 2He + e– + ν which does not happen
(7)
proton, it increases the electric potential energy of the
nucleus. If the increase in the electric potential energy
n p p p + + − is greater than the mass energy released, as we nearly
e− ν
saw in the case of deuterium, then the neutron cannot
If the neutron in deuterium turns into a proton, the decay.
neutron sheds rest mass, but the resulting two proton
nucleus has positive electric potential energy. We can The weak interaction is democratic, it allows a proton
estimate the amount of electric potential energy U pp to decay into a neutron as well as a neutron to decay into
created by using the formula a proton. The proton decay process is
electric potential
e2 p + → n 0 + e+ + ν0 inverse β decay (10)
energy of a 2 U pp = CGS units (8)
proton nucleus 2rp
where e + is the positively charged antielectron
where r p is the proton radius and 2r p = 2 × 10 – 13 is the (positron). This inverse β decay, as it is sometimes
separation of the proton centers. The protons each have called, does not occur for a free proton because energy
a charge + e. Putting numbers into Equation 8 gives is not conserved. The proton rest mass is less than that
of a neutron, let alone that of a neutron and a positron
e2 combined.
Up p =
2rp
However, if you construct a nucleus with too many
2
– 10
4.8 × 10 protons, with too much electric potential energy, then
= ergs the nucleus can get rid of some of its electric potential
2 × 10 – 13 energy by converting a proton into a neutron. This may
(9) happen if enough electric potential energy is released to
= 1.15 × 10 – 6 ergs supply the extra rest mass of the neutron as well as the
= .72 MeV .5 MeV rest mass of the positron.
20-8 Nuclear Matter

α (Alpha) Particles
We can now see the competing processes in an atomic
nucleus. The weak interaction allows neutrons to turn In 1898 Ernest Rutherford, a young research student in
into protons or protons into neutrons. But these β Cambridge University, England, discovered that ra-
decay processes will occur only if energy can be dioactive substances emitted two different kinds of
released. Nuclei with too many neutrons have too rays which he named α rays and β rays after the first
much neutron mass energy, and can get rid of some of two letters of the Greek alphabet. The negatively
the mass energy by turning a neutron into a proton. charged β rays turned out to be beams of electrons, and
Nuclei with too many protons have too much electrical the positive α rays were found to be beams of Helium
potential energy, and can get rid of some of the electri- 4 nuclei. Helium 4 nuclei, consisting of 2 protons and
cal potential energy by turning a proton into a neutron. 2 neutrons, are thus also called α particles. (Later
A stable nucleus is one that has neither an excess of Rutherford observed a third kind of radiation he called
mass energy nor electrical potential energy, a nucleus γ rays, which turned out to be high energy photons.)
that cannot release energy either by turning protons We have seen that β rays or electrons are emitted when
into neutrons or vice versa. a neutron sheds mass by decaying into a proton in a β
decay reaction. But where do the α particles come
To predict precisely which nuclei are stable and which from?
are not requires a more detailed knowledge of the
nuclear force than we have discussed here. But from A nucleus with an excess of electric potential energy
what we have said, you can understand the general can lose energy by converting one of its protons into a
trend. For the light nuclei, the most stable are the ones neutron in an inverse β decay reaction. This, however,
with roughly equal numbers of protons and neutrons, is a relatively rare event. More commonly, the number
nuclei with neither too much neutron mass energy or of protons is reduced by ejecting an α particle. Why the
too much proton electrical potential energy. However, nucleus emits an entire α particle or Helium 4 nucleus,
when nuclei become larger than the range of the instead of simply kicking out a single proton, is a
attractive nuclear force, electric potential energy be- consequence of an anomaly in the nuclear force. It
comes more important . Excess neutrons, with the turns out that a Helium 4 nucleus, with 2 neutrons and
additional attractive nuclear binding force they pro- 2 protons, is an exceptionally stable, tightly bound
vide, are needed to make the nucleus stable. object. If protons are to be ejected, they come out in
pairs in this stable configuration rather than
individually.
 20-9

NUCLEAR BINDING ENERGIES Thus the total binding energy of the deuterium nucleus,
The best way to see the competition between the the energy required to pull the particles apart is
attractive nuclear force and the electric repulsive force
total binding
inside atomic nuclei is to look at nuclear binding
energy of the
energies. Explicitly, we will look at the binding energy = 1877.9 MeV – 1875.1 MeV
deuterium separate deuterium
per nucleon for the most stable nuclei of each element. nucleus nucleus
particles
The binding energy per nucleon (proton or neutron)
represents how much energy we would have to supply
= 2.8 MeV
to pull the nucleus apart into separate free nucleons. (22)
The nuclear force tries to hold the nucleus together— Finally the binding energy per nucleon, there being 2
make it more tightly bound—and therefore increases nucleons, is
the binding energy. The electric force, which pushes
the protons apart, decreases the binding energy. binding energy
= 2.8 MeV
per nucleon 2 nucleons
You calculate the binding energy of a nucleus by
subtracting the rest energy of the nucleus from the sum
of the rest energies of the protons and neutrons that = 1.4 MeV (23)
nucleon
make up the nucleus. If you then divide by the number
of nucleons, you get the binding energy per nucleon.
We will go through an example of this calculation, and Exercise 1
give you an opportunity to work out some yourself. Given that the masses of the Helium 4 (2 protons, 2
Then we will look at a plot of these binding energies to neutrons), the Iron 56 (26 protons, 30 neutrons), and the
see what the plot tells us. Uranium 238 (92 protons, 146 neutrons) nuclei are
MHelium 4 c2 = 3725.95 MeV
Example 1
MIron 56 c2 = 52068.77 MeV (24)
Given that a proton, a neutron, and a deuterium nucleus 2
have the following rest energies, what is the binding MUranium 238 c = 221596.94MeV
energy per particle for the deuterium nucleus? find in each case, the binding energy per nucleon. Your
results should be
m p c2 = 938.3 MeV proton rest energy
7.46 MeV (Helium 4))
m n c2 = 939.6 MeV neutron rest energy bindingenergy
= 9.20 MeV (Iron 56)
m d c2 = 1875.1 MeV deuterium nucleus per nucleon
8.02 MeV (Uranium 238)
rest energy (25)
(20)
Solution
Separately the proton and neutron in a deuterium
nucleus have a total rest energy of
rest energy
a separate
= 938.3 MeV + 939.6 MeV
proton and
neutron
= 1877.9 MeV
(21)
20-10 Nuclear Matter

Figure (4) is a plot of the binding energy, per nucleon, Moving down into the well represents a release of
of the most stable nuclei for each element. We have nuclear energy. There are two ways to do this. We can
plotted increasing binding energy downward so that start with light nuclei and put them together (in a
the plot would look like a well. The deeper down in the process called nuclear fusion), to form heavier nuclei,
well a nucleus is, the more energy per particle that is moving in and down from the left side in Figure (4). Or
required to pull the nucleus out of the well—to pull it we can split apart heavy nuclei (In the process of
apart. The deepest part of the well is at the Iron 56 nuclear fission), moving in and down from the right
nucleus, no other nucleus is more tightly bound. side. Fusion represents the release of nuclear force
potential energy, while fission represents the release of
1 electric force potential energy. When we get to the
2H
bottom, at Iron 56, there is no energy to be released
2 either by fusion or fission.

3He
3
Binding energy (in MeV)

4 fusion fission
release of nuclear release of electric
force potential force potential
5 energy energy
6Li
6 7Li

7
235U
208Pb
23
8 Na 141Ba
27 92Kr
4
He Al 56
Fe
9
bottom of the well at iron 56

0 20 40 60 80 100 120 140 160 180 200 220 240

Number of nucleons (neutrons or protons) in nucleus
Figure 4
The nuclear energy well. The graph shows the
amount of energy, per nucleon, required to pull the
nucleus apart into separate neutrons and protons.
 20-11

The reason Iron 56 is at the bottom of the well is because The importance of knowing the nuclear binding energy
the diameter of an iron nucleus is about equal to the per nucleon is that it tells us whether energy will be
range of the nuclear force. As you build up to Iron 56, released in a particular nuclear reaction. If the some-
adding more nucleons increases the number of attrac- what weakly bound uranium nucleus (7.41 MeV/
tive forces between particles, and therefore increases nucleon) splits into two more tightly bound nuclei like
the strength of the binding. At Iron 56, you have the cesium (8.16 MeV/nucleon) and zirconium (8.41 MeV/
largest nucleus in which every particle attracts every nucleon), energy is released. At the other end of the
other particle. The diameter of the Iron 56 nucleus is the graph, if we combine two weakly bound deuterium
distance over which the attractive nuclear force is nuclei (2.8 MeV/nucleon) to form a more tightly bound
stronger than the repulsive electric force. Helium 4 nucleus (7.1 MeV/nucleon), energy is also
released. Any reaction that moves us toward the Iron
When you build nuclei larger than Iron 56, the protons 56 nucleus releases energy. On the small nucleus side
on opposite sides of the nucleus are far enough apart we get a release of energy by combining small nuclei
that the electric force is stronger and the particles repel. to form bigger nuclei. But once past Iron 56, we get a
Now, adding more particles reduces the binding energy release of energy by splitting nuclei apart for form
per particle and produces less stable nuclei. When a smaller ones.
nucleus become as large as uranium, the impact of a
single neutron can cause the nucleus to split apart into
two smaller, more stable nuclei in the nuclear fission
process.

There are some bumps in the graph of nuclear binding

energies, bumps representing details in the structure of
the nuclear force. The most striking anomaly is the
Helium 4 nucleus which is far more tightly bound than
neighboring nuclei. This tight binding of Helium 4 is
the reason, as we mentioned, that α particles (Helium
4 nuclei) rather than individual protons are emitted in
radioactive decays. But overlooking the bumps, we
have the general feature that nuclear binding energies
increase up to Iron 56 and then decrease thereafter.
20-12 Nuclear Matter

NUCLEAR FUSION touching. Energy is released when the incoming

The process of combining small nuclei to form larger proton falls into the well, but the incoming proton must
ones is called nuclear fusion. From our graph of have enough kinetic energy to get over the electrical
nuclear binding energies in Figure (4), we see that potential energy barrier before the fusion can take
nuclear fusion releases energy if the resulting nucleus place.
is smaller than Iron 56, but costs energy if the resulting Using our formula Q 1Q 2 /r for electric potential en-
nucleus is larger. This fact has enormous significance ergy, we can make a rough estimate of the kinetic
in the life of stars and the formation of the elements. energy and the temperature required for fusion. Con-
Most stars are created from a gas cloud rich in hydrogen sider the fusion of two protons where Q 1 = Q 2 = e .
gas. When the cloud condenses, gravitational potential For the protons to get within a distance R 0, the incom-
energy is released and the gas heats up. If the condens- ing proton must climb a barrier of height
ing cloud is massive enough, if the temperature be- electric potential
2
comes hot enough, the hydrogen nuclei begin to fuse. energy of 2 protons = e
After several reactions they produce Helium 4 nuclei, R0
a distance R0 apart
releasing energy in each reaction. The fusion of 2
hydrogen to form helium becomes the source of energy 4.8 × 10 – 10
for the star for many years to come. = (26)
4 × 1.4 × 10 – 13
Unlike fission, fusion requires high temperatures in
= 4.1 × 10 – 7 ergs
order to take place. Consider the reaction in which two
hydrogen nuclei (protons) fuse to produce a deuterium This number, e 2 /R 0 = 4.1 × 10 – 7 ergs , is the amount
nucleus plus a positron and a neutrino. (When the two of kinetic energy an incoming proton must originally
protons fuse, the resulting nucleus immediately gets rid have in order to get within a distance of approximately
of its electrical potential energy by having one proton R 0 of another proton. Only when it gets within this
turn into a neutron in an inverse β decay process.) distance can the nuclear force take over and fusion take
The fusion of the two protons will take place if the place.
protons get closer together than the range of the nuclear potential energy of approaching proton
approaching proton meets electric
force about—4 proton diameters. Before they get that potential energy
close they repel electrically. Only if the protons were barrier
initially moving fast enough, were hot enough, can they
get close enough to get past the electrical repulsion in
order to feel the nuclear attraction. fixed
proton proton separation r
A good way to picture the situation is to think of Ro
yourself as sitting on one of the protons, and draw a
graph of the potential energy of the approaching pro- nuclear potential
ton, as shown in Figure (5). When the proton separa- energy well
tion r is greater than the range R 0 of the nuclear force,
Figure 5
the protons repel and the incoming proton has to climb
When you shove two protons together, you
a potential energy hill. At R = R 0, the net force turns first have to overcome the electric repulsion
attractive and the potential energy begins to decrease, before the nuclear attraction dominates.
forming a deep well when the particles are near to
 20-13

In our earlier discussion of temperature, we saw that the STELLAR EVOLUTION

average kinetic energy of a particle in a gas of tempera- The story of the evolution of stars provides an ideal
ture T was 3/2 kT. If we had a gas of hydrogen so hot setting to illustrate the interplay of the four basic
that the average proton could enter into a fusion reac- interactions. In this chapter so far we have been
tion, the average kinetic energy of the protons would focusing on the basic consequences of the interplay of
have to be 4.1 × 10 – 7 ergs. The temperature Tf at the nuclear, electric, and weak interactions at the level
which this would happen is found by equating of atomic nucleus. Add gravity and you have the story
4.1 × 10 – 7 ergs to 3/2 kT to give of stellar evolution.
3 kT = 4.1 ×10 –7 ergs A star is born from a cloud of gas, typically rich in
2
hydrogen, that begins to collapse gravitationally. As
4.1 ×10 –7 ergs the cloud collapses, gravitational potential energy is
T = 2
3 1.38 ×10 –16 ergs released which heats the gas. If the temperature does
kelvin not get hot enough to start the fusion of the hydrogen
(27)
nuclei, that’s more or less the end of the story and you
T = 2 ×10 9 kelvin 2 billion
degrees have a proto star, something around the size of the
planet Jupiter or smaller.
This temperature, two billion degrees kelvin, is a huge
overestimate. At this temperature, the average proton If there is more mass in the collapsing gas cloud, more
in the gas would enter into a fusion reaction. If we gravitational potential energy will be released, and the
heated a container of hydrogen to this temperature, the temperature will rise enough to start the fusion of
entire collection of protons would fuse after only a few hydrogen. How hot the center of the star becomes
collisions, and the fusion energy would be released depends on the mass of the star. In a star, like our sun
almost instantaneously. We would have what is known for example, there is a balance between the gravita-
as a hydrogen bomb. tional attraction and the thermal pressure. The greater
the mass, the greater the gravitational attraction, and
In a star, the fusion of hydrogen takes place at the much the stronger the thermal pressure must be.
lower temperatures of about 20 million degrees. At 20
million degrees, only a small fraction of the protons In our discussion of pressure in chapter 17, we saw that
have enough kinetic energy to enter into a fusion when a balloon was cooled by liquid nitrogen, remov-
reaction. At these temperatures the hydrogen is con- ing the thermal energy and pressure of the air molecules
sumed at a slow steady rate in what is known as a inside, the balloon collapsed. ( Figures 17-19.) Similar
controlled fusion reaction. processes occur in a star, except that the confining force
of the rubber is replaced by the confining force of
gravity. The sun is a ball of hot gas. Gravity is trying
to squeeze the gas inward, and the thermal pressure of
the gas prevents it from doing so. There is a precise
balance between the thermal pressure and gravity.

Figure 17-19
Balloon collapsing
in liquid nitrogen.
20-14 Nuclear Matter

From the earth the sun looks more like a solid object There is a range of burning rates between these two
than a ball of gas. The sun has a definite edge, an extremes. As a result, with increasing temperature the
obvious surface with spots and speckles on it. The fusion reaction goes faster, supplies energy at a greater
appearance of a sharp surface is the result of the change rate, and maintains the higher temperature. Thus when
in temperature of the gas with height. The hottest part a new star forms it collapses until there is a balance
of the sun or any star is the center. Here the gas is so hot, between the gravitational force and thermal pressure.
the thermal collisions are so violent, that the electrons Whatever thermal pressure is required is supplied by
are knocked out of the hydrogen atoms and all the the heat generated by the fusion reaction. The more
hydrogen is ionized. The gas is what is called a plasma. thermal pressure needed, the higher the temperature
An ionized gas or plasma is opaque, light is absorbed required and the faster the hydrogen burns.
by the separate charged particles.
One often refers to the region near the center of the star
As you go out from the center of the star, the tempera- where the fusion reaction is taking place as the core of
ture drops. When you go out far enough, when the the star. Our sun, with a temperature in the core of 20
temperature drops to about 3000 kelvins, the electrons million degrees, is burning hydrogen at such a rate that
recombine with the nuclei, you get neutral atoms, and the hydrogen supply will last 10 billion years. Since the
the gas becomes transparent. This transition from an sun is 5 billion years old, about half the available
opaque to transparent gas occurs rather abruptly, giv- hydrogen in the core is used up.
ing us what we think of as the surface of the sun.
A more massive star, like the star that blew up to give
Returning to the balance of gravitational attraction and us the 1987 supernova event, burns its hydrogen at a
thermal pressure, you can see that the more mass in the much shorter time. That star was about 18 times as
star the stronger gravity is, and the greater the thermal massive as the sun, about 40,000 times as bright, and
pressure required to balance gravity. To increase the burned its hydrogen in its core so fast that the hydrogen
thermal pressure, you have to increase the temperature. lasted only about 10 million years.
Thus the more massive a star, the hotter it has to be.
The difference between different mass stars shows up
The proton fusion reaction we have discussed is well most dramatically after the hydrogen fuel is used up.
suited for supplying any required temperature. We What will happen to our sun is relatively calm. When
have seen that the rate at which fusion takes place our sun uses up the hydrogen, the core will start to cool
depends very much on the temperature. At 20 million and collapse. But the collapse releases large amounts
degrees, only a small fraction of the protons have of gravitational potential energy that heats the core to
enough thermal kinetic energy to fuse. At 2 billion higher temperatures than before. This hotter core
degrees, the average proton has enough energy to fuse, becomes very bright, so bright that the light from the
and any hydrogen at this temperature would burn core, when it works its way out to the surface, exerts a
immediately. strong radiation pressure on the gas at the surface. This
radiation pressure will cause the surface of the sun to
expand until the diameter of the sun is about equal to the
diameter of the earth’s orbit. At this point the sun will
have become what is called a red giant star, with the
earth orbiting slightly inside. It is not a very pleasant
picture for the earth, but it will not happen for another
five billion years.
 20-15

Once the sun, as a red giant star, radiates the energy it In successively shorter times at successively higher
got from the gravitational collapse, it will gradually temperatures the more massive elements were created
cool and collapse until the atoms push against each and burned. After neon, there was carbon, then oxygen
other. It will be the electric force between the atomic at 2.1 billion degrees, and finally silicon and sulfur at
electrons that will halt the gravitational collapse of the 3.4 billion degrees. The neon burned in about 12 years,
sun. At that point the sun will become a ball of highly the oxygen in 4 years, and the silicon in just a week.
compressed atomic matter about the size of the earth.
Initially it will be quite bright, an object called a white One of the reasons for the accelerated pace of burning
dwarf star, but eventually it will cool and darken. at the end is that, at temperatures over half a billion
degrees, the star has a more efficient way of getting rid
The story was very different for the star that gave rise of energy than emitting light. At these temperatures
to the supernova explosion. That star, with its mass of some of the photons are energetic enough to create
about 18 times that of the sun, burned its hydrogen in electron-positron pairs which usually annihilated back
10 million years. At that point the star had a core, about into photons but sometimes into neutrinos. Whereas
30 per cent of the star, consisting mostly of Helium 4, photons take thousands of years to carry energy from
the tightly bound nucleus that is the end result of the core of a star to the surface, neutrinos escape
hydrogen fusion. Computer simulations tell us that for immediately. Thus when the star reached half a billion
the next tens of thousands of years, the helium core was degrees, it sprung a neutrino heat leak, and the collapse
compressed from a density of 6 to 1,100 grams per and burning went much faster.
cubic centimeter, and the temperature rose from 40
million to 190 million degrees kelvin. The successive stages of burning took place in smaller
and smaller cores, leaving shells of unburned elements.
The temperature of 190 million degrees is high enough Unburned hydrogen filled the outer volume of the star.
to cause helium nuclei to fuse, forming carbon and Inside was a shell of unburned helium and inside that
oxygen. Higher temperatures are required to fuse successive shells of unburned carbon, oxygen, then a
helium nuclei, because each helium nucleus has two mixture of neon, silicon and sulfur, and a shell of silicon
protons and a charge + 2e. Thus the electric potential and sulfur.
barrier Q 1Q 2 /R 0 is four times as high as it is for proton
fusion, and the helium nuclei need four times as much In the center was iron. Iron was what resulted when the
kinetic energy to fuse. silicon and sulfur burned. And iron is the end of the
road. As we have seen in Figure (4) iron is the most
At these higher temperatures the core radiated more stable atomic nucleus. Energy is released when you
light, causing the outer layers of the star, mostly fuse nuclei to form a nucleus smaller than iron, but it
unburned hydrogen, to expand to about twice the size costs energy to create nuclei larger than iron. The iron
of the earth’s orbit. It had become a red supergiant. core of the star was dead ash not a fuel.
The helium in the core lasted less than a million years,
leaving behind a collapsing core of carbon and oxygen.
The temperature rose to about 740 million degrees
where the carbon ignited to form neon, magnesium and
sodium. When the carbon was used up in about 12,000
years, further collapse raised the temperature to 1.6
billion degrees where neon ignited.
20-16 Nuclear Matter

By the time the star had an iron core, it was rapidly How elements are formed in stars has been a fascinat-
radiating energy in the form of neutrinos, but had run ing detective story carried out over the past 40 years.
out of fuel. At this point the iron core, which had a mass The pioneering work, carried out by William Fowler,
of about 1.4 times the mass of the sun, began to collapse Fred Hoyle, and others, involved a careful study of
due to the lack of support by thermal pressure. When nuclear reactions in the laboratory and then a modeling
the sun runs out of energy, its collapse will be halted by of how stars should evolve based on the known reac-
the electric repulsion between atomic electrons. In the tions. On one occasion a nuclear reaction was pre-
1987 supernova star, the gravitational forces were so dicted to exist because it had to be there for stars to
great that the electrons were essentially crammed back evolve. The reaction was then found in the laboratory,
into the nuclei, the protons converted to neutrons, and exactly as predicted.
the core collapsed into a ball of neutrons about 100
miles in diameter. The modeling of stellar evolution, using experimental
data on nuclear reactions, and large computer pro-
This collapse took a few tenths of a second, and created grams, has quite successfully predicted the relative
a shock wave that rapidly spread to the outer layers of abundance of the various elements , as well as many
the star. Vast quantities of neutrinos were created in the features of stellar evolution such as the expansion of a
collapse, and escaped over the next 10 or so seconds. star into a red giant when the hydrogen fuel in its core
The shock wave reached the surface of the star 3 hours is used up. The success of these models leads us to
later, blowing off the surface of the star and starting a believe that the details we described about what hap-
burst of light 3 hours behind the burst of neutrinos. The pened in the very core of a star about to explode,
light and neutrinos raced each other for 180,000 years, actually happened as described. An exciting conse-
and the neutrinos were still at least 2 hours ahead when quence of the 1987 supernova explosion was that we
they got to the earth. got a glimpse into the core of the exploding star, a
glimpse provided by the neutrinos that took 10 sec-
When a supernova explodes, the outer shells of hydro- onds, as predicted, to escape from the core. The
gen, helium, carbon, neon, magnesium, sodium, sili- neutrinos also arrived three hours before the photons,
con, sulfur and iron are blown out to form a new dust as predicted by the computer models.
cloud. Such a dust cloud—the Crab Nebula, produced
by the 1054 supernova explosion—is shown in Figure
(6). From this cloud new stars and planets will form,
stars and planets rich in the heavier elements created in
the star and recycled into space by the supernova
explosion.

Elements heavier than iron are also in the supernova

remnants. So much energy is released in the collapsing
core of the supernova that elements heavier than iron
are created by fusion, even though this fusion costs
energy. All the silver and gold in your watchband and
ring, the iodine in your medicine cabinet, the mercury
in your thermometer, and lead in your fishing sinker,
all of these elements which lie beyond iron, were
created in the flash of a supernova explosion. Without
supernova explosions, the only raw materials for the
Figure 6a
formation of stars and planets would be hydrogen, The Crab Nebula. The arrow points to the pulsar that
some deuterium, and a trace of other light elements left created the nebula. (Above photo Hale Observatories. 6b:
over from the Big Bang that created the universe. 1950 Photograph by Walter Baade, 1964 by Güigo Munch,
composite by Munch and Virginia Trimble.)
 20-17

NEUTRON STARS In 1967, the graduate student Jocelyn Bell, using

One of the great predictions of astronomy based on the equipment devised by Antony Hewish, observed an
physical properties of matter was made by a young object emitting extremely sharp radio pulses that were
physicist/astronomer S. Chandrasekhar in the 1930s. 1.337 seconds apart. By the end of the year, up to ten
With some relatively straightforward calculations, such pulsing objects were detected, one with pulses
Chandrasekhar predicted that if a cooling, collapsing only 89 milliseconds apart. After eliminating the pos-
star had a mass greater than 1.4 times the mass of the sibility that the radio pulse was communication from an
sun, the force of gravity would be strong enough to advanced civilization, it was determined that the sig-
cram the atomic electrons down into the nuclei, con- nals were most likely from an objects rotating at high
verting the protons to neutrons, leaving behind a ball of speeds. A star cannot rotate that fast unless it is very
neutrons about 10 miles in diameter. compact, less than 100 miles in diameter. The only
candidate for such an object was Chandresekhar’s
Chandrasekhar talked about this idea with his sponsor neutron star.
Sir Arthur Eddington, who was, at the time, one of the
most famous astronomers in the world. In private, Many other pulsing stars—pulsars—have been dis-
Eddington agreed with Chandrasekhar’s calculations, covered. The closest sits at the center of the explosion
but when asked about them in the 1932 meeting of the that created the Crab Nebula seen in Figure (6a). A
Royal Astronomical Society, Eddington replied that he superposition of photographs taken in 1950 and 1964,
did not believe that such a process could possibly occur. Figure (6b), shows that the gas in the Crab Nebula is
Chandrasekhar’s ideas were dismissed by the astro- expanding away from the star marked with an arrow.
nomical community, and a discouraged Chandrasekhar Taking a high speed moving picture of this star, some-
left astronomy and went into the field of hydrodynam- thing that one does not usually do when photographing
ics and plasma physics where he made significant stars, shows that this star turns on and off 33 times a
contributions. second as seen in Figure (7). This is the neutron star left
behind when the supernova exploded.

pulsar

Figure 6b Figure 7
Expansion of the Crab Nebula. Two photographs, taken Neutron star in Crab Nebula, turns on and off 33
14 years apart, the first printed in white, the second times/sec. (Exposure from the Lick Observatory.)
dark, show the expansion centered on the pulsar.
20-18 Nuclear Matter

We have now been able to study many pulsars, and NEUTRON STARS
know that the typical neutron star is a ball of neutrons AND BLACK HOLES
about 10 miles in diameter, rotating at rates up to nearly As predicted by Chandrasekhar, when a cooling star is
1000 revolutions per second! We can detect neutron more than 1.4 times as massive as the sun, the gravita-
stars because they have a bright spot that emits a beam tional attraction becomes strong enough to overcome
of radiation. We see the pulses of radiation when the the electronic structure of matter, shoving the electrons
beam sweeps over us much as the captain of a ship sees into the nuclei and leaving behind a ball of neutrons. A
the bright flash from a lighthouse when the beam neutron star is essentially a gigantic nucleus in which
sweeps past. the attractive gravitational force which holds the ball
Computer models suggest that the bright spot is created together, is balanced by the repulsive component of the
by the magnetic field of the star, a field that was tied to nuclear force which keeps the neutrons from squeezing
the material in the core of the star and was strengthened into each other.
as the core collapsed. Charged particles in the ‘atmo- In our discussion of atomic nuclei, it was the attractive
sphere’ of the neutron star spiral around the magnetic component of the nuclear force that was of the most
field lines striking the star at the magnetic poles. On the interest. It was the attractive part that overcame the
earth, charged particles spiraling around the earth’s electric repulsion between protons. Now in the neutron
magnetic field lines strike the earth’s atmosphere at the star, gravity is doing the attracting and the nuclear force
magnetic poles, creating the aurora borealis and aurora is doing the repelling.
australis, the northern and southern lights that light up
extreme northern and southern night skies. On a Einstein’s special theory of relativity sets a limit on
neutron star, the aurora is much brighter, also the how strong the repulsive part of the nuclear force can
atmosphere much thinner, only a few centimeters be. We can see why with the following qualitative
thick. (In Chapter 28 on Magnetism, we will talk about arguments.
the motion of charged particles in a magnetic field.)
The harder it is to shove two nucleons into each other,
the stronger the repulsive part of the nuclear force, the
more incompressible nuclear matter is. Now in our
beginning discussions of the principle of relativity in
Chapter 1, we saw that the speed of a sound wave
depended upon the compressibility of the material
through which the sound was moving. We used a
stretched Slinky for our initial demonstrations of wave
motion because a stretched Slinky is very easy to
compress, with the result that Slinky waves move very
slowly, about 1 foot per second.

Air is much more incompressible than a Slinky (try

blowing air into a Coke bottle), with the result that
sound waves in air travel about 1000 times faster than
slinky waves. Water is more incompressible yet, and
sound travels through water 5 times faster than in air.
Because steel is even more incompressible than water,
sound travels even faster in steel, about 4 times faster
than in water.
 20-19

The most incompressible substance known is nuclear If one of a pair of binary stars is a black hole, two
matter. It is so incompressible, the repulsive force detectable effects can occur. If the stars are in close
between nucleons is so great that the calculated speed orbit, the black hole will suck off the outer layers of gas
of sound in nuclear matter approaches the speed of from the visible companion, as indicated in the artist’s
light. And that’s the limit. Nothing can be so rigid or conception, Figure (12). According to computer mod-
incompressible that the speed of sound in the substance els, the gas that is drawn off from the companion star
exceeds the speed of light. This fact alone tells us that goes into orbit around the black hole, forming what is
there is a limit to how rigid matter can be, how strong called an accretion disk around the black hole. The gas
repulsive forces can become. The repulsive part of the in the accretion disk is moving very rapidly, at speeds
nuclear force approaches that limit. approaching the speed of light. As a result of the high
speeds, and turbulence in the flow, the particles in the
There is, however, no limit to the strength of attractive accretion disk emit vast quantities of X rays as they
forces. The attractive gravitational force in a neutron spiral down toward the black hole. Strong X ray
star simply depends upon the amount of mass in the emission is thus a signature that an object may be a
star. The more mass, the stronger the force. binary star system with one of the stars being a black
From this we can conclude that we are in serious hole.
trouble if the neutron star gets too big. It is estimated
that in a neutron star with a mass 4 to 6 times the mass
of the sun, the attractive gravitational force will exceed
the repulsive component of the nuclear force, and the
neutrons will begin to collapse into each other.

As the star starts to collapse, gravity gets still stronger.

But gravity has just crushed the strongest known
repulsive force. At some point during the collapse,
gravity will become strong enough to crush any pos-
sible repulsive force. According to the laws of physics,
as we know them, nothing can stop the further collapse
of the star, perhaps down to a mathematical point, or at
least down to a size so small that new laws of physics
take over.

The problem with finding black holes is that, as their

name suggests, they do not emit light. The only way we
have of detecting black holes is by their gravitational
effect on other objects.

It turns out that a good fraction of the stars in the

universe come in binary pairs. Having a pair of stars
form is an effective way of taking up the angular
Figure 8
momentum of a collapsing gas cloud that was initially
Painting of the gas from a blue giant star being
rotating. If Jupiter had been just a bit bigger, igniting sucked into a black hole. (From the May, 1974
its own nuclear reactions, then the earth would have National Geographic, Artist Victor J. Kelley.)
been located in a binary star system.
20-20 Nuclear Matter

Since X rays can be emitted in other ways, a further

check is needed to be sure that a black hole is involved.
By studying the orbit of the visible companion, one can
determine the mass of the invisible one (essentially
using Kepler’s third law, in a slightly modified form).
The test of whether the invisible companion is a black
hole is whether its mass is over 6 solar masses. If it is,
then no dark object could withstand the gravitational
forces involved. The first candidate for an object fitting
this description is the X ray source in the constellation
of Cygnus, an object known as Cygnus X1.

Rather than being scarce, hard–to–find objects, black

holes may play a significant role in the structure of the
universe. There is good evidence, from the study of the
motions of stars, that a gigantic black hole, with a mass
of millions of solar masses, may lie at the heart of our
galaxy and other galaxies as well. And recent studies
have indicated that there may be a black hole at the
center of globular clusters. Wherever black holes may
be, whatever their role in our universe, one fact stands
out—gravity, a force too weak even to be detected on
an atomic scale, can under the right circumstances
become the strongest force of all, strong enough to
crush matter out of existence.
Physics 2000
E. R. Huggins
Dartmouth College

Part 2
E & M,
Quantum Mechanics,
Optics, Calculus

physics2000.com
Chapter 23
Fluid Dynamics

CHAPTER 23 FLUID DYNAMICS

Since the earth is covered by two fluids, air and water, The Current State of Fluid Dynamics
much of our life is spent dealing with the dynamic The ideas that we will discuss here were discovered
behavior of fluids. This is particularly true of the well over a century ago. They are simple ideas that
atmosphere where the weather patterns are governed provide very good predictions in certain restricted
by the interaction of large and small vortex systems, circumstances. In general, fluid flows can become very
that sometime strengthen into fierce systems like torna- complicated with the appearance of turbulent motion.
dos and hurricanes. On a smaller scale our knowledge Only in the twentieth century have we begun to gain
of some basic principles of fluid dynamics allows us to confidence that we have the correct equations to ex-
build airplanes that fly and sailboats that sail into the plain fluid motion. Solving these equations is another
wind. matter and one of the most active research topics in
modern science. Fluid theory has been the test bed of
In this chapter we will discuss only a few of the basic the capability of modern super computers as well as the
concepts of fluid dynamics, the concept of the velocity focus of attention of many theorists. Only a few years
field, of streamlines, Bernoulli’s equation, and the ago, from the work of Lorenz it was discovered that it
basic structure of a well-formed vortex. While these was not possible, even in principle, to make accurate
topics are interesting in their own right, the subject is long-range forecasts of the behavior of fluid systems,
being discussed here to lay the foundation for many of that when you try to predict too far into the future, the
the concepts that we will use in our discussion of chaotic behavior of the system destroys the accuracy of
electric and magnetic phenomena. This chapter is the prediction.
fairly easy reading, but it contains essential material
for our later work. It is not optional. Relative to the current work on fluid behavior, we will
just barely touch the edges of the theory. But even there
we find important basic concepts such as a vector field,
streamlines, and voltage, that will be important through-
out the remainder of the course. We are introducing
these concepts in the context of fluid motion because it
is much easier to visualize the behavior of a fluid than
some of the more exotic fields we will discuss later.
23-2 Fluid Dynamics

THE VELOCITY FIELD As an example of what we will call a velocity field,

Imagine that you are standing on a bridge over a river consider the bathtub vortex shown in Figure (1a). From
looking down at the water flowing underneath you. If the top view the water is going in a nearly circular
it is a shallow stream the flow may be around boulders motion around the vortex core as it spirals down the
and logs, and be marked by the motion of fallen leaves funnel. We have chosen Points A, B, C and D, and at
and specks of foam. In a deep, wide river, the flow each of the points drawn a velocity vector to represent
could be quite smooth, marked only by the eddies that the velocity of the fluid particle at that point. The
trail off from the bridge abutments or the whipping velocity vectors are tangent to the circular path of the
back and forth of small buoys. fluid and vary in size depending on the speed of the
fluid. In a typical vortex the fluid near the core of the
Although the motion of the fluid is often hard to see vortex moves faster than the fluid out near the edge.
directly, the moving leaves and eddies tell you that the This is represented in Figure (1b) by the fact that the
motion is there, and you know that if you stepped into vector at Point D, in near the core, is much longer than
the river, you would be carried along with the water. the one at Point A, out near the edge.

Our first step in constructing a theory of fluid motion is

to describe the motion. At every point in the fluid, we
can think of a small “particle” of fluid moving with a
velocity v . We have to be a bit careful here. If we
vortex with
picture too small a “particle of fluid”, we begin to see
hollow core
individual atoms and the random motion between funnel
atoms. This is too small. On the other hand, if we think
of too big a “particle”, it may have small fluid eddies
inside it and we can’t decide which way this little piece
of fluid is moving. Here we introduce a not completely
justified assumption, namely that there is a scale of
distance, a size of our particle of fluid, where atomic
motions are too small to be seen and any eddies in the
fluid are big enough to carry the entire particle with it.
Figure 1a
With this idealization, we will say that the velocity v of
The "bathtub" vortex is easily seen by filling
the fluid at some point is equal to the velocity of the a glass funnel with water, stirring the water,
particle of fluid that is located at that point. and letting the water flow out of the bottom.

We have just introduced a new concept which we will

call the “velocity field”. At every point in a fluid we
hollow
define a vector v which is the velocity vector of the fluid A
vortex
particle at that point. To formalize the notation a bit,
core
consider the point labeled by the coordinates (x, y, z). D
Then the velocity of the fluid at that point is given by
the vector v (x, y, z), where v (x, y, z) changes as we
B
go from one point to another, from one fluid particle to C
another.

Figure 1b
Looking down from the top, we see the water moving
around in a circular path, with the water near the core
moving faster. The velocity vectors, drawn at four
different points, get longer as we approach the core.
 23-3

The Vector Field vary a bit depending how far down or up you went.
The velocity field, illustrated in Figure (1) is our first (Theoretically the magnitude of g would drop to zero
example of a more general concept called a vector field. at the center of the earth, and drop off as 1/r2 as we went
The idea of a vector field is simply that at every point out away from the earth). This quantity g has a
in space there is a vector with an explicit direction and magnitude and direction at every point, and therefore
magnitude. In the case of the velocity field, the vector qualifies as a vector field. This particular vector field
is the velocity vector of the fluid particle at that point. is called the gravitational field of the earth.
The vector v (x, y, z) points in the direction of motion
of the fluid, and has a magnitude equal to the speed of It is easy to describe how to construct the gravitational
the fluid. field g at every point. Just measure the magnitude and
direction of the gravitational force on a non-accelerated
It is not hard to construct other examples of vector 1 kg mass at every point. What is not so easy is to
fields. Suppose you took a 1 kg mass hung on the end picture the result. One problem is drawing all these
of a spring, and carried it around to different parts of the vectors. In Figure (2) we drew only about five g
earth. At every point on the surface where you stopped vectors. What would we do if we had several million
and measured the gravitational force F = mg = g (for measurements?
m = 1) you would obtain a force vector that points
nearly toward the center of the earth, and has a magni- The gravitational field is a fairly abstract concept—the
tude of about 9.8 m/sec2 as illustrated in Figure (2). If result of a series of specific measurements. You have
you were ambitious and went down into tunnels, or up never seen a gravitational field, and at this point you
on very tall buildings, the vectors would still point have very little intuition about how gravitational fields
toward the center of the earth, but the magnitude would behave (do they “behave”? do they do things?). Later
we will see that they do.

In contrast you have seen fluid motion all your life, and
you have already acquired an extensive intuition about
gravitational the behavior of the velocity field of a fluid. We wish to
g force on a build on this intuition and develop some of the math-
g
one kilogram ematical tools that are effective in describing fluid
g mass motion. Once you see how these mathematical tools
Earth apply to an easily visualized vector field like the
velocity field of a fluid, we will apply these tools to
g g more abstract concepts like the gravitational field we
just mentioned, or more importantly to the electric
field, which is the subject of the next nine chapters.
Figure 2
We can begin to draw a picture of the earth's
gravitational field by carrying a one kilogram
mass around to various points on the surface of
the earth and drawing the vector g representing
the force on that unit mass (m = 1) object.
23-4 Fluid Dynamics

Streamlines
We have already mentioned one problem with vector
water ink
fields—how do you draw or represent so many vec-
tors? A partial answer is through the concept of
streamlines illustrated in Figure (3). In that figure we
have two plates of glass separated by a narrow gap with
object water flowing down through the gap. In order to see the
between path taken by the flowing water, there are two fluid
glass narrow gap
sheets between sheets reservoirs at the top, one containing ink and the other
of glass clear water. The ink and water are fed into the gap in
alternate bands producing the streaks that we see.
glass Inside the gap are a plastic cut-out of both a cylinder and
sheets a cross section of an airplane wing, so that we can
visualize how the fluid flows past these obstacles.
(a) Edge-view of the so called Hele-Shaw cell
The lines drawn by the alternate bands of clear and dark
water are called streamlines. Each band forms a
separate stream, the clear water staying in clear streams
and the inky water in dark streams. What these streams
or streamlines tell us is the direction of motion of the
fluid. Because the streams do not cross and because the
dark fluid does not mix with the light fluid, we know
that the fluid is moving along the streamlines, not
perpendicular to them. In Figure (4) we have sketched
a pair of streamlines and drawn the velocity vectors v1,
v2, v3 and v4 at four points along one of the streams.
(b) Flow around a circular object. What is obvious is that the velocity vector at some point
must be parallel to the streamline at that point, for that
is the way the fluid is flowing. The streamlines give us
a map of the directions of the fluid flow at the various
points in the fluid.

v1
Figure 4
Velocity vectors in a
streamline. Since v2
the fluid is flowing
along the stream, the
velocity vectors are
parallel to the v3
(c) Flow around airplane wing shapes. streamlines. Where
the streamlines are
close together and
Figure 3 the stream becomes Streamlines
In a Hele-Shaw cell, bands of water and ink flow narrow, the fluid
must flow faster and v4
down through a narrow gap between sheets of
glass. With this you can observe the flow around the velocity vectors
different shaped objects placed in the gap. The are longer.
alternate black and clear bands of water and ink
mark the streamlines of the flow.
 23-5

Continuity Equation in the wide sections. This way, the same volume of
When we have a set of streamlines such as that in Figure water per second gets past the constriction as passes per
(4), we have a good idea of the directions of flow. We second past a wide section of the channel. Applying
can draw the direction of the velocity vector at any this idea to Figure (4), we see why the velocity vectors
point by constructing a vector parallel to the streamline are longer, the fluid speed higher, in the narrow sec-
passing through that point. If the streamline we have tions of the streamline channels than in the wide
drawn or photographed does not pass exactly through sections.
that point, then we can do a fairly good job of estimating
the direction from the neighboring streamlines. It is not too hard to go from the qualitative idea that fluid
must flow faster in the narrow sections of a channel, to
But what about the speed of the fluid? Every vector has a quantitative result that allows us to calculate how
both a magnitude and a direction. So far, the stream- much faster. In Figure (5), we are considering a section
lines have told us only the directions of the velocity of streamline or flow tube which has an entrance area
vectors. Can we determine or estimate the fluid speed A1, and exit area A2 as shown. In a short time ∆t, the
at each point so that we can complete our description of fluid at the entrance travels a distance ∆x 1 = v1∆t , while
the velocity field? at the exit the fluid goes a distance ∆x 2 = v2∆t.

When there is construction on an interstate highway The volume of water that entered the stream during the
and the road is narrowed from two lanes to one, the time ∆t is the shaded volume at the left side of the
traffic tends to go slowly through the construction. diagram, and is equal to the area A1 times the distance
This makes sense for traffic safety, but it is just the ∆x 1 that the fluid has moved
wrong way to handle an efficient fluid flow. The traffic
should go faster through the construction to make up Volume of water
= A1∆x 1 = A1v1∆t (1)
for the reduced width of the road. (Can you imagine the entering in ∆t
person with an orange vest holding a sign that says
“Fast”?) Water, when it flows down a tube with a The volume of water leaving the same amount of time
constriction, travels faster through the constriction than is

Volume of water
= A2∆x 2 = A2v2∆t (2)
x2 leaving during ∆t
streamline
v2
A2 If the water does not get squeezed up or compressed
inside the stream between A1 and A2, if we have an
x 2 = v2 t incompressible fluid, which is quite true for water and
x1 in many cases even true for air, then the volume of fluid
A1 v1 entering and the volume of the fluid leaving during the
x1 = v1 t time ∆t must be equal. Equating Equations (1) and (2)
and cancelling the ∆t gives
Figure 5
During the time ∆ t , water entering the small section
of pipe travels a distance v1∆ t , while water leaving A1v1 = A2v2 continuity equation (3)
the large section goes a distance v2 ∆ t . Since the
same amount of water must enter as leave, the
entrance volume A 1 ∆ x 1 must equal the exit volume Equation (3) is known as the continuity equation for
A2 ∆ x 2. This gives A 1 v1∆ t = A2 v2∆ t , or the result incompressible fluids. It is a statement that we do not
A 1 v1 = A2 v2 which is one form of the continuity squeeze up or lose any fluid in the stream. It also tells
equation. us that the velocity of the fluid is inversely proportional
23-6 Fluid Dynamics

to the cross sectional area of the stream at that point. If Velocity Field of a Point Source
the cross sectional area in a constriction has been cut in This is an artificial example that shows us how to apply
half, then the speed of the water must double in order the continuity equation in a somewhat unexpected
to get the fluid through the constriction. way, and leads to some ideas that will be very important
in our later discussion of electric fields.
If we have a map of the streamlines, and know the
entrance speed v1 of the fluid, then we can determine For this example, imagine a small magic sphere that
the magnitude and direction of the fluid velocity v2 at creates water molecules inside and lets the water
any point downstream. The direction of v2 is parallel molecules flow out through the surface of the sphere.
to the streamline at Point (2), and the magnitude is (Or there may be an unseen hose that supplies the water
given by v2 = v1 (A1/A2). Thus a careful map of the that flows out through the surface of the sphere.)
fluid streamlines, combined with the continuity equa-
tion, give us almost a complete picture of the fluid Let the small sphere have a radius r1, area 4π r12 and
motion. The only additional information we need is the assume that the water is emerging radially out through
entrance speed. the small sphere at a speed v1 as shown in Figure (6).
Also let us picture that the small sphere is at the center
of a huge swimming pool full of water, that the sides of
the pool are so far away that the water continues to flow
radially outward at least for several meters. Now
conceptually construct a second sphere of radius
r2 > r1 centered on the small sphere as in Figure (6).

During one second, the volume of water flowing out of

v2
the small sphere is v1A 1, corresponding to ∆t = 1 sec in
Equation (1). By the continuity equation, the volume
of water flowing out through the second sphere in one
v1
point source second, v2 A 2, must be the same in order that no water
of water piles up between the spheres. Using the fact that
v1
v2 r1
v2 A1 = 4π r12 and A2 = 4π r22, we get
v1
continuity
r2 v1 A1 = v2 A2 equation

v1 4π r12 = v2 4π r22
v1 v1
v2 = 12 v1r12 (4a)
r2
v2 v2
Equation 4a tells us that as we go out from the "magic
Figure 6 sphere", as the distance r2 increases, the velocity v2
Point source of water. Imagine that water drops off as the inverse square of r2 , as 1 r22 . We can
molecules are created inside the small sphere write this relationship in the form
and flow radially out through its surface at a
speed v1 . The same molecules will eventually 1 The symbol
flow out through the larger sphere at a lesser v2 ∝ ∝ means (4b)
speed v2 . If no water molecules are created or r22 "proportional to"
destroyed outside the small sphere, then the
continuity equation A 1 v1 = A2 v2 requires A small spherical source like that shown in Figure (6)
that 4 π r 12 v1 = 4 π r 22 v2 . is often called a point source. We see that a point
source of water produces a 1 r2 velocity field.
 23-7

Velocity Field of a Line Source

One more example which we will often use later is the
line source. This is much easier to construct than the
point source where we had to create water molecules.
Good models for a line source of water are the sprin- line source v2
of water
kling hoses used to water gardens. These hoses have a
series of small holes that let the water flow radially
outward. v2 v1 v2
v1 r1 v1
For this example, imagine that we have a long sprinkler
r2
hose running down the center of an immense swim-
ming pool. In Figure (7) we are looking at a cross v1 v1
section of the hose and see a radial flow that looks very
much like Figure (6). The side view, however, is v2 v2
different. Here we see that we are dealing with a line
rather than a point source of water.
a) End view of line source
Consider a section of the hose and fluid of length L.
The volume of water flowing in one second out through
this section of hose is v1 A1 where A 1 = L times (the v2 v2 v2 v2 v2
circumference of the hose) = L(2πr1).
Volume of
water / sec = v1A1 = v1L2π r 1 v1 v1 v1 v1 v1
from a section
L of hose r1 r2
If the swimming pool is big enough so that this water
continues to flow radially out through a cylindrical area
A2 concentric with and surrounding the hose, then the v1 v1 v1 v1 v1
volume of water per second (we will call this the “flux”
of water) out through A2 is
v2 v2 v2 v2 v2
Volume of
water / sec out = v2 A2 = v2 2πr2L L
through A2
Using the continuity equation to equate these volumes b) Side view of line source
of water per second gives Figure 7
v1A1 = v2A2 Line source of water. In a line source, the
water flows radially outward through a
v1L2π r1 = v2L2π r2 ; v1 r1 = v2 r2 cylindrical area whose length we choose as
L and whose circumference is 2 π r .
v1r1 1
v2 = ∝
r2 r2 (5)

We see that the velocity field of a line source drops off

as 1/r rather than 1/r2 which we got from a point source.
23-8 Fluid Dynamics

FLUX The concept of flux can be generalized to irregular

Sometimes simply changing the name of a quantity flows and irregularly shaped surfaces. To handle that
leads us to new ways of thinking about it. In this case case, break the flow up into a bunch of small flow tubes
we are going to use the word flux to describe the separated by streamlines, construct a perpendicular
amount of water flowing per second out of some area for each flow tube as shown in Figure (9), and then
volume. From the examples we have considered, the calculate the total flux by adding up the fluxes from
flux of water out through volumes V1 and V2 are given each flow tube.
by the formulas Total Flux = v1A1 + v2A2 + v2A2 +...
(6)
Flux of Volume of
water flowing
= Σi viA⊥i
water ≡ per second
= v1A1
out of V1 In the really messy cases, the sum over flow tubes
out of V1
becomes an integral as we take the limit of a large
Flux of number of infinitesimal flow tubes.
water = v2A2
out of V2 For this text, we have gone too far. We will not work
with very complicated flows. We can learn all we want
The continuity equation can be restated by saying that from the simple ones like the flow out of a sphere or a
the flux of water out of V1 must equal the flux out of V2 cylinder. In those cases the perpendicular area is
if the water does not get lost or compressed as it flows obvious and the flux easy to calculate. For the spherical
from the inner to the outer surface. flow of Figure (6), we see that the velocity field
dropped off as 1/r2 as we went out from the center of the
So far we have chosen simple surfaces, a sphere and a sphere. For the cylinder in Figure (7) the velocity field
cylinder, and for these surfaces the flux of water is dropped off less rapidly, as 1/r.
simply the fluid speed v times the area out through
which it is flowing. Note that for our cylindrical surface
A1 v1
shown in Figure (8), no water is flowing out through the
ends of the cylinder, thus only the outside area (2πrL)
counted in our calculation of flux. A more general way A2 v2
of stating how we calculate flux is to say that it is the
fluid speed v times the perpendicular area A ⊥ through A3
which the fluid is flowing. For the cylinder, the v3
perpendicular area A ⊥ is the outside area (2πrL); the A4
ends of the cylinder are parallel to the flow and there-
fore do not count.
Figure 9
To calculate the flux of water in an arbitrarily v4
v v v v
no water shaped flow break up the flow into many small flux
r flows out tubes where the fluid velocity is essentially uniform
L
through across the small tube as shown. The flux through
the end the i-th tube is simply vi A i , and the total flux is the
v v v v
sum of the fluxes Σ i vi A i through each tube.

Figure 8
With a line source, all the water flows through the
cylindrical surface surrounding the source and none
through the ends. Thus A ⊥ , the perpendicular area
through which the water flows is 2 π r × L .
 23-9

BERNOULLI’S EQUATION Consider a small tube of flow bounded by streamlines

Our discussion of flux was fairly lengthy, not so much as shown in Figure (10). In a short time ∆t a small
for the results we got, but to establish concepts that we volume of fluid enters on the left and an equal volume
will use extensively later on in our discussion of electric exits on the right. If the exiting volume has more
fields. Another topic, Bernoulli’s law, has a much energy than the entering volume, the extra energy had
more direct application to the understanding of fluid to come from the work done by pressure forces acting
flows. It also has some rather surprising consequences on the fluid in the flow tube. Equating the work done
which help explain why airplanes can fly and how a by the pressure forces to the increase in energy gives us
sailboat can sail up into the wind. Bernoulli's equation.

Bernoulli’s law involves an energy relationship be- To help visualize the situation, imagine that the stream-
tween the pressure, the height, and the velocity of a line boundaries of the flow tube are replaced by fric-
fluid. The theorem assumes that we have a constant tionless, rigid walls. This would have no effect on the
density fluid moving with a steady flow, and that flow of the fluid, but focuses our attention on the ends
viscous effects are negligible, as they often are for of the tube where the fluid is flowing in on the left, at
fluids such as air and water. what we will call Point (1), and out on the right at Point
(2).

x2 = v2 t

flow tube
bounded by
streamlines h2 A2 2 F2
F2 = P2 A2
volume of
x1 = v1 t water exiting
during time t
F1 1 A1 h1
F1 = P1 A1 volume of
water entering
during time t
Figure 10
Derivation of Bernoulli's equation. Select a flow tube bounded by streamlines. For the steady
flow of an incompressible fluid, during a time ∆ t the same volume of fluid must enter on the
left as leave on the right. If the exiting fluid has more energy than the entering fluid, the
increase must be a result of the net work done by the pressure forces acting on the fluid.
23-10 Fluid Dynamics

As a further aid to visualization, imagine that a small Thus the fluid inside the flow tube is subject to external
frictionless cylinder is temporarily inserted into the pressure forces, F1 in from the left and F2 in from the
entrance of the tube as shown in Figure (11a), and at the right. During a time ∆t, the fluid at the entrance moves
exit as shown in Figure (11b). Such cylinders have no a distance ∆x 1 = v1∆t as shown in Figure (10). While
effect on the flow but help us picture the pressure moving this distance, the entering fluid is subject to the
forces. pressure force F1, thus the work ∆W1 done by the
pressure force at the entrance is
At the entrance, if the fluid pressure is P1 and the area
of the cylinder is A 1, then the external fluid exerts a net ∆W1 = F1 ⋅∆x 1 = P1A 1 v 1∆t (8a)
force of magnitude
At the exit, the fluid moves out a distance ∆x 2 = v2∆t,
F1 = P1A 1 (7a) while the external force pushes back in with a pressure
directed perpendicular to the surface of the cylinder as force F2. Thus the pressure forces do negative work on
shown. We can think of this force F1 as the pressure the inside fluid, with the result
force that the outside fluid exerts on the fluid inside the
flow tube. At the exit, the external fluid exerts a ∆W2 = F2 ⋅∆x2 = – P2A2 v 2∆t (8b)
pressure force F2 of magnitude
The net work ∆W done during a time ∆t by external
F2 = P2A 2 (7b) pressure forces on fluid inside the flow tube is therefore
directed perpendicular to the piston, i.e., back toward
∆W = ∆W1 + ∆W2
the fluid inside the tube. (9)
= P1 A1v1∆t – P2 A2v2 ∆t

entrance to Equation (9) can be simplified by noting that

flow tube A 1v1∆t = A 1∆x 1 is the volume ∆V1 of the entering
fluid. Likewise A 2v2∆t = A 2∆x 2 is the volume ∆V2
a) F1 internal fluid of the exiting fluid. But during ∆t, the same volume
P1 ∆V of fluid enters and leaves, thus ∆V1 = ∆V2 = ∆V
A1 and we can write Equation (9) as

exit from
flow tube
work done
by external
pressure
forces ∆W = ∆V P1 – P2 (10)
b) internal fluid F2 on fluid
inside flow
P2 tube
A2
The next step is to calculate the change in energy of the
Figure 11
entering and exiting volumes of fluid. The energy ∆E 1
The flow would be unchanged if we temporarily
inserted frictionless pistons at the entrance and exit. of the entering fluid is its kinetic energy 1 2 ∆m v12
plus its gravitational potential energy ∆m gh1 , where
∆m is the mass of the entering fluid. If the fluid has a
density ρ, then ∆m = ρ∆V and we get
 23-11

∆E1 = 1 ρ∆V v12 + ρ∆V gh1 In this form, an interpretation of Bernoulli’s equation
2 begins to emerge. We see that the quantity
(11a)
P + ρgh + 1 ρv 2 has the same numerical value at the
= ∆V 1 ρv12 + ρgh1 2
2 entrance, Point (1), as at the exit, Point (2). Since we
can move the starting and ending points anywhere
At the exit, the same mass and volume of fluid leave in along the flow tube, we have the more general result
time ∆t, and the energy of the exiting fluid is

∆E 2 = ∆V 1 ρv22 + ρgh2 (11b)

2 constant anywhere
The change ∆E in the energy in going from the P + ρgh + 1 ρv 2 = along a flow tube
2
entrance to the exit is therefore or streamline

∆E = ∆E2 – ∆E1
(15)
= ∆V 1 ρv22 + ρgh2 – 1 ρv12 – ρgh1 (12) Equation (15) is our final statement of Bernoulli’s
2 2
equation. In words it says that for the steady flow of an
Equating the work done, Equation (10) to the change in incompressible, non viscous, fluid, the quantity
energy, Equation (12) gives P + ρgh + 1 2ρv
2 2 has a constant value along a
streamline.
∆V P1 – P2
The restriction that P + ρgh + 1 2ρv 2 2 is constant
= ∆V 1 ρv22 + ρgh2 – 1 ρv12 – ρgh1 (13) along a streamline has to be taken seriously. Our
2 2 derivation applied energy conservation to a plug mov-
Not only can we cancel the ∆Vs in Equation (13), but ing along a small flow tube whose boundaries are
we can rearrange the terms to make the result easier to streamlines. We did not consider plugs of fluid moving
remember. We get in different flow tubes, i.e., along different streamlines.
For some special flows, the quantity
P + ρgh + 1 2ρv
2 2 has the same value throughout
the entire fluid. But for most flows,
P1 + 1 ρv12 + ρgh1 = P2 + 1 ρv22 + ρgh2 (14) P + ρgh + 1 2ρv
2 2 has different values on different
2 2
streamlines. Since we haven’t told you what the special
flows are, play it safe and assume that the numerical
value of P + ρgh + 1 2ρv 2 2 can change when you
hop from one streamline to another.
23-12 Fluid Dynamics

APPLICATIONS OF Pat + ρg 0 = constant

BERNOULLI’S EQUATION
and the constant is Pat. For any depth y = –h, we have
Bernoulli’s equation is a rather remarkable result that
some quantity P + ρgh + 1 2ρv 2 2 has a value that P – ρgy = constant = P at
doesn’t change as you go along a streamline. The terms
inside, except for the P term, look like the energy of a
unit volume of fluid. The P term came from the work P = Pat + ρgy (17)
part of the energy conservation theorem, and cannot
strictly be interpreted as some kind of pressure energy. We see that the increase in pressure at a depth y is ρgy,
As tempting as it is to try to give an interpretation to the a well-known result from hydrostatics.
terms in Bernoulli’s equation, we will put that off for a
while until we have worked out some practical applica- Exercise 1
tions of the formula. Once you see how much the
The density of water is ρ = 103Kg/m3 and atmospheric
equation can do, you will have a greater incentive to pressure is Pat = 1.0 ×105N/ m2 . At what depth does a
develop an interpretation. scuba diver breath air at a pressure of 2 atmospheres?
(At what depth does ρgy = Pat ?) (Your answer should
Hydrostatics be 10.2m or 33 ft.)
Let us start with the simplest application of Bernoulli’s
Exercise 2
equation, namely the case where the fluid is at rest. In
a sense, all the fluid is on the same streamline, and we What is the pressure, in atmospheres, at the deepest
part of the ocean? (At a depth of 8 kilometers.)
have

constant
P + ρgh = throughout (16) Leaky Tank
the fluid For a slightly more challenging example, suppose we
have a tank filled with water as shown in Figure (13).
Suppose we have a tank of water shown in Figure (12). A distance h below the surface of the tank we drill a hole
Let the pressure be atmospheric pressure at the surface, and the water runs out of the hole at a speed v. Use
and set h = 0 at the surface. Therefore at the surface Bernoulli’s equation to determine the speed v of the
exiting water.
Pat Pat
h=0 (1)
h1

water y water
h
P h = –y
streamline
Pat V2
h2
(2)

Figure 12 Figure 13
Hydrostatic pressure at a depth y is Water squirting out through a hole in a leaky tank. A
atmospheric pressure plus ρ gy. streamline connects the leak at Point (2) with some
Point (1) on the surface. Bernoulli's equation tells us
that the water squirts out at the same speed it would
have if it had fallen a height h.
 23-13

Solution: Somewhere there will be a streamline Arguing that the fluid at Point (1) on the top and Point
connecting the free surface of the water (1) to a Point (2) on the bottom started out on essentially the same
(2) in the exiting stream. Applying Bernoulli's equa- streamline (Point 0), we can apply Bernoulli’s equa-
tion to Points (1) and (2) gives tion to Points (1) and (2) with the result

P1 + ρgh 1 + 1 ρv12 = P2 + ρgh 2 + 1 ρv22 P1 + 1 ρv12 + ρgh 1 = P2 + 1 ρv22 + ρgh 2

2 2 2 2
Now P1 = P2 = Pat , so the Ps cancel. The water level We have crossed out the ρgh terms because the differ-
in the tank is dropping very slowly, so that we can set ence in hydrostatic pressure ρgh across the wing is
v1 = 0. Finally h1 – h2 = h, and we get negligible for a light fluid like air.
1 ρv 2 = ρg h – h = ρgh (18) Here is the important observation. Since the fluid speed
2 2 1 2
v1 at the top of the wing is higher than the speed v2 at the
The result is that the water coming out of the hole is bottom, the pressure P2 at the bottom must be greater
moving just as fast as it would if it had fallen freely from than P1 at the top in order that the sum of the two terms
the top surface to the hole we drilled. 2 2 be the same. The extra pressure on the
P + 1 2ρv
bottom of the wing is what provides the lift that keeps
the airplane up in the air.
Airplane Wing
In the example of a leaky tank, Bernoulli’s equation There are two obvious criticisms of the above explana-
gives a reasonable, not too exciting result. You might tion of how airplanes get lift. What about stunt pilots
have guessed the answer by saying energy should be who fly upside down? And how do balsa wood gliders
conserved. Now we will consider some examples that with flat wings fly? The answer lies in the fact that the
are more surprising than intuitive. The first explains shape of the wing cross-section is only one of several
how an airplane can stay up in the air. important factors determining the flow pattern around
a wing.
Figure (14) shows the cross section of a typical airplane
wing and some streamlines for a typical flow of fluid Figure (15) is a sketch of the flow pattern around a flat
around the wing. (We copied the streamlines from our wing flying with a small angle of attack θ. By having
demonstration in Figure 3). an angle of attack, the wing creates a flow pattern where
the streamlines around the top of the wing are longer
The wing is purposely designed so that the fluid has to than those under the bottom. The result is that the fluid
flow farther to get over the top of the wing than it does flows faster over the top, therefore the pressure must be
to flow across the bottom. To travel this greater lower at the top (higher at the bottom) and we still get
distance, the fluid has to move faster on the top of the lift. The stunt pilot flying upside down must fly with a
wing (at Point 1), than at the bottom (at Point 2). great enough angle of attack to overcome any down-
ward lift designed into the wing.
(1)

airplane wing θ
(0)
(2)

Figure 14 Figure 15
Streamline flow around an airplane wing. The wing is A balsa wood model plane gets lift by having the wing
shaped so that the fluid flows faster over the top of the move forward with an upward tilt, or angle of attack.
wing, Point (1) than underneath, Point (2). As a result The flow pattern around the tilted wing gives rise to a
the pressure is higher beneath Point (2) than above faster flow and therefore reduced pressure over the top.
Point (1).
23-14 Fluid Dynamics

Sailboats The operation of a sailboat is easily demonstrated using

Sailboats rely on Bernoulli’s principle not only to an air cart, glider and fan. Mount a small sail on top of
supply the “lift” force that allows the boat to sail into the the air cart glider (the light plastic shopping bags make
wind, but also to create the “wing” itself. Figure (16) excellent sail material) and elevate one end of the cart
is a sketch of a sailboat heading at an angle θ off from as shown in Figure (17) so that the cart rests at the low
the wind. If the sail has the shape shown, it looks like end. Then mount a fan so that the wind blows down and
the airplane wing of Figure (14), the air will be moving across as shown. With a little adjustment of the angle
faster over the outside curve of the sail (Position 1) than of the fan and the tilt of the air cart, you can observe the
the inside (Position 2), and we get a higher pressure on cart sail up the track, into the wind.
the inside of the sail. This higher pressure on the inside
both pushes the sail cloth out to give the sail an airplane If you get the opportunity to sail a boat, remember that
wing shape, and creates the lift force shown in the it is the Bernoulli effect that both shapes the sail and
diagram. This lift force has two components. One pulls propels the boat. Try to adjust the sail so that it has a
the boat forward. The other component , however, good airplane wing shape, and remember that the
tends to drag the boat sideways. To prevent the boat higher speed wind on the outside of the sail creates a
from slipping sideways, sailboats are equipped with a low pressure that sucks the sailboat forward. You’ll go
centerboard or a keel. faster if you keep these principles in mind.

light plastic
sail
tilted air track
air cart
θ

(0)
lift force
forward
component
sail of lift
(2) (1)
fan

post

sail

string
Figure 16
A properly designed sail takes on the shape of an
airplane wing with the wind traveling faster,
creating a lower pressure on the outside of the
sail (Point 1). This low pressure on the outside Figure 17
both sucks the canvas out to maintain the shape Sailboat demonstration. It is easy to rig a mast on an
of the sail and provides the lift force. The air cart, and use a small piece of a light plastic bag for
forward component of the lift force moves the a sail. Place the cart on a tilted air track so that the
boat forward and the sideways component is cart will naturally fall backward. Then turn on a fan
offset by the water acting on the keel. as shown, and the cart sails up the track into the wind.
 23-15

The Venturi Meter cause the fluid level in barometer (B) to be lower than
Another example, often advertised as a simple applica- in the barometer over the slowly moving, high pressure
tion of Bernoulli’s equation, is the Venturi meter stream. The height difference h means that there is a
shown in Figure (18). We have a tube with a constric- pressure difference
tion, so that its cross-sectional area A1 at the entrance
and the exit, is reduced to A2 at the constriction. By the Pressure = P – P = ρgh (20)
difference 1 2
continuity equation (3), we have
v1A 1 If we combine Equations (19) and (20), ρ cancels and
v1A 1 = v2A 2; v2 = we can solve for the speed v1 of the fluid in the tube in
A2
terms of the quantities g, h, A1 and A2. The result is
As expected, the fluid travels faster through the con-
striction since A1 > A2.
2gh
Now apply Bernoulli’s equation to Points (1) and (2). v1 = (21)
A21 A22 –1
Since these points are at the same height, the ρgh terms
cancel and Bernoulli’s equation becomes
Because we can determine the speed v1 of the main
P1 + 1 ρv12 = P2 + 1 ρv22 flow by measuring the height difference h of the two
2 2
Since v2 > v1, the pressure P2 in the constriction must be columns of fluid, the setup in Figure (18) forms the
less than the pressure P1 in the main part of the tube. basis of an often used meter to measure fluid flows. A
Using v2 = v1A 1 /A 2 , we get meter based on this principle is called a Venturi meter.
Pressure
drop in = P1 – P2 Exercise 3
constriction Show that all the terms in Bernoulli's equation have the
same dimensions. (Use MKS units.)
= 1 ρ v22 – v12
2
Exercise 4
= 1 ρ v12 A21 A22 – v12 (19) In a classroom demonstration of a venturi meter shown
2
in Figure (18a), the inlet and outlet pipes had diameters
= 1 ρv12 A21 A22 – 1 of 2 cm and the constriction a diameter of 1 cm. For a
2 certain flow, we noted that the height difference h in the
To observe the pressure drop, we can mount small barometer tubes was 7 cm. How fast, in meters/sec,
tubes (A) and (B) as shown in Figure (18), to act as was the fluid flowing in the inlet pipe?
barometers. The lower pressure in the constriction will
(A) (B)

(3)
h
(4)

v1 v2
(1)
(2)
A2
A1
Figure 18
Venturi meter. Since the water flows faster through
the constriction, the pressure is lower there. By using
vertical tubes to measure the pressure drop, and using Figure 18a
Bernoulli's equation and the continuity equation, you Venturi demonstration. We see about a 7cm drop in
can determine the flow speeds v1 and v2 . the height of the barometer tubes at the constriction.
23-16 Fluid Dynamics

The Aspirator Care in Applying Bernoulli’s Equation

In Figure (18), the faster we move the fluid through the Although the Venturi meter and aspirator are often
constriction (the greater v1 and therefore v2), the greater used as simple examples of Bernoulli’s equation, con-
the height difference h in the two barometer columns. siderable care must be used in applying Bernoulli’s
If we turn v1 up high enough, the fluid is moving so fast equation in these examples. To illustrate the trouble
through section 2 that the pressure becomes negative you can get into, suppose you tried to apply Bernoulli’s
and we get suction in barometer 2. For even higher equation to Points (3) and (4) of Figure (20). You
speed flows, the suction at the constriction becomes would write
quite strong and we have effectively created a crude
vacuum pump called an aspirator. Typically aspira- P3 + ρgh 3 + 1 ρv32 = P4 + ρgh 4 + 1 ρv42
2 2
tors like that shown in Figure (19) are mounted on cold
water faucets in chemistry labs and are used for sucking (22)
up various kinds of fluids. Now P3 = P4 = Patmosphere because Points (3) and (4)
suction
are at the liquid surface. In addition the fluid is at rest
in tubes (3) and (4), therefore v3 = v4 = 0. Therefore
Bernoulli’s equation predicts that
v1 v2
ρgh 3 = ρgh 4
negative
pressure or that h3 = h4 and there should be no height difference!
Figure 19a
If the water flows through the constriction What went wrong? The mistake results from the fact
fast enough, you get a negative pressure that no streamlines go from position (3) to position (4),
and suction in the attached tube. and therefore Bernoulli’s equation does not have to
apply. As shown in Figure (21) the streamlines flow
across the bottom of the barometer tubes but do not go
water
up into them. It turns out that we cannot apply
faucet Bernoulli’s equation across this break in the stream-
spigot lines. It requires some experience or a more advanced
aspirator
knowledge of hydrodynamic theory to know that you
can treat the little tubes as barometers and get the

(3)
suction
(4)
Figure 19b
If the constriction is placed on the end of a water faucet
as shown, you have a device called an aspirator that is
often used in chemistry labs for sucking up fluids. v1 v2

Figure 20
If you try to apply Bernoulli's equation to Points (3)
and (4), you predict, incorrectly, that Points (3) and
(4) should be at the same height. The error is that
Points (3) and (4) do not lie on the same streamline,
and therefore you cannot apply Bernoulli's equation
to them.
 23-17

Hydrodynamic Voltage
(3)
When we studied the motion of a projectile, we found
barometer tube fluid at rest
that the quantity (1/2 mv2 + mgh) did not change as the
moving fluid
ball moved along its parabolic trajectory. When physi-
streamlines
cists discover a quantity like (1/2 mv2 + mgh) that does
not change, they give that quantity a name, in this case
“the ball’s total energy”, and then say that they have
Figure 21
discovered a new law, namely “the ball’s total energy
The water flows past the bottom of the barometer tube,
not up into the tube. Thus Point (3) is not connected to is conserved as the ball moves along its trajectory”.
any of the streamlines in the flow. The vertical tube
acts essentially as a barometer, measuring the pressure With Bernoulli’s equation we have a quantity
of the fluid flowing beneath it. P + ρgh + 1 2ρv
2 2 which is constant along a stream-
line when we have the steady flow of an incompress-
correct answer. Most texts ignore this complication, ible, non viscous fluid. Here we have a quantity
but there are always some students who are clever P + ρgh + 1 2ρv
2 2 that is conserved under special
enough to try to apply Bernoulli’s equation across the circumstances; perhaps we should give this quantity a
break in the flow at the bottom of the small tubes and name also.
then wonder why they do not get reasonable answers.
The term ρgh is the gravitational potential energy of a
There is a remarkable fluid called superfluid helium unit volume of the fluid, and 1 2ρv 2 2 is the same
which under certain circumstances will not have a volume’s kinetic energy. Thus our Bernoulli term has
break in the flow at the base of the barometer tubes. the dimensions and characteristics of the energy of a
(Superfluid helium is liquefied helium gas cooled to a unit volume of fluid. But the pressure term, which
temperature below 2.17 ° K). As shown in Figure (22) came from the work part of the derivation of Bernoulli’s
the streamlines actually go up into the barometer tubes, equation, is not a real energy term. There is no pressure
Points (3) and (4) are connected by a streamline, energy P stored in an incompressible fluid, and
Bernoulli’s equation should apply and we should get Bernoulli’s equation is not truly a statement of energy
no height difference. This experiment was performed conservation for a unit volume of fluid.
in 1965 by Robert Meservey and the heights in the two
barometer tubes were just the same! However, as we have seen, the Bernoulli term is a
useful concept, and deserves a name. Once we name
it, we can say that “ ” is conserved along a streamline
under the right circumstances. Surprisingly there is not
(3) streamlines (4) an extensive tradition for giving the Bernoulli term a
go up
into tubes name so that we have to concoct a name here. At this
point our choice of name will seem a bit peculiar, but
it is chosen with later discussions in mind. We will call
the Bernoulli term hydrodynamic voltage
superfluid helium

Hydrodynamic
≡ P + ρgh + 1 ρv 2 (23)
Figure 22 Voltage 2
In superfluid helium, the streamlines actually go
up into the barometer tubes and Bernoulli's
equation can be applied to Points (3) and (4). and Bernoulli’s equation states that the hydrodynamic
The result is that the heights of the fluid are the voltage of an incompressible, non viscous fluid is
same as predicted. (Experiment by R. Meservey, constant along a streamline when the flow is steady.
see Physics Of Fluids, July 1965.)
23-18 Fluid Dynamics

We obviously did not invent the word voltage; the Town Water Supply
name is commonly used in discussing electrical de- One of the familiar sights in towns where there are no
vices like high voltage wires and low voltage batteries. nearby hills is the water tank somewhat crudely illus-
It turns out that there is a precise analogy between the trated in Figure (23). Water is pumped from the
concept of voltage used in electricity theory, and the reservoir into the tank to fill the tank up to a height h
Bernoulli term we have been discussing. To empha- as shown.
size the analogy, we are naming the Bernoulli term
hydrodynamic voltage. The word “hydrodynamic” is For now let us assume that all the pipes attached to the
included to remind us that we are missing some of the tanks are relatively large and frictionless so that we can
electrical terms in a more general definition of voltage. neglect viscous effects and apply Bernoulli’s equation
We are discussing hydrodynamic voltage before elec- to the water at the various points along the water
trical voltage because hydrodynamic voltage involves system. At Point (1), the pressure is simply atmo-
fluid concepts that are more familiar, easier to visualize spheric pressure Pat, the water is essentially not flow-
and study, than the corresponding electrical concepts. ing, and the hydrodynamic voltage consists mainly of
Pat plus the gravitational term gh 1

Hydrodynamic
= Pat + gh 1
Voltage 1
By placing the tank high up in the air, the gh1 term can
be made quite large. We can say that the tank gives us
“high voltage” water.
water
water tank spraying
(1) (4) up

h1

(2) (3) (5)

hole in pipe
Figure 23
The pressure in the town water supply may be
maintained by pumping water into a water tank as
shown. If the pipes are big enough we can neglect the
viscous effect and apply Bernoulli's equation
throughout the system, including the break in the water
pipe at Point (3), and the top of the fountain, Point (4).
 23-19

Bernoulli’s equation tells us that the hydrodynamic Viscous Effects

voltage of the water is the same at all the points along We said that the hydrodynamic analogy of voltage
the water system. The purpose of the water tank is to involves familiar concepts. Sometimes the concepts
ensure that we have high voltage water throughout the are too familiar. Has your shower suddenly turned cold
town. For example, at Point (2) at one of the closed when someone in the kitchen drew hot water for
faucets in the second house, there is no height left washing dishes; or turned hot when the toilet was
(h2 = 0) and the water is not flowing. Thus all the flushed? Or been reduced to a trickle when the laundry
voltage shows up as high pressure at the faucet. was being washed? In all of these cases there was a
pressure drop at the shower head of either the hot water,
Hydrodynamic the cold water, or both. A pressure drop means that you
= P2
Voltage 2 are getting lower voltage water at the shower head than
At Point (3) we have a break in the pipe and water is was supplied by the town water tank (or by your home
squirting up. Just above the break the pressure has pressure tank).
dropped to atmospheric pressure and there is still no
The hydrodynamic voltage drop results from the fact
height. At this point the voltage appears mainly in the
that you are trying to draw too much water through
form of kinetic energy.
small pipes, viscous forces become important, and
Hydrodynamic Bernoulli’s equation no longer applies. Viscous forces
= Pat + 1 ρv 2 always cause a drop in the hydrodynamic voltage. This
Voltage 3 2
voltage drop can be seen in a classroom demonstration,
Finally at Point (4) the water from the break reaches its Figure (24), where we have inserted a series of small
maximum height and comes to rest before falling down barometer tubes in a relatively small flow tube. If we
again. Here it has no kinetic energy, the pressure is still run a relatively high speed stream of water through the
atmospheric, and the hydrodynamic voltage is back in flow tube, viscous effects become observable and the
the form of gravitational potential energy. If no voltage pressure drops as the water flows down the tube. The
has been lost, if Bernoulli’s equation still holds, then pressure drop is made clear by the decreasing heights
the water at Point (4) must rise to the same height as the of the water in the barometer tubes as we go down-
water at the surface in the town water tank. stream.

In some sense, the town water tank serves as a huge

“battery” to supply the hydrodynamic voltage for the
town water system.

Heights in barometer tubes

dropping due to viscosity

Figure 24
If we have a fairly fast flow in a fairly small tube,
viscosity causes a pressure drop, or as we are
calling it, a "hydrodynamic voltage" drop down the
tube. This voltage drop is seen in the decreasing
heights of the water in the barometer tubes.
(In our Venturi demonstration of Figure (18a), the
heights are lower on the exit side than the entrance
Figure 18a repeated side due to viscosity acting in the constriction.)
Venturi demonstration.
23-20 Fluid Dynamics

VORTICES Outside the core, the fluid goes around in a circular

The flows we have been considering, water in a pipe, pattern, the speed decreasing as the distance from the
air past a sailboat sail, are tame compared to a striking center increases. It turns out that viscous effects are
phenomena seen naturally in the form of hurricanes minimized if the fluid speed drops off as 1/r where r is
and tornados. These are examples of a fluid motion the radial distance from the center of the core as shown
called a vortex. They are an extension, to an atmo- in Figure (26). At some distance from the center, the
spheric scale, of the common bathtub vortex like the speed drops to below the speed of other local distur-
one we created in the funnel seen in Figure (25). bances and we no longer see the organized motion.

Vortices have a fairly well-defined structure which is The tendency of a fluid to try to maintain a 1/r velocity
seen most dramatically in the case of the tornado (see field explains why vortices have to have a core. You
Figures 29 and 30). At the center of the vortex is the cannot maintain a 1/r velocity field down to r = 0, for
core. The core of a bathtub vortex is the hollow tube of then you would have infinite velocities at the center. To
air that goes down the drain. In a tornado or water avoid this problem, the vortex either throws the fluid
spout, the core is the rapidly rotating air. For a out of the core, as in the case of the hollow bathtub
hurricane it is the eye, seen in Figures (27) and (28), vortex, or has the fluid in the core move as a solid
which can be amazingly calm and serene considering rotating object (vθ = r ω) in the case of a tornado, or
the vicious winds and rain just outside the eye. has a calm fluid when the core is large (i.e., viscous
effects of the land are important) as in the case of a
hurricane.

While the tornado is a very well organized example of

a vortex, it has been difficult to do precise measure-
ments of the wind speeds in a tornado. One of the best
measurements verifying the 1/r velocity field was
when a tornado hit a lumber yard, and a television
station using a helicopter recorded the motion of sheets
of 4' by 8' plywood that were scattered by the tornado.
(Using doppler radar, a wind speed of 318 miles per
hour was recorded in a tornado that struck Oklahoma
city on May 3, 1999--a world wind speed record!)

Figure 25
Bathtub vortex in a funnel. We stirred
the water before letting it drain out.

vθ ∝ 1r

core

Figure 27
Figure 26
Eye of hurricane Allen viewed from a satellite.
Vortices tend to have a circular velocity field about
(Photograph courtesy of A. F. Haasler.)
the core, a velocity field vθ whose strength tends to
drop off as 1/r as you go out from the core.
 23-21

Figure 29 Figure 30
Tornado in Kansas. A tornado over water is called a water spout.

Figure 28
Hurricane
approaching
the east
coast of
the U.S.
23-22 Fluid Dynamics

Quantized Vortices in Superfluids Exercise 5

For precision, nothing beats the “quantized” vortex in This was an experiment, performed in the 1970s to
superfluid helium. We have already mentioned that study how platelets form plaque in arteries. The idea
superfluid helium flows up and down the little barom- was that platelets deposit out of the blood if the flow of
eter tubes in a Venturi meter, giving no height differ- blood is too slow. The purpose of the experiment was
ence and nullifying the effectiveness of the device as a to design a flow where one could easily see where the
velocity meter. This happened because superfluid plaque began to form and also know what the velocity
helium has NO viscosity (absolutely none as far as we of the flow was there.
can tell) and can therefore flow into tiny places where The apparatus is shown in Figure (31). Blood flows
other fluids cannot move. down through a small tube and then through a hole in a
circular plate that is suspended a small distance d
More surprising yet is the structure of a vortex in
above a glass plate. When the blood gets to the glass
superfluid helium. The vortex has a core that is about it flows radially outward as indicated in Figure (31c). As
one atomic diameter across (you can’t get much smaller the blood flowed radially outward, its velocity decreases.
than that), and a precise 1/r velocity field outside the At a certain radius, call it rp , platelets began to deposit
core. Even more peculiar is the fact that the velocity on the glass. The flow was photographed by a video
field outside the core is given by the formula camera looking up through the glass.

For this problem, assume that the tube radius was

vθ = κ
h
;
κ = (24) rt = .4mm, and that the separation d between the circular
2πr mHe
plate and the glass was d = .5 mm. If blood were flowing
down the inlet tube at a rate of half a cubic centimeter
where κ , called the “circulation of the vortex”, has the per second, what is the average speed of the blood
precisely known value h mHe , where mHe is the mass
of a helium atom, and h is an atomic constant known as a) inside the inlet tube?
Planck’s constant. The remarkable point is that the
b) at a radius rp = 2cm out from the hole in the circular
strength of a helium vortex has a precise value deter- plate?
mined by atomic scale constants. (This is why we say
that vortices in superfluid helium are quantized.) When (By average speed, we mean neglect fluid friction at
we get to the study of atoms, and particularly the Bohr walls, and assume that the flow is uniform across the
theory of hydrogen, we can begin to explain why radius of the inlet pipe and across the gap as indicated
helium vortices have precisely the strength κ = h mHe. in Figure (31d).
For now, we are mentioning vortices in superfluid
helium as examples of an ideal vortex with a well- Exercise 6
defined core and a precise 1/r velocity field outside. A good review of both the continuity equation and
Bernoulli’s equation, is to derive on your own, without
Quantized vortices of a more complicated structure looking back at the text, the formula
also occur in superconductors and play an important
2gh
role in the practical behavior of a superconducting v1 = (21)
A12/A22 – 1
material. The superconductors that carry the greatest
currents, and are the most useful in practical applica- for the flow speed in a venturi meter. The various
tions, have quantized vortices that are pinned down and quantities v1 , h, A1 and A2 are defined in Figure (18)
cannot move around. One of the problems in develop- reproduced on the opposite page. (If you have trouble
ing practical applications for the new high temperature with the derivation, review it in the text, and then a day
superconductors is that the quantized vortices tend to or so later, try the derivation again on your own.
move and cause energy losses. Pinning these vortices
down is one of the main goals of current engineering
research.
 23-23

blood flowing
in through tube blood flowing
radially outward
a)

c ir c u la r m e t a l p l a t e
glass plate

tube of inner
b) radius rtube gap thickness d

radius at which
platelets form
c) rtube
blood flow
rp

Figure 31 a,b,c
Experiment to measure the blood flow velocity at which platelets stick
to a glass plate. This is an application of the continuity equation.

blood flow (A) (B)

(3)
h
(4)

v1 v2
(1)
(2)
A2
Figure 31d
A1
Neglect fluid friction at walls, and assume
that the flow is uniform across the radius Figure 18
of the inlet pipe and across the gap Venturi meter. Since the water flows faster through
the constriction, the pressure is lower there. By using
vertical tubes to measure the pressure drop, and using
Bernoulli's equation and the continuity equation, you
can determine the flow speeds v1 and v2 .
Chapter 24
Coulomb's Law
and Gauss' Law

CHAPTER 24 COULOMB'S LAW

AND GAUSS' LAW

In our discussion of the four basic interactions we saw COULOMB'S LAW

that the electric and gravitational interaction had very In Chapter 18 we briefly discussed Coulomb’s electric
similar 1 r 2 force laws, but produced very different force law, primarily to compare it with gravitational
kinds of structures. The gravitationally bound struc- force law. We wrote Coulomb’s law in the form
tures include planets, solar systems, star clusters,
galaxies, and clusters of galaxies. Typical electrically
bound structures are atoms, molecules, people, and KQ1Q2 Coulomb,s
F = Law (1)
redwood trees. Although the force laws are similar in r2
form, the differences in the structures they create result
from two important differences in the forces. Gravity where Q 1 and Q 2 are two charges separated by a
is weaker, far weaker, than electricity. On an atomic distance r as shown in Figure (1). If the charges Q 1 and
scale gravity is so weak that its effects have not been Q 2 are of the same sign, the force is repulsive, if they are
seen. But electricity has both attractive and repulsive of the opposite sign it is attractive. The strength of the
forces. On a large scale, the electric forces cancel so electric force decreases as 1 r 2 just like the gravita-
completely that the weak but non-cancelling gravity tional force between two masses.
dominates astronomical structures.
Q1 Q2

r
Figure 1
Two particles of charge Q1 and Q 2 ,
separated by a distance r.
24-2 Coulomb's Law and Gauss' Law

CGS Units Although we work with the familiar quantities volts,

In the CGS system of units, the constant K in Equation amps and watts using MKS units, there is a price we
(1) is taken to have the numerical value 1, so that have to pay for this convenience. In MKS units the
Coulomb’s law becomes constant K in Coulomb’s law is written in a rather
peculiar way, namely
Q 1Q 2
Fe = (2) 1
r2 K =
4πε0
Equation (2) can be used as an experimental definition (3)
of charge. Let Q 1 be some accepted standard charge. ε0 = 8.85 × 10 –12
farads/meter
Then any other charge Q 2 can be determined in terms
of the standard Q 1 by measuring the force on Q 2 when and Coulomb’s law is written as
the separation is r. In our discussion in Chapter 18, we Q 1Q 2 Q1 Q2
took the standard Q 1 as the charge on an electron. Fe = (4)
4πε0r 2 r

This process of defining a standard charge Q 1 and using where Q 1 and Q 2 are the charges measured in cou-
Equation (1) to determine other charges, is easy in lombs, and r is the separation measured in meters, and
principle but almost impossible in practice. These so- the force Fe is in newtons.
called electrostatic measurements are subject to all
sorts of experimental problems such as charge leaking Before we can use Equation (4), we have to know how
away due to a humid atmosphere, redistribution of big a unit of charge a coulomb is, and we would
charge, static charge on the experimenter, etc. Charles probably like to know why there is a 4π in the formula,
Coulomb worked hard just to show that the electric and why the proportionality constant ε0 (“epsilon
force between two charges did indeed drop as 1 r 2. As naught ”) is in the denominator.
a practical matter, Equation (2) is not used to define
electric charge. As we will see, the more easily As we saw in Chapter 18, nature has a basic unit of
controlled magnetic forces are used instead. charge (e) which we call the charge on the electron. A
coulomb of charge is 6.25 × 10 18 times larger. Just as
a liter of water is a large convenient collection of water
MKS Units molecules, 3.34 × 10 25 of them, the coulomb can be
When you buy a 100 watt bulb at the store for use in thought of as a large collection of electron charges,
your home, you will see that it is rated for use at 110– 6.25 × 10 18 of them.
120 volts if you live in the United States or Canada, or
220–240 volts most elsewhere. The circuit breakers in In practice, the coulomb is defined experimentally, not
your house may allow each circuit to carry up to 15 or by counting electrons, and not by the use of Coulomb’s
20 amperes of current in each circuit. The familiar law, but, as we said, by a magnetic force measurement
quantities volts, amperes, watts are all MKS units. The to be described later. For now, just think of the coulomb
corresponding quantities in CGS units are the totally as a convenient unit made up of 6.25 × 10 18 electron
unfamiliar statvolts, statamps, and ergs per second. charges.
Some scientific disciplines, particularly plasma and
solid state physics, are conventionally done in CGS Once the size of the unit charge is chosen, the propor-
units but the rest of the world uses MKS units for tionality constant in Equation (4) can be determined by
describing electrical phenomena. experiment. If we insist on putting the proportionality
constant in the denominator and including a 4π, then ε0
has the value of 8.85 × 10 –12, which we will often
approximate as 9 × 10 –12.
 24-3

Why is the proportionality constant ε0 placed in the Summary

denominator? The kindest answer is to say that this is The situation with Coulomb’s law is not really that bad.
a historical choice that we still have with us. And why We have a 1 r 2 force law like gravity, charge is mea-
include the 4π? There is a better answer to this sured in coulombs, which is no worse than measuring
question. By putting the 4π now, we get rid of it in mass in kilograms, and the proportionality constant just
another law that we will discuss shortly, called Gauss’ happens, for historical reasons, to be written as 1 4πε0.
law. If you work with Gauss’ law, it is convenient to
have the 4π buried in Coulomb’s law. But if you work The units are incomprehensible, so do not worry too
with Coulomb’s law, you will find the 4π to be a much about keeping track of units. After a bit of
nuisance. practice, Coulomb’s law will become quite natural.

Checking Units in MKS Calculations Example 1 Two Charges

If we write out the units in Equation (4), we get Two positive charges, each 1 coulomb in size, are
Q 1(coul)Q 2(coul) placed 1 meter apart. What is the electric force between
F(newtons) = (4a) them?
4πε0r 2(meter)2
In order for the units in Equation (3a) to balance, the Q 1= 1 Q2= 1
proportionality constant ε0 must have the dimensions Fe Fe
1 meter
2
ε0 coulombs (5) Solution: The force will be repulsive, and have a
meter2newton
magnitude
In earlier work with projectiles, etc., it was often useful
to keep track of your units during a calculation as a Q1Q2
check for errors. In MKS electrical calculations, it is Fe = = 1
4πε0r 2 4πε0
almost impossible to do so. Units like 1
2 2
coul meter newton are bad enough as they are. But = = 10 10 newtons
4π×9 × 10 –12
if you look up ε0 in a textbook, you will find its units are
listed as farads/meter. In other words the combination From the answer, 1010 newtons, we see that a coulomb
coul2 newton meter was given the name farad. With is a huge amount of charge. We would not be able to
naming like this, you do not stand a chance of keeping assemble two 1 coulomb charges and put them in the
units straight during a calculation. You have to do the same room. They would tear the room apart.
best you can to avoid mistakes without having the
reassurance that your units check.
24-4 Coulomb's Law and Gauss' Law

Example 2 Hydrogen Atom Since the electron is in a circular orbit, its acceleration
In a classical model of a hydrogen atom, we have a is v 2 /r pointing toward the center of the circle, and we
proton at the center of the atom and an electron travel- get
ing in a circular orbit around the proton. If the radius v2 F Fe
of the electron’s orbit is r = .5 × 10 –10 meters, how a = =
m
=
m
r
long does it take the electron to go around the proton
once? With the electron mass m equal to 9.11×10 –31 kg , we
electron have
(-e)
r e2 rFe .5×10 –10 × 9×10 –8
Fe = v2 = m =
4πε0r 2 9.11×10 –31 kg
proton
(+e) 2
= 4.9×10 12 m2
s
Solution: This problem is more conveniently handled v = 2.2×10 6 m
s
in CGS units, but there is nothing wrong with using
MKS units. The charge on the proton is (+e), on the To go around a circle of radius r at a speed v takes a time
electron (–e), thus the electrical force is attractive and
has a magnitude T = 2π r = 2π × .5×10 –10
v 2.2×10 6
Fe =
(e)(e)
= e2
4πε0r 2 4πε0r 2 = 1.4×10 –16seconds
With e = 1.6 × 10 –19 coulombs and r = .5 × 10 –10m In this calculation, we had to deal with a lot of very
we get small or large numbers, and there was not much of an
2 extra burden putting the 1/4πε0 in Coulomb's law.
1.6 × 10 –19
Fe = 2
4π × 9 × 10 –12 × .5 × 10 –10

= 9 × 10 –8 newtons
 24-5

Exercise 1 Exercise 4

Q
Fe Fg Fg Q Fe
earth
m
θ
M
Q Q
(a) Equal numbers of electrons are added to both the Fe Fe
earth and the moon until the repulsive electric force m m
exactly balances the attractive gravitational force. How
many electrons are added to the earth and what is their
total charge in coulombs? Two styrofoam balls covered by aluminum foil are
suspended by equal length threads from a common
(b) What is the mass, in kilograms of the electrons point as shown. They are both charged negatively by
added to the earth in part (a)? touching them with a rubber rod that has been rubbed
by cat fur. They spread apart by an angle 2 θ as shown.
Exercise 2 Assuming that an equal amount of charge Q has been
placed on each ball, calculate Q if the thread length is
Calculate the ratio of the electric to the gravitational
= 40 cm , the mass m of the balls is m=10 gm, and the
force between two electrons. Why does your answer
angle is θ = 5°. Use Coulomb's law in the form
not depend upon how far apart the electrons are?
Fe = Q1Q2 /4π ε 0 r 2 , and remember that you must use
MKS units for this form of the force law.
Exercise 3

garden peas
stripped of
electrons
Fe Fe

Q protons Q protons

Imagine that we could strip all the electrons out of two

garden peas, and then placed the peas one meter
apart. What would be the repulsive force between
them? Express your answer in newtons, and metric
tons. (One metric ton is the weight of 1000 kilograms.)
(Assume the peas each have about one Avogadro's
number, or gram, of protons.)
24-6 Coulomb's Law and Gauss' Law

FORCE PRODUCED BY A LINE We will see that there is a powerful theorem, discov-
CHARGE ered by Frederick Gauss, that considerably simplifies
In our discussion of gravitational forces, we dealt only the calculation of electric forces produced by extended
distribution of charges. But Gauss’s law involves
with point masses because most practical problems
several new concepts that we will have to develop. To
deal with spherical objects like moons, planets and
stars which can be treated as point masses, or spacecraft appreciate this effort, to see why we want to use
Gauss’s law, we will now do a brute force calculation,
which are essentially points.
using standard calculus steps to calculate the force
The kind of problem we did not consider is the follow- between a point charge and a line of charge. It will be
ing. Suppose an advanced civilization constructed a hard work. Later we will use Gauss’s law to do the
rod shaped planet shown in Figure (2) that was 200,000 same calculation and you will see how much easier it
kilometers long and had a radius of 10,000 km. A is.
satellite is launched in a circular orbit of radius 20,000
km, what is the period of the satellite’s orbit? The setup for our calculation is shown in Figure (3).
We have a negative charge Q T located a distance r from
We did not have problems like this because no one has a long charged rod as shown. The rod has a positive
thought of a good reason for constructing a rod shaped charge density λ coulombs per meter spread along it.
planet. Our spherical planets, which can be treated as We wish to calculate the total force F exerted by the
a point mass, serve well enough. rod upon our negative test particle Q T. For simplicity
we may assume that the ends of the rod are infinitely far
In studying electrical phenomena, we are not restricted away (at least several feet away on the scale of the
to spherical or point charges. It is easy to spread an drawing).
electric charge along a rod, and one might want to know
what force this charged rod exerted on a nearby point To calculate the electric force exerted by the rod, we
charge. In electricity theory we have to deal with will conceptually break the rod into many short seg-
various distributions of electric charge, not just the ments of length dx, each containing an amount of
simple point concentrations we saw in gravitational charge dq = λdx. To preserve the left/right symmetry,
calculations. we will calculate the force between Q T and pairs of dq,
one on the left and one on the right as shown. The left
directional force dF1 and the right directed force dF2
rod shaped add up to produce an upward directional force dF.
planet Thus, when we add up the forces from pairs of dq, the
satellite in orbit
dF's are all upward directed and add numerically.
about the rod
shaped planet The force dF1 between Q T and the piece of charge dq 1
is given by Coulomb’s law as

KQ T dq 1 KQ T dq 1
dF1 = =
Figure 2 R2 x2 + r2
Imagine that an advanced civilization creates
a rod-shaped planet. The problem, which we KλQ T dx (6)
=
have not encountered earlier, is to calculate x2 + r2
the period of a satellite orbiting the planet.
where, for now, we will use K for 1/4πε0 to simplify the
formula. The separation R between Q T and dq 1 is
given by the Pythagorean theorem as R 2 = x 2 + r2.
 24-7

The component of dF1 in the upward direction is Instead we look it up in a table of integrals with the
dF1cos θ , so that dF has a magnitude result
A
dF = 2 dF1 cos θ dx 2 2x A
3
=
x2 + r2 2 4r r 2 + x 2
2
0
KλQ dx r 0
= 2 2 T2 (7)
x +r x + r2
2
= A – 0 (9)
r2 r 2 + A2
The factor of 2 comes from the fact that we get equal
components from both dF1 and dF2 .
For A >> r (very long rod), we can set r 2 + A2 ≈ A
To get the total force on Q T, we add up the forces and we get the results
produced by all pairs of dq starting from x = 0 and going
out to x = ∞. The result is the definite integral A>>r

∞
dx = A = 1 (10)
3 r 2A r2
2KλQT r x2 + r2 2
FQT = 3
dx 0
o x2 + r2 2
Using Equation (10) in Equation (8) we get
∞
dx (8)
= 2KλQT r 32 FQT = 2KλQ T r2
o x2 + r2 r
(11)
2KλQ T
where r, the distance from the charge to the rod, is a FQT = r
constant that can be taken outside the integral.
The important point of the calculation is that the force
The remaining integral dx/ x 2 + r 2 3/2 is not a com- between a point charge and a line charge drops off as
mon integral whose result you are likely to have 1 r rather than 1 r 2 , as long as Q T stays close enough
memorized, nor is it particularly easy to work out. to the rod that the ends appear to be very far away.
dq1 = λdx
dx λ coulombs per meter
x dq
R
=

r
x
2
+

dF dF
r
2

r θ θ dF = dF1 cos θ
cos θ = dF1 dF2 dF1 2
x 2+ r 2
QT
Figure 3
Geometry for calculating the electric force between a point charge
Q T a distance r from a line of charge with λ coulombs per meter.
24-8 Coulomb's Law and Gauss' Law

One of the rules of thumb in doing physics is that if you Exercise 5

have a simple result, there is probably an intuitive Back to science fiction. A rod shaped planet has a mass
derivation or explanation. In deriving the answer in density λ kilograms per unit length as shown in Figure
Equation (11), we did too much busy work to see (4). A satellite of mass m is located a distance r from the
anything intuitive. We had to deal with an integral of rod as shown. Find the magnitude of the gravitational
x 2 + r 2 –3/2, yet we got the simple answer that the force F g exerted on the satellite by the rod shaped
force dropped off as 1/r rather than the 1 r 2 . We will planet. Then find a formula for the period of the satellite
see, when we repeat this derivation using Gauss’s law, in a circular orbit.
that the change from 1 r 2 to a 1/r force results from the
change from a three to a two dimensional problem. λ kilograms/meter
This basic connection with geometry is not obvious in
our brute force derivation. very long
r rod shaped
planet

m Fg = ?

Figure 4
Rod shaped planet and satellite.
 24-9

Short Rod Equation (14) has the advantage that it can handle both
Our brute force calculation does have one advantage, limiting cases of a long rod (L >> r), a short rod or point
however. If we change the problem and say that our charge (L << r), or anything in between.
charge is located a distance r from the center of a finite
rod of length 2L as shown in Figure (5), then Equation For example, if we are far away from a short rod, so that
(8) of our earlier derivation becomes L << r, then r2 + L2 ≈ r . Using the fact that 2Lλ = Q R
is the total charge on the rod, Equation (14) becomes
L
dx KQ RQ T
FQT = 2KλQT r (12) FQ T = L << r (15b)
3 r2
0 x2 + r2 2
which is just Coulomb’s law for point charges. The
The only difference is that the integral stops at L rather more general result, Equation (14), which we obtained
than going out to infinity. by the brute force calculation, cannot be obtained with
simple arguments using Gauss’s law. This formula
From Equation (9) we have was worth the effort.
L
dx = L (13) Exercise 6
3
0 x2 + r2 2 r2 r 2 + L2 Show that if we are very close to the rod, i.e. r << L, then
Equation (14) becomes the formula for the force exerted
And the formula for the force on Q T becomes by a line charge.

K 2Lλ QT
FQ = (14)
r r 2 + L2

L L

λ coulombs/meter

QT
Figure 5
A harder problem is to calculate the force F
exerted on QT by a rod of finite length 2 L.
24-10 Coulomb's Law and Gauss' Law

THE ELECTRIC FIELD KQT QA

FA =
Our example of the force between a point charge and a rA2
line of charge demonstrates that even for simple distri- (16)
butions of charge, the calculation of electric forces can KQT QB
become complex. We now begin the introduction of FB =
rB2
several new concepts that will allow us to simplify
many of these calculations. The first of these is the where rA is the distance from Q T to QA and rB is the
electric field, a concept which allows us to rely more on distance from Q T to QB . For now we are writing
maps, pictures and intuition, than upon formal 1/4πε0 = K to keep the formulas from looking too
calculations. messy.
To introduce the idea of an electric field, let us start with In Figure (7), we have the same distribution of fixed
the simple distribution of charge shown in Figure (6). charge as in Figure (6), namely Q A and Q B, but we are
A positive charge of magnitude Q A is located at Point using a smaller test charge Q′T . For this sketch, Q′T
A, and a negative charge of magnitude –Q B is located is about half as big as Q T of Figure (6), and the resulting
at Point B. We will assume that these charges Q A and force vectors point in the same directions but are about
Q B are fixed, nailed down. They are our fixed charge half as long;
distribution.
KQ′T QA
We also have a positive test charge of magnitude Q T F′A =
rA2
that we can move around in the space surrounding the (17)
fixed charges. Q T will be used to test the strength of KQ′T QB
the electric force at various points, thus the name “test F′B =
charge”. rB2

In Figure (6), we see that the test charge is subject to the The only difference between Equations (16) and (17) is
repulsive force FA and attractive force FB to give a net that Q T has been replaced by Q′T in the formulas. You
force F = FA + FB . The individual forces FA and FB can see that Equations (17) can be obtained from
are given by Coulomb’s law as Equations (16) by multiplying the forces by Q′T QT.
Thus if we used a standard size test charge Q T to
calculate the forces, i.e. do all the vector additions, etc.,
then we can find the force on a different sized test
charge Q′T by multiplying the net force by the ratio
Q′T QT.
QT FA FA
QT
F
F
FB
FB

+ – + –
QA QB QA QB
Figure 6 Figure 7
Forces exerted by two fixed charges If we replace the test charge QT by a smaller
Q A and QB on the test particle QT . test charge Q′T , everything is the same except
that the force vectors become shorter.
 24-11

Unit Test Charge forces it exerts do not disturb anything, but mathemati-
The next step is to decide what size our standard test cally treat the test charge as having a magnitude of 1
charge Q T should be. Physically Q T should be small coulomb. In addition, we will always use a positive
so that it does not disturb the fixed distribution of unit test charge. If we want to know the force on a
charge. After all, Newton’s third law requires that Q T negative test charge, simply reverse the direction of the
pull on the fixed charges with forces equal and opposite force vectors.
to the forces shown acting on Q T .
Figure (8) is the same as our Figures (6) and (7), except
On the other hand a simple mathematical choice is that we are now using a unit test particle equal to 1
Q T = 1 coulomb, what we will call a “unit test charge”. coulomb, to observe the electric forces surrounding our
If Q T = 1, then the force on another charge Q′T is just fixed charge distribution of QA and QB. The forces
Q′T times larger ( F = F Q′T QT = FQ′ Q′T 1 = F Q′T ). acting on QT = 1 coulomb are
The problem is that in practice, a coulomb of charge is
enormous. Two point charges, each of strength +1 KQ A × 1 coulomb
coulomb, located one meter apart, repel each other with FA =
r A2
a force of magnitude (18)
KQ B × 1 coulomb
K × 1 coulomb × 1 coulomb FB =
F = rB2
1 2m2
1 To emphasize that these forces are acting on a unit test
= K = = 9 × 10 9 newtons (17)
4πε0 charge, we will use the letter E rather than F, and write

a result we saw in Example 1. A force of nearly 10

KQ A
billion newtons is strong enough to destroy any experi- EA =
mental structure you are ever likely to see. In practice r A2
(19)
the coulomb is much too big a charge to serve as a KQ B
realistic test particle. EB =
r B2
The mathematical simplicity of using a unit test charge
is too great to ignore. Our compromise is the following. If you wished to know the force F on some charge Q
We use a unit test charge, but think of it as a “small unit located where our test particle is, you would write
test charge”. Conceptually think of using a charge
about the size of the charge of an electron, so that the KQ AQ
FA = = EA Q (21a)
r A2
EA
QT = 1 coulomb KQ BQ
FB = = EB Q (21b)
E r B2
EB
F = QEA +QEB = Q EA +EB
+ –
F = QE (22)
QA QB
where E = EA +EB. Equation (22) is an important
Figure 8
When we use a unit test charge QT = 1 coulomb,
result. It says that the force on any charge Q is Q times
then the forces on it are called "electric field" the force E on a unit test particle.
vectors, EA , EB and E as shown.
24-12 Coulomb's Law and Gauss' Law

ELECTRIC FIELD LINES Mapping the Electric Field

The force E on a unit test particle plays such a central In Figure (10), we started with a simple charge distri-
role in the theory of electricity that we give it a special bution +Q and -Q as shown, placed our unit test particle
name – the electric field E. at various points in the region surrounding the fixed
charges, and drew the resulting force vectors E at each
point. If we do the diagram carefully, as in Figure (10),
force on a a picture of the electric field begins to emerge. Once we
electric ≡ unit test
field E particle have a complete picture of the electric field E, once we
(23) know E at every point in space, then we can find the
force on any charge q by using FQ = QE . The problem
Once we know the electric field E at some point, then we wish to solve, therefore, is how to construct a
the force FQ on a charge Q located at that point is complete map or picture of the electric field E .
(24)
F = QE Exercise 7
In Figure (10), we have labeled 3 points (1), (2), and (3).
If Q is negative, then F points in the direction opposite Sketch the force vectors FQ on :
to E . (a) a charge Q = 1 coulomb at Point (1)

From Equation (24), we see that the electric field E has (b) a charge Q = –1 coulomb at Point (2)
the dimensions of newtons/coulomb, so that QE comes (c) a charge Q = 2 coulombs at Point (3)
out in newtons.
E
E
(3)
E
F = QE
Q
Figure 9
Once we know the electric field E at some E
point, we find the force F acting on a charge Q + E E E –
(1)
at that point by the simple formula F = QE . E

E (2)
E
Figure 10
If we draw the electric field vectors E at various
points around our charge distribution, a picture
or map of the electric field begins to emerge.
 24-13

In part (c) of Exercise (7), we asked you to sketch the Field Lines
force on a charge Q = 2 coulombs located at Point (3). As a first step in simplifying the mapping process, let
The answer is F3 = QE 3 = 2E 3 , but the problem is that us concentrate on showing the direction of the electric
we have not yet calculated the electric field E3 at Point force in the space surrounding our charge distribution.
(3). On the other hand we have calculated E at some This can be done by connecting the arrows in Figures
nearby locations. From the shape of the map that is (10) and (11) to produce the line drawings of Figures
emerging from the E vectors we have drawn, we can (12a) and (12b) respectively. The lines in these draw-
make a fairly accurate guess as to the magnitude and ings are called field lines.
direction of E at Point (3) without doing the calculation.
With a map we can build intuition and make reasonably In our earlier discussion of fluid flow we saw diagrams
accurate estimates without calculating E at every point. that looked very much like the Figures (12a and 12b).
There we were drawing stream lines for various flow
In Figure (10) we were quite careful about choosing patterns. Now we are drawing electric field lines. As
where to draw the vectors in order to construct the illustrated in Figure (13), a streamline and an electric
picture. We placed the points one after another to see field line are similar concepts. At every point on a
the flow of the field from the positive to the negative streamline, the velocity field v is parallel to the stream-
charge. In Figure (11), we have constructed a similar line, while at every point on an electric field line, the
picture for the electric field surrounding a single posi- electric field E is parallel to the electric field line.
tive charge.

The difficulty in drawing maps or pictures of the

electric field is that we have to show both the magnitude
and direction at every point. To do this by drawing a
large number of separate vectors quickly becomes
cumbersome and time consuming. We need a better E

way to draw these maps, and in so doing will adopt

many of the conventions developed by map makers.

Figure 12a
We connected the arrows of Figure 10 to
create a set of field lines for 2 point charges.

E E E

E
Figure 11
Electric field of a point charge.

Figure 12b
Here we connected the arrows in Figure 11
to draw the field lines for a point charge.
24-14 Coulomb's Law and Gauss' Law

Continuity Equation for Electric Fields If we think of the electric field as some kind of an
Figure (14) is our old diagram (23-6) for the velocity incompressible fluid, and think of a point charge as a
field of a point source of fluid (a small sphere that source of this fluid, then the continuity equation ap-
created water molecules). We applied the continuity plied to this electric field gives us the correct 1/r 2
equation to the flow outside the source and saw that the dependence of the field. In a sense we can replace
velocity field of a point source of fluid drops off as 1/r 2. Coulomb’s law by a continuity equation. Explicitly,
we will use streamlines or field lines to map the
Figure (15) is more or less a repeat of Figure (12b) for direction of the field, and use the continuity equation to
the electric field of a point charge. By Coulomb’s law, calculate the magnitude of the field. This is our general
the strength of the electric field drops off as 1/r 2 as we plan for constructing electric field maps; we now have
go out from the point charge. We have the same field to fill in the details.
structure for a point source of an incompressible fluid
and the electric field of a point charge. Is this pure
coincidence, or is there something we can learn from
the similarity of these two fields?
v2
The crucial feature of the velocity field that gave us a v1
1/r 2 flow was the continuity equation. Basically the
idea is that all of the water that is created in the small A1
sphere must eventually flow out through any larger
sphere surrounding the source. Since the area of a
sphere, 4πr2, increases as r2, the speed of the water has
to decrease as 1/r 2 so that the same volume of water per A2
Figure 14
second flows through a big sphere as through a small This is our old Figure 23-6 for the velocity field
one. of a point source of water. The continuity
v1 E1 equation v1 A1 = v2 A2 , requires that the velocity
field v drops off as 1/r2 because the area
through which the water flows increases as π r2 .
v2 E2

E3
v3 E2
Q E1

A1
v4 E4

A2
streamline for field line for Figure 15
velocity field electric field From Coulomb's law, E = KQ/r2 , we see that the
Figure 13 electric field of a point source drops off in exactly the
Comparison of the streamline for the velocity field same way as the velocity field of a point source. Thus
and the field line for an electric field. Both are the electric field must obey the same continuity
constructed in the same way by connecting successive equation E1 A1 = E2 A2 as does the velocity field.
vectors. The streamline is easier to visualize because
it is the actual path followed by particles in the fluid.
 24-15

Flux The continuity equation v1 A1 = v2 A2 requires these

To see how the continuity equation can be applied to fluxes be equal
electric fields, let us review the calculation of the 1/r 2 continuity
velocity field of a point source of fluid, and follow the Φ1 = Φ2 ≡ Φ equation (27)
same steps to calculate the electric field of a point
charge. Using Equations (26) and (27), we can express the
velocity field v2 out at the larger sphere in terms of the
In Figure (16) we have a small sphere of area A 1 in flux of water Φ created inside the small sphere
which the water is created. The volume of water
created each second, which we called the flux of the
water and will now designate by the greek letter Φ , is v2 = Φ = Φ 2 (28)
A2 4πr2
given by

volume of water Let us now follow precisely the same steps for the
created per second ≡ Φ1 = v1A1 (25) electric field of a point charge. Construct a small sphere
in the small sphere of area A1 and a large sphere A2 concentrically sur-
rounding the point charge as shown in Figure (17). At
The flux of water out through a larger sphere of area A2 the small sphere the electric field has a strength E1,
is which has dropped to a strength E2 out at A2.
Let us define E 1A 1 as the flux of our electric fluid
volume of water
flowing out of the smaller sphere, and E 2A 2 as the flux
flowing per second
≡ Φ2 = v2A2 (26) flowing out through the larger sphere
out through a
larger sphere Φ1 = E1A1 (29)
Φ2 = E2A2 (30)

r1
r1
r2
v1
Q r2
A1
E1 A1

v2 E2
A2
Figure 16 A2
The total flux Φ1 of water out of the small sphere is Figure 17
Φ1 = v1 A1⊥⊥ , where A1⊥⊥ is the perpendicular area The total flux Φ1 of the electric field out of the
through which the water flows. The flux through small sphere is Φ1 = E 1 A1⊥⊥ , where A1⊥⊥ is the
the larger sphere is Φ2 = v2 A2⊥⊥ . Noting that no water perpendicular area through which the electric
is lost as it flows from the inner to outer sphere, i.e., field flows. The flux through the larger sphere is
equating Φ1 and Φ2 , gives us the result that the Φ2 = E 2 A2⊥⊥ . Noting that no flux is lost as it flows
velocity field drops off as 1/r2 because the from the inner to outer sphere, i.e., equating Φ1
perpendicular area increases as r2 . and Φ2 , gives us the result that the electric field
drops off as 1/r2 because the perpendicular area
increases as r2 .
24-16 Coulomb's Law and Gauss' Law

Applying the continuity equation E 1A 1 = E 2A 2 to Q Coulomb's

this electric fluid, we get E = Law (34)
4πε0r 2
Φ1 = Φ2 = Φ (31) where now we are explicitly putting in 1/4πεo for the
Again we can express the field at A 2 in terms of the flux proportionality constant K.

E2 = Φ = Φ 2
Comparing Equations (33) and (34), we see that if we
(32)
A2 4πr2 choose

Since A 2 can be any sphere outside, but centered on the

point charge, we can drop the subscript 2 and write Q Flux emerging
Φ = (35)
εo from a charge Q
E(r) = Φ (33)
4πr 2 then the continuity equation (33) and Coulomb's law
(34) give the same answer. Equation (35) is the key that
In Equation (33), we got the correct 1/r 2 dependence allows us to apply the continuity equation to the electric
for the electric field, but what is the appropriate value field. If we say that a point charge Q creates an electric
for Φ ? How much electric flux Φ flows out of a point flux Φ = Q/εo , then applying the continuity equation
charge? gives the same results as Coulomb's law. (You can now
see that by putting the 4π into Coulomb's law, there is
To find out, start with a fixed charge Q as shown in no 4π in our formula (35) for flux.)
Figure (18), place our unit test charge a distance r away,
and use Coulomb's law to calculate the electric force E
on our unit test charge. The result is Negative Charge
If we have a negative charge –Q, then our unit test
particle Q T will be attracted to it as shown in Figure
E (19). From a hydrodynamic point of view, the electric
QT = 1 fluid is flowing into the charge –Q and being destroyed
Φ = EA there. Therefore a generalization of our rule about
electric flux is that a positive charge creates a positive,
= ( 4πεQ0r2 ) 4πr 2 outward flux of magnitude Q/ε o, while a negative
Q r charge destroys the electric flux, it has a negative flux
Q
= – Q/ε o that flows into the charge and disappears.
ε0

A
Figure 18
Using Coulomb's law for the electric field of
E
a point charge Q, we calculate that the total
flux out through any centered sphere –Q
surrounding the point charge is Φ = Q/ ε 0 .

Figure 19
The electric field of a negative charge flows
into the charge. Just as positive charge
creates flux, negative charge destroys it.
 24-17

Flux Tubes A Mapping Convention

In our first pictures of fluid flows like Figure (23-5) If an electric field line represents a small flux tube, the
reproduced here, we saw that the streamlines were little question remains as to how much flux is in the tube?
tubes of flow. The continuity equation, applied to a Just as we standardized on a unit test charge Q T = 1
streamline was Φ = v1A1 = v2A2 . This is simply the coulomb for the definition of the electric field E, we will
statement that the flux of a fluid along a streamline is standardize on a unit flux tube as the amount of flux
constant. We can think of the streamlines as small represented by one electric field line. With this conven-
tubes of flux. By analogy we will think of our electric tion, we should therefore draw Φ = Q/ε0 field lines
field lines as small tubes of electric flux. or unit flux tubes coming out of a positive charge +Q,
or stopping on a negative charge –Q. Let us try a few
examples to see what a powerful mapping convention
Conserved Field Lines
this is.
When we think of the field line as a small flux tube, the
continuity equation gives us a very powerful result, In Figure (20) we have a positive charge Q/ε0 = +5, and
namely the flux tubes must be continuous, must main- a negative charge Q/ε0 = –3, located as shown. By
tain their strength in any region where the fluid is our new mapping convention we should draw 5 unit
neither being created nor destroyed. For the electric flux tubes or field lines out of the positive charge, and
fluid, the flux tubes or field lines are created by, start at, we should show 3 of them stopping on the negative
positive charge. And they are destroyed by, or stop at, charge. Close to the positive charge, the negative
negative charge. But in between the electric fluid is charge is too far away to have any effect and the field
conserved and the field lines are continuous. We will lines must go radially out as shown. Close to the
see that this continuity of the electric field lines is a very negative charge, the lines must go radially in because
powerful tool for mapping electric fields. the positive charge is too far away.
x2
streamline
v2
A2

x 2 = v2 t
x1 Q/ε ο = 5
Q/ε ο = –3
A1 v1
x1 = v1 t Figure 20
We begin a sketch of the electric field by
Figure 23-5 drawing the field lines in close to the charges,
Flux tube in the flow of water. where the lines go either straight in or straight
out. Here we have drawn 3 lines into the
charge –3 ε 0 and 5 lines out of the charge
+ 5 ε 0 . To make a symmetric looking picture,
we oriented the lines so that one will go straight
across from the positive to the negative charge.
24-18 Coulomb's Law and Gauss' Law

Now we get to the interesting part; what happens to the where we cannot see the space between the point
field lines out from the charges? The basic rule is that charges, it looks like we have a single positive charge
the lines can start on positive charge, stop on negative of magnitude Q/ε0 = 2.
charge, but must be continuous in between. A good
guess is that 3 of the lines starting on the positive charge
Summary
go over to the negative charge in roughly the way we
When we started this chapter with the brute force
have drawn by dotted lines in Figure (21). There is no
calculus calculation of the electric field of a line charge,
more room on the minus charge for the other two lines,
you may have thought that the important point of this
so that all these two lines can do is to continue on out
chapter was how to do messy calculations. Actually,
to infinity.
exactly the opposite is true! We want to learn how to
Let us take Figure (21), but step far back, so that the + avoid doing messy calculations. The sketches shown
and the – charge look close together as shown in Figure in Figures (21) and (22) are an important step in this
(22). Between the charges we still have the same heart- process. From what we are trying to get out of this
shaped pattern, but we now get a better view of the two chapter, it is far more important that you learn how to
lines that had nowhere to go in Figure (21). To get a do sketches like Figures (21) and (22), than calcula-
better understanding for Figure (22) draw a sphere tions illustrated by Figure (1).
around the charges as shown. The net charge inside this
With a little experience, most students get quite good at
sphere is
sketching field patterns. The basic constraints are that
Q net Q/ε o lines start on positive charges, or stop on negative
εo = +5–3 = 2
inside sphere charges. Between charges the lines are unbroken and
should be smooth, and any lines left over must either go
Thus by our mapping convention, two field lines
to or come from infinity as they did in Figure (22).
should emerge from this sphere, and they do. Far away

Q/ε ο = –3
Q/ε ο = 5 – 3ε ο + 5ε ο

Figure 21 Figure 22
Once the in close field lines have been drawn, we Distant view of our charge distribution. If we step
can sketch in the connecting part of the lines as way back from the charge distribution of Figure (21),
shown above. Three of the lines starting from the we see a small object whose net charge is
positive charge must end on the negative one. –3 ε 0 + 5 ε 0 = +2 ε 0. Thus 2 lines must finally emerge
The other must go out to infinity. Using from this distribution as shown.
symmetry and a bit of artistic skill, you will
become quite good at drawing these sketches.
 24-19

A Computer Plot starting from the positive charge to the number ending
Figure (23) is a computer plot of the electric field lines on the negative charge is still 5/3. One of the standard
for the +5, -3 charge distribution of Figures (21) and tricks in map making is to change your scale to make
(22). The first thing we noted is that the computer drew the map look as good as possible. Here the computer
a lot more lines than we did. Did it violate our mapping drew 10 lines per unit flux tube rather than 1. We will
convention that the lines represent unit flux tubes, with see that the only time we really have to be careful with
Q/ε o lines starting or stopping on a charge Q? Yes. The the number of lines we draw is when we are using a
computer drew a whole bunch of lines so that we could count of the number of lines to estimate the strength of
get a better feeling for the shape of the electric field. the electric field.
Notice, however, that the ratio of the number of lines

–3 +5

Figure 23
Computer plot of the field lines of a –3 and +5 charge distribution. Rather than drawing
Q/εεo out of a positive charge or into a negative charge, the computer is programmed to
draw enough lines to make the shape of the electric field as clear as possible.
24-20 Coulomb's Law and Gauss' Law

GAUSS’ LAW In Figure (25), we have a point charge Q, and have

The idea of using the continuity equation to map field drawn a spherical Gaussian surface around the charge.
lines was invented by Frederick Gauss and is known as The flux produced by the point charge is
Gauss’ law. A basic statement of the law is as follows. flux produced
Q
Conceptually construct a closed surface (often called a Φtot = ε by the point (37)
o charge
Gaussian surface) around a group of sources or sinks,
as shown in Figure (24). In that figure we have drawn At the Gaussian surface there is an electric field E
the Gaussian surface around three sources and one (which we wish to calculate), and the surface has an
sink. Then calculate the total flux Φtot coming from area A = 4πr2. Therefore the electric flux flowing
these sources. For Figure (24), we have out through the sphere is
Φtot = Φ1 + Φ2 + Φ3 + Φ4 (36) Φout = EA
(38)
where Φ2 happens to be negative. Then if the fluid is flux flowing
incompressible, or we have an electric field, the total = E× 4πr 2 out through
the sphere
flux flowing out through the Gaussian surface must be
equal to the amount of flux Φtot being created inside. Equating the flux created inside (Equation 37) to the
flux flowing out (Equation 38) gives
Gauss' law applies to any closed surface surrounding
our sources and sinks. But the law is useful for Q
E × 4πr 2 = ε
calculations when the Gaussian surface is simple enough o
in shape that we can easily write the formula for the flux (39)
flowing through the surface. To illustrate the way we Q old result
E = obtained
use Gauss' law, let us, one more time, calculate the 4πεo r 2 new way
electric field of a point charge.

Gaussian
source Φ3 surface
r
Φ1 A = 4πr 2
sink
Q
Φ2

closed surface
surrounding several Φ4
sources or sinks
E
(this is called a Figure 25
Gaussian surface) Calculating the electric field of a point charge by
equating the flux Q/ ε 0 created by the point charge to
Figure 24
the flux Φ = E A ⊥ flowing through the Gaussian
If these were sources and sinks in a fluid, it would be
surface. This gives
obvious that the total flux of fluid out through the closed
surface is equal to the net amount of fluid created inside. Q/ ε 0 = E × 4 π r 2
The same concept applies to electric flux. The net flux
Φ tot out through the Gaussian surface is the sum of the E = Q/ 4 π ε 0 r 2
fluxes Φ 1 + Φ 2 + Φ 3 + Φ 4 created inside. as we expect.
 24-21

Electric Field of a Line Charge Equating the flux created inside (Equation 41) to the
As a real test of Gauss’ law, let us calculate the electric flux flowing out through the cylindrical surface (Equa-
field of a line charge and compare the result with our tion 42) gives
E r 2πrL = λL
brute force calculus calculation. In the calculus deriva-
tion, we found that the force on a test charge QT a εo
distance r from our line charge was (from Equation
The L's cancel and we get
(11))
2KλQT 2λQT Er = λ
F = r = 2πεo r
4πεor (43)
where λ is the charge density on the rod as shown in
Figure (26). Setting QT = 1 coulomb, F becomes the Voilà! We get the same result. Compare the calculus
3/2
force E on a unit test charge: derivation with its integral of x 2 + r 2 , to the
simple steps of Equations (41) and (42). We noted that
calculusderived
E = λ
formula for the in physics, a simple answer, like the 1/r dependence of
2πεor electric field of (40) the electric field of a line charge, should have an easy
a line charge derivation. The easy derivation is Gauss’ law. The
To apply Gauss’ law we first construct a cylindrical simple idea is that for a line charge the flux is flowing
Gaussian surface that surrounds a length L of the out through a cylindrical rather than a spherical surface.
charge as shown in Figure (27). Since the charge The area of a cylindrical surface increases as r, rather
density is λ , the total charge Q in , inside our Gaussian than as r2 for a sphere, therefore the electric field drops
surface is off as 1/r rather than as 1/r2 as it did for a point charge.

Q in = λL
L
This amount of charge creates an amount of flux
λ coulombs
Φinside = ε in = λL
Q r per meter
o εo (41) + + + + + + +

Now we can see from Figure (27) that because the

electric field lines go radially outward from the line
side
charge, they only flow out through the curved outer view A = 2πrL
surface of our Gaussian cylinder and not through the
flat ends. This cylindrical surface has an area
A = (2πr)L (circumference × length), and the elec-
tric field out at a distance r is E(r). Thus the flux out E
through the Gaussian cylinder is
Φout = E r 2π r L (42)

λ coulombs per meter r

+ + + + + + +
end
r view

QT = 1 coulomb Figure 27
Using Gauss' law to calculate the electric field of
a line charge. Draw the Gaussian surface
E around a section of the rod. The flux all flows
Figure 26
The force E on a unit test charge near a line charge. out through the cylindrical surface.
24-22 Coulomb's Law and Gauss' Law

Flux Calculations across the tube. The other areas are bigger by a factor
In our calculations of the flux through a Gaussian 1/cos θ where θ is the angle between A and A ⊥. I.e.,
surface, Equation (38) for a spherical surface and
Equation (42) for a cylindrical surface, we multiplied A ⊥ = A cos θ (45)
the strength of the electric field times the area A
Now the flux in the flow tube is the fluid speed v times
through which the field was flowing. In both cases we
the cross-sectional area A ⊥
were careful to construct the area perpendicular to the
field lines, for flux is equal to the strength of the field Φ = vA⊥ (46)
times cross sectional or perpendicular area through
which it is flowing. It would be more accurate to write Using Equation (45) for A ⊥ in Equation (46) gives
the formula for flux as
Φ = vA⊥ = vAcos θ
Φ = E A⊥ (44) But vAcos θ is just the vector dot product of the
velocity vector v (which points in the direction A ⊥) and
where the ⊥ sign reminds us that A⊥ is the perpendicular A , thus we have the more general formula
area.
Φ = vA⊥ = v⋅A (47a)
Area as a Vector By analogy, the electric flux through an area A is
A more formal way to present the formula for flux is to
turn the area A into a vector. To illustrate the procedure,
consider the small flow tube shown in Figure (28). We Φ = EA⊥ = E⋅A (47b)
have sliced the tube with a plane, and the intersection
of the tube and the plane gives us an area A as shown. In extreme cases where the Gaussian surface is not
We turn A into the vector A by drawing an arrow smooth, you may have to break up the surface into
perpendicular to the plane, and of length A. small pieces, calculate the flux dΦi = E⋅dAi for each
piece dA i and then add up all the contributions from
To show why we bothered turning A into a vector, in each piece to get the total flux Φout . The result is called
Figure (29) we have constructed a cross-sectional area a surface integral which we will discuss later. For now
A ⊥ as well as the area A of Figure (28). The cross- we will make sure that our Gaussian surfaces are
sectional area is the smallest area we can construct smooth and perpendicular to the field, so that we can
use the simple form of Equation (47).
plane slicing
the flow tube

A v
A
θ
flow tube flow tube
A
area A
Figure 29
Figure 28 The cross-sectional area, which we have been calling
Definition of the area as a vector. Slice a flow A ⊥ , is the smallest area that crosses the entire tube.
tube by a plane. The area A is the area of the The tilted area is larger than A ⊥ by a factor of
region in the plane bounded by the flow tube. 1/ cos θ , where θ is the angle between A and A ⊥ .
We define the direction of A as pointing As a result we can write
perpendicular to the plane as shown. A⊥ = A cos θ
Φ = v A ⊥ = v A cos θ = v ⋅ A
 24-23

GAUSS' LAW FOR THE Q

ΦE = ε = 1 × 4π Q (51)
GRAVITATIONAL FIELD o 4π εo
In our earliest work with gravitational force problems,
Replacing 1 4π εo by G and Q by M, we expect that
such as calculating the motion of the moon or artificial
a mass M destroys (rather than creates) an amount of
earth satellites, we got the correct answer by replacing
flux ΦG given by
the extended spherical earth by a point mass M e located
at the center of the earth. It is surprising that the ΦG = G(4πM) = 4πGM (52)
gravitational force exerted on you by every rock,
mountain, body of water, the earth's iron core, etc. all
adds up to be equivalent to the force that would be Gravitational Field of a Point Mass
exerted by a point mass M e located 6,000 km beneath Let us check if we have the correct flux formula by first
you. A simplified version of history is that Isaac calculating the gravitational field of a point mass. In
Newton delayed his publication of the theory of gravity Figure (30) we have a point mass M surrounded by a
20 years, and invented calculus, in order to show that Gaussian spherical surface of radius r. The flux flow-
the gravitational force of the entire earth was equivalent ing into the point mass is given by Equation (52) as
to the force exerted by a point mass located at the center. 4πGM. The flux flowing in through the Gaussian
surface is
One can do a brute force calculus derivation to prove
the above result, or one can get the result almost
flux in
immediately from Gauss' law. With Gauss' law, we can through
also find out, almost by inspection, how the gravita- ΦG Gaussian = g ⋅ A = gA⊥ = g 4πr 2
tional force decreases as we go down inside the earth. surface
(53)
Since gravity and electricity are both 1/r 2 forces,
Gauss' law also applies to gravity, and we can get the where the area of the sphere is 4πr2. Equating the flux
formulas for the gravitational version by comparing the in through the sphere (53) to the flux into M (52) gives
constants that appear in the force laws. Defining the
gravitational field g as the force on a unit mass (note that g 4πr 2 = 4πGM
this is also the acceleration due to gravity), we have for
GM
a point charge and a point mass g = (54)
r2

electric E = Q which is the gravitational force exerted on a unit mass.

field Q E (48)
4πεor 2

r
gravitational
field
g = GM M g (49) M g(r)
r2

Aside from the fact that gravitational forces are always mapping
attractive (therefore the field lines always go into a surface
mass m) the only other difference is that 1/4πε o is Figure 30
replaced by G Calculating the gravitational field g
of a point mass using Gauss' law.
G⇔ 1 (50)
4πε o
For the electric forces, a point charge Q produces an
amount of flux
24-24 Coulomb's Law and Gauss' Law

Gravitational Field Gravitational Field Inside the Earth

of a Spherical Mass The gravitational field outside a spherical mass is so
Now let us model the earth as a uniform sphere of mass easy to calculate using Gauss’ law, that you might
as shown in Figure (31). By symmetry the gravitational suspect that we really haven’t done anything. The
field lines must flow radially inward toward the center calculation becomes more interesting when we go
of the sphere. If we draw a Gaussian surface of radius down inside the earth and the field is no longer that of
r outside the earth, we have a total flux flowing in a point mass Me. By determining how the field de-
through the sphere given by creases as we go inside the earth, we can gain some
confidence in the calculational capabilities of Gauss’
Φin = g ⋅ A = gA⊥ law.
(55)
= g4π r 2
In Figure (32), we are representing the earth by a
Now the total amount of mass inside the sphere is Me , uniform sphere of matter of mass Me and radius Re.
so that the total amount of flux that must stop some- Inside the earth we have drawn a Gaussian surface of
where inside the Gaussian surface is radius r, and assume that the gravitational field has a
strength g(r) at this radius. Thus the total flux in
ΦG = 4πGMe (56)
through the Gaussian surface Φin is given by
Since Equations (56) and (55) are identical to Equa-
tions (52) and (53) for a point mass, we must get the Φin = g r ⋅ A = g r A⊥
(57)
same answer. Therefore the gravitational field outside = g r 4π r 2
a spherically symmetric mass Me is the same as the
field of a point mass Me located at the center of the Now the amount of mass inside our Gaussian surface
sphere. is no longer Me, but only the fraction of Me lying below
the radius r. That amount is equal to Me times the ratio
mapping
Me r surface
g(r) g(r)

Me
mapping r
surface
Figure 31
Re
Since 4 π GMe lines of flux go in through
the mapping surface in both Figures 30 and Figure 32
31, the field at the mapping surface must be To calculate the gravitational field inside the earth,
the same for both. This is why, when we are we draw a mapping surface inside, at a radius r less
above the surface of the earth, we can treat than the earth radius Re. The amount of mass Min
the earth as a point mass located at the inside the mapping surface is equal to the earth mass
center of the earth. Me times the ratio of the volume of the mapping
sphere to the volume of the earth.
 24-25

of the volume of a sphere of radius r to the volume of Exercise 8

the entire earth. As shown in Figure (33) a plastic ball of radius R has a
mass = M
Mr below
43 πr 3 total charge Q uniformly distributed throughout it. Use
r e 4 3 πR 3 Gauss’ law to:
e

3 a) Calculate the electric field E(r) outside the sphere

= Me r 3 (58) (r > R). How does this compare with the electric field of
Re a point charge?
The amount of flux that the mass M r absorbs is given b) Calculate the electric field inside the plastic sphere
by Equation (52) as (r < R).
3
ΦG = 4πGMr = 4πGMe r 3 (59) (Try to do this now. The solution is on the next page as
Re an example.)
Equating the flux flowing in through the Gaussian
surface (Equation (57) to the flux absorbed by M r
charge Q uniformly
(Equation 59) gives
spread throughout
3
g r 4π r 2 = 4πGMe r 3 the sphere
Re
R
or
g r = GMe r3 Gravitationalfield
inside the earth
(60)
R
This result, which can be obtained by a much more Figure 33
difficult calculus calculation, shows the earth’s gravi- Diagram for exercises 8&10. A plastic sphere of radius
R has a charge Q spread uniformly throughout. The
tational field dropping linearly (proportional to r), problem is to calculate the electric field inside and
going to zero as r goes to 0 at the center of the earth. outside.
Figure (32) gives an even more general picture of how
the earth’s gravitational field changes as we go down
inside. The flux going in past our Gaussian surface is
determined entirely by the mass inside the surface. The
mass in the spherical shell outside the Gaussian
surface has no effect at all! If we are down inside the
earth, a distance R i from the center, we can accurately
determine the gravitational force on us by assuming
that all the mass below us (r < R i) is located at a point
at the center of the earth, and all the mass above (r > R i)
does not exist.
24-26 Coulomb's Law and Gauss' Law

Solving Gauss' Law Problems 3) The simplest way to calculate the amount of charge
Using Gauss' law to solve for electric fields can be Qin inside our mapping surface is to note that since the
handled in a relatively straightforward way using the charge is uniformly spread throughout the sphere, Qin
following steps: is equal to the total charge Q times the ratio of the
volume inside the mapping surface to the total volume
1) Carefully sketch the problem. of the charged sphere; i.e.,
2) Draw a mapping surface that passes through the 4 3 πr 3 Qr 3
point where you want to solve for the field. Con- Qin = Q × =
4 3 πR 3 R3
struct the surface so that any field lines going
through the surface are perpendicular to the surface.
4) Now use Gauss' law to calculate E:
This way you can immediately spot the perpendicu-
lar area A⊥.
Q
3) Identify Qin , the amount of electric charge inside Φ = EA⊥ = ε in
o
your mapping surface.

4) Solve for E using the Equation Φ = EA⊥ = Qin/εo. Qr 3

E × 4πr 2 =
ε oR 3
5) Check that your answer is reasonable.

As an example, let us follow these steps to solve part (b) Qr

of Exercise 8, i.e., find the electric field E inside a E =
4πε oR 3
uniform ball of charge.

1) Sketch the problem. The sphere has a radius R, and 5) Check to see if the answer is reasonable. At r = 0,
total charge +Q. By symmetry the electric field we get E = 0. That is good, because at the center of
must go radially outward for a positive charge (or the sphere, there is no unique direction for E to
radially inward for a negative charge). point. At r = R, our formula for E reduces to
E = Q/(4πεoR 2) , which is the field of a point charge
Q when we are a distance R away. This agrees with
mapping
the idea that once we are outside a spherical charge,
E surface
the electric field is the same as if all the charge were
at a point at the center of the sphere.
r

2) Since we want the field inside the charged sphere,

we will use a spherical mapping surface of radius
r < R. Because the electric field is everywhere
perpendicular to the mapping surface, the area of the
mapping surface is A⊥ = 4πr 2.
 24-27

Exercise 9 Exercise 12
As shown in Figure (34), the inside of the plastic sphere A hydrogen atom ( H atom) consists of a proton with an
has been hollowed out. The total charge on the sphere electron moving about it. The classical picture is that the
is Q. Use Gauss’ law to electron orbits about the proton much like the earth
orbits the sun. A model that has its origins in quantum
a) Determine the strength of the electric field inside the mechanics and is more useful to chemists, is to picture
hollow cavity. the electron as being smeared out, forming a ball of
b) Calculate the strength of the electric field inside the negative charge surrounding the proton. This ball of
plastic. negative charge is called an "electron cloud ".

For this problem, assume that the electron cloud is a

c) Calculate the strength of the electric field outside
uniform sphere of negative charge, a sphere of radius
the plastic.
R centered on the proton as shown in Figure (35). The
total negative charge (–e) in the electron cloud just
balances the positive charge (+e) on the proton, so that
charge Q spread the net charge on the H atom is zero.
throughout the
a) Sketch the electric field for this model of the H atom.
spherical shell Show the electric field both inside and outside the
R
Ri electron cloud.

b) Calculate the magnitude of the electric field for both

r < R (inside the cloud) and r > R (outside the cloud).

Figure 34
Diagram for Exercises 9 & 11. The charge is now
spread throughout a spherical shell.
Electron
cloud
Exercise 10
Repeat Exercise 8, assuming that Figure (33) repre-
R
sents the end view of a very long charged plastic rod Proton
with a charge of λ coulombs per meter. (A section of
length L will thus have a charge Q = λ L.)

Exercise 11
Figure 35
Repeat Exercise 9, assuming that Figure (34) repre- Picture a hydrogen atom as a proton (of charge +e),
sents the end view of a charged hollow plastic rod with surrounded by an electron cloud. Think of the cloud as
a charge of λ coulombs per meter. a uniform ball of negative charge, with a net charge -e.
24-28 Coulomb's Law and Gauss' Law

Exercise 13 Exercise 14
A butterfly net with a circular opening of radius R, is in Electric fields exist in the earth's atmosphere. (You get
a uniform electric field of magnitude E as shown in lightning if they get too strong.)
Figure (36). The opening is perpendicular to the field.
On a particular day, it is observed that at an altitude of
Calculate the net flux of the electric field through the net
300 meters, there is a downward directed electric field
itself. (The amount of flux through each hole in the net is
of magnitude
E ⋅ dA where dA is the area of the hole.) (This is one of
our favorite problems from Halliday and Resnick.) E 300 m = 70 newton
coulomb

Down at an altitude of 200 meters, the electric field still

points down, but the magnitude has increased to

E 200 m = 100 newton

coulomb

R E How much electric charge is contained in a cube 100

meters on a side in this region of the atmosphere?

altitude
300 m
E

Figure 36 altitude
Butterfly net in a uniform electric field. With 200 m
Gauss' law, you can easily calculate the flux of
the electric field through the net itself.
E
Figure 37
Electric field at two different altitudes.
 24-29

PROBLEM SOLVING Instead, if you cannot get it, just peek at the solution to
One can devise a number of Gauss’ law problems see what point you missed, then put the solution away
where one plugs various numerical values into the and solve it on your own. You may have to peek a
formulas we have derived. But that is not the point of couple of times, but that is OK.
this chapter. Here we are interested, not so much in the If you had to peek at the solution, then wait another day
answers, as in the concepts and techniques used to or so, clean off your desk, and try again. Soon you will
derive them. get the solution without looking, and you will not forget
In this chapter we have introduced two new concepts. how to solve that problem. You will get more out of this
One is electric flux Φ = EA ⊥ , and the other is that the technique than solving 15 numerical examples.
total flux out through a closed surface is equal to When you are studying a new topic with new, unfamil-
Q inside/ε 0 . These two concepts allow us to easily solve iar concepts, the best way to learn the subject is to
for the electric field in certain special cases. Those thoroughly learn a few, well chosen worked out ex-
cases are where A ⊥ is a sphere, a cylinder, or a plane. amples. By learn, we mean problems you can work on
What you need to get from this discussion is the a blank sheet of paper without looking at a solution.
beginning of an intuitive picture of electric flux, how it Pick examples that are relatively simple but clearly
is related to the flow of an incompressible fluid, and illustrate the concepts involved. For this chapter, one
how this concept can be used to handle the few but could pick the example of calculating the electric field
important examples where A ⊥ is either a sphere, inside and outside a uniform ball of charge. If you can
cylinder, or plane. Numerical applications can come do that problem on a clean desk, you can probably do
later, now is the time to develop intuition. most of the other problems in this chapter without too
Most students have some difficulty handling Gauss’s much difficulty.
law problems the first time they see them because the Why learn a sample problem for each new topic? The
concepts involved are new and unfamiliar. Then when reason is that if you know one worked example you will
the problems are solved in a homework session, a find it easy to remember the entire topic. That worked
common reaction is, “Oh, those are not so hard after example reminds you immediately how that concept
all”. One gets the feeling that by just watching the works, how it functions.
problem solved, and seeing that it is fairly easy after all,
they understand it. The rude shock comes at an exam In this text, Chapters 24-32 on electric and magnetic
where suddenly the problem that looked so easy, has fields involve many new concepts. Concepts you will
become unsolvable again. not have seen unless you have already taken the course.
As we go along, we will suggest sample problems,
There is a way to study to avoid this rude shock. Pick what we call “clean desk problems”, which serve as a
one of the problems you could not solve on your own, good example of the way the new concept is used. You
a problem you saw solved in class or on an answer may wish to choose different sample problems, but the
sheet. A problem that looked so easy after you saw it best way to learn this topic is to develop a repertory of
solved. selected sample problems you understand cold.

Wait a day or two after you saw the solution, clean off At this point, go back to some of the problems in this
your desk, take out a blank sheet of paper, and try chapter, particularly Exercises 9 through 14 and see if
solving the problem. you can solve them on a clean desk. If you can, you are
ready for the next chapter.
Something awful may happen. That problem that
looked so easy in the homework review session is now
impossible again. You can’t see how to do it, and it
looked so easy two days ago. You feel really bad—but
don’t, it happens to everyone.
Chapter 25
Field Plots and
Electric Potential

CHAPTER 25 FIELD PLOTS AND ELEC-

TRIC POTENTIAL

Calculating the electric field of any but the simplest THE CONTOUR MAP
distribution of charges can be a challenging task. Figure (1) is a contour map of a small island. The
Gauss’ law works well where there is considerable contour lines, labeled 0, 10, 20, 30 and 40 are lines of
symmetry, as in the case of spheres or infinite lines of equal height. Anywhere along the line marked 10 the
charge. At the beginning of Chapter (24), we were able land is 10 meters above sea level. (You have to look at
to use a brute force calculus calculation to determine some note on the map that tells you that height is
the electric field of a short charged rod. But to handle measured in meters, rather than feet or yards.)
more complex charge distributions we will find it
helpful to apply the techniques developed by map You can get a reasonable understanding of the terrain
makers to describe complex terrains on a flat map. This just by looking at the contour lines. On the south side
is the technique of the contour map which works of the island where the contour lines are far apart, the
equally well for mapping electric fields and mountain land slopes gradually upward. This is probably where
ranges. Using the contour map ideas, we will be lead the beach is located. On the north side where the
to the concept of a potential and equipotential lines or contour lines are close together, the land drops off
surfaces, which is the main topic of this chapter. sharply. We would expect to see a cliff on this side of
the island.

Figure 1
Contour map of a small island with a beach on the
south shore, two hills, and a cliff on the northwest N
side. The slope of the island is gradual where the 40 40
lines are far apart, and steep where the lines are 30 30
20
close together. If you were standing at the point A
labeled (A) and the surface were slippery, you 10
would start to slide in the direction of the arrow. 0
vel
ea le
s
25-2 Field Plots and Electric Potential

Although we would rather picture this island as being Skiers are familiar with this concept. When you want
in the south seas, imagine that it is in the North Atlantic to stop and rest and the slope is icy, you plant your skis
and a storm has just covered it with a sheet of ice. You along a contour line so that they will not slide either
are standing at the point labeled A in Figure (1), and forward or backward. The direction of steepest descent
start to slip. If the surface is smooth, which way would is now perpendicular to your skis, in a direction that ski
you start to slip? instructors call the fall line. The fall line is the direction
you will start to slide if the edges of your skis fail to
A contour line runs through Point A which we have hold.
shown in an enlargement in Figure (2). You would not
start to slide along the contour line because all the In Figure (3), we have redrawn our contour map of the
points along the contour line are at the same height. island, but have added a set of perpendicular lines to
Instead, you would start to slide in the steepest down- show the directions of steepest descent, the direction of
hill direction, which is perpendicular to the contour line the net force on you if you were sitting on a slippery
as shown by the arrow. surface. These lines of steepest descent, are also called
lines of force. They can be sketched by hand, using the
If you do not believe that the direction of steepest rule that the lines of force must always be perpendicular
descent is perpendicular to the contour line, choose any to the contour lines.
smooth surface like the top of a rock, mark a horizontal
line (an equal height line) for a contour line, and
carefully look for the directions that are most steeply
sloped down. You will see that all along the contour
line the steepest slope is, in fact, perpendicular to the
contour line.

direction of steepest
descent contour lines
A
A

Figure 2 Figure 3
Along a contour line the land is level. The You can sketch in the lines of steepest descent
direction of steepest slope or descent is by drawing a set of lines that are always
perpendicular to the contour line. perpendicular to the contour lines. These
lines indicate the direction a ball would start
to roll if placed at a point on the line.
 25-3

In Figure (4) we have the same island, but except for the EQUIPOTENTIAL LINES
zero height contour outlining the island, we show only On a contour map of an island, the contour lines are
the lines of force. The exercise here, which you should lines of equal height. If you walk along a contour line,
do now, is sketch in the contour lines. Just use the rule your height h, and therefore your gravitational poten-
that the contour lines must be drawn perpendicular to tial energy mgh, remains constant. As a result, we can
the lines of force. The point is that you can go either call these the lines of constant or equal potential energy,
way. Given the contour lines you can sketch the lines equipotential energy lines for short.
of force, or given the lines of force you can sketch the
contour lines. This turns out to be a powerful technique Let us apply these mapping concepts to the simpler
in the mapping of any complex physical or mathemati- situation of a spherical mass M shown in Figure (5). As
cal terrain. in Chapter 24, we have drawn the gravitational field
lines, which point radially inward toward the center of
M. We determined these field lines by placing a test
particle of mass m in the vicinity of M as shown in
Figure (6). The potential energy of this test mass m is
given by our old formula (see Equation 10-50a)
attractiveforces
GMm
PE = – have negative (1)
r potentialenergy

In our discussion of electric and gravitational fields, we

defined the fields as the force on a unit test particle.
Setting m = 1 for a unit test mass, we get as the formula
for the potential energy of our unit test mass
Potential energy of GM
= – (2)
a unit test mass r
Figure 4
Here we have removed the contour lines leaving
only the lines of steepest descent and the outline gravitational
of the island. By drawing a set of lines field lines
perpendicular to the lines of steepest descent, r
you can more or less reconstruct the contour
lines. The idea is that you can go back and forth M
from one set of lines to the other.
equipotential
lines

m Figure 5
M The gravitational field lines for a spherical mass
point radially inward. The lines of constant
r potential energy are circles of equal height above
the mass. The equipotential lines are everywhere
Figure 6
perpendicular to the field lines, just as, in the map
The gravitational potential energy of two masses
of the island, the contour lines were everywhere
separated by a distance r is – GMm r .
perpendicular to the lines of steepest descent.
25-4 Field Plots and Electric Potential

But there is one important difference between Figures

From Equation (2) we see that if we stay a constant (5) and (7). In Figure (5) the gravitational force on our
distance r out from M, if we are one of the concentric unit test mass is attractive, in toward the mass M. In
circles in Figure (5), then the potential energy of the Figure (7), the force on our unit positive test particle is
unit test mass remains constant. These circles, drawn out, away from Q if Q is positive. When we have an
perpendicular to the lines of force, are again equal attractive force as in Figure (5), the potential energy
potential energy lines. is negative as in Equation (1). But when the force is
There is a convention in physics to use the word repulsive, as in Figure (7), the potential energy is
potential when talking about the potential energy of a positive. Let us briefly review the physical origin for
unit mass or unit charge. With this convention, then this difference in the sign of the potential energy.
– GM/r in Equation (2) is the formula for the gravita- In any discussion of potential energy, it is necessary to
tional potential of a mass M, and the constant radius define the zero of potential energy, i.e. to say where the
circles in Figure (5) are lines of constant potential. floor is. In the case of satellite motion, we defined the
Thus the name equipotential lines for these circles is satellite’s potential energy as being zero when the
fitting. satellite was infinitely far away from the planet. If we
release a satellite at rest a great distance from the planet,
Negative and Positive it will start falling toward the planet. As it falls, it gains
Potential Energy kinetic energy, which it must get at the expense of
In Figure (7) we have drawn the electric field lines of gravitational potential energy. Since the satellite started
a point charge Q and drawn the set of concentric circles with zero gravitational potential energy when far out
perpendicular to the field lines as shown. From the and loses potential energy as it falls in, it must end up
close analogy between the electric and gravitational with negative potential energy when it is near the
force, we expect that these circles represent lines of planet. This is the physical origin of the minus sign in
constant electric potential energy, that they are the Equation (1). Using the convention that potential
electric equipotential lines. energy is zero at infinity, then attractive forces lead to
negative potential energies.
electric If the force is repulsive as in Figure (7), then we have
field lines to do work on our test particle in order to bring it in from
r infinity. The work we do against the repulsive force is
stored up as positive potential energy which could be
+Q
released if we let go of the test particle (and the test
particle goes flying out). Thus the convention that
equipotential potential energy is zero at infinity leads to positive
lines potential energies for repulsive forces like that shown
in Figure (7).

Figure 7
The electric potential is the potential energy of a
positive unit test charge qtest = + 1 coulomb.
Because a positive charge + Q and a positive test
charge repel this potential energy is positive.
 25-5

ELECTRIC POTENTIAL CONSERVATIVE FORCES

OF A POINT CHARGE (This is a formal aside to introduce a point that we will
Using the fact that we can go from the gravitational treat in much more detail later.)
force law to Coulomb’s law by replacing GMm by
Qq 4πε0 (see Exercise 1), we expect that the formula Suppose we have a fixed charge Q and a small test
for the electric potential energy of a charge q a distance particle q as shown in Figure (8). The potential energy
r from Q is of q is defined as zero when it is infinitely far away from
Q. If we carry q in from infinity to a distance r, we do
electric potential energy Qq an amount of work on the particle
= + (3)
of a charge q 4πε0 r r
Work we do = Fus ⋅ dx (5)
The + sign in Equation (3) indicates that for positive Q ∞
and q we have a repulsive force and positive potential If we apply just enough force to overcome the electric
energy. (If Q is negative, but q still positive, the force repulsive force, if Fus = –qE , then the work we do
is attractive and the potential energy must be negative.) should all be stored as electric potential energy, and
To determine the potential energy of a unit test charge, Equation (5), with Fus = –qE should give us the
we set q = 1 in Equation (3) to get correct electric potential energy of the charge q.

But an interesting question arises. Suppose we bring

electric potential the charge q in along two different paths, paths (1) and
Qq
energy of a unit ≡ electric = (2) shown in Figure (8). Do we do the same amount of
potential 4πε0 r work, store the same potential energy for the two
test charge
different paths?
(4) dr Fe
Fus q pa
Following the same convention we used for gravity, we th
1
will use the name electric potential for the potential
energy of a unit test charge. Thus Equation (4) is the
formula for the electric potential in the region sur- Q r dr
Fus
rounding the charge Q. As expected, the lines of equal Fe h2
q
p at
potential, the equipotential lines are the circles of
constant radius seen in Figure (7). Figure 8
If we bring a test particle q in from infinity to
Exercise 1 a distance r from the charge Q, the electric
potential energy equals Qq/4π εor . But this
Start with Newton's gravitational force law, replace potential energy is the work we do in bringing
GMm by Qq 4π ε0 , and show that you end up with q in from infinity:
Coulomb's electrical force law. r
Qq
Fus ⋅ dr =
4π ε o r
∞

This answer does not depend upon the path we

take bringing q in.
25-6 Field Plots and Electric Potential

If we lift an eraser off the floor up to a height h, and ELECTRIC VOLTAGE

hold it still, then it does not matter what path we took, In our discussion of Bernoulli’s equation, we gave the
the net amount of work we did was mgh and this is collection of terms (P + ρgh + 1/2ρv2 ) the name hy-
stored as gravitational potential energy. When the drodynamic voltage. The content of Bernoulli’s equa-
work we do against a force depends only on the initial tion is that this hydrodynamic voltage is constant along
and final points, and not on the path we take, we say that a stream line when the fluid is incompressible and
the force is conservative. viscous forces can be neglected. Two of the three
In contrast, if we move the eraser over a horizontal table terms, ρgh and 1/2ρv2 represent the energy of a unit
from one point to another, the amount of work we do volume of the fluid, thus we see that our hydrodynamic
against friction depends very much on the path. The voltage has the dimensions of energy per unit volume.
longer the path the more work we do. As a result we Electric voltage is a quantity with the dimensions of
cannot define a friction potential energy because it energy per unit charge that in different situations is
has no unique value. Friction is a non-conservative represented by a series of terms like the terms in
force, and non-conservative forces do not have unique Bernoulli’s hydrodynamic voltage. There is the poten-
potential energies. tial energy of an electric field, the chemical energy
The gravitational fields of stationary masses and the supplied by a battery, even a kinetic energy term, seen
electric fields of stationary charges all produce conser- in careful studies of superconductors, that is strictly
vative forces, and therefore have unique potential 2 2 term in Bernoulli’s equation.
analogous to the 1 2ρv
energies. We will see however that moving charges In other words, electric voltage is a complex concept,
can produce electric fields that are not conservative! but it has one simplifying feature. Electric voltages are
When that happens, we will have to take a very careful measured by a common experimental device called a
look at our picture of electric potential energy. But in voltmeter. In fact we will take as the definition of
dealing with the electric fields of static charges, as we electric voltage, that quantity which we measure using
will for a few chapters, we will have unique electric a voltmeter.
potential energies, and maps of equipotential lines will This sounds like a nebulous definition. Without telling
have an unambiguous meaning. you how a voltmeter works, how are you to know what
the meter is measuring? To overcome this objection,
we will build up our understanding of what a voltmeter
measures by considering the various possible sources
of voltage one at a time. Bernoulli’s equation gave us
all the hydrodynamic voltage terms at once. For
electric voltage we will have to dig them out as we find
them.

Our first example of an electric voltage term is the

electric potential energy of a unit test charge. This has
the dimensions of energy per unit charge which in the
MKS system is joules/coulomb and called volts.
joule
1 ≡ 1 volt (6)
Coulomb
 25-7

In Figure (9), which is a repeat of Figure (8) showing This observation suggests an experimental way to map
the electric field lines and equipotential lines for a point equipotential lines or surfaces. Attach one lead of the
charge Q, we see from Equation (4) that a unit test voltmeter to some particular point, call it Point (A).
particle at Point (1) has a potential energy, or voltage V1 Then move the other lead around. Whenever you get
given by a zero reading on the voltmeter, the second lead must
Q be at another point of the same equipotential line as
electricpotentialor
V1 = Point (A). By marking all the points where the meter
4πε0r1 voltage at Point(1)
reads zero, you get a picture of the equipotential line.
At Point (2), the electric potential or voltage V2 is given The discussion we have just given for finding the
by equipotential lines surrounding a point charge Q is not
Q electricpotentialor practical. This involves electrostatic measurements
V2 = that are extremely difficult to carry out. Just the damp
4πε0r2 voltage at Point(2)
air from your breath would affect the voltages sur-
Voltmeters have the property that they only measure rounding a point charge, and typical voltmeters found
the difference in voltage between two points. Thus if in the lab cannot make electrostatic measurements.
we put one lead of a voltmeter at Point (1), and the other Sophisticated meters in carefully controlled environ-
at Point (2) as shown, then we get a voltage reading V ments are required for this work.
given by
But the idea of potential plotting can be illustrated
voltmeter Q 1 1 nicely by the simple laboratory apparatus illustrated in
V ≡ V2 – V1 = –
reading 4πε0 r2 r1 Figure (10). In that apparatus we have a tray of water
(slightly salty or dirty, so that it is somewhat conduc-
If we put the two voltmeter leads at points equal tive), and two metal cylinders attached by wire leads to
distances from Q, i.e. if r1 = r2 , then the voltmeter a battery as shown. There are also two probes consist-
would read zero. Since the voltage difference between ing of a bent, stiff wire attached to a block of wood and
any two points on an equipotential line is zero, the adjusted so that the tips of the wires stick down in the
voltmeter reading must also be zero when the leads are water. The other end of the probes are attached to a
attached to any two points on an equipotential line. voltmeter so we can read the voltage difference be-
tween the two points (A) and (B), where the probes
touch the water.

volt
– 2
battery
V probes
meter r2
+ 1
r1 A volt
V meter
brass B
cylinders

Figure 9
A voltmeter measures the difference in electrical tap water pyrex dish
voltage between two points.
Figure 10
Simple setup for plotting fields. You plot
equipotentials by placing one probe (A) at a given
position and moving the other (B) around.
Whenever the voltage V on the voltmeter reads
zero, the probes are at points of equipotential.
25-8 Field Plots and Electric Potential

If we keep Probe (A) fixed and move Probe (B) around,

whenever the voltmeter reads zero, Probe (B) will be on
the equipotential line that goes through Point (A).
Without too much effort, one can get a complete plot of
the equipotential line. Each time we move Probe (A)
we can plot a new equipotential line. A plot of a series
of equipotential lines is shown in Figure (11).

Once we have the equipotential lines shown in Figure

(11), we can sketch the lines of force by drawing a set
of lines perpendicular to the equipotential as we did in
Figure (12). With a little practice you can sketch fairly
accurate plots, and the beauty of the process is that you
did not have to do any calculations!
Figure 11
Plot of the equipotential lines from a student project by
B. J. Grattan. Instead of a tray of water, Grattan used a
sheet of conductive paper, painting two circles with
aluminum paint to replace the brass cylinders. (The
conductive paper and the tray of water give similar
results.) We used the Adobe Illustrator program to
draw the lines through Grattan's data points.

Figure 12
It does not take too much practice to sketch in the field lines. Draw smooth lines, always
perpendicular to the equipotential lines, and maintain any symmetry that should be there.
 25-9

Hint—Look at the electric field at Point A in Figure (12),

Exercise 2
enlarged in Figure (13). We know that the field E at Point
The equipotential plot of Figure (11) and the field lines
A is made up of two components, E1 directed away from
of Figure (12) were taken from a student project. The
the left hand cylinder, and E2 directed toward the right
field lines look like the field of two point charges +Q and
hand cylinder, and the net field E is the vector sum of the
-Q separated by a distance r. But who knows what is
two components. If the field is the field of point charges
happening in the shallow tank of water (or a sheet of
then E1 drops off as 1/r12 and E2 as 1/r22 . But if the field
conducting paper)? Perhaps the field lines more nearly
is that of line charges, E1 drops off as 1/r1 and E2 as
represent the field of two line charges + λ and - λ
1/r2 . We have chosen Point (A) so that r1 , the distance
separated by a distance r.
from (A) to the left cylinder is quite a bit longer than the
The field of a point charge drops off as 1/r2 while the field distance r2 to the right cylinder. As a result, the ratio of
of a line charge drops off as 1/r. The point of the exercise E1 to E2 and thus the direction of E , will be quite
is to decide whether the field lines in Figure (12) (or your different for 1 r and 1 r2 forces. This difference is great
own field plot if you have constructed one in the lab) enough that you can decide, even from student lab
more closely represent the field of a point or a line results, whether you are looking at the field of point or
charge. line charges. Try it yourself and see which way it comes
ive) out.
puls
A E1 (re

e1
rd charg
towa
E = E1 + E 2
E2
tow
ard
cha
rge
2

Figure 13
Knowing the direction of the electric field at Point (A)
allows us to determine the relative magnitude of the
fields E1 and E2 produced by charges 1 and 2 alone.
At Point (A), construct a vector E of convenient length
parallel to the field line through (A). Then decompose
E into component vectors E1 and E2 , where E1 lies
along the line from charge 1 to Point (A), and E2
along the line toward charge 2. Then adjust the
lengths of E1 and E2 so that their vector sum is E .
25-10 Field Plots and Electric Potential

A Field Plot Model

The analogy between a field plot and a map maker’s
contour plot can be made even more obvious by
constructing a plywood model like that shown in
Figure (14).

To construct the model, we made a computer plot of the

electric field of charge distribution consisting of a
charge +3 and –1 seen in Figure (15). We enlarged the
computer plot and then cut out pieces of plywood that
had the shapes of the contour lines. The pieces of
plywood were stacked on top of each other and glued
together to produce the three dimensional view of the
field structure.

In this model, each additional thickness of plywood

represents one more equal step in the electric potential
or voltage. The voltage of the positive charge Q = +3
is represented by the fat positive spike that goes up
toward + ∞ and the negative charge q = –1 is repre-
sented by the smaller hole that heads down to – ∞ .
These spikes can be seen in the back view in Figure
(14), and the potential plot in Figure (16).
Figure 14
Model of the electric field in the region of two point
In addition to seeing the contour lines in the slabs of
charges Q+ = + 3, Q– = – 1. Using the analogy to a
topographical map, we cut out plywood slabs in the plywood, we have also marked the lines of steepest
shape of the equipotentials from the computer plot of descent with narrow strips of black tape. These lines of
Figure 15, and stacked the slabs to form a three steepest descent are always perpendicular to the con-
dimensional surface. The field lines, which are tour lines, and are in fact, the electric field lines, when
marked with narrow black tape on the model, always viewed from the top as in the photograph of Figure (15).
lead in the direction of steepest descent on the surface.
Figure (17) is a plywood model of the electric potential
for two positive charges, Q = +5, Q = +2. Here we get
two hills.
.1. 2V

Figure 16 .1.1V

Potential plot along the .1.0V

line of the two charges .9V

+3, –1. The positive .8V

.7V
charge creates an upward
.6V
spike, while the negative
.5V
charge makes a hole.
.4V
.3V
.2V
.1V
–1

+3
–.1V
–.2V Figure 17
–.3V Model of the electric potential in the region
–.4V of two point charges Q = +5 and Q = +2.
 25-11

V = .1

V = .2

V = .3

V = .4

V = .5

.1
V=
0 1
.
V=
–.
V=

–1 +3
Figure 15
Computer plot of the field lines and equipotentials for a charge distribution
consisting of a positive charge + 3 and a negative charge – 1. These lines
were then used to construct the plywood model.
25-12 Field Plots and Electric Potential

Computer Plots Exercise 4

There are now many excellent programs that have Figures (20a) and (20b) are computer plots of the
personal computers draw out field plots for various electric field of opposite charges. One of the plots
charge distributions. In most of these programs you represents the 1/r 2 field of 3 dimensional point charges.
enter an array of charges and the computer draws the The other is the end view of the 1/r field of line charges.
field and equipotential lines. You should practice with You are to decide which is which, explaining how you
one of these programs in order to develop an intuition can tell.
for the field structures various charge distributions
produce. In particular, try the charge distribution
shown in Figure (18) and (19). In Figure (18), we wish
to see the field of oppositely charged plates (a positive
plate on the left and a negative one on the right). This
charge distribution will appear in the next chapter in
our discussion of the parallel plate capacitor.

In Figure (19) we are modeling the field of a circle or

in 3-dimensions a hollow sphere of charge. Something
rather remarkable happens to the electric field lines in
this case. Try it and see what happens! a)

Exercise 3
If you have a computer plotting program available, plot
the field lines for the charge distributions shown in
Figures (18,19), and explain what the significant fea-
tures of the plot are.

+ –
+ –
+ –
+ –
b)
+ –
Figure 18 Figure 20
The idea is to use the computer to develop an intuition Computer plots of 1/r and 1 / r 2 (two dimensional
for the shape of the electric field produced by various and three dimensional) fields of equal and opposite
distributions of electric charge. Here the parallel lines charges. You are to figure out which is which.
of charge simulates two plates with opposite charge.

+
+ +
+ +
+ +
+ + +
Figure 19
We have placed + charges around a circle to
simulate a cylinder or sphere of charge. You
get interesting results when you plot the field
lines for this distribution of charge.
Chapter 26
Electric Fields
and Conductors

CHAPTER26 ELECTRIC FIELDS

AND CONDUCTORS

In this chapter we will first discuss the behavior of ELECTRIC FIELD

electric fields in the presence of conductors, and then INSIDE A CONDUCTOR
apply the results to three practical devices, the Van de If we have a piece of metal a few centimeters across as
Graaff generator, the electron gun, and the parallel illustrated in Figure (1), and suddenly turn on an
plate capacitor. Each of these examples provides not electric field, what happens? Initially the field goes
only an explanation of a practical device, but also helps right through the metal. But within a few pico seconds
build an intuitive picture of the concept of electric (1 pico second = 10 –12 seconds) the electrons in the
voltage. metal redistribute themselves inside the metal creating
their own field that soon cancels the external applied
electric field, as indicated in Figure (2).

metal

E=?
Figure 1
What is the electric field inside a chunk of metal?
Metals have conduction electrons that are free to move.
If there were an electric field inside the metal, the
conduction electrons would be accelerated by the field.

– +
– +
– +
– +
Figure 2
If you place a chunk of metal in an
external electric field, the electrons move
until there is no longer a force on them.
26-2 Electric Fields and Conductors

The very concept of an electrical conductor requires If there is no charge inside the conductor, then the only
that, in the steady state, there be no electric field inside. place any charge can exist is in the surface. If there is
To see why, imagine that there is a field inside. Since a redistribution of charge, the redistributed charge must
it is a conductor, the electrons in the conductor are free lie on the surface of the conductor.
to move. If there is a field inside, the field will exert a
force on the electrons and the electrons will move. Figure (4) is a qualitative sketch of how surface charge
They will continue to move until there is no force on can create a field that cancels the applied field.
them, i.e., until there is no field remaining inside. The
electrons must continue to move until the field they
create just cancels the external field you applied.
applied
metal electric
Surface Charges field
Where does the redistributed charge have to go in order E
to create an electric field that precisely cancels the
applied electric field? Gauss’ law provides a remark-
ably simple answer to this question. The redistributed Figure 4a
charge must reside on the surface of the conductor. An external field is applied to a block of metal.
This is because Gauss’ law requires that there be no net
charge inside the volume of a conductor.

To see why, let us assume that a charge Q is inside a – +

conductor as shown in Figure (3). Draw a small
Gaussian surface around Q. Then by Gauss’ law the – +
E
flux Φ = E⋅A coming out through the Gaussian sur- – E' +
face must be equal to Q in ε0 where Q in is the net charge
inside the Gaussian surface. But if there is no field
– +
inside the conductor, if E = 0, then the flux E⋅A out
through the Gaussian surface must be zero, and there- Figure 4b
fore the charge Q in must be zero. In response to the electric field, the electrons move to
the left surface of the metal, leaving behind positive
charge on the right surface. These two surface charges
have their own field E ′ that is oppositely directed to E .

– +
+Q
– +
E – E + E' = 0 + E
Figure 3 – +
Is there any electric charge inside a conductor? To – +
find out, draw a Gaussian surface around the
suspected charge. Since there is no electric field Figure 4c
inside the conductor, there is no flux out through Inside the block of metal the fields cancel. The
the surface, and therefore no charge inside. result is that the external field on the left stops on
the negative surface charge. The field on the right
starts again on the positive surface charge.
 26-3

In Figure (4a) we see the electric field just after it has Gauss’ law can be used to calculate how much charge
been turned on. Since the electrons in the metal are must be at the surface if a field of strength E is
negatively charged (q = -e), the force on the electrons impinging as shown in Figure (5). In that figure we
F = (-e) E is opposite to E and directed to the left. have drawn a small pill box shaped Gaussian surface,
with one end in the conductor and the other outside in
In Figure (4b), electrons have been sucked over to the the field E . If the area of the end of the pill box is dA,
left surface of the metal, leaving positive charge on the then the flux out of the pill box on the right is
right surface. The negative charge on the left surface Φout = EdA .
combined with the positive charge on the right pro-
duced the left directed field E′ shown by the dotted Let σ coulombs/meter2 be the charge density on the
lines. The oppositely directed fields E and E′ cancel in surface. The amount of the conductor’s surface sur-
Figure (4c) giving no net field inside the metal. rounded by the pill box is dA, thus the amount of charge
inside the pill box is
Surface Charge Density
Amount of charge
When a field E impinges on the surface of a conductor,
inside the ≡ Qin = σ dA
it must be oriented at right angles to the conductor as
Gaussian surface
shown in Figure (5). The reason for this is that if E had
a component E|| parallel to the surface, E|| would pull By Gauss’ law, the flux Φout = EdA must equal 1/ε0
the movable charge along the surface and change the times the total charge inside the Gaussian surface and
charge distribution. The only direction the surface we get
charge cannot be pulled is directly out of the surface of
the conductor, thus for a stable setup the electric field Qin σdA
at the surface must be perpendicular as shown. Φout = = = EdA
ε0 ε0

surface + The dA's cancel and we are left with

charge +
σ coul/m 2 +
+ σ E = electric field at the conductor
+ E =
E ε0 σ = charge densityat the surface (1)
+
+
metal +
+ Equation (1) gives a simple relation between the strength
area of the electric field at the surface of a conductor, and the
dA surface charge density σ at that point. Just remember
that the field E must be perpendicular to the surface of
E
the conductor. (If the applied field was not originally
perpendicular to the surface, surface charges will slide
along the surface, reorienting the external field to make
Charged surface inside pillbox. it perpendicular.)
The amount of charge on this
surface is σdA To appreciate how far we have come with the concepts
Figure 5
of fields and Gauss’ law, just imagine trying to derive
To calculate the surface charge density, we draw a Equation (1) from Coulomb’s law. We wouldn’t even
small cylindrical pill box of cross-sectional area know how to begin.
dA. We then equate the flux of electric field out
through the right surface of the pill box to We will now work an example and assign a few
1/εεotimes the charge inside the pill box. exercises to build an intuition for the behavior of fields
and conductors. Then we will apply the results to some
practical devices.
26-4 Electric Fields and Conductors

Example: Field metal

in a Hollow Metal Sphere
Suppose we have the hollow metal sphere shown in
Figure (6). A total charge Q is placed on the sphere.
What are the electric fields outside and inside the
sphere?

One key to solving this problem is to realize that since Figure 6

We place a charge Q on a hollow metal sphere.
the sphere is symmetric, the fields it produces must also Where do the charge and the field lines go?
be symmetric. We are not interested in fields that do
one thing on the left side and something else on the
Gaussian
right, for we do not have any physical cause for such an
surface
asymmetry. E
In Figure (7) we have drawn a Gaussian surface metal
surrounding the metal sphere as shown. Since there is
a net charge +Q on the sphere, and therefore inside the ro
Gaussian surface, there must be a net flux Q ε0 out
through the surface. Since the Gaussian surface has an
area 4π r2 , Gauss' law gives

2 Q
Φ = EoutA⊥ = Eout × 4πr =
ε0

Q Figure 7
Eout = (2) If we place a Gaussian surface around and outside
4πε0 r2 the sphere, we know that the charge Q must be
inside the Gaussian surface, and therefore Q / ε o
which happens to be the field of a point charge. lines must come out through the surface

In Figure (8) we have drawn a Gaussian surface inside

the metal at a radius ri. Since there is no field inside the Gaussian
metal, EA⊥ = 0 and there is no flux flowing out surface
through the Gaussian surface. Thus by Gauss' law there E
can be no net charge inside the Gaussian surface. metal
Explicitly this means that there is no surface charge on
the inside of the conductor. The charge Q we spread on
ri
the conducting sphere all went to the outside surface!

Finally in Figure (9) we have drawn a Gaussian surface

inside the hollow part of the hollow sphere. Since there
is no charge—only empty space inside this Gaussian
surface, there can be no flux out through the surface,
and the field E inside the hollow part of the sphere is
exactly zero. This is a rather remarkable result consid- Figure 8
ering how little effort was required to obtain it. If we place our Gaussian surface inside the metal
where E = 0 , no lines come out through the Gaussian
surface and therefore there must be no net charge Q
inside the Gaussian surface. The fact that there is no
charge within that surface means all the charge we
placed on the sphere spreads to the outside surface.
 26-5

Exercise 1 Exercise 2
A positive charge +Q is surrounded concentrically by a A chunk of metal has an irregularly shaped cavity inside
conducting sphere with an inner radius ra and outer as shown in Figure (11). There are no holes and the
radius rb as shown in Figure (10). The conducting cavity is completely surrounded by metal.
sphere has no net charge. Using Gauss’ law, find the
electric field inside the hollow section (r < ra) , inside the The metal chunk is struck by lightning which produces
conducting sphere (ra < r < rb) and outside the sphere huge electric fields and deposits an unknown amount of
(r > rb) . Also calculate the surface charge densities on charge on the metal, but does not burn a hole into the
the inner and outer surfaces of the conducting sphere. cavity. Show that the lightning does not create an
Show that Equation (1) applies to the charge densities electric field inside the cavity. (For a time on the order of
you calculate. pico seconds, an electric field will penetrate into the metal, but
if the metal is a good conductor like silver or copper, the
distance will be very short.)

(What does this problem have to do with the advice to

stay in a car during a thundershower?)
ra
+Q
rb
metal

Figure 10
cavity
Start with an uncharged hollow metal sphere and place
a charge +Q inside. Use Gauss' law to determine the
electric field and the surface charges throughout the
region. Figure 11
A chunk of metal with a completely enclosed hollow
cavity inside is struck by lightning.

Exercise 3
A positive charge +Q placed E
on a conducting sphere of ra- R
Gaussian dius R, produces the electric
surface field shown.
E
a) What is the charge density
metal σ on the surface of the sphere?
b) Use Equation (1) to find a formula for the magnitude
ri of the electric field E produced by the surface charge
density σ .

c) How does the field calculated in part b) compare with

the strength of the electric field a distance R from a point
charge Q?

Exercise 4
Repeat Exercise 1 assuming that the conducting sphere
Figure 9 has a net charge of – Q. Does the charge on the
If the Gaussian surface is drawn inside the hollow cavity conducting sphere have any effect on the fields inside
as shown, then there is no charge inside the Gaussian the sphere? Why is there no field outside the sphere?
surface. Thus no field lines emerge through the
Gaussian surface, and E must be zero inside the cavity.
26-6 Electric Fields and Conductors

VAN DE GRAAFF GENERATOR Acting like a conveyor belt, the motorized belt carries
The Van de Graaff generator is a conceptually straight- the charge up and into the inside of the hollow metal
forward device designed to produce high voltages. A sphere. If there is already charge on the sphere, then,
sketch of the apparatus is shown in Figure (12), where as we have seen in Example (1), there will be an electric
we have a hollow metal sphere with a hole in the field outside the sphere as shown in Figure (13). (For
bottom, and a conveyer belt whose purpose is to bring this example we are assuming that the belt is carrying
charge up into the sphere. The belt is driven by a motor positive charge.) But inside the sphere there will be no
at the bottom. field. (The hole in the bottom of the sphere lets a small
amount of electric field leak inside, but not enough to
The first step is to get electric charge onto the belt. This worry about.)
is done electrostatically by having an appropriate ma-
terial rub against the belt. For example, if you rub a As the charge is being carried up by the belt, the electric
rubber rod with cat fur, you leave a negative charge on field outside the sphere pushes back on the charge, and
the rubber rod. If you rub a glass rod with silk, a positive the belt has to do work to get the charge up to the sphere.
charge will be left on the glass rod. I do not know what The more charge that has built up on the sphere, the
sign of charge is left on a comb when you run it through stronger the electric field E, and the more work the belt
your hair on a dry day, but enough charge can be left on has to do. In a typical Van de Graaff generator used in
the comb to pick up small pieces of paper. We will lecture demonstration, you can hear the motor working
leave the theory of creating electrostatic charges to harder when a large charge has built up on the sphere.
other texts. For our discussion, it is sufficient to
visualize that some kind of rubbing of the belt at the
bottom near the motor deposits charge on the belt. (As
an example of charging by rubbing, run a comb through + +
your hair several times. The comb becomes electrically
+ +
charged and will pick up small pieces of paper.)

metal + + E
sphere E=0
motor
pulley + wire
+
+
drive
+ + +
+
+
+
charge removed conveyor +
from the belt belt E

charge placed motor Figure 13

on the belt pulley It takes work to carry the charge up to the sphere
drive against the electric field that is pushing down on the
charge. But once inside the sphere where there is
Figure 12 almost no field, the charge freely moves off the belt,
The Van de Graaff generator. Electric charge is onto the wire, charging up the sphere. The more
carried up the belt and dumped inside the hollow charge on the sphere, the stronger the electric field E
metal sphere. Since there are no electric fields outside the sphere, and the more work required to
inside the sphere, the electric charge freely flows bring new charge up into the sphere. (In the
off the belt to the sphere, where it then spreads demonstration model, you can hear the motor slow
evenly to the outside surface of the sphere. down as the sphere becomes charged up.)
 26-7

When the charge gets to the sphere how do we get it off Electric Discharge
the belt onto the sphere? When the sphere already has When a large amount of charge has accumulated on the
a lot of positive charge on it, why would the positive metal sphere of the Van de Graaff generator, we can
charge on the belt want to flow over to the sphere? produce some very strong fields and high voltages. We
Shouldn’t the positive charge on the belt be repelled by can estimate the voltage by bringing a grounded sphere
the positive sphere? up to the Van de Graaff generator as shown in Figure
(14). A voltage of about 100,000 volts is required to
Here is where our knowledge of electric fields comes make a spark jump about an inch through air. Thus if
in. As illustrated in Figure (13), there may be very we get a spark about 2 inches long between the Van de
strong electric fields outside the sphere, but inside there Graaff generator and the grounded sphere, we have
are none. Once the conveyor belt gets the charge inside brought enough charge onto the generator sphere to
the sphere, the charge is completely free to run off to the create a voltage of about 200,000 volts. (The length of
sphere. All we need is a small wire that is attached to the sparks acts as a crude voltmeter!)
the inside of the sphere that rubs against the belt. In fact,
the neighboring + charge on the belt helps push the As an exercise, let us estimate how many coulombs of
charge off the belt onto the wire. charge must be on the Van de Graaff generator sphere
to bring it up to a voltage of 200,000 volts.
Once the charge is on the wire and flows to the inside
of the sphere, it must immediately flow to the outside Outside the Van de Graaff generator sphere, the electric
of the sphere where it helps produce a stronger field E field is roughly equal to the electric field of a point
shown in Figure (13). charge. Thus the voltage or electric potential of the
sphere should be given by Equation (25-4) as

spark grounded Q
V=
metal sphere 4πε0r (25-4)
where r is the radius of the Van de Graaff generator
insulated sphere. (Remember that r is not squared in the formula
support
for potential energy or voltage.)

grounding
Let us assume that r = 10 cm or .1 m, and that the voltage
motor wire V is up to 200,000 volts. Then Equation (25-4) gives
metal grounding plate
Q = 4πε 0rV
wire to water pipe
= 4π × 9 × 10 -12 × .1 × 200,000

Figure 14 Q ≈ 2 × 10 -6 coulombs
We can discharge the Van de Graaff generator
by bringing up a grounded sphere as shown. A couple millionth’s of a coulomb of charge is enough
Since about 100,000 volts are required to make a to create 200,000 volt sparks. As we said earlier, a
spark one inch long, we can use the maximum whole coulomb is a huge amount of charge!
length of sparks to estimate the voltage produced
by the Van de Graaff generator.
26-8 Electric Fields and Conductors

Grounding THE ELECTRON GUN

The grounded sphere in Figure (14) that we used to In Figure (15) we have a rough sketch of a television
produce the sparks, provides a good example of the tube with an electron gun at one end to create a beam
way we use conductors and wires. of electrons, deflection plates to move the electron
beam, and a phosphor screen at the other end to produce
Beneath the Van de Graaff generator apparatus we
a bright spot where the electrons strike the end of the
have placed a large sheet of aluminum called a ground-
tube.
ing plane that is attached to the metal pipes and the
electrical ground in the room. (Whenever we have Figure (16) illustrates how a picture is drawn on a
neglected to use this grounding plane during a demon- television screen. The electron beam is swept horizon-
stration we have regretted it.) We have attached a tally across the face of the tube, then the beam is moved
copper wire from the grounding plane to the “grounded” down one line and swept horizontally again. An
sphere as shown. American television picture has about 500 horizontal
lines in one picture.
Thus in Figure (14), the grounding plane, the room’s
metal pipes and electrical ground wires, and the As the beam is swept across, the brightness of the spot
grounded sphere are all attached to each other via a can be adjusted by changing the intensity of the elec-
conductor. Now there can be no electric field inside a tron beam. In Figure (16), line 3, the beam starts out
conductor, therefore all these objects are at the same bright, is dimmed when it gets to the left side of the
electric potential or voltage. (If you have a voltage letter A, shut off completely when it gets to the black
difference between two points, there must be an electric line, then turned on to full brightness to complete the
field between these two points to produce the voltage line. In a standard television set, one sweep across the
difference.) It is common practice in working with tube takes about 60 microseconds. To draw the fine
electricity to define the voltage of the water pipes (or a details you see on a good television set requires that the
metal rod stuck deeply into the earth) as zero volts or intensity of the beam can be turned up and down in little
“ground”. (The ground wires in most home wiring are more than a tenth of a microsecond.
attached to the water pipes.) Any object that is con-
nected by a wire to the water pipes or electrical ground bright
spot
wire is said to be grounded. The use of the earth as the
definition of the zero of electric voltage is much like
deflection
using the floor of a room as the definition of the zero of plates bea
m
on
the gravitational potential energy of an object. e lectr

In Figure (14), when the grounded sphere is brought up

phosphor
to the Van de Graaff generator and we get a 2 inch long screen
spark, the spark tells us that the Van de Graaff sphere electron
had been raised to a potential of at least 200,000 volts gun
above ground.
Figure 15
Van de Graaff generators are found primarily in two Cathode ray tubes, like the one shown above, are
applications. One is in science museums and lecture commonly used in television sets, oscilloscopes, and
demonstration to impress visitors and students. The computer monitors. The electron beam (otherwise
other is in physics research. Compared to modern known as a "cathode ray") is created in the electron
accelerators, the 200,000 volts or up to 100 million gun, is aimed by the deflection plates, and produces a
bright spot where it strikes the phosphor screen.
volts that Van de Graaff generators produce is small.
But the voltages are very stable and can be precisely
controlled. As a result the Van de Graaff’s make
excellent tools for studying the fine details of the
structure of atomic nuclei.
 26-9

The heart of this system is the electron gun which The Filament
creates the electron beam. The actual electron gun in As shown in Figure (17), the source of the electrons in
a television tube is a complex looking device with an electron gun is the filament, a piece of wire that has
indirect heaters and focusing rings all mounted on the been heated red-hot by the passage of an electric
basic gun. What we will describe instead is a student- current. At these temperatures, some of the electrons
built gun which does not produce the fine beam of a in the filament gain enough thermal kinetic energy to
commercial gun, but which is easy to build and easy to evaporate out through the surface of the wire. The
understand. white coating you may see on a filament reduces the
amount of energy an electron needs to escape out
through the metal surface, and therefore helps produce
a more intense beam of electrons. At standard tempera-
ture and pressure, air molecules are about 10- molecu-
lar diameters apart as indicated in Figure (18). There-
fore if the filament is in air, an electron that has
evaporated from the filament can travel, at most, a few
hundred molecular diameters before striking an air
molecule. This is why the red-hot burner on an electric
stove does not emit a beam of electrons. The only way
we can get electrons to travel far from the filament is to
place the filament in a vacuum as we did in Figure (17).
Figure 16 The better the vacuum, the farther the electrons can
The letter A on a TV screen. To construct an travel.
image the electron beam is swept horizontally, and
turned up where the picture should be bright and
turned down when dark. The entire image consists
of a series of these horizontal lines, evenly spaced,
one below the other.
air
c tron molecule
ele
heated glass test filament s
filament tube
urface

to source
of heating current • • •
• •
•
• • •
•
•
•
vacuum
• •
• • • • •

electrons boiled
off surface of
filament
Figure 18
Figure 17
Whenever we heat a metal to a high enough
Source of the electrons. The tungsten filament
temperature, electrons boil out of the surface. But if
is heated by an electric current. When it
there is air at standard pressure around, the electrons
becomes red-hot, electrons boil out through the
do not get very far before striking an air molecule.
surface. The white coating on the filament
makes it easier for the electrons to escape.
26-10 Electric Fields and Conductors

Accelerating Field A Field Plot

Once the electrons are out of the filament we use an A field plot of the electric field lines inside the electron
electric field to accelerate them. This is done by placing gun cap gives a more precise picture of what is happen-
a metal cap with a hole in the end over the end of the ing. Figure (23) is a computer plot of the field lines for
filament as shown in Figure (19). The filament and cap a cylindrical filament inside a metal cap. We chose a
are attached to a battery as shown in Figure (20) so that cylindrical filament rather than a bent wire filament
the cap is positively charged relative to the filament. because it has the cylindrical symmetry of the cap and
is therefore much easier to calculate and draw. But the
Intuitively the gun works as follows. The electrons are fields for a wire filament are not too different.
repelled by the negatively charged filament and are
attracted to the positively charged cap. Most of the First notice that the field lines are perpendicular to both
electrons rush over, strike, and are absorbed by the cap metal surfaces. This agrees with our earlier discussion
as shown in Figure (21). But an electron headed for the that an electric field at the surface of a conductor cannot
hole in the cap discovers too late that it has missed the have a parallel component for that would move the
cap and goes on out to form the electron beam. charge in the conductor. The second thing to note is that
due to the unfortunate fact that the charge on the
A picture of the resulting electron beam is seen in electron is negative, the electric field points oppositely
Figure (22). The beam is visible because some air to the direction of the force on the electrons. The force
remains inside the tube, and the air molecules glow is in the direction of -E.
when they are struck by an electron.
positive cap
filament + + + + + + +
– – – – – – +
negative filament
– +
–
–
– + beam of electrons
– – – – – – + that missed the cap
flow of electrons + + + + + + + and went out
metal cap through the hole
Figure 19
Figure 21
To create a beam of electrons, we start by placing a
Electrons flow from the negative filament
metal cap with a hole in it, over the filament.
to the positive cap. The beam of electrons
is formed by the electrons that miss the cap
+
–
+
–
+
–
+
––
+
+ beam of and go out through the hole.
– – – – +
+ + + + +
electrons

– +

Figure 20
We then attach a battery to the metal cap so that the
cap has a positive voltage relative to the filament.

Figure 22
Resulting electron beam.
 26-11

The electrons, however, do not move along the -E field Equipotential Plot
lines. If they boil out of the filament with a negligible Once we know the field lines, we can plot the equipo-
speed they will start moving in the direction -E. But as tential lines as shown in Figure (24). The lines are
the electrons gain momentum, the force -eE has less labeled assuming that the filament is grounded (0 volts)
and less effect. (Remember, for example, that for a and that the cap is at 100 volts . The shape of the
satellite in a circular orbit, the force on the satellite is equipotentials, shown by dashed lines, does not change
down toward the center of the earth. But the satellite when we use different accelerating voltages, only the
moves around the earth in an orbit of constant radius.) numerical value of the equipotentials changes.
In Figure (23), the dotted lines show a computer plot of
the trajectories of the electrons at several points. The The reason that the equipotential lines are of such
most important trajectories for our purposes are those interest in Figure (24) is that they can also be viewed as
that pass through the hole in the cap and go out and form a map of the electron’s kinetic energy.
the electron beam.
Remember that the voltage V is the potential energy of
a unit positive test charge. A charge q has a potential
Exercise 5 energy qV, and an electron, with a charge – e , has an
Describe two other examples where an object does not electric potential energy – eV .
move in the direction of the net force acting on it.
In our electron gun, the electrons evaporate from the
trajectories of filament with very little kinetic energy, call it zero. By
individual electrons the time the electrons get to the 10-volt equipotential,
their electric potential energy has dropped to (–e × 10)
joules, and by conservation of energy, their kinetic
energy has gone up to (+e × 10) joules. At the 50 volt
equipotential the electron’s kinetic energy has risen to
(e × 50) joules, and when the electrons reach the 100
volt cap, their energy is up to (e × 100) joules. Thus the
electric
equipotential lines in Figure (24) provide a map of the
field
kinetic energy of the electrons.

equipotential lines
cylindrical
filament 80 V
50 V
cap 20 V
Figure 23
Plot of the electric field in the region between the
filament and the cap. Here we assume that we have
a cylindrical filament heated by a wire inside.

Figure 24
Equipotential plot. We see that by the time the
electrons have reached the hole in the cap, they
have crossed the same equipotential lines and
therefore have gained as much kinetic energy as
the electrons that strike the cap. (From a student
project by Daniel Leslie and Elad Levy.)
26-12 Electric Fields and Conductors

ELECTRON VOLT For example, if we have a 100 eV electron, its kinetic

AS A UNIT OF ENERGY energy 1 2 mv 2 is given by
What is perhaps most remarkable about the electron
KE = 1 2 mv 2
gun is that every electron that leaves the filament and
strikes the cap gains precisely the same kinetic energy. joules
If we use a battery that produces 100 volt accelerating = 100 eV × 1.6 × 10 -19 (5)
eV
voltage, then every electron gains precisely (e × 100)
joules of kinetic energy. This is also true of the Using the value m = 9.11 × 10 -31 kg for the electron
electrons that miss the cap and go out and form the mass in Equation (5) gives
electron beam. -19
2 × 100 × 1.6 × 10
The amount of energy gained by an electron that falls v = -31
9.11 × 10
through a 1 volt potential is (e × 1 volt) = 1.6 x 10-19 (6)
joules. This amount of energy is called an electron volt meters
6
and designated by the symbol eV. = 6 × 10
sec
energy gained by an electron which is 2% the speed of light.
1eV =
falling through a 1 volt potential
In studies involving atomic particles such as electrons
= (e coulombs) × (1 volt) (3) and protons, the electron volt is both a convenient and
very commonly used unit. If the electron volt is too
-19
= 1.6 x 10 joules small, we can measure the particle energy in MeV
(millions of electron volts) or GeV (billions of electron
The dimensions in Equation (3) make a bit more sense
volts or Gigavolts).
when we realize that the volt has the dimensions of
joule/coulomb, so that 1 MeV ≡ 10 6 eV
joule
1eV = e coulombs × 1 1 GeV ≡ 10 9 eV (6)
coulomb
For example, if you work the following exercises, you
= (e) joules (3a) will see that the rest energies m0c2 of an electron and a
proton have the values
The electron volt is an extremely convenient unit for
describing the energy of electrons produced by an electron rest energy = .51 MeV
electron gun. If we use a 100 volt battery to accelerate
the electrons, we get 100 eV electrons. Two hundred proton rest energy = .93 GeV (7)
volt batteries produce 200 eV electrons, etc.
The reason that it is worth remembering that an electron’s
To solve problems like calculating the speed of a 100 rest energy is about .5 MeV and a proton’s about 1 GeV,
eV electron, you need to convert from eV to joules. The is that when a particle’s kinetic energy gets up toward
conversion factor is its rest energy, the particle’s speed becomes a signifi-
cant fraction of the speed of light and nonrelativistic
joules formulas like 1/2 mv2 for kinetic energy no longer
1.6 × 10 -19 conversion
factor (4)
eV apply.
 26-13

Example About Computer Plots

Calculate the rest energy of an electron in eV. One final note in our discussion of the electron gun.
You might feel that by using the computer plots in
Solution: Figures (23) and (24) we have cheated a bit. We
m0c 2 joules haven’t done the work ourselves, we let somebody (or
E = something) else do the calculations for us and we are
joules
1.6 × 10 -19 just using their answers. Yes and no!
eV
2 First of all, with a little bit of practice you can learn to
9.11 × 10 –31 × 3 × 10 8 draw sketches that are quite close to the computer plots.
= Use a trick like noting that field lines must be perpen-
1.6 × 10 -19
dicular to the surface of a conductor where they touch
= .51 × 10 6 eV the conductor. If two conductors have equal and
opposite charge – if they were charged by a battery – all
the field lines that start on the positive conductor will
Exercise 6
stop on the negative one. Use any symmetry you can
Calculate the rest energy of a proton in eV and GeV.
find to help sketch the field lines and then sketch the
Exercise 7
equipotential lines perpendicular to the field lines.
Some places it is easier to visualize the equipotential
What accelerating voltage must be used in an electron
lines, e.g., near the surface of a conductor, and then
gun to produce electrons whose kinetic energy equals
draw in the perpendicular field lines.
their rest energy?
The other point is that, for a number of practical
80 V problems the geometry of the conductors is compli-
100 V
cated enough that only by using a computer can we
accurately plot the field lines and equipotentials. But
once a computer plot is drawn, we do not have to worry
about how it was calculated. Like a hiker in a new
territory, we can use the computer plot as our contour
map to tell us the shape and important features of the
terrain. For example in our field plots of the electron
50 V gun, we see that there is virtually no field out in front of
20 V the hole where the electrons emerge, therefore from the
0V time the electrons leave the hole they coast freely at
equipotential lines constant speed and energy down the tube.

Figure 24a
Another field plot by Leslie and Levy, showing the
electric field and equipotential lines in a gun with a
shorter cap.
26-14 Electric Fields and Conductors

THE PARALLEL PLATE CAPACITOR Let us take a close look at what we have in Figure (27).
Our final example in this chapter of fields and conduc- The electric field lines E leave the positively charged
tors is the parallel plate capacitor. Here we will work top plate and go straight down to the negatively charged
with a much simpler field structure than for the electron bottom plate. Since all the lines starting at the top plate
gun, and will therefore be able to calculate field strengths stop at the bottom one, there must be an equal and
and voltages. The parallel plate capacitor serves as the opposite charge +Q and -Q on the two plates. There is
prototype example of a capacitor, a device used through- no net charge on the capacitor, only a separation of
out physics and electrical engineering for storing elec- charge. And because the field lines go straight down,
tric fields and electric energy. nowhere do they get closer together or farther apart, the
field must have a uniform strength E between the
Suppose we take two circular metal plates of area A, plates.
separate them by a distance d, and attach a battery as
shown in Figure (25). This setup is called a parallel We can use Gauss’ law to quickly calculate the field
plate capacitor, and the field lines and equipotential for strength E. The top plate has a charge Q, therefore the
this setup are shown in the computer plot of Figure (26). total flux out of the top plate must be Φ = Q/ε0 . But we
also have a field of strength E flowing out of a plate of
Except at the edges of the plates, the field lines go area A. Thus flux of E flowing between the plates is
straight down from the positive to the negative plate, Φ = EA . Equating these two formulas for flux gives
and the equipotentials are equally spaced horizontal
lines parallel to the plates. If the plate separation d is Q
Φ = EA =
small compared to the diameter D of the plates, then we ε0
can neglect the fringing of the field at the edge of the
plates. The result is what we will call an ideal parallel Q
plate capacitor whose field structure is shown in Figure E= (8)
ε0A
(27). The advantage of working with this ideal capaci-
tor is that we can easily derive the relationship between We can relate the voltage V and the field strength E by
the charging voltage V, and the charge Q. remembering that E is the force on a unit test charge
and V is the potential energy of a unit test charge. If
battery I lift a unit positive test charge from the bottom plate
a distance d up to the top one, I have to exert an
+ A upward force of strength E for a distance d and
d therefore do an amount of work E × d. This work is
– stored as the electric potential energy of the unit test
capacitor plates charge, and is therefore the voltage V:
Figure 25
The parallel plate capacitor. The capacitor is charged
up by connecting a battery across the plates as shown. V= Ed (9)

plate of area A
+ + + + + + + + + + + + + + + + + + + + + + + + + +
d
– – – – – – – – – – – – – – – – – – – – – – – – – –
Figure 27
Figure 26 In our idealized parallel plate capacitor the field
The electric field between and around lines go straight from the positive to the negative
the edge of the capacitor plates. plate, and the field is uniform between the plates.
 26-15

It may seem surprising, but V is also the voltage of the

battery (see Figure 25) used to charge up the capacitor.

There is also a simple relationship between the charge

Q on the capacitor plates and the voltage difference V
between them. Substituting the value of E from
Equation (8) into Equation (9) gives gold
d leaves
V = Q (10)
ε0A
Equation (10) makes an interesting prediction. If we
have a fixed charge Q on the capacitor (say we charged Figure 28
up the capacitor and removed the battery), then if we Gold leaf electrometer attached
increase the separation d between the plates, the volt- to a parallel plate capacitor.
age V will increase.

One problem with trying to measure this increase in

voltage is that if we attach a common voltmeter be-
tween the plates to measure V, the capacitor will
quickly discharge through the voltmeter. In order to see
this effect we must use a special voltmeter called an
electrometer that will not allow the capacitor to dis-
charge. The classic electrometer, used in the 1800's, is
the gold leaf electrometer shown in Figures (28) and
(29). When the top plate of the electrometer is charged,
some of the charge flows to the gold leaves, forcing the
leaves apart. The greater the voltage, the greater the
charge and the greater the force separating the leaves. Figure 29a
Looking through the electrometer at the
Thus the separation of the leaves is a rough measure of edge of the charged capacitor plates.
the voltage.

In Figure (28), we see a gold leaf electrometer attached

to two metal capacitor plates. When the plates are
charged, the gold leaves separate, indicating that there
is a voltage difference between the plates.

In Figures (29a,b), we are looking through the elec-

trometer at the edge of the capacitor plates. In going
from (29a) to (29b), we moved the plates apart without
changing the charge on the plates. We see that when the
plates are farther apart, the gold leaves are more
separated, indicating a greater voltage as predicted by
equation (10).
Figure 29b
Without changing the charge, the plates
are moved further apart. The increased
separation of the gold leaves shows that
the voltage difference between the
capacitor plates has increased.
26-16 Electric Fields and Conductors

Exercise 8 eE = eVp
A = m
Two circular metal plates of radius 10 cm are separated e med
by microscope slide covers of thickness d = . 12 mm. A where Vp, is the voltage and d the separation of the
voltage difference of 5 volts is set up between the plates
deflection plates.
using a battery as shown in Figure (25). What is the
charge Q on the plates? If a particle is subjected to a downward acceleration for
a time T, and initially has no downward velocity, its
Deflection Plates final downward velocity vfy is from the constant accel-
eration formulas as
A fitting conclusion to this chapter is to see how the
fields in parallel plate capacitor can be used to deflect eVp
the beam of electrons produced by an electron gun. In vfy = AyT = T
med (11)
Figure (30) the beam of electrons from an electron gun
is aimed between the plates of a parallel plate capacitor. If the electrons emerge from the electron gun at a speed
The upward directed electric field E produces a down- v, then the time T it takes them to pass between the
ward directed force -eE on the electrons, so that when plates is
the electrons emerge from the plates, they have been
deflected downward by an angle θ as shown. We wish T = D
v (12)
to calculate this angle θ which depends on the strength
of the deflection voltage Vp, the length D of the plates, The tangent of the deflection angle θ is given by the
and on the speed v of the electrons. ratio vfy /v which we can get from Equations (11) and
(12):
While the electrons are between the plates, their accel-
eration is given by v
eVpD θ
vfy = × vfy
F = -eE
A = m med v
e me
where me is the electron mass. This acceleration is
constant and directed downward, just as in our old vfy eVpD
tan θ = = (13)
projectile motion studies. Using Equation (9) E = V/d v m edv 2
for the magnitude of E, we find that the downward
acceleration A of the electrons has a magnitude The final step is to note that the speed v of the electrons
is determined by the electron gun accelerating voltage
D Vacc by the relationship
–

electron
gun d θ 1 2eVacc
m v 2 = eVacc or v2 = (14)
+V deflection voltage VP 2 e me
Figure 30
To deflect the beam of electrons, we place what is Equations (13) and (14) finally give
essentially a parallel plate capacitor in the path of the
beam as shown. The electrons are deflected by the
eVp D 1 D Vp
electric field between the capacitor plates. tanθ = = (15)
m ed 2eVacc m e 2 d Vacc

which is a fairly simple result considering the steps we

went through to get it. It is reassuring that tanθ comes
out as a dimensionless ratio, which it must.
 26-17

Exercise 9
In an electron gun, deflection plates 5 cm long are
separated by a distance d = 1.2 cm. The electron beam
is produced by a 75 volt accelerating voltage. What
deflection voltage Vp is required to bend the beam 10
degrees?

Exercise 10
In what is called the Millikan oil drop experiment, shown
in Figure (31), a vapor of oil is sprayed between two
capacitor plates and the oil drops are electrically
charged by radioactive particles.

Consider a particular oil drop of mass m that has lost one

electron and therefore has an electric charge q = + e.
(The mass m of the drop was determined by measuring
its terminal velocity in free fall in the air. We will not worry
about that part of the experiment, and simply assume
that the drop's mass m is known.) To measure the
charge q on the oil drop, and thus determine the
electron charge e, an upward electric field E is applied
to the oil drop. The strength of the field E is adjusted until
the upward electric force just balances the downward
gravitational force. When the forces are balanced, the
drop, seen through a microscope, will be observed to Fe
come to rest due to air resistance. m E d
Fg
The electric field E that supports the oil drop is pro-
power supply
duced by a parallel plate capacitor and power supply microscope of voltage V
that can be adjusted to the desired voltage V. The
separation between the plates is d. Figure 31
Millikan oil drop apparatus
a) Reproduce the sketch of Figure (31), Then put a +
sign beside the positive battery terminal and a – sign
beside the negative one.

b) Find the formula for the voltage V required to

precisely support the oil drop against the gravitational
force. Express your answer in terms of the geometry of
the capacitor (plate separation d, area A, etc.) the
drop's mass m, the acceleration due to gravity g, and
the electron charge e.
Chapter 27
Basic Electric Circuits

CHAPTER 27 Basic Electric Cir-

cuits

In the modern age (post 1870) we have been sur-

rounded by electric circuits. House wiring is our most
familiar example, but we have become increasingly
familiar with electric circuits in radio and television
sets, and even the digital watch you may be wearing. In
this chapter we will discuss the basic electric circuits in
order to introduce the concepts of electric current,
resistance, and voltage drops around the circuit. We
will restrict ourselves to devices like batteries, resis-
tors, light bulbs, and capacitors. The main purpose is
to develop the background needed to work with electric
circuits and electronic measuring equipment in the
laboratory.
27-2 Basic Electric Circuits

ELECTRIC CURRENT Household wiring is protected by fuses or circuit

An electric current in a wire is conceptually somewhat breakers that shut off the current if it exceeds 15 or 20
like the current of water in a river. We can define the amps. (You can see why you do not want to run a hair
current in a river as the amount of water per second dryer and an electric heater on the same circuit.)
flowing under a bridge. The amount of water could be There is a common misconception that the electrons in
defined as the number of water molecules, but a more a wire travel very fast when a current is flowing in the
convenient unit would be gallons, liters, or cubic wire. After all when you turn on a wall light switch the
meters. light on the other side of the room appears to turn on
An electric current in a wire is usually associated with instantly. How did the electrons get there so fast?
the flow of electrons and is measured as the amount of The answer can be seen by an analogy to a garden hose.
charge per second flowing past some point or through When you first attach an empty hose to a spigot and turn
some cross-sectional area of the wire, as illustrated in on the water, it takes a while before the hose fills up with
Figure (1). We could measure the amount of charge by water and water comes out of the other end. But when
counting the number of electrons crossing the area, but the hose is already full and you turn on the spigot, water
it is more convenient to use our standard unit of charge, almost instantly comes out of the other end. Not the
the coulomb, and define an electric current as the water that just went in, but the water that was already
number of coulombs per second passing the cross- in the hose.
sectional area. The unit of current defined this way is
called an ampere. A copper wire is analogous to the hose that is already
full of water; the electrons are already there. When you
turn on the light switch, the light comes on almost
1 coulomb per instantly because all the “electric fluid” in the wire
second passing starts moving almost at once.
1 ampere = (1)
a cross–sectional
area of wire To help build an intuition, let us estimate how fast the
electrons must move in a copper wire with a 1 millime-
From your experience with household wiring you ter cross-sectional area carrying an electric current of
should already be familiar with the ampere (amp) as a one ampere. This is not an unreasonable situation for
unit of current. A typical light bulb draws between 1 2 household wiring.
and 1 ampere of current, and so does the typical motor A copper atom has a nucleus containing 29 protons
in an electric appliance (drill, eggbeater, etc.). A surrounded by a cloud of 29 electrons. Of the 29
microwave oven and a toaster may draw up to 6 amps, electrons, 27 are tightly bound to the nucleus and 2 are
and hair dryers and electric heaters up to 12 amps. in an outer shell, loosely bound. (All metal atoms have
Household wiring is limited in its capability of carrying one, two, and sometimes 3 loosely bound outer elec-
electric current. If you try to carry too much current in trons.) When copper atoms are collected together to
a wire, the wire gets hot and poses a fire hazard. form a copper crystal, the 27 tightly bound electrons
remain with their respective nuclei, but the two loosely
moving bound electrons are free to wander throughout the
electric
crystal. In a metal crystal or wire, it is the loosely bound
charge
electrons (called conduction electrons) that form the
electric fluid that makes the wire a conductor.

cross sectional area

Figure 1
An electric current is defined as the amount of charge
per second flowing past a cross-sectional area.
 27-3

Copper has an atomic weight of 63.5, thus there are Positive and Negative Currents
63.5 grams of copper in a mole. And the density of If you are using a hose to fill a bucket with water, there
copper is 9 gm /cm3, thus a mole of copper has a is not much question about which way the current of
volume water is flowing—from the hose to the bucket. But
with electric current, because there are two kinds of
63.5gm / mole 3
volume of one = = 7 cm electric charge, the situation is not that simple. As
mole of copper 9gm / cm3 mole
shown in Figure (2), there are two ways to give an
Since a mole of a substance contains an Avogadro’s object a positive charge, add positive charge or remove
number 6 × 1023 of particles of that substance, and negative charge. If a wire connected to the object is
since there are 2 conduction electrons per copper atom, doing the charging, it may be difficult to tell whether
7 cm3 of copper contain 12 × 1023 conduction elec- there is a current of positive charge into the object or a
trons. Dividing by 7, we see that there are 1.7 × 1023 current of negative charge out of the object. Both have
conduction electrons in every cubic centimeter of essentially the same effect.
copper and 1.7 × 1020 in a cubic millimeter. Convert- You may argue that at least for copper wires a current
ing this to coulombs, we get of positive charge doesn’t make sense because the
electric current is being carried by the negative conduc-
number of tion electrons. But a simple model of an electric current
coulombs of will clearly demonstrate that a positive current flowing
conduction 1.7 × 10 20 electrons mm3
= one way is essentially equivalent to a negative current
electrons 6.25 × 10 18electrons coulomb
in 1mm3 flowing the other way.
of copper

= 27 coulombs mm3 +i
+
In our 1 millimeter cross-sectional area wire, if the
electrons flowed at a speed of 1 millimeter per second,
27 coulombs of charge would flow past any point in the
wire per second, and we would have a current of 27 –i
amperes. To have a current of 1 ampere, the electrons +
would have to move only 1/27 as fast, or 1/27 of a
Figure 2
millimeter per second! This slow speed results from
A current of positive charge into an
the huge density of conduction electrons. object, or a current of negative charge
out, leaves the object positively charged.
27-4 Basic Electric Circuits

In Figure (3) we have tried to sketch a picture of a example the average speed of the conduction electrons
copper wire in which the conduction electrons are was 1/27 of a millimeter per second, which we will take
moving to the left producing a left directed negative as the speed v of the moving negative rod in Figure (4).
current. The problem with Figure (3) is that it is hard
to show the conduction electrons flowing through the Figure (5a) is the same as Figure (4), except we have
lattice of stationary positive copper nuclei. The picture drawn a stick figure representing a person walking to
is difficult to draw, and Figure (3) is not particularly the left at a speed v. The person and the negatively
informative. charged rod are both moving to the left at the same
speed.
To more clearly show that the positive charge is at rest
and that it is the negative charge that is moving, we have Figure (5b) is the same situation from the point of view
in Figure (4) constructed a model of a copper wire in of the stick figure person. From her point of view, the
which we have two separate rods, one moving and one negative rod is at rest and it is the positive rod that is
at rest. The stationary rod has the positive copper nuclei moving to the right. Our left directed negative current
and the moving rod has the negative conduction elec- in Figure (5a) is seen by the moving observer to be a
trons. This model is not a very good representation of right directed positive current (Figure 5b). Whether
what is going on inside the copper wire, but it does we have a left directed negative current or a right
remind us clearly that the positive charge is at rest, and directed positive current just depends upon the point of
that the current is being carried by the moving negative view of the observer.
charge.
But how fast was our moving observer walking? If
When you see this model, which we will use again in Figure (5) is a model of a 1 mm2 copper wire carrying
later discussions, think of the two rods as merged a current of 1 ampere, the speed v in Figure (5) is 1/27
together. Picture the minus charge as flowing through of a millimeter per second. This is about 2 millimeters
the lattice of positive charge. Remember that the only per minute! Although faster than the continental drift,
reason that we drew them as separate rods was to this motion should certainly have little effect on what
clearly show which charge was carrying the electric we see. If the wire is leading to a toaster, the toast will
current. come out the same whether or not we walk by at a speed
of 2 mm per minute. For most purposes, we can take
Using the results of the previous section, we can make a left directed negative current and a right directed
our model of Figure (4) more specific by assuming it positive current as being equivalent. Relatively sophis-
represents a copper wire with a 1 millimeter cross ticated experiments, such as those using the Hall effect
section carrying a current of one ampere. In that (to be discussed later) are required to tell the difference.

positive copper moving conduction positively charged

ions at rest electrons rod at rest
– – – – + + + + + + +
+ + + + + + + + – – – – – – –
– – – –
v
copper wire
moving negatively
Figure 3 charged rod
A copper wire at rest with the conduction Figure 4
electrons moving to the left. This gives us Model of a copper wire carrying an electric current.
a left-directed negative current. We are representing the positive copper ions by a
positively charged rod at rest, and the conduction
electrons by a moving, negatively charged rod.
 27-5

A Convention just seen that the flow of electrons in one direction is

It was Ben Franklin who made the assignment of almost completely equivalent to the flow of positive
positive and negative charge. The charge left on a glass charge in the other. If we do our calculations for
rod rubbed by silk was defined as positive, and that left positive currents, then we know that the electrons are
on a rubber rod rubbed by cat fur as negative. This has simply moving in the opposite direction.
often been considered a tragic mistake, for it leaves the
electron, the common carrier of electric current, with a In order to maintain sanity and not get tangled up with
negative charge. It also leads to the unfortunate intui- minus signs, in this text we will, whenever possible,
tive picture that an atom that has lost some electrons talk about the flow of positive currents, and talk about
ends up with a positive charge. the force on positive test charges. If the problem we are
working on involves electrons, we will work every-
Some physics textbooks written in the 1930s redefined thing assuming positive charges and positive currents,
the electron as being positive, but this was a disaster. and only at the end of the problem we will take into
We cannot undo over two centuries of convention that account the negative sign of the electron. With some
leads to the electron as being negative. practice, you will find this an easy convention to use.

The worst problem with Franklin’s convention comes

when we try to handle the minus signs in problems
involving the flow of electrons in a wire. But we have

v
+ + + + + +
– – – – – –
a) observer walking along with the moving
negatively charged rod

v
+ + + + + +
– – – – – –
b) from the observer's point of view the negative
rod is at rest and the positive charge is moving
to the right

Figure 5 a, b
In (a) we have a left directed negative current, while in (b) we have a right directed
positive current. The only difference is the perspective of the observer. (You can turn a
negative current into an oppositely flowing positive one simply by moving your head.)
27-6 Basic Electric Circuits

CURRENT AND VOLTAGE from a town water tank, when the water was at the top
Students first studying electricity can have difficulty of the tank it was at atmospheric pressure and not
conceptually distinguishing between the concepts of moving, but was at a great height h. In the town water
current and voltage. This problem can be handled by tank the hydrodynamic voltage comes mainly from the
referring back to our hydrodynamic analogy of Chap- ρgh term.
ter 23. Let us focus our attention on the high pressure in a
In Chapter 23 we were discussing Bernoulli’s equation faucet that is shut off. In this case we have high voltage
which stated that the quantity (P + ρgh + 1/2 ρv2) water but no current. We can get a big current if we turn
was constant along a stream line if we could neglect the faucet on, but the voltage is there whether or not we
viscous effects in the fluid. Because of the special have a current.
nature of this collection of terms, we gave them the In household wiring, the electrical outlets may be
name hydrodynamic voltage. thought of as faucets for the electrical fluid in the wires.
hydrodynamic The high voltage in these wires is like the high pressure
= P + ρgh + 1 ρv 2 (23-23)
voltage 2 in the water pipes. You can have a high voltage at the
outlet without drawing any current, or you can connect
(The second and third terms in the hydrodynamic
an appliance and draw a current of this high pressure
voltage are the potential energy of a unit volume of
electrical fluid.
fluid and the kinetic energy. The pressure term, while
not a potential energy, is related to the work required to
move fluid into a higher pressure region.) Resistors
In an electric heater the electrical energy supplied by
Many features of hydrodynamic voltage should al- the power station is converted into heat energy by
ready be familiar. If you live in a house with good water having electric current flow through a dissipative or
pressure, when you turn on the faucet the water comes resistive material. The actual process by which electri-
out rapidly. But if someone is running the washing cal energy is turned into heat energy is fairly complex
machine in the basement or watering the garden, the but not unlike the conversion of mechanical energy to
water pressure may be low, and the water just dribbles heat through friction. One can think of resistance as an
out of the faucet. We will think of the high pressure internal friction encountered by the electric current.
water as high voltage water, and the low pressure water
as low voltage water. In our discussion of Bernoulli’s equation we saw that
the hydrodynamic voltage P + ρgh + 1/2 ρv2 was
Let us look more carefully at high voltage water in a constant along a stream line if there were no viscous
faucet. When the faucet is shut off, the water is at rest effects. But we also saw in Figure (23-24) that when
but the pressure is high, and the main contribution to the there were viscous effects this hydrodynamic voltage
hydrodynamic voltage is the P term. When the faucet dropped as we went along a stream line.
is on, the water that has just left the faucet has dropped
back to atmospheric pressure but it is moving rapidly. Heights in barometer tubes
Now it is the 1/2 ρv2 that contributes most to the dropping due to viscosity
hydrodynamic voltage. If the water originally comes

v
Figure 23-24
Hydrodynamic voltage drop due to viscous effects.
 27-7

In fluid flows, we get the most dissipation where the and is proportional to that current. When the voltage
fluid is moving rapidly through a narrow constriction. drop ∆ V is proportional to the current i, the resistor
This is seen in our venturi demonstration of Figure is said to obey Ohm’s law. This can be written as the
(23-18), reproduced here in Figure (6). Here we have a equation
large tube with a constriction. The glass barometer
tubes show us that the pressure remains relatively ∆V = iR Ohm's law (1)
constant before the constriction, but does not return to The proportionality constant R is called the resistance
its original value afterward. There is a net pressure drop R of the resistor. From Equation (1) you can see that R
of ρgh, where h is the height drop indicated in the has the dimensions volt/amp. This unit is called an
figure. ohm, a name which is convenient in practice but which
Consider the points in the fluid at the dots labeled (2) further complicates the problem of following dimen-
and (9), in the center of the stream below tubes 2 and 9. sions in electrical calculations.
These points are at the same heights ( h 2 = h 9 ), and the
fluid velocities are the same ( v2 = v9 ) because the flow R = ∆V amps
volts = ∆V ohms
i i
tube has returned to its original size. Because of the
Resistors are the most common element in electronic
pressure drop ( P9 < P2 ), the hydrodynamic voltage
( P9 + ρgh 9 + 1/2 ρv92 ) at point (9) is less than that at circuits. They usually consist of a small cylinder with
wire pigtails sticking out each end as shown in Figure
point (2) by an amount equal to P2 – P9 = ρgh. The
(7). The material inside the cylinder which creates the
barometer tubes 2 and 9 are acting as hydrodynamic
voltmeters showing us where the voltage drop occurs. voltage drop, which turns electrical energy into heat
energy, is usually carbon.
Just as in fluid flows, dissipation in electric currents are
The resistors you find in an electronics shop come in a
associated with voltage drops, in this case electrical
huge selection of values, with resistances ranging from
voltage drops. In general, the amount of the voltage
about 0.1 ohm up to around 109 ohms in a standard
drop depends on the amount of current, the geometry of
series of steps. The physical size of the resistor depends
the flow path, on the material through which the current
not on the value of the resistance but on the amount of
is flowing, and on the temperature of the material. But
in a special device called a resistor, the voltage drop ∆V electrical energy the resistor is capable of dissipating
without burning up. The value of the resistance is
depends primarily on the current i through the resistor
usually indicated by colored stripes painted on the
resistor, there being a standard color code so that you
h can read the value from the stripes.

(A light bulb is a good example of an electrical device

that dissipates energy, in this case mostly in the form of
heat and some light. The only problem with a light bulb
is that as the filament gets hot, its resistance increases.
If we wish to use Ohm’s law, we have to add the
qualification that the bulb’s resistance R increases with
temperature.)
resistor
wire wire

(2) • v water • (9) Figure 7

The resistor, found in most electronic circuits. The
(1) (3) (4) (5) (6) (7) (8) purpose of the resistor is to cause an electric voltage
Figure 6 drop analogous to the hydrodynamic voltage drop we
The hydrodynamic voltage, as measured by the saw in Figure 6 across the restriction in the flow tube.
barometer tubes, drops by an amount ρ gh in going
across the constriction from Point (2) to Point (9).
27-8 Basic Electric Circuits

A Simple Circuit Point (2) at the upper end of the resistor, is connected
To get some intuition for how resistors are used, to the positive terminal of the battery, Point (1), by a
consider the circuit shown in Figure (8) containing a wire. In our circuit diagrams we always assume that
battery and a resistor connected by wires. In drawing our wires are good conductors, having no electric fields
circuits, it is convention to use a line for a inside them and therefore no voltage drops along them.
wire, the symbol for a resistor, and Thus
+ – for a battery. In the symbol for a battery, no voltage drop
the short perpendicular line represents the negative V2 = V1 = Vb along a wire (4)
terminal of the battery and the long side the positive
The bottom of the resistor is connected to the negative
terminal. When we have a current i flowing through the
terminal of the battery by a wire, therefore
wire we draw an arrow indicating the direction of flow
of positive charge i and label the current no voltage drop
V3 = V4 = 0 along a wire (5)
with a letter such as i, i 1, etc.

In Figure (9), we have labeled the voltages V1, V2, V3 Equations (4) and (5) determine the voltage drop ∆V
and V4 at four points around the circuit. By definition that must be occurring at the resistor
we will take the negative side of the battery as being ∆V = V2 – V3 = Vb (6)
zero volts, or what we call ground
And by Ohm’s law, Equation (1), this voltage drop is
V4 = 0 volts by definition (2) related to the current i through the resistor by
On the positive side of the battery, the voltage is up to ∆V = iR = Vb Ohms law (7)
the battery voltage Vb which is 1.5 volts for a common
flashlight battery and up to 9 volts for many transistor Solving for the current i in the circuit gives
radio batteries
Vb
V1 = Vb the battery voltage (3) i = (8)
R
current i In future discussions of circuits we will not write out all
battery resistor the steps as we have in Equations (2) through (8), but
+ R the first time through a circuit we wanted to show all the
V
By convention, the – details. i
negative side of the
battery is usually
V2
considered to be at V1
0 volts (ground). wire + V=iR
– Vb
Figure 8 V4
About the simplest electrical circuit consists of a V3
battery connected to a resistor. If the resistor
Figure 9
were a light bulb, you would have a flashlight.
Voltages around the circuit.
 27-9

Equation (8) is the one that really shows us how Power

resistors are used in a circuit. We can see from As one of the roles of a resistor is electrical power
Equation (8) that if we use a small resistor, we get a big dissipation, let us determine the power that is being
current, and if we use a large resistor we get a small dissipated when a current is flowing through a resistor.
current. In most applications resistors are used to Recall that power is the amount of energy transferred
control the flow of current. or dissipated per unit time. In the MKS system power
has the dimensions of joules per second which is called
In modern electronics such as radios and computers, a watt
typical battery voltages are around 5 volts and typical
currents a milliampere (10-3 amps). What size resistor joules
Power = = watt (10)
R do we have to use in Equation (8) so that we get a one second
milliampere current from a 5 volt battery? The answer
is Now suppose we have a current flowing through a
resistor R as shown in Figure (10). The voltage drop
Vb 5 volts across the resistor is V, from a voltage of V volts at the
R = = top to 0 volts at the bottom as shown.
i 10 -3 amps
= 5000 ohms ≡ 5000 Ω (9) Because V is the electric potential energy of a unit
charge (the coulomb), every coulomb of charge flow-
where we used the standard symbol Ω for ohms. Many ing through the resistor loses V joules of electric
of the resistors in electronics circuits have values like potential energy which is changed to heat.
this in the 1,000 Ω to 10,000 Ω range.
If we have a current i, then i coulombs flow through
The Short Circuit the resistor every second. Thus the energy lost per
second is the number of coulombs (i) times the energy
Equation (8) raises an interesting problem. What if
lost per coulomb (V) or (iV):
R = 0 ? The equation predicts an infinite current! We
could try to make R = 0 by attaching a wire rather than joules
Power = i coul
sec × V coul
a resistor from Points (2) to (3) in Figure (9). What (11)
would happen is that a very large current would start to joules
= iV sec = iVwatts
flow and either melt the wire, start a fire, drain the
battery, or destroy the power supply. (A power supply Ohm’s law, Equation (1), can be used to express the
is an electronic battery.) When this happens, you have power in terms of R and either i or V
created what is called a short circuit. The common
lingo is that you have shorted out the battery or power V2
2
Power = iV = i R = (11a)
supply and this is not a good thing to do. R
i
V

V=0
Figure 10
The voltage drops from V to 0 as the current i flows
through the resistor. The power dissipated is the
current i coulombs/second times the voltage drop V
joules/coulomb, which is iV joules/second, or watts.
27-10 Basic Electric Circuits

Exercise 1 KIRCHOFF’S LAW

These are some simple exercises to have you become Imagine that you are going for an afternoon hike on a
familiar with the concepts of volts and amps. nearby mountain. You drive up to the base lodge, park
your car, and start up the trail. The trail goes up over a
a) Design a circuit consisting of a 9 volt battery and a
ridge, down into a ravine, up to the peak of the
resistor, where the current through the resistor is 25
milliamperes ( 25×10– 3 amps).
mountain, down the other side and then around the
mountain back to the base lodge. When you get back
b) A flashlight consists of a 1.5 volt battery and a 1 watt to your car, how much gravitational potential energy
light bulb. How much current flows through the bulb have you gained from the trip? The answer is clearly
when the flashlight is on? zero—you are right back where you started.
c) When you plug a 1000 watt heater into a 120 volt If you defined gh, which is the potential energy of a unit
power line, how much current goes through the heater?
mass, as your gravitational voltage, then as you went up
What is the resistance R of the heater when the filament
is hot?
the ridge, there was a voltage rise as h increased. Going
down into the ravine there was a voltage drop, or what
d) In most households, each circuit has a voltage of 120 we could call a negative voltage rise. The big voltage
volts and is fused for 20 amps. (The circuit breaker rise is up to the top of the mountain, and the big negative
opens up if the current exceeds 20 amps). What is the voltage rise is down the back side of the mountain.
maximum power you can draw from one circuit in your When you add up all the voltage rises for the complete
house? trip, counting voltage drops as negative rises, the sum
e) An electric dryer requires 3000 watts of power, yet it is zero.
has to be plugged into wires that can handle only 20
amps. What is the least voltage you can have on the
Consider our Figure (9) redrawn here. If we start at
circuit? Point (4) where the voltage is zero, and “walk” around
the circuit in the direction of the positive current i, we
f) In many parts of the world, the standard voltage is 240 first encounter a voltage rise up to V = Vb due to the
volts. The wires to appliances are much thinner. Explain battery, then a voltage drop back to zero at the resistor.
why. When we get back to the starting point, the sum of the
voltage rises is zero just as in our trip through the
mountains. Even in more complicated circuits with
many branches and different circuit elements, it is
usually true that the sum of the voltage rises around any
complete path, back to your starting point, is zero. It
turns out that this is a powerful tool for analyzing
electric circuits, and is known as Kirchoff’s law.
(Kirchoff’s law can be violated, we can get a net
i voltage rise in a complete circuit, if changing magnetic
V2 fields are present. We will treat this phenomenon in a
V1
+ later chapter. For now we will discuss the usual
V=iR situation where Kirchoff’s law applies.)
– Vb
V4
V3

Figure 9 (redrawn)
Voltages around the circuit.
 27-11

Application of Kirchoff’s Law

Sum of the voltage rises
There are some relatively standard, cookbook like
going clockwise around = Vb + VR
procedures that make it easy to apply Kirchoff’s law to
the analysis of circuits. The steps in the recipe are as the circuit of Figure 12 (12)
follows: = Vb + (-iR)

(1) Sketch the circuit and use arrows to show the = 0

Equation (12) gives
direction of the positive current in each loop as we did
in Figure (11). Do not be too concerned about getting i = Vb (13)
the correct direction for the current i. If you have the R
i which is the result we had back in Equation (8).
Figure 11
Labeling the Series Resistors
direction of the By now we have beaten to death our simple battery
current. resistor circuit. Let us try something a little more
challenging—let us put in two resistors as shown in
arrow pointing the wrong way, then when you finish Figure (13). In that figure we have drawn the circuit
solving the problem, i will turn out to be negative. and labeled the direction of the current (Step 1), and
drawn in the arrows representing the voltage rises (Step
(2) Label all the voltage rises in the circuit. Use arrows 2). Setting the sum of the voltage rises equal to zero
to indicate the direction of the voltage rise as we did in (Step 3) gives
Figure (12). Note that if we go through the resistor in
the direction of the current, we get a voltage drop. Vb + (-iR1) + (-iR2) = 0 (14)
Therefore the arrow showing the voltage rise in a
resistor must point back, opposite to the direction of the i = Vb
(15)
(R1 + R2 )
i
The two resistors in Figure (13) are said to be connected
Figure 12 in series. Comparing Equation (13) for a single resistor
Vb VR = i R Labeling the and Equation (15) for the series resistors, we see that if
voltage rises.
R1 + R2 = R (series resistors) (16)
then we get the same current i in both cases (if we use
current i in the resistor. (The analogy is to a rock strewn the same battery). We say that if R1 + R2 = R then
waterfall where the water loses hydrodynamic voltage the series resistors are equivalent to the single resistor
as it flows down through the rocks. The direction of the R.
voltage rise is back up the waterfall, in a direction
opposite to that of the current.) i

(3) The final step is to “walk” around the loop in the

R1 V1 = i R 1
direction of i (or any direction you choose), and set the
sum of the voltage rises you encounter equal to zero. If
you encounter an arrow that points in the direction you Vb
are walking, it counts as a positive voltage rise (like Vb
in Figure 12). If the arrow points against you (like VR), R2 V2 = i R 2
then it is a negative rise. Applying this rule to Figure
(12) gives
Figure 13
Two resistors in series.
27-12 Basic Electric Circuits

Parallel Resistors The main problem with using Kirchoff’s laws for
A bit more challenging is the circuit of Figure (14) complex circuits is that we can get more equations than
where the resistors are wired in “parallel”. In Step (1), we need or want. For our current example, if you solve
we drew the circuit and labeled the currents. But here Equation (18) for Vb = i1R1, then put that result in
we have something new. When the current gets to the Equation (20), you get i1R1 - i2R2 = 0 which is
point labeled (A), it is like a fork in the stream and the Equation (19). In other words Equation (19) does not
current divides. We have labeled the two branch tell us anything that we did not already know from
currents i1 and i 2 , and have the obvious subsidiary Equations (18) and (20). The mathematicians would
condition (conservation of current, if you like). say that Equations (18), (19), and (20) are not linearly
independent.
i1 + i 2 = i (17)
Let us look at the situation from a slightly different
There is no problem with Step (2), the voltage rises are point of view. To completely solve the circuit of Figure
Vb, i1R1 and i 2R 2 as shown. But we get something (15), we have to determine the currents i, i1 and i2. We
new when we try to write down Kirchoff’s law for the have three unknowns, but four equations, Equations
sum of the voltage rises around a complete circuit. (17), (18), (19) and (20). It is well known that you need
Now we have three different ways we can go around a as many equations as unknowns to solve a system of
complete circuit, as shown in Figures (15 a, b, c). equations, and therefore we have one too many equa-
tions.
Applying Kirchoff’s law to the path shown in Figure
(15a) we get i
i1
Vb + (-i1R1 ) = 0 (18)
Vb R1 R2 (a)
For Figure (15b) we get

(-i2R 2) + (i1R 1) = 0 (19)

i2
and for Figure (15c) we get i1

Vb + (-i2R2 ) = 0 (20) R1 R2 (b)

i A i2
i1

Vb R1 i 1R 1 i 2R 2 i i2
i1

Vb R1 R2 (c)
Figure 14
Two resistors in parallel.

Figure 15
Three possible loops for analyzing the
parallel resistance circuit. They give
more equations than needed.
 27-13

We cannot arbitrarily throw out one of the equations for Exercise 2

the remaining three must be linearly independent. For You are given a device, sealed in a box, with electrical
example, if we threw out Equation (17) and tried to leads on each end. (Such a device is often referred to
solve Equations (18), (19) and (20) for i1, i2, and i, we as a "black box", the word black referring to our lack of
couldn’t get an answer because Equation (19) contains knowledge of the contents, rather than the actual color
no information not already in Equation (18) and (20). of the device.) You use an instrument called an
When you are working with a system of linear equa- ohmmeter to measure the electrical resistance be-
tions, the hardest problem is to decide which is a set of tween the two terminals and find that it's resistance R is
linear independent equations. Then you can use a 470 ohms (470 Ω ).
standard set of procedures that mathematicians have R = 470 Ω
for solving linear equations. These procedures involve
determinants and matrices, which are easily handled on
a computer, but are tedious to work by hand. a) Sketch a circuit, containing the black box and one
resistor, where the total resistance of the circuit is
In our treatment of circuit theory we will limit our
500 Ω .
discussion to simple circuits where we can use grade
school methods for solving the equations. Problems of b) Sketch a circuit, containing the black box and one
linear independence, determinants and matrices will be resistor, where the total resistance of the circuit is
left to other treatments of the topic. 400 Ω .

To solve our parallel resistor circuit of Figure (14), we Exercise 3 The Voltage Divider
have from Equation (18) We wish to measure the voltage Vb produced by a high
voltage power supply, but our voltmeter has the limited
Vb range of +2 to -2 volts. To make the measurement we
i1 =
R1 use the voltage divider circuit shown below, containing
a big resistor R1 and a small resistor R2 . If, for example,
and from Equation (20) R2 is 1000 times smaller than R1 , then the voltage
across R2 is 1000 times smaller than that across R1 . By
Vb measuring the small voltage across the small resistor
i2 =
R2 we can use this result to determine the big voltage Vb .
Substituting these values in Equation (17) gives a) What current i flows through the circuit. Express your
answer in terms of Vb .
Vb Vb b) Find the formula for Vb in terms of V2 , the voltage
i = i1 + i2 = +
R1 R2 measured across the small resistor.
(21)
1 1 c) Find a formula for Vb in terms of V2 , R1 and R2 ,
= Vb ( + )
R1 R2 assuming R1 > > R2 , so that you can replace ( R1 + R2 )
by R1 in the equation for i.
Comparing Equation (21) for parallel resistors, and
Equation (13) for a single resistor d) Our voltmeter reads V2 = .24 volts. What was Vb ?

i = Vb ( 1 ) (13)
R 6
R1 = 10 Ω
V1
We see that two parallel resistors R1 and R2 are equiva- Voltage
lent to a single resistor R if they obey the relationship divider Vb
circuit
3 volt
equivalent V2 R2 = 10 Ω
1 1 1 meter
= + parallel (22)
R R1 R2 resistors
27-14 Basic Electric Circuits

CAPACITANCE AND CAPACITORS Let us denote by the letter Q the quantity or volume of
In addition to the resistor, another common circuit water stored in the tank. If we talk only about cylindri-
element is the capacitor. A resistor dissipates energy, cal tanks (of cross-sectional area A), then this volume
causes a voltage drop given by Ohm’s law V = iR, and is proportional to the height h and therefore the hydro-
is often used to limit the amount of current flowing in dynamic voltage Vh
a section of a circuit. A capacitor is a device for storing Volume of water A
electrical charge and maintains a voltage proportional ≡ Q = Ah = ρgh
in cylindrical tank ρg
to the charge stored. We have already seen one explicit
example of a capacitor, the parallel plate capacitor
studied in the last chapter. Here we will abstract the A
Q = V (24)
general features of capacitors, and see how they are ρg h
used as circuit elements.
If we define the proportionality constant A/ρg in
Equation (24) as the capacitance C of the tank
Hydrodynamic Analogy
Before focusing on the electrical capacitor, it is instruc- capacitance of
tive to consider an accurate hydrodynamic analogy— A a cylindrical tank
the cylindrical water tank shown in Figure (16). If the C = ≡ (25)
ρg with a cross –
tank is filled to a height h, then all the water in the tank sectional area A
has a hydrodynamic voltage
1 2 then we get
Vh = P + ρgh + ρv = ρgh (23)
2
Q = CVh (26)
For water at the top of the tank, y = h, the voltage is all
in the form of gravitational potential energy ρgh. (We as the relation between the hydrodynamic voltage and
will ignore atmospheric pressure.) At the bottom of the volume Q of water in the tank.
tank where y = 0, the voltage is all in the pressure term
P = ρgh. The dynamic voltage term 1 2 ρv 2 does not
play a significant role.

area A
Figure 16
Analogy between a cylindrical tank of water and an
electrical capacitor. In the tank, all the water in the
tank is at a hydrodynamic voltage Vh = ρ gh , and the
quantity Q of water in the tank, given by Q = Ah =
A/ ρ g ρ gh = A/ ρ g Vh is proportional to Vh .
 27-15

Cylindrical Tank as a
Constant Voltage Source
One of the main uses of a water storage tank is to
maintain a water supply at constant hydrodynamic
voltage.

Figure (17) is a schematic diagram of a typical town

water supply. Water is pumped from the reservoir up
into the water tank where a constant height h and
therefore constant voltage ρgh is maintained. The
houses in the town all draw constant voltage water from
this tank.

Let us see what would happen if the water tank was too
small. As soon as several houses started using water,
the level h in the tank would drop and the pump at the
reservoir would have to come on. The pump would
raise the level back to h and shut off. Then the level
would drop again and the pump would come on again.
The result would be that the hydrodynamic voltage or
water pressure supplied to the town would vary and
customers might complain.

On the other hand if the town water tank has a large

cross-sectional area and therefore large capacitance C,
a few houses drawing water would have very little
effect on the level h and therefore voltage ρgh of the
water. The town would have a constant voltage water
supply and the water company could pump water from
the reservoir at night when electricity rates were low.

We will see that one of the important uses of electrical

capacitors in electric circuits is to maintain constant or tank
nearly constant electric voltages. There is an accurate h
analogy to the way the town water tank maintains
constant voltage water. If we use too small a capacitor,
the electrical voltage will also fluctuate when current is
reservoir pump
drawn.
Figure 17
Town water supply. By maintaining a
constant height h of water in the storage
tank, all the water supplied to the town has a
constant hydrodynamic voltage Vh ρ gh .
27-16 Basic Electric Circuits

Electrical Capacitance Using Equation (28) for E and Equation (27) for σ, we
Figure (18) is a repeat of the sketches of the parallel get the relationship
V = σd =
plate capacitor discussed in Chapter 26. The important Q
d
features of the capacitor are the following. We have ε0 ε0 A
two metal plates of area A separated by a distance d. or
The positive plate shown on top has a charge + Q, the
bottom plate a charge – Q. Since the area of the plates ε0A
is A, the surface charge density on the inside of the plate Q = V (30)
d
is
Q
σ = (27) which is our old Equation (26-10).
A
As in our hydrodynamic analogy, we see that the
In Chapter 26, page 26-3, we saw that a charge density quantity of charge Q stored in the capacitor is propor-
σ on the surface of a conductor produced an electric tional to the voltage V on the capacitor. Again we call
field of strength the proportionality constant the capacitance C
E = σ (28)
ε0 definitionof
perpendicularly out of the conductor. In Figure (18) Q = CV electrical (31)
capacitance
this field starts at the positive charge on the inside of the
upper plate and stops at the negative charge on the
inside surface of the bottom plate. Comparing Equations (30) and (31) we see that the
formula for the capacitance C of a parallel plate capaci-
Recall that one form of electric voltage is the electric tor is
potential energy of a unit test charge. To lift a positive
capacitanceof a
unit test charge from the bottom plate to the top one ε0 A parallel plate
requires an amount of work equal to the force E on a C = capacitor of area A, (32)
d
unit charge times the distance d the charge was lifted. plateseparation d
This work E * d is equal to the increase of the potential
energy of the unit charge, and therefore to the increase For both the parallel plate capacitor and the cylindrical
in voltage in going from the bottom to the top plate. If water tank, the capacitance is proportional to the cross-
we say that the bottom plate is at a voltage V = 0, then sectional area A. The new feature for the electrical
the voltage at the top plate is capacitor is that the capacitance increases as we make
the plate separation d smaller and smaller.
V = Ed (29)
Our parallel plate capacitor is but one example of many
+Q kinds of capacitors used in electronic circuits. In some,
V = Ed the geometry of the metal conductors is different, and
E= σ/ε0 in others the space between the conductors is filled with
d
V=0 a material called a dielectric which increases the effec-
plates of area A –Q tive capacitance. But in all common capacitors the
Figure 18
amount of charge Q is proportional to voltage V across
The parallel plate capacitor. If we place charges + Q the capacitor, i.e. Q = CV, where C is constant indepen-
and – Q on plates of area A, the charge density on dent of the voltage V and in most cases independent of
the plates will be σ = Q/ A , the electric field will be the temperature.
E = σ / ε 0 and the voltage between the plates V = Ed.
 27-17

The dimensions of capacitance C are coulombs per Exercise 4 - Electrolytic Capacitor

volt, which is given the name farad in honor of Michael In an electrolytic capacitor, one of the plates is a thin
Faraday who pioneered the concept of an electric field. aluminum sheet and the other is a conducting dielectric
Although such an honor may be deserved, this is one liquid surrounding the aluminum. A nonconducting
more example of the excessive use of names in the oxide layer forms on the surface of the aluminum and
MKS system that make it hard to follow the dimensions plays the same role as the air gap in the parallel plate
in a calculation. capacitors we have been discussing. The fact that the
oxide layer is very thin means that you can construct a
To get a feeling for the size of a farad, suppose that we capacitor with a very large capacitance in a small
have two metal plates with an area A = 0.1 meter2 container.
and make a separation d = 1 millimeter = 10-3 meters.
These plates will have a capacitance C given by dielectric liquid
ε 0A 9 × 10 -12 × .1
C = = aluminum oxide
d 10 -3 layer

= 9 × 10 -10 farads For this problem, assume that you have a dielectric
capacitor whose total capacitance is 1 farad, and that
which is about one billionth of a farad. If you keep the the oxide layer acts like an air gap 10 – 7 meters thick in
separation at 1 millimeter you would need plates with a parallel plate capacitor. From this, estimate the area
an area of 100 million square meters (an area 10 of the aluminum surface in the capacitor.
kilometers on a side) to have a capacitance of 1 farad.

Commercial capacitors used in electronic circuits come

in various shapes like those shown in Figure (19), and
in an enormous range of values from a few farads down
to 10 –14 farads.

Our calculation of the capacitance of a parallel plate

capacitor demonstrates that it is not an easy trick to
produce capacitors with a capacitance of 10 – 6 farads
or larger. One technique is to take two long strips of
metal foil separated by an insulator, and roll them up
into a small cylinder. This gives us a large plate area
with a reasonably small separation, stuffed into a
relatively small volume.

In a special kind of a capacitor called an electrolytic

capacitor, the effective plate separation d is reduced to
almost atomic dimensions. Only this way are we able Figure 19
to create the physically small 1 farad capacitor shown Examples of capacitors used in electronic circuits. The
in Figure (19). The problem with electrolytic capaci- one on the right is a variable capacitor whose plate
tors is that one side has to be positive and the other area is changed by turning the knob. The square black
capacitor is a 4 farad electrolytic. Its capacitance is one
negative, as marked on the capacitor. If you reverse the
million times greater than the tall regular capacitor
voltage on an electrolytic capacitor, it will not work and behind it.
may explode.
27-18 Basic Electric Circuits

ENERGY STORAGE IN CAPACITORS Now let us take a charge dQ out of the bottom plate,
In physics, one of the important uses of capacitors is leaving a charge – (Q + dQ) behind, and lift it to the top
energy storage. The advantage of using capacitors is plate, leaving (Q + dQ) there. The work dW we do to
that large quantities of energy can be released in a very lift the charge is equal to dQ times the work required to
short time. For example, Figure (20) is a photograph lift a unit test charge, namely dQ times the voltage V
of the Nova laser at the Lawrence Livermore National dW = VdQ
Laboratory. This laser produces short, but very high
energy pulses of light for fusion research. The laser is or replacing V by Q/C, we have
powered by a bank of capacitors which, for the short Q
dW = dQ (33)
length of time needed, can supply power at a rate about C
200 times the power generating capacity of the United +Q
States. V volts
The easiest way to determine the amount of energy
dQ
stored in a capacitor is to calculate how much work is 0 volts
required to charge up the capacitor. In Figure (21) we –Q
have a capacitor of capacitance C that already has a Figure 21
charge + Q on the positive plate and – Q on the negative Charging up a capacitor. If the capacitor is
already charged up to a voltage V, the amount
plate. The voltage V across the capacitor is related to
of work required to lift an additional charge dQ
Q by Equation (12), Q = CV. from the bottom to top plate is dW = VdQ.

Figure 20
The Nova laser, powered by a bank of capacitors. While the laser
is being fired, the capacitors supply 200 times as much power as
the generating capacity of the United States.
 27-19

You can see from Equation (33) that when the capacitor Since the energy stored in the capacitor is proportional
is uncharged and we lift the first dQ, no work is required to the volume occupied by the electric field, we see that
because there is no field yet in the capacitor. However the energy per unit volume, the energy density, is
once there is a big charge on the capacitor, much work simply given by
is required to lift an additional dQ. The total amount of
work to charge the capacitor from zero charge to a final
charge Q f is clearly given by the integral Energy ε0 E 2
= (36)
Qf
Q density 2
Work = dW = dQ
0 C
This result, that the energy density in an electric field is
The fact that the capacitance C is a constant, means that proportional to the square of the strength of the field,
we can take it outside the integral and we get turns out to be a far more general result than we might
Qf expect from the above derivation. It applies not just to
1 Q2f the uniform electric field in an idealized capacitor, but
W = Q dQ = (34)
C 2C to electric fields of arbitrary shape.
0
Since it is easier to measure the final voltage V rather Exercise 5
than the charge Qf in a capacitor, we use Qf = CV to A parallel plate capacitor consists of two circular alumi-
rewrite Equation (34) in the form num plates with a radius of 11 cm separated by a
distance of 1 millimeter. The capacitor is charged to a
voltage of 5 volts.
Energy stored CVf 2
= a) What is the capacitance, in farads, of the capacitor?
in a capacitor 2 (35)
b) Using Equation 35, calculate the energy stored in the
capacitor.
The energy stored is proportional to the capacitance C
of the capacitor, and the square of the voltage V. c) What is the magnitude of the electric field E between
the plates?

Energy Density in an Electric Field d) Using equation 36, calculate the energy density in the
Equation 35 can be written in a form that shows that the electric field.
energy stored in a capacitor is proportional to the e) What is the volume of space, in cubic meters,
square of the strength of the electric field. Substituting between the plates?
Vf = E × d and C = ε0 A d into Equation 35 gives
f) From your answers to parts d) and e), calculate the
total energy in the electric field between the plates.
Energy stored C Vf 2 1 ε A
Compare your answer with your answer to part b-
= = 0
× E2 d 2
in a capacitor 2 2 d .

g) Using Einstein's formula E = mc2 , calculate the mass,

ε E2 ε E2 Volume
= 0 ×Ad = 0 × Inside in kilograms, of the electric field between the plates.
2 2 capacitor
h) The mass of the electric field is equal to the mass of
where we note that A × d is the volume inside the how many electrons?
capacitor.
27-20 Basic Electric Circuits

CAPACITORS AS Applying Kirchoff’s law to Figure (22), i.e., setting the

CIRCUIT ELEMENTS sum of the voltage rises around the circuit equal to zero,
we get
Figure (22) is a simple circuit consisting of a battery of
voltage Vb and a capacitor of capacitance C. The Vb + (-Vc) = Vb -Q/C = 0
standard circuit symbol for a capacitor is ,
which is a sketch of a parallel plate capacitor. Q = CVb (37)
When the battery is attached to the capacitor, the upper Thus we get a relatively straightforward result for the
plate becomes positively charged and the lower one amount of charge stored by the battery.
negatively charged as shown. The upper plate could For something a little more challenging, we have
actually become positively charged either by positive connected two capacitors in parallel to a battery as
charge flowing into it or negative charge flowing out— shown in Figure (23). Because single wires go all the
it does not matter. We have followed our convention way across the top and across the bottom, the three
of always showing the direction of positive currents, voltages Vb, V1 and V2 must all be equal, and we get
thus we show i flowing into the positive plate and out
of the negative one. Q1 = C1Vb Q2 = C2Vb
We have also followed our convention of labeling the The total charge Q stored on the two capacitors in
voltage rises with an arrow pointing in the direction of parallel is therefore
the higher voltage. The voltage Vc on the capacitor is
related to the charge Q stored by the definition of Q = Q1 + Q2 = (C1 + C2 )Vb
capacitance, Vc = Q/ C .
Comparing this with Equation (37), we see that two
+i capacitors in parallel store the same charge as a single
capacitor C given by
+Q Q capacitors
Vb Vc =
C C = C1 + C2 attached in
(38)
–Q parallel

Comparing this result with Equation (16), we find that

+i
Figure 22 for capacitors in parallel or resistors in series, the
A battery and a capacitor in a circuit. We have drawn effective capacitance or resistance is just the sum of the
the diagram showing positive current flowing into the values of the individual components.
top plate and out of the bottom plate. The upper plate
could have become positively charged by having a i i2
negative current flowing out of it. The arrow i1
designating the voltage on the capacitor points in the Q2
direction of the voltage rise. Vb C1 V1 = Q1 C2 V2 =
C1 C2

Figure 23
Capacitors connected in parallel. The
three voltages Vb , V1 and V2 must all
be level because the wires go all the
way across the three elements.
 27-21

In Figure (24) we have two capacitors in series. The It is interesting to note that for storing charge, parallel
trick here is to note that all the charge that flowed out capacitors are more efficient because the charge can
of the bottom plate of C1 flowed into the top plate of C2, flow into both capacitors as seen in Figure (23). When
as indicated in the diagram. But if there is a charge the capacitors are in series, charge flowing out of the
– Q on the bottom plate of C1, there must be an equal bottom of one capacitor flows into the top of the next,
and opposite charge + Q on the top and we have and we get no enhancement in charge storage capabil-
Q1 = Q. Similarly we must have Q2 = Q. ity. What we do get from series capacitors is higher
voltages, the total voltage rise across the pair is the sum
To apply Kirchoff’s law, we set the sum of the voltage of the voltage rise on each.
rises to zero to get
–Q1 –Q2 Exercise 6
Vb + + = 0 You have a 5 microfarad (abbreviated 5µf) capacitor
C1 C2
and a 10 µf capacitor. What are all the values of
Setting Q1 = Q2 = Q gives capacitor you can make from these two?

1 1
Vb = Q + (39)
C1 C2

Comparing Equation (39) with Equation (37) in the

form Vb = Q/C we see that

1 1 1 capacitors
= + attached in (40)
C C1 C2 series

is the formula for the effective capacitance of capaci-

tors connected in series. This is analogous to the
formula for parallel resistors.
i

C1 V1 = Q1
C1

Vb i

C2 V2 = Q2
C2

i
Figure 24
Capacitors in series. In this case the sum of V1
and V2 must be equal to the battery voltage Vb .
27-22 Basic Electric Circuits

mercury switch THE RC CIRCUIT

The capacitor circuits we have discussed so far are not
too exciting. When you are working with an electronic
Vb C R circuit you do not hitch capacitors together in series or
parallel, you simply go to the parts drawer and select a
capacitor of the desired value.

mercury switch
If we add a resistor to the circuit as shown in Figure
closed (25), we begin to get some interesting results. The
i = Vb /R circuit is designed so that if the mercury switch is
closed, the capacitor is charged up to a voltage Vb by the
Vb C VC = Vb R VR = Vb
battery. Then, at a time we will call t = 0, the switch is
opened, so that the capacitor will discharge through the
resistor. During the discharge, the battery is discon-
nected and the only part of the circuit that is active is that
mercury switch shown in Figure (26). (The reason for using a mercury
open switch was to get a clean break in the current. Mechani-
i cal switches do not work well.)

Figure (27) shows the capacitor voltage just before and

Vb C VC = Q R VR = iR
C for a while after the switch was opened. We are looking
at the experimental results of discharging a
C = 10 -6 farad (one microfarad) capacitor through an
R = 10 4 ohm resistor. We see that a good fraction of
Figure 25 the capacitor voltage has decayed in about 10 millisec-
An RC circuit. When the mercury switch is
onds (10-2 seconds) .
closed, the capacitor quickly charges up to a
voltage VC = Vb . When the switch is opened, the
To analyze the capacitor discharge, we apply Kirchoff’s
capacitor discharges through the resistor.
law to the circuit in Figure (26). Setting the sum of the
voltage rises around the circuit equal to zero gives
i
VC – VR = 0

Q
VC = C C R VR = iR Q
– iR = 0
C
which can be written in the form
Figure 26 Q
Capacitor discharge. When the switch is open, the –i + = 0 (41)
RC
only part of the circuit we have to look at is the
capacitor discharging through the resistor.
The problem with Equation (41) is that we have two
unknowns, i and Q, and only one equation. We need to
find another relationship between these variables in
order to predict the behavior of the circuit.
 27-23

The additional relationship is obtained by noting that There are two principle ways of solving a differential
the current i is the number of coulombs per second equation. One is to use a computer, and the other, the
flowing out of the capacitor. In a short time dt, an so-called analytic method, is to guess the answer and
amount of charge dQ that leaves the capacitor is given then check to see if you have made the correct guess.
by We will first apply the analytic method to Equation
dQ = idt (42) (44). In the supplement we will show how a computer
solution is obtained.
Dividing Equation (42) through by dt, and including a
The important thing to remember about a differential
minus sign to represent the fact that i is causing a
equation is that the solution is a shape or a curve, not a
decrease in the charge Q in the capacitor, we get
number. The equation x2 = 4 has the solutions
dQ x = ±2, the solution to Equation (44) is the curve
discharge of
= –i (43) shown in Figure (27). One of the advantages of
dt a capacitor
working with electric circuits is that the theory gives
you the differential equation, and the equipment in the
Substituting Equation (43) in (41) gives one equation
lab allows you to look at the solution. The curve in
for the unknown Q
Figure (27) is the voltage on the capacitor recorded by
equation for the computer based oscilloscope we used to record the
dQ Q
+ = 0 the discharge (44) motion of air carts and do the analysis of sound waves.
dt RC of a capacitor We are now using the device as a voltmeter that draws
a picture of the voltage.
Exponential Decay
The curve in Figure (27) is well known to scientists in
The next problem is that Equation (44) is a differential
many fields as an exponential decay. Exponential
equation, of a type we have not yet discussed in the text.
decays are best known in studies of radioactive decay
We met another kind of differential equation in our
and are associated with the familiar concept of a half
discussion of harmonic motion, a differential equation
life. Let us first write down the formula for an exponen-
that involved second derivatives and had oscillating
tial decay, check that the formula is, in fact, a solution
sinusoidal solutions. Equation (44) has only a first
to Equation (44) and then discuss the special properties
derivative, and produces a different kind of solution.
of the curve.

Vb
Figure 27
Experimental results
from discharging the
one microfarad
10– 6 f capacitor
i through a 10 k ohm
104 Ω resistor.
-6 The switch, shown
C = 10 f R = 104 Ω in Figure 26, is
thrown at time t = 0.
27-24 Basic Electric Circuits

If Figure (27) represents an exponential decay of the The Time Constant RC

capacitor voltage Vc, then Vc must be of the form Substituting Equation (49) for α back into our formula
VC = V0 e–α t for VC gives
(45)
where V0 and α are constants to be determined. Since VC = V0 e– t RC
(50)
Equation (44) is in terms of Q rather than VC , we can Since the exponent (t/RC) must be dimensionless, the
use the definition of capacitance Q = CV to rewrite quantity RC in the denominator must have the dimen-
Equation (45) as sions of time. Since R is in ohms and C in farads, we
must have
CVC = CV0 e– α t
(46) ohms * farads = seconds (51)
– αt
Q = Q0 e
We have mentioned that units in electrical calculations
where Q0 = CV0. are hard to follow, and this is a prime example. We
leave it as a challenge to go back and actually show,
Differentiating Equation (46) with respect to time
from the definition of the ohm and of the farad, that the
gives
product ohms times farads comes out in seconds.
dQ
= – αQ0 e–α t = – α Q (47) The quantity RC that appears in Equation (50) is known
dt
as the time constant for the decay. At the time t = RC,
This result illustrates one of the properties of an expo- the voltage VC has the value
nential decay, namely that the derivative of the function
is proportional to the function itself (here dQ dt = –αQ ). VC (at t = RC) = Vbe –RC
RC RC
(52)
Substituting Equation (47) into our differential Equa- Vb
= Vbe – 1 = e
tion (44) gives
I.e., in one time constant RC, the voltage has decayed
dQ Q to 1/e = 1/2.7 of its initial value.
+ = 0
dt RC (44)
To see if this analysis works experimentally, we have
Q gone back to Figure (27) and marked the time RC. In
–αQ + = 0 (48) that experiment
RC
R = 104 ohms
The Qs cancel in Equation (48) and we get C = 10-6 farads
thus
1
α = (49)
RC RC = 10 -2 ohm farads
= 10 -2 seconds (53)
Thus the coefficient of the exponent is determined by
the differential equation. = 10 milliseconds
We see that at a time T = RC, the voltage dropped from
Exercise 7 Vb = 4 volts to V(t=RC) = 1.5 volts, which is down by
Determine the constant V0 in Equation (45) from Figure a factor 1/e = 1/2.7.
(27), by noting that at t = 0, e– α t = e0 = 1 .
 27-25

If we wait another time constant, until t = 2RC, we have Initial Slope

Vb One of the special features of a time constant is the fact
VC (at t = 2RC) = Vb e–2 = that the initial slope of the curve intercepts the zero
e2 value one time constant later, as illustrated in Figure
and we get another factor of e = 2.7 in the denominator. (30). It does not matter where we take our initial time
In Figure (28) the voltage is down to 4/(2.7*2.7) = .55 to be. Pick any point on the curve, draw a tangent line
volts at t = 2RC. After each succeeding time constant at that point, and the tangent line intercepts the V = 0
RC, the voltage drops by another factor of 1/e. line one time constant RC later. This turns out to be the
most convenient way to determine the time constant
from an experimental curve. Try it yourself in the
Half-Lives
following exercise.
When you first studied radioactive decay, you learned
initial slope
about half-lives. A half-life was the time it took for half
of the remaining radioactive particles to decay. Wait
another half-life and half of those are gone. In our Vb
description of the exponential decay, the time constant
RC is similar to a half-life, but just a bit longer. When
we wait for a time constant, the voltage decays down to
Vb e
1/2.7 of its initial value rather than 1/2 of its initial /
value. In Figure (29) we compare the half-life t 1/2 and Vb e2
/
the time constant RC. Although the half-life is easier 0
to explain, we will see that the time constant RC RC 2RC
provides a more convenient unit of time for the analysis Figure 30
of the exponential decay curve. The initial slope of the discharge curve intersects the
VC = 0 origin at a time t = RC, one time constant later.
This fact provides an easy way to estimate the time
Vb constant for an exponential curve.

/
Vb e

b/
V e 2

RC 2RC
Figure 28
Exponential decay of the voltage in the capacitor. In a
time t = RC the voltage drops by a factor 1/e = 1/2.7. In
the next time interval RC, the voltage drops by another
factor of 1/e.

1/e = 1/2.7
Vb Figure 29
The time it takes the voltage to
drop to half its initial value, what
Vb 2
/ we could call the "half life" of
the voltage, is a bit shorter than
/
Vb e the time constant RC.
0

t(1/2) RC
27-26 Basic Electric Circuits

Exercise 8 In Figure (32) we have drawn the circuit diagram and

indicated the voltage rises around the circuit. Kirchoff’s
In Figure (31) we have the experimental voltage decay
law there gives
for an RC circuit where R is known to be 10 5 ohms.
Determine the time constant for this curve and from that Vb + (–VR) + (–VC) = 0
find out what the value of the capacitance C must have
been. Q
Vb – iR – = 0 (54)
C
One difference between Figure (32) for the charge up
The Exponential Rise
and Figure (27) for the discharge, is that for the charge
Figure (32) is a circuit in which we use a battery to
up, the current i is flowing into, rather than out of, the
charge a capacitor through a resistance R. The experi-
capacitor. Therefore in a time dt the charge in the
mental result, for our capacitor C = 10-6 farads, R = 104
capacitor increases by an amount dQ = idt, and we
ohms is shown in Figure (33). Here the capacitor starts
have
charging relatively fast, then the rate of charging slows
until the capacitor voltage finally reaches the battery dQ
= +i (55)
voltage Vb . dt

If the shape of the curve in Figure (33) looks vaguely Using Equation (55) in (54), dividing through by R, and
familiar, it should. Turn the curve over and it looks like rearranging a bit, gives
our exponential decay curve with time increasing dQ Q
toward the left. If that is true, then the initial slope of + = Vb
dt RC R (56)
the charge up curve should intercept the VC = Vb line
It is the term on the right that makes this differential
one time constant later as shown in Figure (34), and it
equation harder to solve. A simple guess like the one
does.
we made in Equation (46) does not work, and we have
Because we have a battery in the circuit while the to try a more complicated guess like
capacitor is charging up, the analysis is a bit more Q = A + Be– α t (57)
messy than for the capacitor discharge. (You should be
able to repeat the analysis of the capacitor discharge on When you take a course in solving differential equa-
your own, and at least be able to follow the steps for tions, much of the time is spent learning how to guess
analyzing the charge up.) the form of solutions. For now, let us just see if the
guess in Equation (57) can be made to work for some
value of the constants A, B, and α .
i

C R = 10 5Ω
R VR = iR

Vb

C VC = Q/C

Figure 31
Experimental results for the discharge of Figure 32
a capacitor through a 105 Ω resistor. Charging up a capacitor through a resistor.
 27-27

Differentiating Equation (57) with respect to time We can express this result in terms of voltages if we
gives divide through by C and use VC = Q/C, we get
dQ
= – α Be– α t (58) charging a
dt VC = Vb 1 – e– t/RC capacitor (63)
Substituting Equations (57) and (58) into (56) gives
A B – αt Vb
– αBe– α t + + e = (59) (We warned you that the addition of just one more term
RC RC R to our differential equation would make it messier to
The only way we can satisfy Equation (59) is have the solve.)
two terms with an e-αt cancel each other. I.e., we must The answer, Equation (63) is the standard form for an
have exponential rise. It is in fact just our exponential decay
curve turned upside down, and Figure (34) represents
- αB + B = 0
RC the easy way to determine the time constant RC from
1 experimental data.
α = (60)
RC
which is a familiar result.
Vb
The remaining terms in Equation (59) give
A Vb
= A = CVb (61)
RC R
Putting the values for A and α (Equations 61 and 60)
back into our guess (Equation 57), we get
Q = CVb + Be– t/RC (62) RC
Figure 34
If you continue along the initial slope line of the
The final step is to note that at time t = 0, Q = 0, so that charge-up curve, you intersect the final voltage Vb
at a time t = RC, one time constant later. Turn this
0 = CVb + Be0 B = – CVb diagram upside down and it looks like Figure 31.
thus our final result is

Q = CVb 1 – e– t/RC (63a)

Vb

4
R = 10 Ω
–6
C = 10 farads

Figure 33
Plot of the capacitor voltage versus time for the charging up of a capacitor through a resistor.
If you turn the diagram upside down, you get the curve for the discharge of a capacitor.
27-28 Basic Electric Circuits

Exercise 9 THE NEON BULB OSCILLATOR

Figure (35) shows the voltage across a capacitor C We will end this chapter on basic electric circuits with
being charged through a resistance R. Given that a discussion of an electronic device called a neon bulb
C = 1.0 x 10– 8 farads, estimate the value of R. oscillator. This is conceptually the world’s simplest
electronic device that does something useful—it oscil-
lates and its frequency of oscillation can be adjusted.
The device is not practical, for it is hard to adjust, its
waveform is far from being a pure sinusoidal shape,
and it requires a relatively high voltage power supply.
But when you work with this apparatus, you will begin
to get a feeling for the kind of tricks we pull in order to
make useful apparatus.

The Neon Bulb

R=? The new circuit element we will add to our neon
Vb oscillator circuit is the common neon bulb which glows
–8
C = 1.0 10 farads orange and is often used as a night light. The bulb,
which we will designate by the symbol is
simply a small glass tube with neon gas inside and two
wires as shown in Figure (36). The bulb turns on when
there is a large enough voltage difference between the
Figure 35 wires that the neon gas becomes ionized and starts to
Given the experimental results for the charge-up of glow. For typical neon bulbs, the glow starts when the
a capacitor, determine the value of the resistance R. voltage reaches approximately 100 volts. When the
(Answer R = 68K) bulb is glowing the neon gas is a good conductor and
the bulb is like a closed switch. When the bulb is not
glowing, the gas is inert and the bulb is like an open
switch.

A given neon bulb has a rather consistent voltage Vf

(firing voltage) at which it turns on, and voltage Vq
(quenching voltage) at which it shuts off. In a typical
bulb Vf may be 100 volts and Vq equal to 40 volts.
These numbers will, however, vary from bulb to bulb.
When a neon bulb is included in a circuit, it acts like an
glass
bulb automatic switch, closing (turning on) when the volt-
age across it reaches Vf and opening (shutting off)
when the voltage drops to Vq.
neon gas

wires

Figure 36
A neon bulb. When the voltage across the wires
reaches a threshold value, typically around 100 volts,
the neon gas starts to glow, and the gas suddenly
changes from an insulator to a conductor.
 27-29

The Neon Oscillator Circuit recording

We can make a neon oscillator using the circuit shown R VR volt
in Figure (37). The left hand part of the circuit is just meter
Vb
the RC circuit we used in Figure (32) to charge up a neon
capacitor. The only really new feature is the neon bulb C VC bulb V
in parallel with the capacitor. The recording voltmeter,
indicated by the symbol V is there to record the
capacitor voltage VC . Figure 37
Neon bulb oscillator circuit.
The output of the neon oscillator circuit is shown in
Figure (38). Initially the capacitor is charging up just
as it did in Figure (33). During the charge up, the neon R VR
bulb is off; it is like an open switch and might as well Vb
not be there. The effective circuit is shown in Figure
(39). When the voltage on the capacitor (and on the C VC V
neon bulb) reaches the firing voltage Vf, the neon bulb
turns on and acts like a short circuit as shown in Figure
(40). It is not exactly a short circuit, the neon bulb and Figure 39
the wire leads have some small resistance. But the Effective circuit while the neon bulb is off.
resistance is so small that the capacitor rapidly dis-
charges through the bulb, and the capacitor voltage
drops almost instantly. R VR lighted
neon bulb
When the capacitor and neon bulb voltage Vc drops to Vb
the bulb quenching voltage Vq, the bulb shuts off, and C V
the capacitor starts charging up again. As seen in
Figure (38), this process keeps repeating and we get the
oscillating voltage shown.
Figure 40
Effective circuit while the neon bulb is glowing.
For the last cycle in Figure (38), we opened a switch to
disconnect the neon bulb, allowing the capacitor to
charge up all the way to the power supply voltage Vb .
This allowed us to display all three voltages Vb , Vf ,
and Vq on one experimental plot.
Figure 38
Experimental output of a neon oscillator
R = 10Meg C = .5µf circuit. The capacitor charges up until the
voltage reaches the neon bulb firing
Vb voltage Vf , at which point the neon bulb
Vf turns on and the voltage rapidly drops.
When the voltage has fallen to the quench
voltage Vq , the neon bulb shuts off and the
capacitor voltage starts to rise again. On the
last cycle, we opened a switch to disconnect
the neon bulb, allowing the capacitor to
Vq charge up all the way to the power supply
voltage Vb .
(A voltage divider was used to measure these
high voltages. In the figure, the voltage scale
has been corrected to represent the actual
voltage on the capacitor.)
27-30 Basic Electric Circuits

Period of Oscillation Dividing Equation (66) by Equation (67) gives

To calculate the period of oscillation, we start with the 1 –Vq /Vb Vb –Vq
diagram of Figure (41) showing a cycle of the oscilla- e (t2 – t1)/RC = = (68)
1 –Vf /Vb Vb –Vf
tion superimposed upon the complete charge up curve
which starts at VC = 0 and goes up to VC = Vb. This where we used the fact that
curve is given by the formula e–α
–β
= eβ – α
VC = Vb 1 – e–t/RC (63) e
Taking the logarithm of Equation (68), using
What we want to calculate is the time T = (t2 - t1) it takes
for the capacitor to charge up from a voltage Vq to Vf. ln(eα) = α, we have
t2 – t1 Vb –Vq
At time t1, VC = Vq and Equation (63) gives = ln
RC Vb –Vf
Vq = Vb 1 – e–t1/RC (64)
or using T = t2 - t1 we get for the period of oscillation
At time t = t2, VC = Vf and we have
Vb–Vq
Vf = Vb 1 – e– t2/RC (65) T = RC ln
period of
(69)
Vb–Vf neon oscillator
Equation (64) and (65) can be rearranged to give
Vq Equation (69) was a bit messy to derive, and it is not a
e–t1/RC = 1 – (66) fundamental result that you need to memorize. You are
Vb
unlikely to meet a neon oscillator except in an introduc-
Vf tory physics lab. But we have used the theory devel-
e–t2/RC = 1 – (67) oped in this chapter to make an explicit prediction that
Vb
can be tested in the laboratory.
Vb
Exercise 10
Vf
See how well Equation (69) applies to the experimental
data of Figure (38). (The marked values on resistors and
capacitors are usually accurate only to within ± 10%.)
Vq
T
t1 t2

Figure 41
Determining the period T of the oscillation. The
formulas for t1 and t2 are obtained from the
capacitor charge up equations
Vf = Vb 1 – e– t1/ RC
Vq = Vb 1 – e– t2/ RC
The messy part is extracting the period T = t2 – t1
from these equations.
 27-31

Equation (69) provides clear instructions on how to to –5 volts or less, and should not be attached to a
change the period or frequency of a neon oscillator. voltage source of the order of 100 volts. This problem
The easiest way to make major changes in the period is can be solved by using the voltage divider circuit
to change the time constant RC. Because of the ease discussed in Exercise (3), and shown in Figure (42).
with which we can select different values of R and C
(typical values of R ranging from 102 ohms to 108 For a standard laboratory experiment, we have found it
ohms, and typical values of C from 10-4 to 10-12 convenient to mount, in one box, the voltage divider,
farads), a large range of time constants RC are avail- neon bulb, and switch - the components shown inside
able. However high frequencies are limited by the the dotted rectangle of Figure (42). This reduces stu-
characteristics of the neon bulb. We found it difficult to dent exposure to high voltages and guarantees that the
get the circuit to oscillate faster than 30 cycles per voltmeter will be exposed to voltages 1000 times
second. smaller than those across the capacitor.

Adjusting the battery voltage Vb changes the shape of In Figure (43), we have recorded the entire voltage
the neon oscillator wave and also allows fine adjust- range of the experiment, starting from an uncharged
ments in the period . capacitor. We first opened the switch above the neon
bulb in Figure (42), and let the capacitor charge up to
the full voltage Vb . Then closing the switch allowed
Experimental Setup the capacitor voltage to oscillate between Vf and Vq as
An experimental problem you face while working with seen in Figure (38). While the actual capacitor voltage
the neon oscillator circuit, is that the voltages of interest ranges from 0 to 100 volts, the recording voltmeter
range up to 100 volts or more. Modern oscilloscopes or shows a range of 0 to 100 millivolts because of the
recording voltmeters tend to operate in the range of +5 1000 to 1 voltage divider.

box containing
switch, neon bulb
and voltage divider Figure 42
Neon oscillator circuit with
voltage divider. The switch
R VR 1000 to 1
voltage divider above the neon bulb allows
switch 8 us to disconnect the bulb
Vb 10 Ω
from the circuit. It is
neon recording convenient to mount, in a
C VC bulb 5 volt
10 Ω V single box, the components
meter within the dotted rectangle.

Figure 43
Full range of voltages from
the neon oscillator circuit.
The voltage scale is in
millivolts because of the
voltage divider.
27-32 Basic Electric Circuits

Exercise 11 Review Problem

Figure (44a) shows the circuit used to observe the
discharge of a capacitor. The capacitor is made from
the two circular aluminum plates shown in Figure (44b).
The plates have a diameter of 22 cm and are separated
a distance (d) by small pieces of glass. In Figure (44c),
we are observing the discharge of the capacitor through
a 10kΩ 104 Ω resistor. For this discharge, what is the
separation (d) of the plates?

mercury switch

Vb C R scope

Figure 44a Figure 44b

Circuit for observing the discharge of a capacitor. The capacitor plates.

Figure 44c
Voltage during discharge.
Chapter 28
Magnetism

CHAPTER 28 MAGNETISM

In our discussion of the four basic interactions, we saw millimeters per minute. One would not expect phenom-
that electric forces are very strong but in most circum- ena like the Lorentz contraction or time dilation to play
stances tend to cancel. The strength of the forces are any observable role whatever in such electrical phe-
so great, but the cancellation is so nearly complete that nomena.
the slightest imbalance in the cancellation leads to
important effects such as molecular forces. As illus- But, as we shall see, observable effects do result from
trated in Figure (18-6) reproduced here, a positively the tiny imbalance in electric forces caused by the
charged proton brought up to a neutral hydrogen atom Lorentz contraction. Since these effects are not de-
experiences a net attractive force because the negative scribable by Coulomb’s law, they are traditionally
charge in the atom is pulled closer to the proton. This given another name—magnetism. Magnetism is one of
net force is the simplest example of the type of molecu- the consequences of requiring that the electrical force
lar force called a covalent bond. law and electric phenomena be consistent with the
principle of relativity.
In this chapter we will study another way that the
precise balance between attractive and repulsive elec- Historically this point of view is backwards. Magnetic
tric forces can be upset. So far in our discussion of effects were known in the time of the ancient Greeks.
electrical phenomena such as the flow of currents in Hans Christian Oersted first demonstrated the connec-
wires, the charging of capacitors, etc., we have ignored tion between magnetic and electric forces in 1820 and
the effects of special relativity. And we had good James Clerk Maxwell wrote out a complete theory of
reason to. We saw that the conduction electrons in a electromagnetic phenomena in 1860. Einstein did not
wire move at utterly nonrelativistic speeds, like two discover special relativity until 1905. In fact, Einstein
used Maxwell’s theory as an important guide in his
discovery.
electron

proton If you follow an historical approach, it appears that

–
+
special relativity is a consequence of electricity theory,
+
and a large number of physics texts treat it that way.
center of proton Seldom is there a serious discussion of special relativ-
electron cloud
ity until after Maxwell’s theory of electricity has been
developed. This is considered necessary in order to
Figure 18-6 explain the experiments and arguments that lead to the
The net attraction between a positive discovery of the special theory.
charge and a neutral atom is caused by a
redistribution of charge in the atom.
28-2 Magnetism

But as we know today, electricity is one of but several Two Garden Peas
basic forces in nature, and all of them are consistent In preparation for our discussion of relativistic effects
with special relativity. Einstein’s famous theory of in electricity theory, let us review a homely example
gravity called general relativity can be viewed as a that demonstrates both how strong electric forces actu-
repair of Newton’s theory of gravity to make it consis- ally are, and how complete the cancellation must be for
tent with the principle of relativity. (This “repair” the world to act the way it does. Suppose we had two
produced only minor corrections when applied to our garden peas, each with a mass of about 2 grams,
solar system, but has sweeping philosophical implica- separated by a distance of 1 meter. Each pea would
tions.) If the principle of relativity underlies the contain about one mole (6 × 1023 ) of protons in the
structure of all forces in nature, if all known phenom- atomic nuclei, an equal number of electrons surround-
ena are consistent with the principle, then it is not ing the nuclei. Thus each pea has a total positive charge
especially necessary to introduce special relativity in +Q in the protons given by
the context of its historical origins in electromagnetic
theory. total positive
In this chapter we are taking a non-historical point of charge in a ≈ 6 × 10 23e
view. We already know about special relativity (from garden pea
chapter one), and have just studied Coulomb’s electri-
cal force law and some simple applications like the = 6 × 10 23 × 1.6 × 10 -19
electron gun and basic circuits. We would now like to (1)
5
see if Coulomb's law is consistent with the principle of = 10 coulombs
relativity. In some sense, we would like to do for
Coulomb’s law of electricity what Einstein did to and there is an equal and opposite amount of negative
Newton’s law of gravity. charge in the electrons.
 28-3

When two peas are separated by a distance of 1 meter To put this answer in a more recognizable form, note
as shown in Figure (1), we can think of there being four that the weight of one metric ton (1000 kg) of matter is
pairs of electric forces involved. The positive charge in
pea (1) repels the positive charge in pea (2) with a force
of magnitude Fg (1 metric ton) = mg

3 m
repulsive force = 10 kg × 9.8
between positive QQ sec2
= (2)
charge in 4πε0 r2 3
= 9.8 × 10 newtons
the two peas
Expressing the force between our two positively charged
which gives rise to one pair of repulsive forces. The peas in metric tons we get
negative charges in each pea also repel each other with
a force of the same magnitude, giving rise to the second repulsive force
between two 8.8 × 10 19 newtons
repulsive pair of electric forces. But the positive charge =
positive peas 9.8 × 10 3 newtons/ton
in Pea (1) attracts the negative charge in Pea (2), and the
1 meter apart
negative charge in Pea (1) attracts the positive charge
in Pea (2). This gives us two pairs of attractive forces
that precisely cancel the repulsive forces. ≈ 10 15 tons !

Let us put numbers into Equation (2) to see how big (4)
these cancelling electric forces are. Equation (2) can be If we stripped the electrons from two garden peas, and
viewed as giving the net force if we removed all the placed them one meter apart, they would repel each
electrons from each garden pea, leaving just the pure other with an electric force of 1015 tons!! Yet for two
positive charge of the protons. The result would be real garden peas, the attractive and repulsive electric
force cancel so precisely that the peas can lie next to
Q2 each other on your dinner plate.
F =
4πε0r2
Exercise 1
2 Calculate the strength of the gravitational force between
(10 5 coulombs)
= the peas. How much stronger is the uncancelled
4π × 9 × 10 -12 × (1)2 electric force of Equation (3)?

= 8.8 × 10 19 newtons (3) With forces of the order of 1015 tons precisely cancel-
ing in two garden peas, we can see that even the tiniest
pea pea imbalance in these forces could lead to striking results.
An imbalance of one part in 1015, one part in a million
billion, would leave a one ton residual electric force.
1 2
This is still huge. We have to take seriously imbalances
Figure 1
Electric forces between two garden peas. On that are thousands of times smaller. One possible
pea #1, there is the attractive force between the source of an imbalance is the Lorentz contraction, as
protons in pea #1 and the electrons in pea #2, seen in the following thought experiment.
and between the electrons in pea #1 and the
protons in pea #2. The two repulsive forces are
between the electrons in the two peas and the
protons in the two peas. The net force is zero.
28-4 Magnetism

A THOUGHT EXPERIMENT In another calculation, we saw that if a millimeter cross

In our previous discussion of electric currents, we had section copper wire carried a steady current of one
difficulty drawing diagrams showing the electrons ampere, the conduction electrons would have to move
flowing through the positive charge. To clarify the role at the slow speed of 1/27 of a millimeter per second, a
of the positive and negative charge, we suggested a motion so slow that it would be hard to detect. As a
model of a copper wire in which we think of the positive result there should be no important physical difference
and negative charge as being attached to separate rods between the two points of view, and a left directed
as shown in Figure (27-5a) repeated here. In that model negative current should be physically equivalent to a
the rods have equal and opposite charge to represent the right directed positive current.
fact that the copper wire is electrically neutral, and the A closer examination of Figure (27-5) shows that we
negative rod is moving to represent the electric current have left something out. The bottom figure, (27-5b) is
being carried by a flow of the negative conduction not precisely what the moving observer sees. To show
electrons. what has been left out, we have in Figure (2a) redrawn
The point of the model in Figure (27-5) was to show Figure (27-5a) and carefully labeled the individual
that a left directed negative current, seen in (a) is charges. To maintain strict overall charge neutrality we
essentially equivalent to a right directed positive cur- have used charges +Q on the positive rod, charges -Q
rent seen in (b). In Figure (27-5a), we drew a stick on the negative rod, and both sets of charges have equal
figure diagram of a person walking to the left at the separations of centimeters.
same speed v as the negative rod. Figure (27-5b) is the From the point of view of the moving observer in
same setup from the point of view of the stick figure Figure (2b), the negative rod is at rest and the positive
person. She sees the negative rod at rest and the rod is moving to the right as we saw back in Figure (27-
positive rod moving to the right as shown. 5b). But, due to the Lorentz contraction, the spacing
between the charges is no longer ! Since the positive
rod was at rest and is now moving, the length of the
v positive spacing must be contracted to a distance
1 - v2 /c2 as shown.
v
+ + + + + +
– – – – – – On the other hand the negative rod was moving in
a) observer walking along with the moving Figure (2a), therefore the negative spacing must ex-
negatively charged rod pand to / 1 - v2 /c2 when the negative rod comes to
rest. (Start with a spacing / 1 - v2 /c2 for the negative
charges at rest in Figure (2b), and go up to Figure (2a)
where the negative rod is moving at a speed v. There the
v spacing must contract by a factor 1 - v2 /c2 , and the
new spacing is / 1 - v2 /c2 × 1 - v2 /c2 = as
+ + + + + +
– – – – – –
shown.)
b) from the observer's point of view the negative
rod is at rest and the positive charge is moving
to the right
As a result of the Lorentz contraction, the moving
observer will see that the positive charges on her
Figure 27-5 a,b moving rod are closer together than the negative charges
In (a) we have a left directed negative current, while in on her stationary rod. (We have exaggerated this effect
(b) we have a right directed positive current. The only in our sketch, Figure (2b)). Thus the moving observer
difference is the perspective of the observer. (You can of Figure (2b) sees not only a right directed positive
turn a negative current into an oppositely flowing
positive one simply by moving your head.) current, but also a net positive charge density on her
two rods. The Lorentz contraction has changed a
neutral wire in Figure (2a) into a positively charged one
in Figure (2b)!
 28-5

v + + + + + + +
– – – – – – –

a) Observer walking along with the moving negatively

charged rod.

1– v 2 /c 2

+ + + + + + + +
v
– – – – – – –

/ 1– v 2/c 2
b) Charged rods from the observer's point of view. Now
that the positive charge is moving, the spacing
between positive charges has contracted from
to 1– v 2 /c 2 . The negative rod is now at rest, the
Lorentz contraction is undone, and the negative
spacing has expanded from to / 1– v 2/c 2.

Figure 2
An electric current from two points of view.
28-6 Magnetism

Charge Density on the Two Rods The net charge density λ is obtained by adding λ + and
Our next step will be to calculate the net charge density λ – of Equations (5) and (7) to get
λ on the pair of rods shown in Figure (2b). Somewhat Q
messy algebra is required for this calculation, but the λ = λ + + λ– = 1– (1– v 2 /c2)
2 2
result will be used in much of the remainder of the text. 1– v /c
The effort will be worth it.
Q v2 v2
λ = = λ+ (8)
If we have a rod with charges spaced a distance d apart 1– v 2 /c2 c2 c2
as shown in Figure (3), then a unit length of the rod, 1
meter, contains 1/d charges. (For example, if d = .01 Equation (8) can be simplified by noting that the
meter, then there will be 1/d = 100 charges per meter.) current i carried by the positive rod in Figure (2b) is
If each charge is of strength Q, then there is a total equal to the charge λ + on 1 meter of the rod times the
charge Q/d on each meter of the rod. Thus the charge speed v of the rod
density is λ = Q/d coulombs per meter. Applying this
current i
result to the positive rod of Figure (2b) gives us a
positive charge density i = λ+ v carried by the (9)
positive rod
Q Q
λ+ = = (5)
(In one second, v meters of rod move past any fixed
d+ 2 2
1 - v /c cross-sectional area, and the charge on this v meters of
rod is λ + v.) Using Equation (9), we can replace λ + and
And on the negative rod the charge density is
one of the v’s in Equation (8) by i to get the result
–Q –Q
λ– = =
d– 1 – v 2/c2 iv
/ λ = 2
(10)
c
(6)
–Q 1 – v 2/c2
= Due to the Lorentz contraction , the moving observer in
Figure (2b) sees a net positive charge density λ = iv c2
Multiplying the top and bottom of the right side of on the wire which from our point of view, Figure (2a)
Equation (6) by 1 - v2 /c2 , we can write λ – as was precisely neutral.
–Q 1 – v 2/c2 1 – v 2/c2 Although Equation (10) may be formally correct, one
λ– = ×
2 2
1 – v /c has the feeling that it is insane to worry about the
Lorentz contraction for speeds as slow as 2 millimeters
–Q (7) per minute. But the Lorentz contraction changes a
= 1 – v 2/c2
1 – v 2/c2 precisely neutral pair of rods shown in Figure (2a), into
a pair with a net positive charge density λ = iv/c2 in
Figure (2b). We have unbalanced a perfect cancella-
d λ coulombs/meter = Q/d tion of charge which could lead to an imbalance in the
+ + + + + + + cancellation of electrostatic forces. Since we saw from
Q Q Q Q Q Q Q our discussion of the two garden peas that imbalances
Figure 3
as small as one part in 1018 or less might be observable,
If the charges are a distance d apart, then there are let us see if there are any real experiments where the
1/d charges per meter of rod. (If d = .1 meters, then charge density λ is detectable.
there are 10 charges/meter.) If the magnitude of
each charge is Q, then λ , the charge per meter is Q
times as great, i.e., λ = Q * (1/d).
 28-7

A Proposed Experiment wire as shown in Figure (4). This test particle should
How would we detect the charge imbalance in Figure experience a force
(2b)? If there is a net positive charge density l on the two
rods in Figure (2b), repeated here again in Figure (4), F = qE (12)
then the net charge should produce a radial electric field
which would be repulsive if the test particle q is positive
whose strength is given by the formula
and attractive if q is negative. Using Equations (10) for
E = λ (11) λ and (11) for E, Equation (12) gives for the predicted
2πε0r magnitude of F:
We derived this result in our very first discussions of
Coulomb’s law in Chapter 24. (Remember that the two qλ q iv
separate rods are our model for a single copper wire F = qE = = (13)
2πε0r 2πε0rc2
carrying a current. The rods are not physically sepa-
rated as we have had to draw them, the negative Rearranging the terms on the right side of Equation
conduction electrons and positive nuclei are flowing (13), we can write F in the form
through each other.)
i
We can test for the existence of the electric field F = qv× (14)
produced by the positive charge density λ = iv/c2 by 2πrε0c2
placing a test particle of charge q a distance r from the
Why we have written Equation (14) this way will
become clear shortly.

1– v 2 /c 2

+ + + + + + + + v
– – – – – –

r / 1– v 2/c 2

+q

F = qE
Figure 4
To test for the net charge density, as seen by the observer at rest
relative to the minus charge, the observer places a test charge q a
distance r from the wire as shown. If there is a net charge λ on
the wire, the charge will produce an electric field E, which will
exert a force F= qE on the test particle as shown.
28-8 Magnetism

Origin of Magnetic Forces speed v. The positive rod is at rest, the Lorentz
You might think that the next step is to put reasonable contractions are undone, and there is no net charge on
numbers into Equation (14) and see if we get a force F our rods. All we have is a negative current flowing to
that is strong enough to be observed. But there is an the left.
important thought experiment we will carry out first.
The idea is to look at the force on a test particle from two We can also see the test particle. It is now moving to
different points of view, one where the wire appears the left at a speed v, and it is still attached to the spring.
charged as in Figures (4 & 2b), and where the wire Here is the crucial point of this discussion. We also
appears neutral as in Figure (2a). The two points of see that the spring is stretched. We also see that the
view are shown in Figure (5). end of the spring has been pulled beyond the mark
Figure (5b), on the left, is the situation as observed by indicating the unstretched length. We also detect the
the moving observer. She has a copper wire carrying force F on the test particle!
a positive current directed to the right. Due to the Why do we see a force F on the test particle? Our
Lorentz contraction, her copper wire has a charge copper wire is electrically neutral; we do not have an
density λ which creates an electric field E. To observe electric field E to produce the force F . Yet F is there.
E, she mounts a test particle -q at one end of a spring If we cut the spring, the test particle would accelerate
whose other end is fixed, nailed to her floor. She detects toward the copper wire, and both we and the moving
the force F = –qE by observing how much the spring observer would see this acceleration.
has been stretched.
At this point, we have come upon a basic problem.
Our point of view is shown in Figure (5a). It is exactly Even if the Lorentz contraction is very small and the
the same setup, we have touched nothing! It is just force F in Figure (5b) is very small, we at least predict
viewed by someone moving to the right relative to her. that F exists. In Figure (5a) we predict that a neutral
In our point of view, the moving observer, the negative wire, that is carrying a current but has absolutely no net
rod, and the test particle are all moving to the left at a charge on it, exerts an attractive force on a moving
negative charge as shown.

1– v 2 /c 2 v

v
+ + + + + + + + + + + + + +
v
– – – – – – – – – – – – –

r / 1– v 2/c 2
F=–qE F
v
–q –q
spring
unstretched v unstretched
length of nail length of
spring spring

(b) her view (a) our view

Figure 5
Two views of the same experiment. For the observer moving with the electrons, she sees a positively charged wire
exerting an attractive force on the negative charge at rest. We see an electrically neutral wire carrying a negative
current, and a moving negative charge. The spring is still stretched, meaning the attractive force is still there.
 28-9

With a few modifications, the experiment shown in battery to produce a healthy current) we will see the
Figure (5a) is easy to perform and gives clear results. electron beam deflected toward the wire if there is an
Instead of a negative test particle attached to a spring, observable force. The experimental result is shown in
we will use a beam of electrons in an electron gun as Figure (6c). There is a large, easily observed deflec-
shown in Figure (6). In Figure (6a) we see the setup of tion. The force F is easily seen.
our thought experiment. In Figure (6b) we have
replaced the two charged rods with a neutral copper Exercise 2
wire carrying a current -i, and replaced the test particle In Figure (7) we reversed the direction of the current in
with an electron beam. the wire and observe that the electron beam is deflected
away from the wire. Devise a thought experiment,
According to Equation (14), the force F on the test
analogous to the one shown in Figure (5a,b) that ex-
particle -q should have a strength proportional to the plains why the electron beam is repelled from the wire
current i in the wire. Thus when we turn on a current by this setup. (This is not a trivial problem; you may have
(shorting the wire on the terminals of a car storage to try several charge distributions on moving rods
before you can imitate the situation shown in Figure (7a).
But the effort is worth it because you will be making a
+ + + + + + +
v physical prediction that is checked by the experimental
– – – – – – –
results of Figure (7b).
F

(a) thought v –q
experiment Figure 6d Movie
Movie showing
magnetic deflection.

–i –i
copper wire copper wire

(b) proposed electron proposed electron

experiment beam experiment beam

electron gun electron gun

Figure 6c
Figure 7
For an experimental test of the results of the thought
experiment, we replace the moving negative charge If we reverse the direction of the current in the wire,
with a beam of electrons in an electron gun. The the electrons in the beam are repelled
electrons are attracted to the wire as predicted.
28-10 Magnetism

Magnetic Forces MAGNETIC FORCE LAW

Historically an electric force was defined as the force From our Coulomb’s law calculation of the electric
between charged particles and was expressed by force in Figure (5a), we were able to obtain the formula
Coulomb’s law. The force in Figure (5a) between a for the magnetic force in Figure (5b). The result,
moving test charge and an uncharged wire does not Equation (14) repeated here, is
meet this criterion. You might say that for historical
reasons, it is not eligible to be called an electric force. i
F = qv× 2
(14)
2πrε0c
The forces we saw in Figures (6c) and (7b), between a
moving charge and a neutral electric current, were where q is the charge on the test particle, i the current
known before special relativity and were called mag- in the wire, and r the distance from the wire to the charge
netic forces. Our derivation of the magnetic force in as shown in Figure (8). The only thing our derivation
Figure (5a) from the electric force seen in Figure (5b) does not make clear is whether v in Equation (14) is the
demonstrates that electric and magnetic forces in this speed of the test charge or the speed of the electrons in
example are the same thing just seen from a different the wire. We can’t tell because we used the same speed
point of view. v for both in our thought experiment. A more complex
thought experiment will show that the v in Equation
When we go from Figure (5b) to (5a), which we can do
(14) is the speed of the test particle.
by moving our head at a speed of 2 millimeters per
minute, we see essentially no change in the physical
setup but we have an enormous change in perspective. The Magnetic Field B
We go from a right directed positive current to a left In Equation (14) we have broken the somewhat com-
directed negative current, and the force on the test plex formula for the magnetic force into two parts. The
particle changes from an electric to a magnetic force. first part qv is related to the test charge (q is its charge
and v its speed), and the second part in the curly
brackets, which we will designate by the letter B
B ≡ i
(15)
2π rε0c2
is related to the wire. The wire is carrying a current i and
located a distance r away.

The quantity B in Equation (15) is called the magnitude

of the magnetic field of the wire, and in terms of B the
–i magnetic force becomes

Fmagnetic = qvB (16)

F r

v –q Equation (16) is almost a complete statement of the

magnetic force law. What we have left to do for the law
Figure 8 is to assign a direction to B, i.e. turn it into the vector B,
Force on a charge -q moving at a speed v parallel to and then turn Equation (16) into a vector equation for
a negative current -i a distance r away. the force Fmagnetic.
 28-11

There is one more definition. In the MKS system of Direction of the Magnetic Field
units, it is traditional to define the constant µ0 by the We will temporarily leave our special relativity thought
equation experiment and approach magnetism in a more tradi-
tional way. Figure (9) is a sketch of the magnetic field
1
µ0 = definition of µ0 (17) of the earth. By convention the direction of the
ε 0c 2 magnetic field lines are defined by the direction that a
compass needle points. At the equator the magnetic
Using this definition of µ0 in Equation (15) for B, we field lines point north (as does a compass needle) and
get the field lines are parallel to the surface of the earth. As
we go north from the equator the magnetic field lines
µ 0i begin to point down into the earth as well as north. At
B = magnetic field of a wire (18) the north magnetic pole the magnetic field lines go
2π r
straight down.
as the formula for the magnetic field of a wire.
Figure (9) is drawn with the magnetic north pole at the
It turns out to be quite an accomplishment to get top. The earth’s rotational axis, passing through the
Equations (16), (17), and (18) out of one thought true north pole, is at an angle of 11.5 degrees as shown.
experiment. These equations will provide the founda- Over time the location of the earth’s magnetic pole
tion for most of the rest of our discussion of electric and wanders, and occasionally flips down to the southern
magnetic (electromagnetic) theory. hemisphere. Currently the north magnetic pole is
located in north central Canada.

North magnetic pole N Earth's rotational axis

Magnetic field lines

pointing north

S
Figure 9
Magnetic field of the earth. The magnetic field lines show that the direction a freely floating compass
needle would point at any location outside the earth. For example, at the equator the compass needle
would be parallel to the surface of the earth and point north. At the north magnetic pole, the compass
needle would point straight down (and thus not be very useful for navigation).
28-12 Magnetism

As we mentioned, it is by long standing convention that deflect a compass needle. This was the first evidence
the direction of the magnetic field is defined by the of the connection between the subject of electricity
direction a compass needle points. We can therefore with its charges and currents, and magnetism with its
use a set of small compasses to map the direction of the magnets and compasses.
magnetic field.
The fact that a wire carrying a current deflects a
In 1820, while preparing a physics lecture demonstra- compass needle means that the current must be produc-
tion for a class of students, Hans Christian Oersted ing a magnetic field. We can use the deflected compass
discovered that an electric current in a wire could needles to show us the shape of the magnetic field of
a wire. This is done in Figures (10a,b) where we see a
ring of compasses surrounding a vertical wire. In (10a)
there is no current in the wire, and all the compass
needles all point north (black tips). In (10b) we have
turned on an upward directed current in the wire, and
the compass needles point in a circle around the wire.
Using the north pole of the compass needle to define
the direction of the magnetic field, we see that the
magnetic field goes in a counterclockwise circle around
the wire.

In Figure (11) we have replaced the compasses in

Figure (10) with a sprinkle of iron filings. When the
current in the wire is turned on, the iron filings align
themselves to produce the circular field pattern shown.
What is happening is that each iron filing is acting as a
small compass needle and is lining up parallel to the
Figure 10a
With no current flowing in the wire,
magnetic field. While we cannot tell which way is
all the compass needles point north. north with iron filings, we get a much more complete

Figure 10b Figure 11

When an upward directed current is turned on, Iron fillings sprinkled around a current form a circular
the compass needles point in a pattern. Each iron filing lines up like a compass needle,
counterclockwise circle about the wire. giving us a map of the magnetic field.
 28-13

picture or map of the direction of the magnetic field. The Right Hand Rule for Currents
Figure (11) is convincing evidence that the magnetic Iron filings give us an excellent picture of the shape of
field surrounding a wire carrying a current is in a the magnetic field, but do not tell us which way the field
circular field, not unlike the circular flow pattern of is pointing. For that we have to go back to compasses
water around the core of a vortex. as in Figure (9), where B is defined as pointing in the
direction of the north tip of the compass needle. In that
The use of iron filings turns out to be a wonderfully figure we see that when a positive current i is flowing
simple way to map magnetic field patterns. In Figure toward us, the magnetic field goes in a counter clock-
(12), a sheet of cardboard was placed on a bar magnet wise direction as illustrated in Figure (14).
and iron filings sprinkled on the cardboard. The result,
with two poles or points of focus resembles what is The above description for the direction may be hard to
called a dipole field. remember. A more concise description is the follow-
ing. Point the thumb of your right hand in the
In Figure (13) we have thrown iron filings at an old iron direction of the current as shown in Figure (14), then
magnet and created what one young observer called a your fingers will curl in the direction of the magnetic
“magnet plant.” Here we see the three dimensional field. This mnemonic device for remembering the
structure of the magnetic field, not only between the direction of B is one of the right hand rules. (This is
pole pieces but over the top half of the magnet. the version we used in Figure 2-37 to distinguish right
and left hand threads.) If we had used compasses that
pointed south, we would have gotten a left hand rule.

B
Figure 12
A sheet of cardboard is placed over the poles of a
magnet and sprinkled with iron filings. From the
pattern of the filings we see the shape of the more
complex magnetic field of the magnet.

Figure 14
Right hand rule for the magnetic field of a
current i. Point the thumb in the direction of
the positive current and your fingers curl in
the direction of the magnetic field.

Figure 13
You get a three dimensional picture of the magnetic
field if you pour the iron filings directly on the magnet.
Our young daughter called this a Magnet Plant.
28-14 Magnetism

Parallel Currents Attract The Magnetic Force Law

While we are in the business of discussing mnemonic Now that we have a direction assigned to the magnetic
rules, there is another that makes it easy to remember field B we are in a position to include directions in our
whether a charge moving parallel to a current is at- formula for magnetic forces. In Figure (16) which is
tracted or repelled. In Figure (6) we had a beam of the same as (15) but also shows the magnetic field, we
negative electrons moving parallel to a negative cur- have a positive charge moving parallel to a positive
rent -i, and the electrons were attracted to the current. current, and therefore an attractive force whose magni-
In Figure (7) the current was reversed and the electrons tude is given by Equation (16) as
were repelled. One can work out a thought experiment
Fmag = qvB (16)
similar to the ones we have done in this chapter to show
that a positive charge moving parallel to a positive There are three different vectors in Equation (16),
current as shown in Figure (15) is attracted. Fmag, v, and B. Our problem is to see if we can combine
these vectors in any way so that something like Equa-
The simple, yet general rule is that parallel currents
tion (16) tells us both the magnitude and the direction
attract, opposite currents repel. A positive charge
of the magnetic force Fmag. That is, can we turn
moving in the direction of a positive current, or a
Equation (16) into a vector equation?
negative charge moving along with a negative current
are attracting parallel currents. When we have negative +i
charges moving opposite to a negative current as in
Figure (7) we have an example of opposite currents that
repel.
+i
B v
+q
Fmag

v
+q
Fmagnetic (a) side view

Figure 15 +q
A positive charge, moving parallel to a positive Fmag
i up
current, is attracted by the current. Thinking of
the moving positive charge as a positive upward
directed current, we have the rule that parallel
currents attract, opposite currents repel.

(b) top view

Figure 16
The directions of the vectors Fmag, v and B for a
positive charge moving parallel to a positive current.
 28-15

The right hand side of Equation (16) involves the Although the formula for Fmag, Equation (19), was
product of the vectors v and B. So far in the text we derived for a special case, the result is general. When-
have discussed two different ways of multiplying ever a particle of charge q is moving with a velocity v
vectors; the dot product A⋅B which gives a scalar through a magnetic field B, no matter what the relative
number C, and the cross product A× B which gives the directions of v and B, the magnetic force is correctly
vector C. Since we want the product of v and B to give given as qv × B.
us the vector Fmag, the cross product appears to be the
better candidate, and we can try Exercise 3
magnetic Using the magnetic force law Fmag = qv × B and the
Fmag = qv × B force law
(19) right hand rule for the magnetic field of a current, show
that:
as our vector equation.
(a) An electron moving parallel to a negative current -i
To see if Equation (19) works, look at the three vectors is attracted (Figure 6)
v, B, and Fmag of Figure (16) redrawn in Figure (17).
The force Fmag is perpendicular to the plane defined by (b) An electron moving opposite to a negative current is
v and B which is the essential feature of a vector cross repelled (Figure 7)
product. To see if Fmag is in the correct direction, we
use the cross product right hand rule. Point the fingers Lorentz Force Law
of your right hand in the direction of the first vector in
Since electric and magnetic forces are closely related,
the cross product, in this case v, and curl them in the
it makes sense to write one formula for both the electric
direction of the second vector, now B. Then your
and the magnetic force on a charged particle. If we have
thumb will point in the direction of the cross product
a charge q moving with a velocity v through an electric
v × B. Looking at Figure (17), we see that the thumb of
field E and a magnetic field B, then the electric force is
the right hand sketch does point in the direction of Fmag,
qE, the magnetic force qv × B, and the total “electro-
therefore the direction of Fmag is correctly given by the
magnetic” force is given by
cross product v × B. (If the direction had come out
wrong, we could have used B × v instead.) Lorentz
F = qE + qv × B force law
(20)

B points back v Fmag = q v X B

into paper Equation (20), which is known as the Lorentz force
B law, is a complete description of the electric and
magnetic forces on a charged particle, provided E and
B are known.
Fmag +q

Figure 17
Right hand rule for the vector cross product v × B.
Point the fingers of your right hand in the direction
of the first vector v, and then curl them in the
direction of the second vector B. Your thumb ends
up pointing in the direction of the vector v × B.
28-16 Magnetism

Dimensions of the worked only with tesla, you would have a hard time
Magnetic Field, Tesla and Gauss communicating with much of the scientific commu-
The dimensions of the magnetic field can be obtained nity. What we will do in this text is use either gauss or
from the magnetic force law. In the MKS system we tesla depending upon which is the more convenient
have unit. When we come to a calculation, we will convert
any gauss to tesla, just as we convert any distances
meters measured in centimeters to meters.
F newtons = q coulombs v ×B
second
which gives us B in units of newton seconds per Uniform Magnetic Fields
coulomb meter. This set of dimensions is given the Using the magnetic force law Fmag = qv × B to cal-
name tesla culate magnetic forces is often the easy part of the
MKSunits
problem. The hard part can be to determine the
newton second for magnetic field B. For a current in a straight wire, we
≡ tesla (21) were able to use a thought experiment and the Lorentz
coulomb meter magnetic
fields contraction to get Equation (20) for the strength of B.
But in more complicated situations, where we may
Although most MKS electrical quantities like the volt have bent wires, thought experiments become too
and ampere are convenient, the tesla is too large. Only difficult and we need other techniques for calculating
the strongest electromagnets, or the new superconduct- B.
ing magnets used in particle accelerators or magnetic
resonance imaging apparatus, can produce fields of the One of the other techniques, which we will discuss in
order of 1 tesla or more. Fields produced by coils of the next chapter, is called Ampere’s law. This law will
wire we use in the lab are typically 100 times weaker, give us the ability to calculate the magnetic field of
and the earth’s magnetic field is 100 times weaker still. simple current distribution much the same way that
Gauss’ law allowed us to calculate the electric field of
In the CGS system of units, magnetic fields are mea- simple charge distributions. But until we get to Ampere’s
sured in gauss, where law in the next chapter, we will confine our study of the
magnetic force law to the simplest of all possible
1 gauss = 10 –4 tesla magnetic fields, the uniform magnetic field.

The gauss is so much more convenient a unit that there

is a major incentive to work with CGS units when
studying magnetic phenomena. For example the earth’s
magnetic field has a strength of about 1 gauss at the
earth’s surface, and the magnetic field that deflected
the electrons in Figures (6) and (7) has a strength of
about 30 gauss at the electron beam. Refrigerator
magnets have comparable strengths.

We could be pedantic, insist on using only MKS units,

and suffer with numbers like .00021 tesla in discus-
sions of the earth’s magnetic field. But if someone
wants you to measure a magnetic field, they hand you Figure 18
a “gauss meter” not a tesla meter. Magnetic-type Between the poles of this magnet there
instruments are usually calibrated in gauss. If you is a relatively uniform magnetic field.
 28-17

Working with uniform magnetic fields, fields that are

constant in both magnitude and direction, is so conve-
nient that physicists and engineers go to great lengths
to construct them. One place to find a uniform field is
between the flat pole pieces of a magnet, as seen in
Figure (18) which is our “magnet plant” of Figure (13) d
with fewer iron filings.

If we bend a wire in a loop, then a current around the

loop produces the fairly complex field pattern shown in
Figure (19). When we use two loops as seen in Figure
(20), the field becomes more complicated in some
places but begins to be more uniform in the central i i
region between the coils. With many loops, with the
coil of wire shown in Figure (21a), we get a nearly
uniform field inside. Such a coil is called a solenoid,
and will be studied extensively in the next chapter. An
iron filing map of the field of a large diameter, tightly
wound solenoid is seen in Figure (21b).

i
Figure 20
The magnetic field in the region between a pair of
coils is relatively uniform. We can achieve the
greatest uniformity by making the separation d
between the coils equal to the radius of the coils.
Such a setup is called a pair of Helmholtz coils.
Figure 19
The magnetic field of a current loop is fairly complex.

Section of coil
Figure 21a
Magnetic field in the upper half of a section of a
Figure 21b
coil of wire. When you have many closely spaced
Iron filing map of the magnetic field of a
coils, the field inside can become quite uniform
large diameter coil. (Student project,
through most of the length of the coil.
Alexandra Lesk and Kirsten Teany.)
28-18 Magnetism

Helmholtz Coils For the Helmholtz coils in Figure (23), it was observed
that when a current of one amp flowed through the
For now we will confine our attention to the reasonably
coils, the strength of the magnetic field in the central
uniform field in the central region between two coils
regions was 8 gauss. A current of 2 amps produced a
seen in Figure (20). Helmholtz discovered that when
16 gauss field. Thus the field strength, for these coils,
the coils are spaced a distance d apart equal to the coil
is related to the current i by
radius r (Figure 22), we get a maximally uniform field
B between the coils. This arrangement, which is called for the Helmholtz
a pair of Helmholtz coils, is commonly used in physics B gauss = 8i amps coils of
and engineering apparatus. Figure (23) shows a pair of Figure (23)only
Helmholtz coils we use in our undergraduate physics
labs and which will be used for several of the experi- In the lab we measure the strength of B simply by
ments discussed later. An iron filing map of the field reading (i) from an ampmeter and multiplying by 8. Of
produced by these coils is seen in Figure (24a), and one course, if you are using a different set of coils,(i) will be
of the experiments will give us a field plot similar to multiplied by a different number. (Do not worry about
Figure (24b). the mixed units, remember that we convert gauss to
tesla before doing MKS calculations.)
In our derivation of the magnetic field of a current in a
straight wire, we saw that the strength of the magnetic One can derive a formula for an idealized set of
field was proportional to the current i in the wire. Helmholtz coils. The derivation is complicated and the
This is true even if the wire is bent to form coils, or even answer B = 8µ0 N i/ 5 5 r is rather a mess. (N is the
twisted into a complex tangle. That means that once number of turns in each coil.) The simple feature which
you have mapped the magnetic field for a given current we expected, is that B is proportional to the strength of
(i) in a set of wires, doubling the current produces the the current (i) in the coils. Because another law, called
same shape map with twice as strong a field. Faraday’s law, can be used to give us a more accurate
calibration of real Helmholtz coils, we will leave the
d derivation of the Helmholtz formula above to other
texts.

d=r
Figure 22
For Helmholtz coils, the separation d equals the coil
radius r.

Figure 23
Helmholtz coils used in a number of lab
experiments discussed in the text. Each coil
consists of 60 turns of fairly heavy magnet wire.
 28-19

MOTION OF CHARGED PARTICLES

IN MAGNETIC FIELDS
In physics, one of the primary uses of magnetic fields
is to control the motion of charged particles. When the
magnetic field is uniform, the motion is particularly
simple and has many practical applications from par-
ticle accelerators to mass spectrometers. Here we will
discuss this motion and several of the applications.

The main feature of the magnetic force law,

Fmagnetic = qv × B (19)
is that because of the cross product v × B, the magnetic
force is always perpendicular to the velocity v of the
charged particle. This has one important immediate
consequence. Magnetic forces do no work! The
Figure 24a
formula for the power, i.e., the work done per second,
Iron filing map of the magnetic field of the
Helmholtz coils. (Student project, Alexandra is
Lesk and Kirsten Teany.)
Work done
per second = power = F⋅v (22)
by a force F
Since the magnetic force Fmag is always perpendicular
to v, we have

Work done by a
= Fmagnetic ⋅ v
magnetic force
(23)
= qv ×B ⋅v ≡ 0
magnetic fields do not change the energy of a particle,
they simply change the direction of motion.

Figure 24b
Plot from a student experiment, of the magnetic
field in the region between and around the coils.
28-20 Magnetism

Helmholtz Motion in a Uniform Magnetic Field

coils
When we have a charged particle moving through a
uniform magnetic field, we get a particularly simple
kind of motion—the circular motion seen in Figure
electron gun (25b). In Figure (25a) we sketched the experimental
setup where an electron gun is placed between a pair of
v
Helmholtz coils so that the magnetic field B is perpen-
dicular to the electron beam as shown. Figure (25b) is
B F = (– e) v X B a photograph of the electron beam deflected into a
circular path. In Figure (25c) we have a sketch of the
forces on an electron in the beam. The magnetic field
B in this diagram is up out of the paper, thus v × B points
radially out from the circle. But the electron has a
side view negative charge, thus the magnetic force FB
this way
Figure 25a FB = –e v × B
Top view looking down on the electron gun placed
between the Helmholtz coils. The electrons in the points in toward the center of the circle as shown.
beam move perpendicular to the magnetic field B.
The magnetic force F = (– e) v × B is directed up,
out of the paper in this drawing. v

–q

B directed
out of paper
F

F = (–q) v X B

Figure 25c
Figure 25b As the electrons move along a curved path, the
Side view of the electron beam, as seen through the magnetic force F = – q v × B always remains
lower coil in Figure 25a. In this view the magnetic perpendicular to the velocity and therefore cannot
field is directed out of the paper toward the reader. change the speed of the electrons. The resulting
motion is uniform circular motion where the
force and the acceleration are directed toward the
center of the circle.

Figure 25d
Movie of the
experiment.
 28-21

To apply Newton’s second law to the electrons in Next convert B from gauss to the MKS tesla
Figure (25), we note that a particle moving in a circle
accelerates toward the center of the circle, the same 70 gauss = 70 × 10 – 4 tesla (27)
direction as Fmag in Figure (25c). Thus FB and ma are Substituting Equations (26) and (27) in (25) gives
in the same direction and we can use the fact that for
circular motion a = v2 /r to get FB = m a or mv .911 × 10 -30 × 6.9 × 10 6
r = =
qB 1.6 × 10 -19 × 7 × 10 -3
qvB = mv2 /r (24)
Solving for r, we predict from Equation (24) that the r = 5.6 × 10 -3 m = .56 cm (28)
electron beam will be bent into a circle of radius r given
by Exercise 4
The scale of distance shown in Figure (25b) was drawn
mv knowing the dimensions of the cap in the electron gun.
r= (25) Use this scale to estimate the radius of curvature of the
qB
electron beam and compare the result with the predic-
tion of Equation (28).
Equation (25) is an important result that we will use
often. But it is so easy to derive, and it is such good Exercise 5
practice to derive it, that it may be a good idea not to
Use the experimental results shown in Figure (26) (B =
memorize it.
70 gauss) to estimate the accelerating voltage used for
Let us use the experimental numbers provided with the electrons in the beam. (The experimental answer is
included in the homework answer section.)
Figure (25b) as an example of the use of Equation (25).
In that figure, the strength of the magnetic field is B =
70 gauss, and the electrons were accelerated by an
accelerating voltage of 135 volts. The constants m and
q are the mass and charge of an electron.

The first step is to calculate the speed v of the electrons

using the fact that the electrons have 135 eV of kinetic
energy. We begin by converting from eV to joules
using the conversion factor 1.6 × 10 -19 joules per eV.
This gives

1 2 -19 joules
mv = 135 eV × 1.6 × 10
2 eV
-19 Figure 26
2 2 × 135 × 1.6 × 10 Use the fact that the magnetic field for this
v = -30
.911 × 10 example was 70 gauss to estimate the accelerating
voltage that produced the electron beam.
2
12 m
= 47.4 × 10 2
s

6 m
v = 6.9 × 10
s (26)
28-22 Magnetism

Particle Accelerators The pulse of electrons enter an evacuated circular track

Our knowledge of the structure of matter on a sub- shown in the top view. To keep the pulse of electrons
atomic scale, where we study the various kinds of moving in the circular track, large electromagnets
elementary particles, has come from our ability to shown in the cross-sectional view are used to provide
accelerate particles to high energies in particle accel- a perpendicular magnetic field. In this example the
erators such as the synchrotron. In a synchrotron, an magnetic field B points downward so that the magnetic
electric field E is used to give the particle’s energy, and force qv × B = -ev × B points inward toward the
a magnetic field B is used to keep the particles confined center of the track.
to a circular track.
We saw that a magnetic field cannot do any work on the
Figure (27a) is a schematic diagram of a small electron electrons since the magnetic force is perpendicular to
synchrotron. At the top is an electron gun that is used the particle’s velocity. Therefore to give the electrons
to produce a beam of electrons. In practice the gun is more energy, we use an electric field E. This is done by
quickly turned on then off to produce a pulse of inserting into one section of the path a device that
electrons. produces an electric field so that the electric force -eE
points in the direction of the motion of the electrons.
electron evacuated circular doughnut
gun through which electrons move
One might think of using a charged parallel plate
capacitor to create the electric field E, but that is not
feasible. Later we will see that radio waves have an
electric field E associated with them, and it is a radio
path of wave electric field in a so-called “resonant cavity” that
electrons
is used to produce the required strong fields. For now
top view it does not matter how E is produced, it is this electric
field that adds energy to the electrons.

electric field
accelerates
electrons

cross-section
of doughnut

electro-
B magnets B

cross-sectional view
Figure 27a
Diagram of a synchrotron, in which the electrons,
produced by the electron gun, travel through a Figure 27b
circular evacuated doughnut. The electrons are The Berkeley synchrotron shown here, accelerated
accelerated by an electric field, gaining energy on protons rather than electrons. It was the first machine
each trip around. The electrons are kept in a with enough energy to create anti protons. After this
circular orbit by an increasingly strong magnetic machine was built, ways were devised for focusing the
field produced by the electromagnets. particle beam and using an evacuated doughnut with a
much smaller cross-sectional area.
 28-23

When electrons gain energy, their momentum p = mv

increases. Writing Equation (25) in the form
p
r = mv = (25a)
qB qB
we see that an increase in the electron’s momentum p
will cause the orbital radius r to increase. The radius r
will increase unless we compensate by increasing the
strength B of the magnetic field. The rate at which we
increase B must be synchronized with the rate at which
we increase the particle’s momentum p in order to
keep r constant and keep the electrons in the circular
path. Because of this synchronization, the device is
called a synchrotron.
Figure 28
You can see that the amount of energy or momentum The Fermi Lab accelerator has two accelerating
we can supply to the particles is limited by how strong rings, one on top of the other. In each, the
evacuated doughnut is only 2 inches in diameter,
a field B we can make. Iron electromagnets can create and four miles in circumference. The bottom ring
fields up to about 1 tesla (here the MKS unit is useful) uses superconducting magnets (painted yellow),
or 10,000 gauss. The superconducting magnets, being while the older upper ring has iron magnets
used in the latest accelerator designs, can go up to (painted red and blue.)
around 5 tesla.

Noting that B is limited to one or a few tesla, Equation

(25) tells us that to get more momentum or energy, we
must use accelerators with a bigger radius r. This
explains why particle accelerators are getting bigger
and bigger. The biggest particle accelerator now
operating in the United States is the proton accelerator
at the Fermi National Accelerator Laboratory in Batavia,
Illinois shown in Figures (28) and (29).
In Figure (28), we see a section of tunnel and the
magnets that surround the 2 inch diameter evacuated
pipe which carries the protons. Originally there was
one ring using iron magnets (painted red and blue in the
photograph). Later another ring with superconducting
magnets was installed, in order to obtain stronger Figure 29
magnetic fields and higher proton energies. The ring of Aerial view of the Fermi Lab particle accelerator.
superconducting magnets (painted yellow) is beneath
the ring of iron magnets.

Figure (29) is an aerial view showing the 4 mile

circumference of the accelerator. Currently the largest
accelerator in the world is at the European Center for
Particle Physics (CERN). The 27 kilometer path of that
accelerator is seen in Figure (30) on the next page.
28-24 Magnetism

RELATIVISTIC ENERGY edge of the magnetic field and orbital radius immedi-
AND MOMENTA ately tells us the momentum of the particles in the large
synchrotrons.
Even the smallest synchrotrons accelerate electrons
and protons up to relativistic energies where we can no To determine the energy of the particles in these
longer use the non relativistic formula 1/2 mv2 for machines, we need a relationship between a particle’s
kinetic energy. For any calculations involving the large energy E and momentum p. The relationship can be
accelerators we must use fully relativistic calculations obtained by writing out E and p in the forms
like E = mc2 for energy and p = mv for momentum
where m = m0/ 1 - v2 /c2 is the relativistic mass. m0
p = mv = v (29)
2 2
Equation (25) or (25a) for a charged particle moving in 1 - v /c
a circular orbit of radius r, can be written in the form m0
2 2
(25b) E = mc = c (30)
p = qBr 2 2
1 - v /c
where B is the strength of the uniform magnetic field
It is then straightforward algebraic substitution
and p the particle momentum. It turns out that Equation
to show that
(25) is correct even at relativistic energies provided
An exact
p = mv is the relativistic momentum. Thus a knowl- E2 = p 2 c2 + m02 c4 (31)
relationship

Figure 30
Path for the 8 kilometer circumference Super Proton Synchrotron (SPS, solid circle) and
the 27 kilometer Large Electron-Positron collider (LEP, dashed circle) at CERN, on the
border between France and Switzerland. The Geneva airport is in the foreground.
 28-25

Exercise 6 How good was our approximation that we could ne-

glect the particle’s rest energy and use the simple
Directly check Equation (31) by plugging in the values
Equation (32)? Recall that the rest energy of a proton
of p and E from Equations (29) and (30).
is about 1GeV. Thus the SPS accelerator produced
In the big particle accelerators the kinetic energy protons with a kinetic energy 430 times greater! For
supplied by the accelerators greatly exceeds the particle’s these particles it is not much of an error to neglect the
rest energy m0 c2, so that the m0 c2 2 term in Equation rest energy.
(31) is completely negligible. For these “highly relativ-
istic” particles, we can drop the m02 c4 term in Equation Exercise 7
(31) and we get the much simpler formula a) The Fermi lab accelerator, with its radius of
1 kilometer, uses superconducting magnets to produce
E ≈ pc If E >> m0c2 (32) beams of protons with a kinetic energy of 1000 GeV
1012 eV . How strong a magnetic field is required to
produce protons of this energy?
Equation (32) is an accurate relationship between
energy and momentum for any particle moving at a b) Iron electromagnets cannot produce magnetic
speed so close to the speed of light that its total energy fields stronger than 2 tesla, which is why superconduct-
E greatly exceeds its rest energy m0c2. ing magnets were required to produce the 1000 GeV
protons discussed in part a). Before the ring with
For the high energy particle accelerators we can com- superconducting magnets was constructed, a ring
bine Equations (25) and (32) to get using iron magnets already existed in the same tunnel.
The iron magnets could produce 1.5 tesla fields. What
E = p c = qBrc (33) was the maximum energy to which protons could be
Consider CERN's Super Proton Synchrotron or SPS, accelerated before the superconducting magnets were
shown by the smaller solid circle in Figure (30), which installed? (You can see both rings of magnets in Figure
was used to discover the particles responsible for the 28.)
weak interaction. In this accelerator, the magnets pro-
Exercise 8
duced fields of B = 1.1 tesla, and the radius of the ring
The large electron-positron (LEP) collider, being con-
was r = 1.3km (for a circumference of 8km). Thus we
structed at CERN, will create head on collisions be-
have
tween electrons and positrons. (Electrons will go around
E = qBrc one way, and positrons, having the opposite charge,
will go around the other.) The path of the LEP accelera-
= 1.6 × 10
-19
coulombs × 1.1 tesla tor, which will have a circumference of 27 km, is shown
in Figure (30), superimposed on the countryside north
of Geneva, Switzerland.
× 1.3 × 10 3m × 3 × 10 8 m s
a) Assuming that the LEP accelerator will use 3 tesla
–8 superconducting magnets, what will be the maximum
= 6.9 × 10 joules kinetic energy, in eV, of the electrons and positrons that
Converting this answer to electron volts, we get will be accelerated by this machine?

–8 b) What will be the speed of these electrons and

6.9 × 10joules 9
E = = 430 × 10 eV positrons? (How many 9’s in v/c?)
joules
1.6 × 10 -19
eV
= 430GeV (34)
28-26 Magnetism

stereo BUBBLE CHAMBERS

camera In the study of elementary particles, it is just as impor-
tant to have adequate means of observing particles as it
is to have accelerating machines to produce them. One
liquid hydrogen beam of of the more useful devices for this purpose is the bubble
charged chamber invented by Donald Glaser in 1954.
particles
from It may not be true that Glaser invented the bubble
accelerator chamber while looking at the streaks of bubbles in a
glass of beer. But the idea is not too far off. When a
charged particle like an electron, proton or some exotic
elementary particle, passes through a container of
liquid hydrogen, the charged particle tends to tear
light electrons from the hydrogen atoms that it passes,
leaving a trail of ionized hydrogen atoms. If the
Figure 31a
pressure of the liquid hydrogen is suddenly reduced the
Schematic diagram of the Berkeley
10-inch hydrogen bubble chamber. liquid will start boiling if it has a “seed”—a special
location where the boiling can start. The trail of ionized
hydrogen atoms left by the charged particle provides a
trail of seeds for boiling. The result is a line of bubbles
showing where the particle went.

In a typical bubble chamber, a stereoscopic camera is

used to record the three dimensional paths of the
particles. It is impressive to look at the three dimen-
sional paths in stereoscopic viewers, but unfortunately
all we can conveniently do in a book is show a flat two
dimensional image like the one in Figure (32). In that
picture we see the paths of some of the now more
common exotic elementary particles. In the interest-
ing part of this photograph, sketched above, a negative
π- meson collides with a positive proton to create a
neutral Λ0 and a neutral Κ0 meson. The neutral Λ0 and
Κ0 do not leave tracks, but they are detected by the fact
that the Κ0 decayed into a π+ and a π- meson, and the
Λ0 decayed into a π- and a proton p+, all of which are
charged particles that left tracks.

To analyze a picture like Figure (32) you need more

information than just the tracks left behind by particles.
You would also like to know the charge and the
Figure 31b momentum or energy of the particles. This is done by
The 10-inch Bubble chamber at the Lawrence
placing the bubble chamber in a magnetic field so that
Radiation Laboratory, University of California,
Berkeley. (Photograph copyright The Ealing positive particle tracks are curved one way and nega-
Corporation, Cambridge, Mass.) tives ones the other. And, from Equation (25b), we see
that the radii of the tracks tell us the momenta of the
particles.
 28-27

Another example of a bubble chamber photograph is Exercise 9

Figure (33) where we see the spiral path produced by Calculate the energy, in eV of the electron as it entered
an electron. The fact that the path is spiral, that the the photograph in Figure (33). Since you do not know
radius of the path is getting smaller, immediately tells off hand whether the particle was relativistic or not, use
us that the electron is losing momentum and therefore the exact relation
energy as it moves through the liquid hydrogen. The
2 2 2 2 4
magnetic field used for this photograph had a strength E = p c + m 0c (31)
B = 1.17 tesla, and the initial radius of the spiral was 7.3
cm. From this we can determine the momentum and to determine E from p. From your answer decide
energy of the electron. whether you could have used the non relativistic formula
2
KE = 1/2 m0v or the fully relativistic formula E = pc, or
whether you were in an intermediate range where
neither approximation works well.

+
π− π
−
π
K0
Λ0 p
−
π
H atom

B = 1.17 tesla

Figure 32
Bubble chamber photograph showing the creation of a
K 0 meson and a Λ 0 particle, and their subsequent
R i = 7.3 cm e- annihilations. We now know that the K meson is a
quark/anti quark pair, and the Λ 0 particle contains 3
Figure 33 quarks as does a proton and a neutron. The K and Λ
Spiraling electron. An electron enters the chamber at particles last long enough to be seen in a bubble
the lower left and spirals to rest as it loses momentum. chamber photograph because they each contain a
The spiral track is caused by the magnetic field applied strange quark which decays slowly via the weak
to the chamber which deflects a charged particle into a interaction. (Photo copyright The Ealing Corporation
curved path with a radius of curvature proportional to Cambridge, Mass.)
the particle's momentum. The straight track crossing
the spiral is a proton recoiling from a collision with a
stray neutron. Because the proton has much greater
mass than the electron, its track is much less curved.
28-28 Magnetism

The Mass Spectrometer being the same. Thus ions of the two isotopes will have
A device commonly seen in chemistry and geology slightly different masses, and land at slightly different
labs is the mass spectrometer which is based on the distances down the photographic plate. If an isotope is
circular orbits that a charged particle follows in a missing in one sample the corresponding line on the
uniform magnetic field. Figure (34) is a sketch of a photographic plate will be absent. The analogy be-
mass spectrometer which consists of a semi circular tween looking at the lines identifying isotopes, and
evacuated chamber with a uniform magnetic field B looking at a photographic plate showing the spectrum
directed up out of the paper. The direction of B is of light, suggested the name mass spectrometer.
chosen to deflect positive ions around inside the cham-
ber to a photographic plate on the right side. The ions Exercise 10
to be studied are boiled off a heated filament and Suppose that you wish to measure the mass of an iodine
accelerated by a negative cap in a reversed voltage atom using the apparatus of Figures (34) and (35). You
electron gun shown in Figure (35). By measuring the coat the filament of the gun in Figure (35) with iodine,
position where the ions strike the photographic plate, and heat the filament until iodine atoms start to boil off.
we know the radius of the orbit taken by the ion. In the process, some of the iodine atoms lose an
Combine this with the knowledge of the field B of the electron and become positive ions with a charge +e.
spectrometer, and we can determine the ion’s momen- The ions are then accelerated in the gun by a battery of
tum p if the charge q is known. The speed of the ion is voltage Vb and then pass into the evacuated chamber.
determined by the accelerating voltage in the gun, thus (a) Assuming that Vb = 125 volts (accelerating voltage)
knowing p gives us the mass m of the ions. Non and B = 1000 gauss (0.1 tesla), and that the iodine
relativistic formulas work well and the calculations are atoms follow a path of radius r = 18.2 cm, calculate the
nearly
electrons
identical
in Figure
to (25).
our analysis
(See Equations
of the path
26 to of
28.th- mass m of the iodine atoms.

Mass spectrometers are used to identify elements in (b) How many times more massive is the iodine ion than
small sample of material, and are particularly useful i a proton? From the fact that protons and neutrons have
being able to separate different isotopes of an element about the same mass, and that an electron is 2000 times
Two different isotopes of an element have differen lighter, use your result to estimate how many nuclear
numbers of neutrons in the nucleus, everything else particles (protons or neutrons) are in an iodine nucleus.

uniform magnetic field

directed out of paper
evaculated atoms
beam of chamber
atoms coating of atoms
can whose mass is to
be measured

r
battery hot filament
d=2r
gun photographic film Figure 35
When the substance to be studied is heated by a
Figure 34
filament, atoms evaporate and some lose an electron
Top view of a mass spectrograph. A uniform
and become electrically charged positive ions. The
magnetic field B rises directly up through
ions are then accelerated by an electric field to
the chamber. The beam of atoms is produced
produce a beam of ions of known kinetic energy.
by the accelerating gun shown in Figure 35.
 28-29

Magnetic Focusing B
In the magnetic force examples we have considered so end
far, the velocity v of the charged particle started out view
perpendicular to B and we got the circular orbits we
have been discussing.
side
If we place an electron gun so that the electron beam is view
aimed down the axis of a pair of Helmholtz coils, as Figure 37a
shown in Figure (36), the electron velocity v is parallel In reality the electron beam spreads out when it leaves
to B , v × B = 0 and there is no magnetic force. the cap. Most of the electrons are not moving parallel
to B , and there will be a magnetic force on them.
Figure (36) is a bit too idealized for the student built
electron gun we have been using in earlier examples.
Some of the electrons do come out straight as shown in
FB = (– e) v X B
Figure (36), but many come out at an angle as shown in
Figure (37a). In Figure (37b) we look at the velocity v
v v
components v⊥ and v|| of an electron emerging at an
angle q. Because v|| × B = 0 only the perpendicular θ
component v⊥ contributes to the magnetic force v FB

Fmag = qv⊥ × B (36) B

B out of paper
This force is perpendicular to both v⊥ and B as shown
in the end view of the electron gun, Figure (37b). In this side view end view
end view, where we can’t see v|| , the electron appears
to travel around the usual circular path. Figure 37b
Consider an electron emerging from the cap at an
angle θ from the center line as shown in the side
view above. Such an electron has a component of
velocity v⊥ perpendicular to the magnetic field. This
produces a magnetic force FB = – e v⊥ × B which
B points toward the axis of the gun. The magnetic
force FB can be seen in the end view above. From
the end view the electron will appear to travel in a
circle about the axis of the gun. The stronger the
magnetic field, the smaller the radius of the circle.

Figure 36
Electron gun inserted so that the beam of
electrons moves parallel to the magnetic field of
the coils. If the beam is truly parallel to B ,
there will be no magnetic force on the electrons.
28-30 Magnetism

It is in the side view, Figure (38) that we see the effects The tightening of the helix is seen in Figure (39) where
of v|| . Since there is no force related to v|| , this in (a) we see an electron beam with no magnetic field.
component of velocity is unchanged and simply carries The electrons are spraying out in a fairly wide cone. In
the electron at a constant horizontal speed down the (b) we have a 75 gauss magnetic field aligned parallel
electron gun. The quantity v|| is often called the drift to the axis of the gun and we are beginning to see the
speed of the particle. (The situation is not unlike helical motion of the electrons. In (c) the magnetic field
projectile motion, where the horizontal component vx is increased to 200 gauss and the radius of the helix has
of the projectile’s velocity is unaffected by the vertical decreased considerably. As B is increased, the elec-
acceleration a y .) trons are confined more and more closely to a path
along the magnetic field lines. In our electron gun, the
When we combine the circular motion, seen in the end magnetic field is having the effect of focusing the
view of Figure (37c), with the constant drift speed v|| , electron beam.
down the tube seen in Figure (38a) the net effect is a
helical path like a stretched spring seen in Figure (38b).
The electron in effect spirals around and travels along
the magnetic field line. The stronger the magnetic
field, the smaller the circle in Figure (37c), and the
tighter the helix.

B
a) No magnetic field
v
v v
θ v
v
Figure 38a
In the side view of the motion of the
electron, we see that v|| is unchanged, v||
b) B = 75 gauss
just carries the electron down the tube.

helical motion
of the electron
c) B = 200 gauss
Figure 38b
Oblique view of the helical motion of the electron.
When you combine the uniform motion of the electron
down the tube with the circular motion around the
axis of the tube, you get a helical motion with the
same shape as the wire in a stretched spring.

d) Movie

Figure 39
Focusing an electron beam with a parallel
magnetic field. The beam travels along a
helical path which becomes tighter as the
strength of the magnetic field is increased.
 28-31

SPACE PHYSICS The Magnetic Bottle

Even in non-uniform magnetic fields there is a ten- If a magnetic field has the correct shape, if the field lines
dency for a charged particle to move in a spiral path pinch together as shown at the left or the right side of
along a magnetic field line as illustrated in Figure (40). Figure (41), then the magnetic force Fmag on a charged
This is true as long as the magnetic field is reasonably particle has a component that is directed back from the
uniform over a distance equal to the radius r of the spiral pinch. For charged particles with the correct speed, this
(from Equation (25), r = mv⊥/qvB). Neglecting the back component of the magnetic force can reflect the
spiral part of the motion, we see that the large scale particle and reverse v|| . If the magnetic field is pinched
effect is that charged particles tend to move or flow at both ends, as in Figure (41) the charged particle can
along magnetic field lines. This plays an important role reflect back and forth, trapped as if it were in a magnetic
in space physics phenomena which deals with charged bottle.
particles emitted by the sun (the “solar wind”) and the
In the subject of plasma physics, one often deals with
interaction of these particles with the magnetic field of
hot ionized gasses, particularly in experiments de-
the earth and other planets.
signed to study the possibility of creating controlled
There are so many interesting and complex effects in fusion reactions. These gases are so hot that they would
the interaction of the solar wind with planetary mag- melt and vaporize any known substance they touch.
netic fields that space physics has become an entire The only known way to confine these gases to do
field of physics. Seldom are we aware of these effects experiments on them is either do the experiments so
unless a particularly powerful burst of solar wind fast that the gas does not have time to escape (inertial
particles disrupts radio communications or causes an confinement), or use magnetic fields and devices like
Aurora Borealis to be seen as far south as the temperate the magnetic bottle shown in Figure (41) (magnetic
latitudes. The Aurora are caused when particles from confinement).
the solar wind spiral in along the earth’s magnetic field magnetic "bottle"
lines and end up striking atoms in the upper atmo-
sphere. The atoms struck by the solar wind particles
emit light just like the residual air atoms struck by the
electrons in an electron gun. B B

FB FB
B path of charged
particle

Figure 41
Magnetic bottle. When the magnetic field lines pinch
together, the charged particles can be reflected back in
a process called magnetic mirroring. (At the two ends
Figure 40 of the magnetic bottle above, the magnetic force FB
When charged particles from the sun enter the has a component back into the bottle.)
earth's magnetic field, they spiral around the
magnetic field lines much like the electrons in
the magnetic focusing experiment of Figure 39.
28-32 Magnetism

Van Allen Radiation Belts In Figure (43), a student, Jeff Lelek, started with the
The earth’s magnetic field shown in Figure (9) and formula for a “dipole magnetic field”, namely
repeated in Figure (42) forms magnetic bottles that can
trap charged particles from the solar wind. The ends of B0
B = – * Z – 3 * Z⋅R *R (37)
the bottles are where the field lines come together at the R3
north and south magnetic poles, and the regions where
significant numbers of particles are trapped are called which is a reasonably accurate representation of the
the Van Allen radiation belts shown in Figure (42). earth’s magnetic field, and calculated some electron
Protons are trapped in the inner belt and electrons in the orbits for this field. The result is fairly complex, but we
outer one. do get the feeling that the electron is spiraling around
the magnetic field lines and reflecting near the mag-
It is not feasible to do hand calculations of the motion netic poles.
of charged particles in non-uniform magnetic fields.
The motion is just too complicated. But computer To provide a simpler interpretation of this motion, the
calculations, very similar to the orbit calculations dis- student let the calculation run for a long time, saving up
cussed in Chapter 4, work well for electric and mag- the particle coordinates at many hundreds of different
netic forces. As long as we have a formula for the shape points along the long orbit. These points are then
of E or B , we can use the Lorentz force law (Equation plotted as the dot pattern shown in Figure (43). (In this
20) picture, the latitude of the particle is ignored, the points
are all plotted in one plane so we can see the extent of
F = qE + qv × B the radial and north-south motion of the particles.) The
result gives us a good picture of the distribution of
as one of the steps in the computer program. The particles in a Van Allen radiation belt. This and similar
computer does not care how complicated the path is, calculations are discussed in the supplement on com-
but we might have trouble drawing and interpreting the puter calculations with the Lorentz force law.
results.

Figure 42 Figure 43
Charged particles, trapped by the earth's magnetic field, Computer plot of the motion of a proton in a dipole
spiral around the magnetic field lines reflecting where magnetic field. The formula for this field and the
the lines pinch together at the poles. The earth's computer program used to calculate the motion of
magnetic field thus forms a magnetic bottle, holding the the proton are given in the Appendix. As you can
charged particles of the Van Allen radiation belts. see, the motion is relatively complex. Not only does
the proton reflect back and forth between the poles,
but also precesses around the equator.
 28-33

Figure 44
In this computer plot, all the data points from Figure 43 are plotted as dots in one plane. From this we see the
shape of a Van Allen radiation belt emerge. (Figures (43) and (44) from a student project by Jeff Lelek.)
Chapter 29
Ampere's Law

CHAPTER 29 AMPERE'S LAW

In this chapter our main focus will be on Ampere’s law,

a general theorem that allows us to calculate the
magnetic fields of simple current distributions in much
the same way that Gauss’ law allowed us to calculate
the electric field of simple charge distributions.

As we use them, Gauss’ and Ampere’s laws are integral

theorems. With Gauss’ law we related the total flux out
through a closed surface to 1/ε0 times the net charge
inside the surface. In general, to calculate the total flux
through a surface we have to perform what is called a
surface integral. Ampere’s law will relate the integral
of the magnetic field around a closed path to the total
current flowing through that path. This integral around
a closed path is called a line integral.

Until now we have concentrated on examples that did

not require us to say much about integration. But as we
discuss Ampere’s law in this chapter and the remaining
Maxwell equations in the next few chapters, it will be
convenient to draw upon the formalism of the surface
and line integral. Therefore we will take a short break
to discuss the mathematical concepts involved in these
integrals.
29-2 Ampere's Law

THE SURFACE INTEGRAL that goes completely across the stream, from bank to
In our discussion of Gauss’ law near the end of Chapter bank, from the surface to the bottom. The total flux ΦT
24, we defined the flux Φ of a fluid in a flow tube as the of water flowing through this net is therefore equal to
amount of water per second flowing past the cross- the total current in the stream.
sectional area of the tube as shown in Figure (1). This To calculate the total flux ΦT , we break the stream
is equal to the velocity v times the cross-sectional area flow up into a number of small flow tubes bounded by
A ⊥ of the tube as given in Equation (24-46) stream lines as shown in (3). Focusing our attention on
Φ = vA⊥ flux in a flow tube (24-46) the i th flow tube, we see that the tube intersects an area
dA i of the fish net. The flux through the fish net due
As seen in Figure (2), if we slice the flow tube by a plane to the i th tube is
that is not normal to the flow tube, the area A of the
intersection of the tube and plane is larger than the dΦi = vi ⋅dA i (1)
cross-sectional area A ⊥. The relationship is
A⊥ = A cos θ where θ is the angle between v and A where vi is the velocity of the water at the intersection
(see Equation 24-45). Defining the vector A as having of the tube and the net. The total flux or current of water
a magnitude A and direction normal to the plane, we in the stream is simply the sum of the fluxes in each flow
have tube, which can be written
v ⋅ A = vA cos θ = vA ⊥ ΦT = ∑ Φi = ∑ vi ⋅dA i (2)
all flow i
and the formula for the flux in the flow tube is tubes

Φ = v⋅A
where the dA i are just those areas on the fish net
(24-47a)
marked out by the flow tubes.
In Chapter 24 we considered only problems where A⊥ If we go to infinitesimal sized flow tubes, the sum in
was something simple like a sphere around a point Equation (2) becomes an integral which can be written
source, or a cylinder around a line source, and we could as
easily write a formula for the total flux. We now wish
to consider how we should calculate, at least in prin-
ciple, the flux in a more complex flow like the stream ΦT = v⋅dA surface integral (3)
shown in Figure (3). area of
fish net

To give our flux calculation a sense of reality, suppose

plane slicing
that we wish to catch all the salmon swimming up a
the flow tube
stream to spawn. As shown in Figure (3), we place a net

cross-sectional A
area A A

θ
v
area A
flow tube v

flux Φ = v A flux Φ = v A = v.A

Figure 1 Figure 2
The flux of water Φ through a flow tube is the If we have an area A that is not normal to the stream,
amount of water per second flowing past a then the cross-sectional area is A⊥ = A cos θ , and the
cross-sectional area A⊥ . flux is Φ = vA⊥ = vA cos θ , which can also be written
Φ = v ⋅ A.
 29-3

where the dA's are infinitesimal pieces of area on the Gauss’ Law
fish net and our sum or integration extends over the The statement of Gauss’ law applied to electric fields
entire submerged area of the net. Because we are in Chapter (24) was that the total electric flux ΦT out
integrating over an area or surface in Equation (3), this through a closed surface was equal to 1/ε0 times the
integral is called a surface integral. total charge Q in inside the surface. Our surface integral
of Equation (3) allows us to give a more formal (at least
Think of Equation (3), not as an integral you “do”, like more mathematical sounding) statement of Gauss’
x 2dx = x 3/3, but more as a formal statement of the law.
steps we went through to calculate the total flux ΦT.
Suppose we have a collection of charged particles as
Suppose, for example, someone came up to you and shown in Figure (4), which are completely surrounded
asked how you would calculate the total current in the by a closed surface S. (Think of the closed surface as
stream. If you were a mathematician you might being the surface of an inflated balloon. There cannot
answer, “I would calculate the integral be any holes in the surface or air would escape.) The
ΦT = v⋅dA (3a)
total flux of the field E out through the surface S is
formally given by the surface integral
S
where S is a surface cutting the stream.”
ΦT = E⋅dA
(4)
If you were a physicist, you might answer, “Throw a S
fish net across the stream, making sure that there are no where the dA are small pieces of the surface, and E is
gaps that the fish can get through. (This defines the the electric field vector at each dA.
mathematician's surface S). Then measure the flux of
water through each hole in the net (these are the v⋅dA's
of Equation 3a), and then add them up to get the total
Q1 Q4
flux (do the integral).”
Q3
Basically, the mathematician’s statement in Equation
Q2
(3a) is short hand notation for all the steps that the
physicist would carry out.
fish net
across stream Q5
surface S

Figure 4
Closed surface S completely surrounding a collection
of charges. The flux of E out through the closed
surface is equal to 1 ε0 times the total charge inside.
dA i

vi

i th flow
tube

Figure 3
To calculate the flux of water through a fish net, we can
first calculate the flux of water through each hole in the
net, and then add up the fluxes to get the total flux.
29-4 Ampere's Law

The total charge Q in inside the surface S is obtained by Inside the surface the total charge is Q. Thus Gauss’
adding up all the charges we find inside. Any charges law, Equation (5), gives
outside do not count. (We have to have a completely
closed surface so that we can decide whether a charge Q in
E ⋅dA =
is inside or not.) Then equating the total flux ΦT to ε0
S
Q in/ε 0 we get the integral equation
Q
E(r)4π r2 =
Q formal ε0
E⋅dA = ε in statementof (5) Q
0 E(r) =
closed Gauss' law 4πε0r2 (6)
surface S

The only thing that is new here is the use of the notation
There is nothing really new in Equation (5) that we did E ⋅dA for total flux ΦT. When we actually wish to
not say back in Chapter (24). What we now have is a S
calculate ΦT, we look for a surface that is perpendicular
convenient short hand notation for all the steps we to E so that we can use the simple formula EA⊥.
discussed earlier.
If the charge distribution were complex, more like
We will now use Equation (5) to calculate the electric Figure (4), we could calculate E ⋅dA by casting a fish
field of a point charge. Although we have done this S
net all around the charges and evaluating Ei ⋅dAi for
same calculation before, we will do it again to remind each hole in the net. The formal expression of the
us of the steps we actually go through to apply Equation surface integral at least gives us a procedure we can
(5). A formal equation like this becomes real or useful follow if we are desperate.
only when we have an explicit example to remind us
how it is used. When you memorize such an equation, surface S
also memorize an example to go with it.

In Figure (5), we have a point charge +Q that produces r

a radial electric field E as shown. To apply Gauss’ law
we draw a spherical surface S of radius r around the Q E
charge. For this surface we have

ΦT = E ⋅dA = EA ⊥ = E(r)4πr 2
S
Figure 5
For the electric field of a point charge, we know
immediately that the total flux Φ T out through the
spherical surface is the area 4 π r2 times the strength
E r of the field. Thus

ΦT ≡ E ⋅ dA = E r × 4 π r2
S
 29-5

THE LINE INTEGRAL The first thing we have to worry about in discussing
Another formal concept which we will use extensively Equation (7), is what path the particle takes in going
in the remaining chapters on electromagnetic theory is from Point (1) to Point (2). If we are moving an eraser
the line integral. You have already been exposed to the over a blackboard, the longer the path, the more work
idea in earlier discussions of the concept of work. If we we do. In this case, we cannot do the line integral until
exert a force F on a particle while the particle moves the path has been specified.
from Point (1) to Point (2) as shown in Figure (6), then On the other hand, if we are carrying the particle around
the work we do is given by the integral the room, exerting a force F = – Fg that just over-
2 comes the gravitational force, then the work we do is
Work W = F⋅dx (7) stored as gravitational potential energy. The change in
1 potential energy, and therefore the line integral of
Equation (7), depends only on the end points (1) and (2)
where we are integrating along the path in Figure and not on the path we take. When the line integral of
(6). a force does not depend upon the path, we say that the
Equation (7) is short hand notation for many steps. force is conservative.
What it really says is to draw the path taken by the A formal statement that a force is conservative is that
particle in going from Point (1) to (2), break the path up the line integrals are equal for any two paths -- for
into lots of little steps dxi, calculate the work dWi we example, path (a) and path (b) in Figure (7).
do during each step, dWi = Fi ⋅dx i, and then add up
all the dWi to get the total work W. 2 2
F⋅dx = F⋅dx (8)
Fi 2
1(path a) 1(path b)

b
th
i dx i pa (2)

1 dWi = Fi dx i
2
W
tot
= Σ Fi dx i F dx
i
1 (1) a
Figure 6 p ath
To calculate the total work done moving a particle from
point (1) to point (2) along a path , first break the path Figure 7
up into many short displacements dx . The work dWi is If the work done in carrying the particle from point (1)
dWi = Fi ⋅ dxi . The total work W is the sum of all the dWi . to point (2) does not depend upon which path we take,
we say that the force is conservative.
29-6 Ampere's Law

Let us write Equation (8) in the form Equation (10) does not really depend upon the Points
2 2 (1) and (2). More generally, it says that if you go out and
then come back to your starting point, and the sum of
F ⋅dx – F ⋅dx = 0 all your Fi⋅dxi is zero, then the force is conservative.
1(path a) 1(path b) This special case of a line integral that comes back to
Now take the minus sign inside the integral over Path the starting point as in Figure (9) is called the line
(b) so that we have a sum of Fi⋅ – dxi integral around a closed path, and is denoted by an
integral sign with a circle in the center
2 2
F ⋅dx + F ⋅ – dx = 0 (9) the line integral
1(path a) 1(path b)
F⋅dx ≡ around a closed path (11)
as in Figure 9
For the path (b) integral, we have reversed the direction
of each step. The sum of the reversed steps is the same With the notation of Equation (11), we can formally
as going back, from point (2) to point (1) as illustrated define a conservative force F as one for which
in Figure (8). Thus Equation (9) becomes
2 1 F⋅dx = 0 definition of a (12)
F ⋅dx + F ⋅dx = 0 (10) for any conservative force
closed path
1(path a) 2(path b)

If Equation (10) applies for any Path (a) and (b), then This line integral around a closed path will turn out to
the force F is conservative. be an extremely useful mathematical tool. We have
already seen that it distinguishes a conservative force
th b like gravity, where F⋅dx = 0 , from a non conserva-
pa
– dx (2) tive force like friction on a blackboard eraser, for which
–

F⋅dx ≠ 0 . In another case, namely Ampere’s law to

be discussed next, the line integral of the magnetic field
around a closed path tells us something about the
dx currents that flow through the path.

(1) a
p ath

Figure 8
If we take path b backwards, i.e. go from point (2) to
point (1), the dxi on path b are reversed and the i dx i
integral along path b changes sign. If the line integral
from (1) to (2) does not depend upon the path, then the
line integral for any return trip must be the negative of
Fi
Figure 9
the integral for the trip out, and the sum of the two
integrals must be zero. For a conservative force, the line integral F ⋅ dx , that
goes completely around a closed path, must be zero. It
is not necessary to specify where the calculation starts.
 29-7

AMPERE’S LAW and Equation (13) becomes

Figure (10), which is similar to Figure (28-14), is a B⋅ d = Bd
sketch of the magnetic field produced by a current in a (15)
straight wire. In this figure the current i is directed up In addition, our path has a constant radius r, so that
and out of the paper, and the magnetic field lines travel B = µ0 i/2π r is constant all around the path. We can
in counter clockwise circles as shown. We saw from take this constant outside the integral in Equation (15)
equation (28-18) that the strength of the magnetic field to get
is given by
µ0i B⋅ d = B d (16)
B = (28-18)
2πr Next we note that d is just the sum of the lengths of
In Figure (11) we have drawn a circular path of radius our steps around the circle; i.e., it is just the circumfer-
r around the wire and broken the path into a series of ence 2πr of the circle, and we get
steps indicated by short vectors d i. We drew the stick
figure to emphasize the idea that this is really a path and B⋅ d = B d = B × 2πr (17)
that d i shows the length and direction of the i th step.
Finally substituting the value of B from Equation (28-
For each of the steps, calculate the dot product Bi⋅ d i 18) we get
where Bi is the magnetic field at that step, and then
add up the Bi⋅ d i for all the steps around the path to get
µ0i
× 2πr = µ0 i
Σ Bi ⋅d
all steps
i → B⋅d
(13)
B⋅ d =
2πr
(18)

around path
The result is the line integral of B around the closed
path.

Why bother calculating this line integral? Let us put in d i

the value for B given by Equation (28-18) and see r Bi
why. We happen to have chosen a path where each step
d i is parallel to B at that point, so that
Bi⋅ d i = Bid i (14) iup

iup
Σi Bi ⋅d i → B⋅d = B r × 2πr
Figure 11
µi Circular path of radius r around the wire. As we
B = 0 walk around the path, each step represents a
2πr
displacement d . To calculate the line integral
B ⋅ d , we take the dot product of d with B at
each interval and add them up as we go around
Figure 10 the entire path. In this case the result is simply
Circular magnetic field of a wire. B(r) times the circumference 2 π r of the path.
29-8 Ampere's Law

There are several points we want to make about Equa- µ0 i µ0 iθ1

tion (18). First we made the calculation easy by B⋅ d = B1 r1 θ1 = r1 θ1 =
2πr1 2π
choosing a circular path that was parallel to B all the arc 1
way around. This allowed us to replace the dot product
B⋅ d by a numerical product Bd , pull the constant B µ0 i µ0 iθ2
outside the integral, and get an answer almost by B⋅ d = B2 r2 θ2 = r2 θ2 =
2πr2 2π
inspection. This should be reminiscent of our work arc 2
with Gauss’ law where we chose surfaces that made it
easy to solve the problem. µ0 i µ0 iθ3
B⋅ d = B3 r3 θ3 = r3 θ3 =
2πr3 2π
The second point is that we get an exceptionally simple arc 3
answer for the line integral of B around the wire,
namely Adding the contribution from each arc segment we get
the line integral around the closed path

B⋅ d = µ0 i (18a) µ0 iθ1 µ0 iθ2 µ0 iθ3

B⋅ d = + +
2π 2π 2π
The line integral depends only on the current i through µ0 i
the path and not on the radius r of the circular path. = θ1 + θ2 + θ3
2π
What about more general paths that go around the
But θ 1 + θ 2 + θ 3 is the sum of the angles around the
wire? To find out, we have to do a slightly harder
circle, and is therefore equal to 2π. Thus we get for the
calculation, but the answer is interesting enough to
path of Figure (12)
justify the effort.

In Figure (12) we have constructed a closed path made B is perpendicular to d here

up of three arc sections of lengths r1 θ 1, r2 θ 2, and r3 θ 3
connected by radial sections as shown. These arcs are
sections of circles of radii r1, r2 and r3, respectively. We
2
rθ
2

wish to calculate B⋅ d for this path and see how the

answer compares with what we got for the circular
path. r 1θ 1
θ2 θ1
The first thing to note as we go around our new path is
θ3
that in all the radial sections, B and d are perpendicu-
r 3θ
3
lar to each other, so that B⋅ d = 0. The radial sections
do not contribute to our line integral and all we have to
B
do is add up the contributions from the three arc
segments. These are easy to calculate because
B⋅ d = Bd and B is constant over each arc, so that the
integral of B⋅ d over an arc segment is just the value of Figure 12
B times the length rθ of the arc. We get A somewhat arbitrary path around the wire is made of
arc sections connected by radial sections. Since B is
perpendicular to d in the radial sections, the radial
sections do not contribute to the B ⋅ d for this path.
In the arcs, the length of the arc increases with r, but B
decreases as 1/r, so that the contribution of the arc
does not depend upon how far out it is. As a result the
B ⋅ d is the same for this path as for a circular path
centered on the wire.
 29-9

µ0 i On the outer segment we are going in the same direc-

B⋅ d = 2π = µ0 i (19) tion as B so that B⋅ d is positive and we get
2π
which is the same answer we got for the circular path. µ0 i µ0 i
B⋅ d = B1 r1 θ = r1 θ = θ
2πr1 2π
The result in Equation (19) did not depend upon how arc 1
many line segments we used, because each arc contrib-
uted an angle θ, and if the path goes all the way around, On the inner arc we are coming back around in a
the angles always add up to 2π. In Figure (13) we have direction opposite to B, the quantity B⋅ d is negative,
imitated a smooth path (the dotted line) by a path and we get
consisting of many arc sections. The more arcs we use
–µ0 i –µ0 i
the closer the imitation. We can come arbitrarily close B⋅ d = –B 2 r2 θ = r2 θ = θ
to the desired path using paths whose integral B⋅ d 2πr2 2π
arc 2
is µ0 i. In this sense we have proved that Equation (19)
applies to any closed path around the wire. Adding up the two contributions from the two arcs, we
get
It is another story if the path does not go around the
wire. In Figure (14) we have such a path made up of two µ0iθ – µ0iθ
arc and two radial segments as shown. As before, we B⋅ d = + = 0 (20)
2π arc 1
2π arc 2
can ignore the radial segments because B and d are Path of
Figure 14
perpendicular and B⋅ d = 0.
For this closed path which does not go around the
current, we get B⋅ d = 0. This result is not changed
if we add more arcs and radial segments to the path. As
long as the path does not go around the current, we get
zero for B⋅ d .

dθ
r 1θ

r 2θ

θ
B

Figure 13
We can approximate an arbitrary path (dotted line) Figure 14
by a series of connected radial and arc sections. The In this example, where the path does not go around the
smaller the angle d θ marking the arc sections, the wire, the sections labeled r1 θ and r2 θ contribute equal
better the approximation. In calculating B ⋅ d , the and opposite amounts to the line integral B ⋅ d . As a
radial sections do not count, and we can bring all the result B ⋅ d is zero for this, or any path that does not go
arc sections back to a single circle centered on the around the wire.
wire. As a result, B ⋅ d does not depend upon the
shape of the path, as long as the path goes around
the wire.
29-10 Ampere's Law

Several Wires FT = qv × B (22)

It is relatively straightforward to generalize our results
to the case where we have several wires as in Figures where B is the effective field acting on the test particle,
(15 a,b). Here we have three currents i1, i2, and i3 each then Equations (21) and (22) give
alone producing a magnetic field B1 , B2 and B3
respectively. B = B1 + B2 + B3 (23)
The first step is to show that the net field B at any point
is the vector sum of the fields of the individual wires. The fact that magnetic fields add vectorially is a
We can do this by considering the force on a test particle consequence of the vector addition of forces and our
of charge q moving with a velocity v as shown in Figure use of the magnetic force law to define B.
(15a). Our earlier results tell us that the current i1 exerts
a force With Equation (23) we can now calculate B⋅ d for
the field of several wires. Let us draw a path around
F1 = qv × B1 two of the wires, as shown in Figure (15b). For this
path, we get
Similarly i2 and i3 exert forces

F2 = qv × B2 B⋅ d = B1 + B2 + B3 ⋅ d
Closed path
F3 = qv × B3 of Fig. 29–15
(24)
Newton’s second law required us to take the vector sum = B1 ⋅ d + B2 ⋅ d + B3 ⋅ d
of the individual forces to get the total force F acting on
an object
Since the closed path goes around currents i1 and i2, we
F = F1 + F2 + F3 get from Equation (19)
(21)
= qv × B 1 + B 2 + B 3 closed path

If we write this total force in the form

i1 i2 i1 i2
v

i3 q i3
Figure 15a Figure 15b
A charge q moving in the vicinity of three currents Calculating B ⋅ d for a path that
i1 , i2 and i3 . If the magnetic field B at the charge goes around two of the wires.
is the vector sum of the fields B1 , B2 and B3 of the
three wires, then the net magnetic force FB on q is
given by
FB = qv × B = qv × B1 + B2 + B3
= qv × B1 + qv × B2 + qv × B3
= FB1 + FB2 + FB3
and we get the desired result that the net force on q
is the vector sum of the forces exerted by each wire.
 29-11

Field of a Straight Wire

B 1 ⋅ d = µ 0i 1
Our first application of Ampere’s law will be to calcu-
late the magnetic field of a straight wire. We will use
B 2 ⋅ d = µ 0i 2 this trivial example to illustrate the steps used in
applying Ampere’s law.
Since the path misses i3, we get
First we sketch the situation as in Figure (16), and then
B 3⋅ d = 0 write down Ampere’s law to remind us of the law we
are using
and Equation (24) gives
B⋅d = µ 0i enclosed
current
B⋅ d = µ0 i1 + i2 = µ0 × enclosed (25) Next we choose a closed path that makes the line
Closed path
by path integral as simple as possible. Generally the path
of Fig. 15 should either be along B so that B⋅ d = Bd , or
perpendicular so that B⋅ d = 0. The circular path of
Equation (25) tells us that B⋅ d around a closed path Figure (16) gives B⋅ d = Bd with B constant, thus
is equal to µ0 times the total current i = i1+ i2 encircled
by the path. This has the flavor of Gauss’ law which B⋅ d = Bd = B d = B * 2πr = µ0 i
said that the total flux or surface integral of E out
through a closed surface was 1/ε0 times the total charge The result is
Qin inside the surface. Just as charge outside the closed
surface did not contribute to the surface integral of E,
currents outside the closed path do not contribute to the µ0 i
B =
line integral of B. 2π r
We derived Equation (25) for the case that all our
which we expected. When you memorize Ampere’s
currents were in parallel straight wires. It turns out that
it does not matter if the wires are straight, bent, or form law, memorize an example like this to go with it.
a hideous tangle. As a general rule, if we construct a
closed path, then the line integral of B around the closed circular path
path is µ0 times the net current ienclosed flowing of radius r
through the path
B
r
(26)
B⋅ d = µ0 ienclosed Ampere's iup
Law
any closed
path

This extremely powerful and general theorem is known Figure 16

as Ampere’s law. Using Ampere's law to calculate the magnetic field of a
wire. We have B⋅⋅ d = B × 2 π r around the path.
So far in this chapter we have focused on mathematical Thus Ampere's law B⋅⋅ d = µ 0i gives B = µ 0i/ 2 π r .
concepts. Let us now work out some practical applica-
tions of Ampere’s law to get a feeling for how the law
is used.
29-12 Ampere's Law

Exercise 1 Exercise 3
Each of the indicated eight conductors in Figure (17) Show that a uniform magnetic field B cannot drop
carries 2.0A of current into (dark) or out of (white) the abruptly to zero as one moves at right angles to it, as
page. Two paths are indicated for the line integral suggested by the horizontal arrow through point a in
B ⋅ d . What is the value of the integral for (a) the Figure (19). (Hint: Apply Ampere's law to the rectangu-
dotted path? (b) the dashed path? lar path shown by the dashed lines.) In actual magnets
"fringing" of the lines of B always occurs, which means
that B approaches zero gradually.

Figure 17 B
a
Exercise 2
Eight wires cut the page perpendicularly at the points
shown in Figure (18). A wire labeled with the integer k S
(k = 1, 2..., 8) bears the current ki0 . For those with odd
k, the current is up, out of the page; for those with even Figure 19
k it is down, into the page. Evaluate B ⋅ d along the
closed path shown, in the direction shown.
Exercise 4
Figure (20) shows a cross section of a long cylindrical
conductor of radius a, carrying a uniformly distributed
current i. Assume a = 2.0 cm, i = 100A, and sketch a
3 plot of B(r) over the range 0 < r < 4 cm.
4 6
2 8

5 7 a

1
r

Figure 18
i total = 100 amps

Figure 20

(The above are some choice problems from Halliday

and Resnick.)
 29-13

Exercise 5 Exercise 6
Figure (21) shows a cross section of a hollow cylindrical Figure (22) shows a cross section of a long conductor
conductor of radii a and b, carrying a uniformly distrib- of a type called a coaxial cable. Its radii (a, b, c) are
uted current i. shown in the figure. Equal but opposite currents i exist
in the two conductors. Derive expressions for B(r) in the
a) Show that B(r) for the range b < r < a is given by ranges
µ0i r2 – b2 a) r < c,
B(r) =
2π r a2 – b2
b) c < r < b,
b) Test this formula for the special cases of r = a, r = b, c) b < r < a, and
and r = 0.
d) r > a.
c) Assume a = 2.0 cm, b = 1.8 cm, and i = 100 A. What
is the value of B at r = a? (Give your answer in tesla and e) Test these expressions for all the special cases that
gauss.) occur to you.
a
a

r r
b b c

coaxial cable
Figure 21 Figure 22

Exercise 6 is a model of a coaxial cable, where the

current goes one way on the inner conductor and back
the other way on the outside shield. If we draw any
circuit outside the cable, there is no net current through
the circuit, thus there is no magnetic field outside. As
a result, coaxial cables confine all magnetic fields to
the inside of the cable. This is important in many
electronics applications where you do not want fields
to radiate out from your wires. The cables we use in the
lab, the ones with the so called BNC connectors, are
coaxial cables, as are the cables that carry cable televi-
sion.
29-14 Ampere's Law

FIELD OF A SOLENOID Right Hand Rule for Solenoids

As with Gauss’ law, Ampere’s law is most useful when The direction of the field inside the solenoid is a bit
we already know the field structure and wish to calcu- tricky to figure out. As shown in Figure (24), up near
late the strength of the field. The classic example to the wires and in between the turns, the field goes in a
which ampere’s law is applied is the calculation of the circle around the wire just as it does for a straight wire.
magnetic field of a long straight solenoid. As we go out from the wire the circular patterns merge
to create the uniform field in the center of the solenoid.
A long solenoid is a coil of wire in which the length L
of the coil is considerably larger than the diameter d of We see, from Figure (24), that if the current goes
the individual turns. The shape of the field produced around the coil in such a way that the current is up out
when a current i flows through the coil was illustrated of the paper on the right side and down into the paper
in Figure (28-21) and is sketched here in Figure (23). on the left, then the field close to the wires will go in
Iron filings gave us the shape of the field and Ampere’s counterclockwise circles on the right and clockwise
law will tell us the strength. circles on the left. For both these sets of circles, the field
inside the coil points down. As a result the uniform
The important and useful feature of a solenoid is that we field inside the coil is down as shown.
have a nearly uniform magnetic field inside the coil and
nearly zero field outside. The longer the solenoid, There is a simple way to remember this result without
relative to the diameter d, the more uniform the field B having to look at the field close to the wires. Curl the
inside and the more nearly it is zero outside. fingers of your right hand in the direction of the flow of
the current i in the solenoid, and your thumb will point
in the direction of the magnetic field inside the sole-
d noid. We will call this the right hand rule for solenoids.
N turns
in coil
i up

n = N/L is
the number (1)
of turns per
unit length idown iup
L h

nh turns
enclosed B
Figure 24
If you know the direction of the current in the wire,
i
you can determine the direction of the magnetic field
by looking very close to the wire where the field goes
B around the wire. You get the same answer if you curl
Figure 23 the fingers of your right hand around in the direction
Calculating the magnetic field of a long solenoid. the current in the coil is flowing. Your thumb then
Around the path starting at point (1) we have points in the direction of the field.
B ⋅ d = 0 + Bh + 0 + 0 . The amount of current
enclosed by the path is itot = (nh)i where n is the number
of turns per unit length. Thus Ampere's law
B⋅⋅ d = µ 0 itot gives Bh = µ 0nhi or B = µ 0ni .
 29-15

Evaluation of the Line Integral Calculation of i encl os ed

Figure (25) is a detail showing the path we are going to From Figure (25) we see that we get a current i up
use to evaluate B⋅ d for the solenoid. This path goes through our path each time another turn comes up
down the solenoid in the direction of B (side 1), and out through the path. On the left side of the coil the current
through the coil (side 2), up where B = 0 (side 3) and goes down into the paper, but these downward currents
back into the coil (side 4). lie outside our path and therefore are not included in our
evaluation of ienclosed. Only the positive upward
We can write B⋅ d as the sum of four terms for the currents count, and ienclosed is simply i times the
four sides number of turns that go up through the path.

To calculate the number of turns in a height h of the coil,

B⋅ d = B⋅d + B⋅d we note that if the coil has a length L and a total of N
side 1 side 2 turns, then the number of turns per unit length n is given
by
+ B⋅d + B⋅d
side 3 side 4
number of turns N
On sides 2 and 4, when the path is inside the coil, B and ≡ n = (28)
per unit length L
d are perpendicular and we get B⋅d = 0. Outside
the coil it is still 0 because there is no field there. and in a height h there must be nh turns
Likewise B⋅d = 0 for side 3 because there is no field
there. The only contribution we get is from side 1 inside number of turns
the coil. If h is the height of our path, then = nh (29)
in a height h
With nh turns, each carrying a current i, going up
B⋅ d = B⋅d = Bh (27) through our path, we see that ienclosed must be
side 1
ienclosed = inh (30)

(1) Using Ampere's law

path for B• d We are now ready to apply Ampere’s law to evaluate
(4) the strength B of the field inside the solenoid. Using
Equation (27) for B⋅ d , and Equation (31) for
ienclosed, we get
h i up (3)
B⋅ d = µ 0i enclosed

Bh = µ0 nih
(2) thumb pointing in
same direction as magnetic field
B = µ0 ni inside a solenoid
(31)
current through path
B
Figure 25 The uniform magnetic field inside a long solenoid is
Right hand rule for using Ampere's law. We define the proportional to the current i in the solenoid, and the
positive direction around the path as the direction you number of turns per unit length, n.
curl the fingers of your right hand when the thumb is
pointing in the direction of the current through the
path. (As you see, we can come up with a right hand
rule for almost anything.)
29-16 Ampere's Law

Exercise 7 One More Right Hand Rule

We will so often be using solenoids later in the course, If we really want to be careful about minus signs (and
that you should be able to derive the formula B = µ0 ni , it is not always necessary), we have to say how the sign
starting from Ampere’s law without looking at notes. of ienclosed is evaluated in Figure (25). If, as in Figure
This is a good time to practice. Take a blank sheet of (26) we reversed the direction of our path, then on side
paper, sketch a solenoid of length L with N turns. Then (1) B⋅d is negative because our path is going in the
close the text and any notes, and derive the formula for opposite direction to B. Thus for this path the complete
B. integral B⋅d is negative, and somehow our ienclosed
must also be negative, so that we get the same answer
We have mentioned that equations like
we got for Figure (25).
B1 ⋅ d = µ0 ienclosed are meaningless hen scratching
until you know how to use them. The best way to do that If we curl the fingers of our right hand in the direction
is learn worked examples along with the equation. Two
that we go around the path, then in Figure (25) our
good examples for Ampere’s law are to be able to
thumb points up parallel to the current through the path,
calculate the magnetic field inside a wire (Exercise 4),
and to be able to derive the magnetic field inside a
and in Figure (26) our thumb points down, opposite to
solenoid. If you can do these two derivations without the current. If we define the direction indicated by our
looking at notes, you should have a fairly good grasp of right hand thumb as the positive direction through the
the law. path, as shown in Figure (27), then the current is going
in a positive direction in Figure (25) but in a negative
direction in Figure (26). This gives us a negative
ienclosed for Figure (26) which goes along with the
(1) minus sign we got in the evaluation of B⋅d .

By now you should be getting the idea of how we define

(4) directions in magnetic formula. Always use your right
hand. After a while you get so used to using your right
hand that you do not have to remember the individual
right hand rules.
h i up (3)
positive direction
thumb pointing through path
down into
(2) paper

Figure 26
B
If we go around the wrong way, we just get two minus
signs and all the results are the same. Here we went
around the path so that our thumb pointed opposite to Figure 27
the direction of the current through the path. As a In general, we use the right hand convention to
result the magnetic field in the solenoid points opposite associate a positive direction around a path to a
to the direction of the path in the solenoid. positive direction through a path.
 29-17

The Toroid If there are N turns of wire in the toroid, and the wire
If we take a long solenoid, bend it in a circle and fit the carries a current i, then all N turns come up through the
ends together, we get what is called a toroid shown in path on the inside of the solenoid, and i enclosed is given
Figure (28). The great advantage of a toroid is that there by
are no end effects. In the straight solenoid the magnetic
ienclosed = Ni (33)
field at the ends fanned out into space as seen in our iron
filing map of Figure (28-23). With the toroid there are Using Equations (32) and (33) in Ampere’s law gives
no ends. The field is completely confined to the region
inside the toroid and there is essentially no field outside.
B ⋅ d = µ0 ienclosed
For this reason a toroid is an ideal magnetic field
storage device.
B × 2 π r = µ0 Ni
It is easy to use Ampere’s law to calculate the magnetic
field inside the toroid. In Figure (28) we have drawn a
µ0 N i magnetic field
path of radius r inside the toroid. Going around this path B = (34)
2πr of a toroid
in the same direction as B, we immediately get

B⋅d = B ×2πr (32) Note that N 2πr is the number of turns per unit length,
n, so that Equation (34) can be written B = µ0 ni
which is the solenoid formula of Equation (31). To a
because B is constant in magnitude and parallel to d . good approximation the field in a toroid is the same as
in the center of a straight solenoid.

The derivation of Equation (34) is so easy and such a

good illustration of the use of Ampere’s law that it
should be remembered as an example of Ampere’s
law.

i
r

B⋅⋅ d = B * 2 π r µ 0Ni
⇒B=
µ 0 i tot = µ 0Ni 2πr

B
Figure 28
When the solenoid is bent into the shape of a toroid, there are no end effects. The
magnetic field is confined to the region inside the toroid, and Ampere's law is easily
applied. (You should remember this as an example of the use of Ampere's law.)
29-18 Ampere's Law

Exercise 8
Figure 29 shows a 400-turn solenoid that is 47.5 cm long
and has a diameter of 2.54 cm. (The 10 turns of wire
wrapped around the center are for a later experiment.)
Calculate the magnitude of the magnetic field B near the
center of the solenoid when the wire carries a current of
3 amperes. (Give your answer in tesla and gauss.)

Exercise 9
Figure 30 shows the toroidal solenoid that we use in
several experiments later on. The coil has 696 turns
wound on a 2.6 cm diameter plastic rod bent into a circle
of radius 21.5 cm. What is the strength of the magnetic
field inside the coil when a current of 1 amp is flowing
through the wire? (Give your answer in tesla and gauss.)

Figure 29
A 400 turn straight solenoid 47.5 cm long,
wound on a 2.54 cm diameter rod.

Figure 30
A 696 turn toroidal solenoid wound on
a 2.6 cm diameter plastic rod bent into
a circle of radius 21.5 cm.
Chapter 30
Faraday's Law

FARADAY'S LAW

In this chapter we will discuss one of the more remark-

able, and in terms of practical impact, important laws
of physics – Faraday’s law. This law explains the
operation of the air cart speed detector we have used
in air track experiments, the operation of AC voltage
generators that supply most of the electrical power in
the world, and transformers and inductors which are
important components in the electronic circuits in
radio and television sets.

In one form, Faraday’s law deals with the line integral

E⋅ d of an electric field around a closed path. As an
introduction we will begin with a discussion of this line
integral for electric fields produced by static charges.
(Nothing very interesting happens there.) Then we will
analyze an experiment that is similar to our air cart
speed detector to see why we get a voltage proportional
to the speed of the air cart. Applying the principle of
relativity to our speed detector, i.e., riding along with
the air cart gives us an entirely new picture of the
behavior of electric fields, a behavior that is best
expressed in terms of the line integral E⋅ d . After a
discussion of this behavior, we will go through some
practical applications of Faraday’s law.
30-2 Faraday's Law

ELECTRIC FIELD Now consider the distribution of fixed point charges

OF STATIC CHARGES shown in Figure (2). Let E1 be the field of Q1, E2 of Q2,
etc. Because an electric field is the force on a unit test
In this somewhat formal section, we show that
charge, and because forces add as vectors, the total
E⋅ d = 0 for the electric field of static charges. With
electric field E at any point is the vector sum of the
this as a background, we are in a better position to
individual fields at that point
appreciate an experiment in which E⋅ d is not zero.

In Figure (1), we have sketched a closed path through E = E1 + E2 + E3 + E4 + E5 (1)

the electric field E of a point charge, and wish to
calculate the line integral E⋅ d for this path. To We can now use Equation (1) to calculate E⋅ d
simplify the calculation, we have made the path out of around the closed path in Figure (2). The result is
arc and radial sections. But as in our discussion of
Figure 29-13, we can get arbitrarily close to any path
using arc and radial sections, thus what we learn from E⋅d = E1 + E2 + ... + E5 ⋅ d
the path of Figure (1) should apply to a general path. (2)
= E1 ⋅d + + E5 ⋅d
Because the electric field is radial, E is perpendicular
to d and E⋅ d is zero on the arc sections. On the radial But E 1 ⋅d = 0 since E1 is the field of a point charge,
sections, for every step out where E⋅ dr is positive and the same is true for E2 ... E5. Thus the right side
there is an exactly corresponding step back where E⋅ dr of Equation (2) is zero and we have
is negative. Because we come back to the starting
point, we take the same steps back as we took out, all for the field E of
the radial E⋅d r cancel and we are left with E⋅ d = 0 E⋅d = 0 any distribution of
(3)
for the electric field of a point charge. static charges
Equation (3) applies to any distribution of static charges,
a point charge, a line charge, and static charges on
E
conductors and in capacitors.
d closed
path
Q2

Q1 Q4
Q
3

E
Q5

Figure 1
Closed path through the electric field of a Figure 2
point charge. The product E⋅⋅ d is zero Closed path in a region of a distribution of point
on the arc sections, and the path goes out charge. Since E⋅⋅ d = 0 is zero for the field of
just as much as it comes in on the radial each point charge alone, it must also be zero for
sections. As a result E⋅⋅ d = 0 when we the total field E = E1 + E2 + E3 + E4 + E5
integrate around the entire path.
 30-3

A MAGNETIC FORCE EXPERIMENT Since the gap in the loop does not allow the conduction
Figures (3a,b) are two views of an experiment designed electrons to flow along the wire, how are we going to
to test for the magnetic force on the conduction elec- detect the magnetic force on them? There is no net
trons in a moving copper wire. We have a wire loop force on the wire because the magnetic field exerts an
with a gap and the loop is being pulled out of a magnet. equal and opposite force on the positive copper ions in
At this instant only the end of the loop, the end opposite the wire.
the gap, is in the magnetic field. It will soon leave the Our conjecture is that this magnetic force on the
field since it is being pulled out at a velocity v as shown. conduction electrons would act much like the gravita-
In our earlier discussions we saw that a copper atom has tional force on the water molecules in a static column
two loosely bound conduction electrons that are free to of water. The pressure at the bottom of the column is
flow from one atom to another in a copper wire. These higher than the pressure at the top due to the gravita-
conduction electrons form a negatively charged elec- tional force. Perhaps the pressure of the negatively
tric fluid that flows in a wire much like water in a pipe.

Because of the gap we inserted in the wire loop of

Figure (3), the conduction electrons in this loop cannot B v
flow. If we move the loop, the conduction electrons
must move with the wire. That means that the conduc-
tion electrons have a velocity v to the right as shown,
perpendicular to the magnetic field which is directed
into the page. Thus we expect that there should be a
magnetic force
Figure 3a
Wire loop moving through magnetic
Fmag = – ev × B (4)
field of iron magnet.
acting on the electrons. This force will be directed
down as shown in Figure (3b).
moving wire loop

magnetic field
X X X X X X directed into paper
v
X X X X X X

X X FB X X X X
h v gap V
X X X X X X
FB voltmeter
X X X X X X
v
X X X X X X
FB v
magnetic force conduction electrons
on conduction
electrons
Figure 3b
When you pull a wire loop through a magnetic field, the electrons, moving at a velocity v with the wire, feel
a magnetic force FB = –e v × B if they are in the field. This force raises the pressure of the electron fluid
on the bottom of the loop and reduces it on the top, creating a voltage V across the gap. The arrow next to
the voltmeter indicates a voltage rise for positive charge, which is a voltage drop for negative charge.
30-4 Faraday's Law

charged electric fluid is higher at the bottom of the loop The next step is to calculate the magnitude of the
than the top due to the magnetic force. voltage reading we expect to see. As you follow this
calculation, do not worry about the sign of the voltage
To find out if this is true, we use an electrical pressure V because many sign conventions (right hand rules,
gauge, which is a voltmeter. A correctly designed positive charge, etc.) are involved. Instead concentrate
voltmeter measures an electrical pressure drop without on the basic physical ideas. (In the laboratory, the sign
allowing any current to flow. Thus we can place the of the voltage V you read on a voltmeter depends on
voltmeter across the gap and still not let the conduction how you attached the leads of the voltmeter to the
electrons flow in the loop. apparatus. If you wish to change the sign of the voltage
reading, you can reverse the leads.)
If our conjecture is right, we should see a voltage
reading while the magnetic force is acting. Explicitly Since voltage has the dimensions of the potential
there should be a voltage reading while the wire is energy of a unit test charge, the magnitude of the
moving and one end of the loop is in the magnetic field voltage in Figure (3) should be the strength of the force
as shown. The voltage should go to zero as soon as the on a unit test charge, – e v × B with – e replaced by
wire leaves the magnetic field. If we reverse the 1, times the height h over which the force acts. This
direction of motion of the loop, the velocity v of the height h is the height of the magnetic field region in
conduction electrons is reversed, the magnetic force Figure (3). Since v and B are perpendicular, v×B = vB
-ev × B should also be reversed, and thus the sign of the and we expect the voltage V to be given by
voltage on the voltmeter should reverse. If we oscillate
the wire back and forth, keeping one end in the mag-
netic field, we should get an oscillating voltage reading force on unit distance over
V = ×
on the meter. test charge which force acts

The wonderful thing about this experiment is that all

these predictions work precisely as described. There voltage V on loop
are further simple tests like moving the loop faster to get V = vB × h moving at speed (5)
a stronger magnetic force and therefore a bigger volt- v through field B
age reading. Or stopping the wire in the middle of the
magnetic field and getting no voltage reading. They all Figure 3c
work! Pulling the coil out of the magnet
 30-5

AIR CART SPEED DETECTOR where we enter another region, from (4) to (5), where
The air cart velocity detector we have previously the magnetic force is increasing the fluid pressure. We
discussed, provides a direct verification of Equation get another increase of vBh, and then go to Position (6)
(5). The only significant difference between the air cart at the bottom of the voltmeter. In Figure (4) we have
speed detector and the loop in Figure (3) is that the two voltage rises as we go around the two loops, and we
speed detector coil has a number of turns (usually 10). should get twice the reading on the voltmeter.
In order to see the effect of having more than one turn
in the coil, we show a two turn coil being pulled out of voltage reading
V = 2vBh
a magnetic field in Figure (4). for 2 loops

Figure (4) is beginning to look like a plumbing diagram It is an easy abstraction to see that if our coil had N turns,
for a house. To analyze the diagram, let us start at the voltage rise would be N times as great, or
Position (1) at the top of the voltmeter and follow the
wire all the way around until we get to Position (6) at
voltage on an N turn
the bottom end of the voltmeter. When we get to
Position (2), we enter a region from (2) to (3) where the V = NvBh coil being pulledout (6)
of a magneticfield
magnetic force is increasing the electron fluid pressure
by an amount vBh, as in Figure (3).
Adding more turns is an easy way to increase or
Now instead of going directly to the voltmeter as in amplify the voltage.
Figure (3), we go around until we get to Position (4)

magnetic field
directed into paper two turn coil

v
(4)
(2)
v
X X X X X

X X F X FB X X
B (1)
X X X X X
V
X X FB X FB X X (6) volt
X X X X X v meter
(3) F
B FB (5)
v

Figure 4
A two turn loop being pulled through a magnetic field. With
two turns we have twice as much force pushing the electric
fluid toward the bottom of the gap giving twice the voltage V.
30-6 Faraday's Law

The setup for the air cart speed detector is shown in The situation most closely corresponding to Figure (4)
Figure (6). A multi turn coil, etched on a circuit board is position (d) where the coil is leaving the magnet.
as shown in Figure (5), is mounted as a sail on top of an According to Equation (6), the voltage at this point
air cart. Suspended over the air cart are two angle iron should be given by V = N vBh, where N = 10 for our
bars with magnets set across the top as shown. This 10 turn coil, v is the speed of the carts, B is the strength
produces a reasonably uniform magnetic field that goes of the magnetic field between the angle iron bars, and
across from one bar to the other as seen in the end view h is the average height of the coils. (Since the coils are
of Figure (6). drawn on a circuit board the outer loop has the greatest
height h and the inner loop the least.) The first time you
In Figure (7), we show the experiment of letting the cart use this apparatus, you can directly measure V, N, v and
travel at constant speed through the velocity detector. h and use Equation (6) to determine the magnetic field
In the initial position (a), the coil has not yet reached the strength B. After that, you know the constants N, B and
magnetic field and the voltage on the coil is zero, as h, and Equation (6) written as
indicated in the voltage curve at the bottom of the
figure. 1
v = V× (6a)
NBh
multiple turn coil
gives you the cart’s speed in terms of the measured
voltage V. Equation (6a) explains why the apparatus
acts as a speed detector.

electrical circuit board Let us look at the voltage readings for the other cart
connectors positions. The zero readings at Positions (a) and (e) are
Figure 5 easily understood. None of the coil is in the magnetic
The multi turn coil that rides on the air cart.
(Only 5 turns are shown.)
field and therefore there is no magnetic force or volt-
age.

magnet
magnets
N turn
coil
angle iron

N turn coil
magnetic
field air cart

air track air cart angle iron

(a) (b)

Figure 6
The Faraday velocity detector. The apparatus is reasonably easy to build. We first constructed a 10
turn coil by etching the turns of the coil on a circuit board. This was much better than winding a coil,
for a wound coil tends to have wrinkles that produce bumps in the data. Light electrical leads, not
shown, go directly from the coil to the oscilloscope. The coil is mounted on top of an air cart and
moves through a magnetic field produced by two pieces of angle iron with magnets on top as shown.
Essentially we have reproduced the setup shown in Figures 3 and 4, but with the coil mounted on an
air cart. As long as the coil remains with one end in the magnetic field and the other outside, as shown
in (b), there will be a voltage on the leads to the coil that is proportional to the velocity of the cart.
 30-7

Figure 6c
Velocity detector apparatus. The magnetic
field goes across, between the two pieces of
angle iron. The coil, mounted on a circuit
board, is entering the magnetic field.

Figure 7 magnets
Voltage on the coil as it moves at
constant speed through the magnetic
V=0
field. At position (a ) the coil has not yet
reached the field and there is no voltage. v
At position (b) one end of the cart is in
the field, the other outside, and we get a (a)
voltage proportional to the speed of the
cart. At (c) there is no voltage because
both ends of the cart are in the magnetic
field and the magnetic force on the two v
ends cancel. (There is no change of
magnetic flux at this point.) At (d), the (b)
other end alone is inside the field, and
we get the opposite voltage from the one V=0
we had at (b). (Due to the thickness of
v
the coil and fringing of the magnetic
field, the voltage rises and falls will be
somewhat rounded.) (c)

(d)

V=0
v

(e)

V voltage
on coil
(c)

t
(a) (b) (d) (e)
position of cart
30-8 Faraday's Law

magnetic field
We need a closer look to understand the changes in
directed into paper
voltage, when all or part of the coil is inside the
magnetic field. This situation, for a one turn coil, is X X X X X X X X X

illustrated in Figure (8). For easier interpretation we X X X X

X X X X X X X X
have moved the gap and voltmeter to the bottom of the X X X X X X X X
coil as shown. It turns out that it does not matter where X X X X X X X X
v
the gap is located, we get the same voltage reading. We X X X X X X X X

have also labeled the figures (b), (c), and (d) to corre- X X X X X X X X

spond to the positions of the air cart in Figure (7). X X X X X X X X

X X X X X X X X
V
In Figure (8c) where both ends of the coil are in the X X X X

X X X X X X X X X
magnetic field, the conduction electrons are being
pulled down in both ends and the fluid is balanced. The (b) coil entering magnetic field
electron fluid would not flow in either direction if the
gap were closed, thus there is no pressure across the gap
and no voltage reading. In contrast, in Figure (8d) X X X X (3) X X X X

X X
where only the left end of the coil is in the magnetic X X X X X X X X
field, the magnetic force on the left side would cause the X X X X X X X X

conduction electrons to flow counterclockwise around X X X X X X X

(1) (2) v
the loop if it were not for the gap. There must be an X X X X X X X

electric pressure or voltage drop across the gap to X X X X X X X X

X X X X X X X X
prevent the counterclockwise flow. This voltage drop X X X (4) X X X X
is what we measure by the voltmeter. V
X X

X X X X X X X X X
oscilloscope
In Figure (8b), where the coil is entering the magnetic
field, the magnetic force on the right side of the coil (c) coil completely in magnetic field
would try to cause a clockwise flow of the conduction
electrons. We should get a pressure or voltage opposite
X X X X X X X X X
to Figure (8d) where the coil is leaving. This reversal X X X X
in voltage is seen in the air cart experiment of Figure X X X X X X X X

(7), as the cart travels from (b) to (d). X X X X X X X X

X X X X X X X X v
Note that in Figure (8), where the horizontal sections of X X X X X X X X

the coil are also in the magnetic field, the magnetic X X X X X X X X

force is across rather than along the wire in these X X X X X X X X

X X X X X X X X
sections. This is like the gravitational force on the fluid V
X X X X
in a horizontal section of pipe. It does not produce any X X X X X X X X X

pressure drops.
(d) coil leaving magnetic field

Figure 8
When the coil is completely in the magnetic field, the
magnetic force on the electrons in the left hand leg (1)
is balanced by the force on the electrons in the right
hand leg (2), and there is no net pressure or voltage
across the gap. When the coil is part way out, there is
a voltage across the gap which balances the magnetic
force on the electrons. The sign of the voltage depends
upon which leg is in the magnetic field.
 30-9

A RELATIVITY EXPERIMENT The answer she needs lies in the Lorentz force law that
Now that we have seen, from Figure (7), extensive we discussed in Chapter 28. This law tells us the total
experimental evidence for the magnetic force on the electromagnetic force on a charge q due to either
conduction electrons in a wire, let us go back to Figure electric or magnetic fields, or both. We wrote the law
(3) where we first considered these forces, and slightly in the form
modify the experiment. Instead of pulling the coil out
of the magnet, let us pull the magnet away from the coil F = qE + qv × B (28-20)
as shown in Figure (9b). where E and B are the electric and magnetic fields
In Figure (9a) we have redrawn Figure (3), and added acting on the charge.
a stick figure to represent a student who happens to be
walking by the apparatus at the same speed that we are magnetic field
pulling the coil out of the magnet. To this moving directed into paper v
observer, the coil is at rest and she sees the magnet X X X X X X X X X

moving to the left as shown in (9b). In other words, X

X
X

X
X

X
X

X
X

X X X X
v
pulling the magnet away from the coil is precisely the X X X X X X X X

same experiment as pulling the coil from the magnet, X X X X X X X X

X X X X X X X X V
except it is viewed by a moving observer. X X X X X X X X

X X X X X X X X
The problem that the moving observer faces in Figure X X X X X X X X

(9b) is that, to her, the electrons in the coil are at rest. For X X X X X v
X X X X X X X X X
her the electron speed is v = 0 and the magnetic force
FB , given by magnetic
(a) moving coil, magnets at rest
force
FB = (– e) v X B
FB = –e v × B = 0 for Figure 9b (7)
on electrons
is zero! Without a magnetic force to create the pressure
in the electrical fluid in the wire, she might predict that
there would be no voltage reading in the voltmeter.
X X X X X X X X X

But there is a voltage reading on the voltmeter! We v

X X X X X

have used this voltage to build our air cart velocity X

X
X

X
X

X
X

X
X

X
X

X
X

X
X

X
detector. If the voltmeter had a digital readout, for X X X X X X X X

example, then it is clear that everyone would read the X X X X X X X X V

X X X X X X X X
same number no matter how they were moving, whether X X X X X X X X
they were like us moving with the magnet (9a), or like X X X X X X X X

her moving with the coil (9b). In other words, she has v X X X X X

X X X X X X X X X
to find some way to explain the voltage reading that she
electrons in
must see. (b) moving magnet, coil at rest wire at rest;
no magnetic
force
Figure 9
The only difference between (a) and (b) is the point
of view of the observer. In (a) we see a magnetic
force FB = –e v × B because the electrons are
moving at a speed v through a magnetic field B .
To the observer in (b), the magnet is moving, not
the electrons. Since the electrons are at rest, there
is no magnetic force on them. Yet the voltmeter
reading is the same from both points of view.
30-10 Faraday's Law

Let us propose that the Lorentz force law is generally In Figure (9c) we have redrawn Figure (9b) showing an
correct even if we change coordinate systems. In electric field causing the force on the electrons. Be-
Figure (9a) where we explained everything in terms of cause the electrons have a negative charge, the electric
a magnetic force on the conduction electrons, there was field must point up in order to cause a downward force.
apparently no electric field and the Lorentz force law
gave That the magnetic force of Figure (9a) becomes an
electric force in Figure (9c) should not be a completely
surprising result. In our derivation of the magnetic
F = qE + qv × B force law, we also saw that an electric force from one
point of view was a magnetic force from another point
in Figure 9a,
= –e v×B (8a) of view. The Lorentz force law, which includes both
E=0 electric and magnetic forces, has the great advantage
that it gives the correct electromagnetic force from any
In Figure (9b), where v = 0, we have point of view.

Exercise 1
F = qE + qv × B
Equation (9) equates E in Figure (9c) with v × B in Figure
in Figure 9b, (9a). Show that E and v × B point in the same direction.
= –e E (8b)
v =0
upward electric field E
causes downward force
In other words, we will assume that the magnetic force on electrons
of Figure (9a) has become an electric force in Figure
(9b). Equating the two forces gives X X X X X X X X X

X X X X X
v X X X X X X X

That should be From

= v×B
X X X X X X X X
E (9) X X X X X X X X
in Figure 9b Figure 9a X X X X X X X X V
X X X X X X X X

X X X X X X X X

X X X X X X X X

v X X X X X

X X X X X X X X X

electrons in
(c) moving magnet, coil at rest wire at rest
feel an
electric force
Figure 9c
From the point of view that the coil is at rest, the
downward force on the electrons in the coil must be
produced by an upward directed electric field.
 30-11

FARADAY'S LAW Magnetic Flux

An experiment whose results may be surprising, is In our discussion of velocity fields and electric fields,
shown in Figure (10). Here we have a magnetic field we used the concept of the flux of a field. For the
produced by an electromagnet so that we can turn B on velocity field, the flux Φv of water was the volume of
and off. We have a wire loop that is large enough to water flowing per second past some perpendicular area
surround but not lie in the magnetic field, so that B = 0 A⊥. For a uniform stream moving at a speed v, the flux
all along the wire. Again we have a gap and a voltmeter was Φv = vA⊥. For the electric field, the formula for
to measure any forces that might be exerted on the flux was ΦE = EA⊥.
conduction electrons in the wire.
In Figures (9 and 10), we have a magnetic field that
We have seen that if we pull the wire out of the magnet, "flows" through a wire loop. Following the same
Figure (9a), we will get a voltage reading while the loop convention that we used for velocity and electric fields,
is leaving the magnetic field. We have also seen, Figure we will define the magnetic flux ΦB as the strength of
(9c), that we get a voltage reading if the magnetic field the field B times the perpendicular area A⊥ through
is pulled out of the loop. In both cases we started with which the field is flowing
a magnetic field through the loop, ended up with no
magnetic field through the loop, and got a reading on ΦB = BA⊥ Definition of (10)
the voltmeter while the amount of magnetic field magnetic flux
through the loop was decreasing. In both figures (9) and (10), the flux ΦB through the
Now what we are going to do in Figure (10) is simply wire loop is decreasing. In Figure (9), ΦB decreases
shut off the electromagnet. Initially we have a mag- because the perpendicular area A⊥ is decreasing as the
netic field through the loop, finally no field through the loop and the magnet move apart. In Figure (10), the
loop. It may or may not be a surprise, but when we shut flux ΦB is decreasing because B is being shut off. The
off the magnetic field, we also get a voltage reading. important observation is that whenever the flux ΦB
We get a voltage reading if we pull the loop out of the through the loop decreases, whatever the reason for the
field, the field out of the loop, or shut off the field. We change may be, we get a voltage reading V on the
are seeing that we get a voltage reading whenever we voltmeter.
change the amount of magnetic field, the flux of
magnetic field, through the loop.
magnetic field
pointing down
and being coil at rest
shut off

X X X X X X
X X X X X X
X X X X X X
X X X X X X
X X X X X X V
X X X X X X
X X X X X X
X X X X X X

electromagnet
at rest
Figure 10
Here we have a large coil that lies completely outside
the magnetic field. Thus there is no magnetic force on
any of the electrons in the coil wire. Yet when we turn
the magnet on or off, we get a reading in the volt meter.
30-12 Faraday's Law

One Form of Faraday's Law dΦB = –BdA = –Bhdx

The precise relationship between the voltage and the
change in the magnetic flux through the loop is found = –Bhvdt (13)
from our analysis of Figure (9) where the loop and the
magnet were pulled apart. We got a voltage given by where the – sign indicates a reduction in flux, and we
Equation (5) as used Equation (11) to replace dx by vdt .

Dividing both sides of Equation (13) by dt gives

V = vBh (5)
dΦB
Let us apply Equation (5) to the case where the magnet = –Bhv (14)
is being pulled out of the loop as shown in Figure (11). dt
In a time dt, the magnet moves to the left a distance dx But Bhv is just our voltmeter reading. Thus we get the
given by surprisingly simple formula
dx = vdt (11)
d ΦB Oneform of
and the area of magnetic field that has left the loop, V = – Faraday's law (15)
dt
shown by the cross hatched band in Figure (11), is

area of magnetic Equation (15) is one form of Faraday’s law.

dA = hdx = field that has Equation (15) has a generality that goes beyond our
(12)
left the loop original analysis of the magnetic force on the conduc-
tion electrons. It makes no statement about what causes
This decrease in area causes a decrease in the magnetic
the magnetic flux to change. We can pull the loop out
flux ΦB = BA ⊥ through the loop. The change in flux
of the field as in Figure (9a), the field out of the loop as
dΦB is given by in Figure (9b), or shut the field off as in Figure (10). In
all three cases Equation (15) predicts that we should see
large coil at rest
a voltage, and we do.
moving
electromagnet
x If we have a coil with more than one turn, as we had
X X X X X X
flux leaving back in Figure (4), and put a voltmeter across the ends
X X X X X X
in a time dt of the coil, then we get N times the voltage, and
X X X X X X
v
X X X X X X
h V Equation (15) becomes
X X X X X X

X X X X X X d ΦB for a coil
dx V = N – with N turns
(15a)
magnetic field dt
pointing down
provided dΦB /dt is the rate of change of magnetic flux
Figure 11 in each loop of the coil.
As the magnet and the coil move away from each
other, the amount of magnetic flux through the Exercise 2
coil decreases. When the magnet has moved a
distance dx, the decrease in area is hdx, and the Go back to Figure (7) and explain the voltage plot in
magnetic flux decreases by B×hdx . terms of the rate of change of the flux of magnetic
field through the coil riding on top of the air cart.
 30-13

A Circular Electric Field To produce the same kind of voltage V that we have
In Figure (10), where we shut the magnet off and got a seen in the previous experiments, the electric field at
voltage reading on the voltmeter, there must have been the wire must be directed up on the left hand side, as it
some force on the electrons in the wire to produce the was in Figure (9c). But because of the circular symme-
voltage. Since there was no magnetic field out at the try of the setup in Figure (12), the upwardly directed
wire, the force must have been produced by an electric electric field on the left side, which is parallel to the
field. We already have a hint of what that electric field wire, must remain parallel to the wire as we go around
looks like from Figure (9c). In that figure, we saw that the wire loop. In other words, the only way we can have
the moving magnetic field created an upwardly di- an upwardly directed electric field acting on the elec-
rected electric field acting on the electrons on the left trons on the left side of the loop, and maintain circular
side of the wire loop. symmetry, is to have the electric field go in a circle all
the way around the loop as shown in Figure (12).
To figure out the shape of the electric field produced
when we shut off the magnet, consider Figure (12), We can determine the strength of this circular electric
where we have a circular magnet and a circular loop of field, by figuring out how strong an electric field must
wire . We chose this geometry so that the problem act on the electrons in the wire, in order to produce the
would have circular symmetry (except at the gap in the voltage V across the gap. We then use Equation (15) to
loop). relate this voltage to the rate of change of the magnetic
flux through the loop.

downward pointing electric field

magnetic field around decreasing
being turned off magnetic flux

E
X X
X X X X
X X X X X X
X X X X X X
E X X X X X X
V
X X X X X X
X X X X d ΦB
E V = –
X X dt

circular electric field

upwardly directed pushes on electrons
electric field exerts downward all the way around
electric force on electrons in the the wire loop
left side of wire loop as in Fig. 9b
Figure 12
When the magnetic field in the magnet is turned off, a circular electric field is
generated. This electric field exerts a force on the electrons in the wire, creating
a pressure in the electric fluid that is recorded as a voltage pulse by the voltmeter.
30-14 Faraday's Law

Recall that the definition of electric voltage used in Line Integral of E around a Closed Path
deriving Equation (5) was In Figure (13) we have removed the wire loop and volt
meter from Figure (12) so that we can focus our
force on unit distance over attention on the circular electric field produced by the
V = ×
test charge which force acts decreasing magnetic flux. This is not the first time we
have encountered a circular field. The velocity field of
For Figure (12), the force on a unit test charge is the a vortex and the magnetic field of a straight current
electric field E, and this force acts over the full circum- carrying wire are both circular. We have redrawn
ference 2πr of the wire loop. Thus the voltage V across Figure (29-10) from the last chapter, showing the
the gap is circular magnetic field around a wire.

The formula for the strength of the magnetic field in

V = E × 2πr Figure (29-10) is
Equating this voltage to the rate of change of magnetic
flux through the wire loop gives B × 2πr = µ0 i (28-18)

dΦB a result we derived back in Equation 28-18. This

V = E × 2πr = – (16) should be compared with the formula for the strength
dt
of the electric field in Figure (13)
Equation (16) tells us that the faster the magnetic field
dies, i.e. the greater dΦB dt, the stronger the electric dΦB
field E produced. E × 2πr = – (16)
dt

electric magnetic field

downward pointing electric field
current around electric
magnetic field around decreasing
magnetic flux pointing up current
being turned off
E

X X
B
X X X X
X X X X X X
X X X X X X
E X X X X X X
iup
X X X X X X
X X X X
X X E

dΦB
E × 2π r = – B × 2πr = µ0 i
dt
Figure 13 Figure 29-10
Circular electric field around Circular magnetic field
a changing magnetic flux. around an electric current.
 30-15

In our discussion of Ampere’s law, we called µ 0 i the USING FARADAY'S LAW

“source” of the circular magnetic field. By analogy, we Up until now we have been looking for arguments
should think of the rate of change of magnetic flux, leading up to Faraday’s law. Let us now reverse the
– dΦB /dt , as the “source” of the circular electric field. procedure, treating Equation 17 as a basic law for
electric fields, and see what the consequences are.
In Chapter 29, we generalized Ampere’s law by replac-
ing B * 2πr by the line integral B⋅d along a closed
path around the wire. The result was Electric Field of an Electromagnet
As a beginning exercise in the use of Faraday's law, let
Ampere's law us use Equation (17) to calculate the electric field of the
B⋅ d = µ0 i for (29-18) electromagnet in Figure (13). We first argue that
magnetic fields because of the circular symmetry, the electric field
should travel in circles around the decreasing magnetic
where the line integral can be carried out along any field. Thus we choose a circular path, shown in Figure
closed path surrounding the wire. Because of close (13a), along which we will calculate E⋅ d . Then
analogy between the structure and magnitude of the using the assumption (because of circular symmetry)
magnetic field in Figure (29-10) and the electric field in that E is parallel to d and has a constant magnitude all
Figure (13), we expect that the more general formula the way around the circular path, we can write
for the electric field produced by a changing magnetic
flux is
E⋅ d = E d = E d = E 2πr (18)

dΦB Faraday's law

E⋅ d = – for (17) Using this result in Equation (17) gives
dt electric fields
dΦB
E⋅ d = E 2πr = – (19)
Equation 17 is the most general form of Faraday’s law. dt
It says that the line integral of the electric field around
any closed path is equal to (minus) the rate of change which is the result we had in Equation (16).
of magnetic flux through the path.
Right Hand Rule for Faraday's Law
downward pointing path for We can get the correct direction for E with the follow-
magnetic field calculating ing right hand rule. Point the thumb of your right hand
being turned off E⋅ d in the direction of the magnetic field. If the magnetic
flux is decreasing (if – dΦB /dt is positive), then the
fingers of your right hand curl in the direction of E. If
X X the magnetic flux is increasing, then E points the other
X X X X way. Please practice this right hand rule on Figures
X X X X X X (13a), (9c), and (15).
X X X X X X
X X X X X X
X X X X X X
X X X X
X X r

Figure 13a
Using Faraday's law to calculate E .
30-16 Faraday's Law

Electric Field of Static Charges THE BETATRON

If all we have around are static electric charges, then As we have mentioned before, when you encounter a
there are no magnetic fields, no magnetic flux, and no new and strange equation like Faraday’s law, it is
changing magnetic flux. For this special case, essential to have an example that you know inside out
dΦB /dt = 0 and Faraday’s law gives that illustrates the equation. This transforms the equa-
tion from a collection of symbols into a set of instruc-
for electric fields tions for solving problems and making predictions.
E⋅ d = 0 produced by (20) One of the best examples to learn for the early form of
static charges
Faraday’s law, Equation (15a), was the air cart speed
detector experiment shown in Figure (7). (You should
When the line integral of a force is zero around any
have done Exercise 2 analyzing the experiment using
closed path, we say that the force is conservative. (See
Equation (15a).
Equation 29-12.) Thus we see that if we have only
static electric charge (or constant magnetic fields), the The most direct example illustrating Faraday’s law for
electric field is a conservative field. electric fields, Equation (17), is the particle accelerator
called the betatron. This device was used in the 1950s
In contrast, if we have changing magnetic fields, if
for study of elementary particles, and later for creating
dΦB /dt is not zero, the electric field is not conserva-
electron beams for medical research.
tive. This can lead to some rather interesting results
which we will see in our discussion of a device called A cross-sectional view of the betatron is shown in
the betatron. Figure (14a). The device consists of a large electro-
magnet with a circular evacuated doughnut shaped
chamber for the electrons. The circular shape of the
electromagnet and the evacuated chamber are more
evacuated doughnut for
clearly seen in the top view, Figure (14b). In that view
charged particles we show the strong upward directed magnetic field B0
in the gap and the weaker upward directed magnetic
field out at the evacuated doughnut.

The outer magnetic field Br is required to keep the

Br Bo Br electrons moving along a circular orbit inside the
evacuated chamber. This field exerts a force
FB = –e v × B r that points toward the center of the
circle and has a magnitude mv2 /r in order to produce
electromagnet the required radial acceleration. Thus Br is given by

Br = mver
Figure 14a
Cross-sectional view of a betatron, showing which is our familiar formula for electrons moving
the central field B0 and the field Br out at the along a circular path in a magnetic field. (As a quick
evacuated doughnut. The relative strength of
review, derive the above equation.)
B0 and Br can be adjusted by changing the
shape of the electromagnet pole pieces.
Since a magnetic field does no work we need some
means of accelerating the electrons. In a synchrotron,
shown in Figure (28-27), a cavity which produces an
electric accelerating field is inserted into the electron’s
path. As an electron gains energy and momentum (mv)
each time it goes through the cavity, the magnetic field
 30-17

B was increased so that the electron’s orbital radius r = field around the flux as shown in Figure (15), just as in
mv/eB remains constant. (The synchronizing of B with Figure (13). This electric field is exactly parallel to the
the momentum mv leads to the name synchrotron.) orbit of the electrons and accelerates them continu-
ously as they go around.
In the betatron of Figure (14), we have a magnetic field
Br to keep the electrons in a circular orbit, and as the What is elegant about the application of Faraday's law
electrons are accelerated, Br is increased to keep the to the electrons in the betatron, is that E ⋅ d , which
electrons in an orbit of constant radius r. But what has the dimensions of voltage, is the voltage gained by
accelerates the electrons? There is no cavity as in a an electron going once around the circular orbit. The
synchrotron. energy gained is just this voltage in electron volts

Suppose that both B0 and Br are increased simulta-

neously. In the design shown in Figure (14a), B0 and energy gained
Br are produced by the same electromagnet, so that we (in eV) by electron = E⋅d (21)
can increase both together by turning up the electro- going around once
magnet. If the strong central field B0 is increased, we
have a large change in the magnetic flux through the This voltage is then related to dΦB/dt by Faraday’s
electron orbit, and therefore by Faraday’s law law.
E ⋅ d = – dΦB/dt we must have a circular electric

magnetic field B r path of

at the electron path electrons

FB E Bo directed
magnetic field B o
m up and increasing E
directed up

evacuated
doughnut
Figure 14b Figure 15
Top view of the betatron showing the When the strong central field B0 in the
evacuated doughnut, the path of the betatron is rapidly increased, it produces a
electrons, and the magnetic fields B0 in the circular electric field that is used to
center and Br out at the electron path. In accelerate the electrons. The electric field E
order to keep the electrons moving on a is related to the flux Φ B of the central field
circular path inside the doughnut, the B0 by Faraday's law E ⋅ d = – d Φ B / dt .
magnetic force FB = –e v × Br must have a
magnitude FB = mv2 r where r is the radius of
the evacuated doughnut.
30-18 Faraday's Law

Let us consider an explicit example to get a feeling for TWO KINDS OF FIELDS
the kind of numbers involved. In the 100 MeV betatron At the beginning of the chapter we showed that the line
built by General Electric, the electron orbital radius is
integral E ⋅ d around a closed path was zero for any
84 cm, and the magnetic field B0 is cycled from 0 to
.8 tesla in about 4 milliseconds. (The field B0 is then electric field produced by static charges. Now we see
dropped back to 0 and a new batch of electrons are that the line integral is not zero for the electric field
accelerated. The cycle is repeated 60 times a second.) produced by a changing magnetic flux. Instead it is
given by Faraday’s law E ⋅ d = –dΦB /dt. These
The maximum flux Φm through the orbit is results are shown schematically in Figure (16) where
we are looking at the electric field of a charged rod in
2 2 (16a) and a betatron in (16b).
Φm = B0 πr = .8 tesla × π × (.84m)
max
In Figure (17), we have sketched a wire loop with a
Φm = 1.8 tesla m
2 voltmeter, the arrangement we used in Figure (12) to
measure the E ⋅ d . We will call this device an
If this amount of flux is created in 4 milliseconds, then “ E ⋅ d meter ”. If you put the E ⋅ d meter over
the average value of the rate of change of magnetic flux the changing magnetic flux in Figure (16b), the voltme-
ΦB is ter will show a reading of magnitude V = dΦB/dt . If
we put the E ⋅ d meter over the charged rod in
dΦB Φm 1.8 Figure (16a), the meter reads V = 0. Thus we have a
= = = 450 volts
dt .004 sec .004 simple physical device, our E ⋅ d meter, which can
distinguish the radial field in Figure (16a) from the
Thus each electron gains 450 electron volts of kinetic circular field in Figure (16b). In fact it can distinguish
energy each time it goes once around its orbit. the circular field in (16b) from any electric field E
whatsoever that we can construct from static charges.
Exercise 3 Our E ⋅ d meter allows us to separate all electric
(a) How many times must the electron go around to fields into two kinds, those like the one in (16b) that can
reach its final voltage of 100 MeV advertised by the give a non zero reading, and those, produced by static
manufacturer? charges, which give a zero reading.
(b) For a short while, until the electron’s kinetic energy
Fields which register on our E ⋅ d meter generally
gets up to about the electron’s rest energy m0c2, the
electron is traveling at speeds noticeably less than c.
close on themselves like the circular fields in (16b).
After that the electron’s speed remains very close to c. Since these fields do not appear to have sources, they
How many orbits does the electron have to make before are called sourceless or “solenoidal” fields. An
its kinetic energy equals its rest energy? What fraction E ⋅ d meter is the kind of device we need to detect
of the total is this? solenoidal fields.
(c) How long does it take the electron to go from the point The conservative fields produced by static charges
that its kinetic energy equals its rest energy, up to the never close on themselves. They always start on
maximum of 100 MeV? Does this time fit within the 4 positive charge, end on negative charge, or come from
milliseconds that the magnetic flux is being increased? or go to infinity. These fields diverge from point
charges and thus are sometimes called “divergent”
fields. Our E ⋅ d meter does not work on the
divergent fields because we always get a zero reading.
 30-19

integration Although the E ⋅ d meter does not work on diver-

path E⋅d = 0 gent fields, Gauss’ law with the surface integral does.
In a number of examples we used Gauss' law

Q in
E ⋅ dA = (29-5)
ε0
closed
surface
E
to calculate the electric field of static charges. We are
seeing now that we use a surface integral to measure
charged
divergent fields, and a line integral to measure sole-
rod
noidal fields. There are two kinds of electric fields, and
we have two kinds of integrals to detect them.

It turns out to be a general mathematical theorem that

any vector field can be separated into a purely divergent
(a) Electric field of a static charge distribution
part and a purely solenoidal part. The field can be
has the property E ⋅ d = 0 uniquely specified if we have both an equation involv-
ing a Gauss’ law type surface integral to tell us the
divergent part, and an equation involving a Faraday’s
integration law type line integral to tell us the solenoidal part.
path E⋅d ≠ 0

E Bo directed E
up and increasing E⋅d meter V

Figure 17
(b) Electric field produced by a changing Wire loop and a volt meter can be used directly to
measure E ⋅ d around the loop. We like to call
magnetic flux has E ⋅ d = – dΦ B dt this apparatus an E ⋅ d meter.
Figure 16
Two kinds of electric field. Only the
field produced by the changing magnetic
flux has a non zero line integral.
30-20 Faraday's Law

Exercise 4 Note on our E⋅d meter

a) Maxwell’s equations are a set of equations that
completely define the behavior of electric fields E and Back in Figure (17) we used a wire loop and a voltmeter
magnetic fields B. One of Maxwell’s equations is as an E ⋅ d meter. I.e., we are saying that the
Faraday’s law voltage reading V on the voltmeter gives us the integral
of E around the closed path defined by the wire loop.
E⋅d = – dΦB /dt This is strictly true for a loop at rest, where the conduc-
tion electrons experience no magnetic force and all
which gives the line integral for the electric field. How forces creating the electric pressure are caused by the
many Maxwell equations are there? (How many equa- electric field E.
tions will it take to completely define both E and B?)
Earlier, in Figure (9), we had two views of an E ⋅ d
b) Are any of the other equations for electric and meter. In the bottom view, (9b) the loop is at rest and
magnetic fields we have discussed earlier, candidates
the voltage must be caused by an electric force. The
to be one of Maxwell’s equations?
moving magnetic field must have an electric field
c) At least one of Maxwell’s equations is missing – we associated with it. But in Figure (9a) where the magnet
have not discussed it. Can you guess what the equation is at rest, there is no electric field and the voltage
is and write it down? Explain what you can about your reading is caused by the magnetic force on the conduc-
guess. tion electrons in the moving wire. Strictly speaking, in
d) Back in our early discussion of velocity fields and
Figure (9a) the wire loop and voltmeter are measuring
Gauss’ law, we said that a point source for the velocity a pressure caused by magnetic forces and not an
field of an incompressible fluid like water, was a small E ⋅ d . The wire loop must be at rest, the path for
“magic” sphere in which water molecules were created. our line integral cannot move, if we are measuring
Suppose we do not believe in magic and assume that E⋅d .
for real water there is no way that water molecules can
be created or destroyed. Write down an integral equa- In practice, however, it makes little difference whether
tion for real water that expresses the fact that the we move the magnet or the loop, because the principle
vreal water has no sources (that create water molecules) of relativity requires that we get the same voltage V.
or sinks (that destroy them).

Do the best you can on these exercises now. Keep a

record of your work, and see how well you did when we
discuss the answers later in chapter 32.
 30-21

APPLICATIONS OF The AC Voltage Generator

FARADAY’S LAW In Figure (18) we have inserted a wire loop of area A in
The last few sections have been somewhat heavy on the magnetic field B of a magnet. We then rotate the
theory. To end this chapter on a more practical note, we coil at a frequency ω about an axis of the coil as shown.
will consider some simple applications of Faraday’s We also attach a voltmeter to the coil, using sliding
law, one that has immense practical applications and contacts so that the voltmeter leads do not twist as the
another that we can use in the laboratory. First we will coil spins.
discuss the AC voltage generator which is used by most As shown in Figure (19), as the loop turns, the magnetic
power stations throughout the world. We will also flux changes sinusoidally from a maximum positive
describe a field mapping experiment in which we use flux in (19a) to zero flux in (c) to a maximum negative
our E ⋅ d meter to map the magnetic field of a pair flux in (d) to zero in (e). In (18c), we have shown the
of Helmholtz coils. In the next chapter Faraday’s law vector A representing the area of the coil (A points
is used to explain the operation of transformers and
inductors that are common circuit elements in radio and rotating coil
television sets. of area A
ω
coil of wire
B a) θ=0 ΦB = BA
ω
magnet magnet
B

a) end view of a coil of wire

θ Φ B = B.A
rotating in a magnetic field b) small θ
= BA cosθ
B
rotating coil
of area A A
π
c) θ= ΦB = 0
2
magnet
B

V d) θ = 2π Φ B = –BA
b) top view showing the coil of area A
B
A

e) θ = 3π ΦB = 0
2
c) Vector A representing the
area of the loop B

Figure 18 Figure 19
An electric generator consists of a coil The changing magnetic flux through the rotating loop.
of wire rotating in a magnetic field. The general formula for Φ B is B A cos θ where θ is the
angle shown in (b), between the magnetic field and the
normal to the loop. If the coil is rotating uniformly,
then θ = ω t , and Φ B = B A cos ω t
30-22 Faraday's Law

perpendicular to the plane of the coil) and we can use Equation 26 predicts that the voltage amplitude V0
our usual formula for magnetic flux to get produced by an N turn coil of area A rotating in a
magnetic field B is
ΦB = B⋅A = BAcos θ (22)
V0 = ωNBA (27)
If the coil is rotating at a constant angular velocity ω ,
then θ = ωt and we have where the angular frequency ω radians per second is
related to the frequency f cycles per second and the
ΦB = BAcos ωt (23) period T seconds per cycle by

Differentiating Equation (23) with respect to time rad rad cycle rad 1
gives ω sec = 2π cycle × f sec
= 2π cycle × sec
T cycle
dΦB
= –ωBA sinωt (24)
dt Exercise 5
Suppose that you have a magnetic field B = 1 tesla, and
Finally we use Faraday’s law in the form you rotate the coil at 60 revolutions (cycles) per second.
Design a generator that will produce a sine wave
dΦB voltage whose amplitude is 120 volts.
V = – (15)
dt
Exercise 6
to predict that the voltage V on the voltmeter will be Figures (21a,b) show the voltage produced by a coil of
wire rotating in a uniform magnetic field of a fairly large
V = ωBA sin ωt (25) electromagnet. (The setup is similar to that shown in
Figures 18 and 19.) The coil was square, 4 cm on a side,
If we use a coil with N turns, we get a voltage N times and had 10 turns. To go from the results shown in Figure
as great, or (21a) to those shown in Figure (21b), we increased the
rotational speed of the motor turning the coil. In both
V = ωNBA sin ωt = V0 sin ωt (26) diagrams, we have selected one cycle of the output
wave, and see that the frequency has increased from 10
where V0 is the amplitude of the sine wave as shown in cycles per second to nearly 31 cycles per second.
Figure (20). Equation (26) shows that by rotating a coil
in a magnetic field, we get an alternating or “AC” a) Explain why the amplitude of the voltage signal
voltage. Power stations use this same principle to increased in going from Figure (21a) to (21b). Is the
increase what you expected?
generate AC voltages.
b) Calculate the strength of the magnetic field of the
electromagnet used. Do you get the same answer using
Vo
Figure (21a) and using Figure (21b)?

–V
o
T

Figure 20
Amplitude and period of a sine wave.
 30-23

Gaussmeter
Exercise 6 demonstrates one way to measure the
strength of the magnetic field of a magnet. By spinning
a coil in a magnetic field, we produce a voltage ampli-
tude given by Equation 27 as V0 = ωNBA . Thus by
measuring V0 , ω, N, and A, we can solve for the
magnetic field B.

A device designed to measure magnetic fields is called

a gaussmeter. A commercial gaussmeter, used in our
plasma physics lab, had a small coil mounted in the tip
of a metal tube as shown in Figure (22). A small motor
also in the tube spun the coil at high speed, and the
a) amplitude V0 of the coil voltage was displayed on a
meter. The meter could have been calibrated using
Equation (27), but more likely was calibrated by insert-
ing the spinning coil into a known magnetic field.

In an attempt to measure the magnetic field in the

Helmholtz coils used for our electron gun experiments,
students have also built rotating coil gaussmeters.
Despite excellent workmanship, the results were uni-
formly poor. The electrical noise generated by the
sliding contacts and the motor swamped the desired
signal except when B was strong. This approach turned
out not to be the best way to measure B in the Helmholtz
coils.

b) rotating coil
Figure 21 motor
Voltage output from a coil rotating in a uniform
magnetic field. The coil was 4 cm on a side, and
had 10 turns. In each figure we have selected
one cycle of the output wave, and see that the
frequency of rotation increased from 10 cycles
per second in a) to nearly 31 cycles per
second in b).
IIIIIIIIIII
meter

Figure 22
A commercial gauss meter, which measures the
strength of a magnetic field, has a motor and a
rotating coil like that shown in Figure 18. The
amplitude V0 of the voltage signal is displayed on
a meter that is calibrated in gauss.
30-24 Faraday's Law

A Field Mapping Experiment This is essentially the same formula we had for the
To measure the magnetic field in the Helmholtz coils, rotating coil gaussmeter, Equation (27). The differ-
it is far easier to “rotate the field” than the detector loop. ence is that by “rotating the field” rather than the coil,
That is, use an alternating current in the Helmholtz we avoid sliding contacts, motors, electrical noise, and
coils, and you will get an alternating magnetic field in can make very precise measurements.
the form
A feature of Equation (31) that we did not have when
B = B 0 sin ωt (28) we rotated the coil is the dot product B0 ⋅A. When the
where w is the frequency of the AC current in the coils. detector coil is aligned so that its area vector A (which
Simply place a stationary detector loop in the magnetic is perpendicular to the plane of the detector coil) is
field as shown in Figure (23) and the magnetic flux parallel to B0 , the dot product B0 ⋅A is a maximum.
through the detector loop will be Thus we not only measure the magnitude of B0 , we also
get the direction by reorienting the detector coil until
ΦB = B⋅A = B0 ⋅A sin ωt (29) the V0 is a maximum.

where A is the area of the detector loop. By Faraday’s As a result, a small coil attached to an oscilloscope,
law, the voltage in the voltmeter or oscilloscope at- which is our E⋅d meter, can be used to accurately
tached to the detector loop is given by map the magnitude and direction of the magnetic field
of the Helmholtz coils, or of any coil of wire. Unlike
dΦB
V = – = – ωB0 ⋅A cos ωt (30) our earlier electric field mapping experiments, there
dt are no mysteries or unknown constants. Faraday’s law,
If our detector loop has N turns of wire, then the voltage through Equation (31), gives us a precise relation
will be N times as great, and the amplitude V0 we see between the observed voltage and the magnetic field.
on the oscilloscope screen will be The experimental setup is seen in Figure (24).

Still another way to measure magnetic fields is illus-

V0 = Nω B0 ⋅A (31) trated in Exercise 7.

Helmholtz coils 10 turn

loop

2
1 cm
B area
θ

small
stick

V detector loop

oscilloscope
Figure 23 Figure 24
If you use an alternating current in the Helmholtz coils, Experimental setup for the magnetic field mapping
then B has an alternating amplitude B = B0 cos ω t . experiment. A 60 cycle AC current is running through
You can then easily map this field with the detector loop the Helmholtz coils, producing an alternating magnetic
shown above. If you orient the loop so that the signal on flux through the 10 turn search coil. The resulting
the oscilloscope is a maximum, then you know that B is induced voltage is seen on the oscilloscope screen.
perpendicular to the detector loop and has a magnitude
given by
V = V0 sin ω t = dΦ B / dt = d / dt NABcos ω t .
 30-25

magnets
Exercise 7
The point of this experiment is to determine the strength
of the magnetic field produced by the small magnets
that sat on the angle iron bars in the velocity detector
apparatus. We placed a short piece of wood between
two magnets so that there was a small gap between the magnetic 1 turn coil
ends as seen in the actual size computer scan of Figure field
(25). The pair of magnets were then suspended over air cart
the air track as shown in Figure (26).
air track
On top of the air cart we mounted a single turn coil.
When the air cart passes under the magnets, the single
turn coil passes through the lower gap between the
stationary
magnets as shown. The dimensions of the single turn 1 turn
magnets
coil
coil are shown in Figure 27. We also show the dimen-
sions and location of the lower end of one of the magnets v
at a time when the coil has passed part way through the
gap. You can see that, at this point, all the magnetic flux moving air cart
across the lower gap is passing completely through the
single turn coil. Figure 28 is a recording of the induced
voltage in the single turn coil as the coil passes com-
Figure 26
pletely through the gap. The left hand blip was pro-
A single turn coil, mounted on an air cart, moves
duced when the coil entered the gap, and the right hand through the lower gap between the magnets.
blip when the coil left the gap. The air track was
horizontal, so that the speed of the air cart was constant
as the coil moved through the gap. Determine the
strength of the magnetic field B in the gap. Show and
explain your work.

Figure 25
Two C
Magnets with
wood spacer.

end of magnet
1 turn coil Figure 28
2.5 cm .95 cm Voltage induced in the single turn coil.
2.54 cm

25.3 cm

Figure 27
Dimensions of the single turn coil. We also
show the dimensions of the end of the
magnets through which the coil is passing.
30-26 Faraday's Law

1" diameter
Exercise 8 V1 i1(t) inner coil
As shown in Figure (29), we started with a solenoid with
219 turns wrapped in a 1" diameter plastic tube. The coil R = .1Ω
is 45.4 cm long. The current going through the coil first
goes through a .1 Ω resistor. By measuring the voltage 150 turns
V1 across that resistor, we can determine the current
through the solenoid. V1 is shown as the lower curve in
Figure (30). 45.4 cm V2

a) Using V1 from Figure 30, calculate the magnitude B

of the magnetic field in the solenoid.
219 turns
We then wound 150 turns of wire around the center
section of the solenoid, as indicated in Figure (29). You
can see that the entire flux Φ1 of the Magnetic field of the
solenoid, goes through all the turns of the outer coil.
B
Figure 29
b) Use this fact to predict the voltage V2 across the outer
The inner (primary) coil 1 is 45.4 cm long, has 219
coil, and then compare your prediction with the experi-
turns and is wound on a 2.54 cm (1") diameter tube.
mental V2 shown in the upper curve of Figure (30). The outer (secondary) coil consists of 150 turns
wound tightly around the center section of the primary
coil. The current through the primary coil goes
through a .1Ω resistor, and the voltage V1 is measured
across that resistor. V2 is the voltage induced in the
secondary coil.

voltage V2 across
the outer coil

both voltages are

to the same scale

voltage V1 across
the .1Ω resistor

Figure 30
The voltage V1 across the .1Ω resistor measures the current in the primary (219
turn) coil. V2 is the voltage induces in the secondary (outer 150 turn) coil.
Chapter 31
Induction and
Magnetic Moment

CHAPTER 31 INDUCTION AND MAG-

NETIC MOMENT
In this chapter we discuss several applications of
Faraday’s law and the Lorentz force law. The first is
to the inductor which is a common electronic circuit
element. We will pay particular attention to a circuit
containing an inductor and a capacitor, in which an
electric current oscillates back and forth between the
two. Measurements of the period of the oscillation and
dimensions of the circuit elements allows us to predict
the speed of light without looking at light. Such a
prediction leads to one of the basic questions faced by
physicists around the beginning of the 20th century:
who got to measure this predicted speed? The answer
was provided by Einstein and his special theory of
relativity.

In the second part of this chapter we will discuss the

torque exerted by a magnetic field on a current loop,
and introduce the concept of a magnetic moment.
This discussion will provide some insight into how the
presence of iron greatly enhances the strength of the
magnetic field in an electromagnet. However the main
reason for developing the concept of magnetic moment
and the various magnetic moment equations is for our
later discussion of the behavior of atoms and elemen-
tary particles in a magnetic field. It is useful to clearly
separate the classical ideas discussed here from the
quantum mechanical concepts to be developed later.
31-2 Inductors and Magnetic Moment

THE INDUCTOR produces a nearly uniform magnetic field inside the

In our discussion of Faraday’s law and the betatron in coil whose strength is given by the formula
Chapter 30, particularly in Figure (30-15), we saw that B = µ 0ni (29-31)
an increasing magnetic field in the core of the betatron
and whose direction is given by the right hand rule as
creates a circular electric field around the core. This
shown in the side view, Figure (1a).
electric field was used to accelerate the electrons.
If the coil has a cross-sectional area A, as seen in the top
A more common and accessible way to produce the
view Figure (1b), then the amount of magnetic flux ΦB
same circular electric field is by turning up the current
“flowing” up through the coil is given by
in a solenoid as shown in Figure (1). As we saw in our
discussion of Ampere’s law in Chapter 29, a current i ΦB = BA = µ 0nAi (1)
in a long coil of wire with n turns per unit length, And if we are increasing the current i in the coil, then
B the rate of increase of this flux is (since µ0, n and A are
constants)
dΦB di
= µ 0nA (2)
dt dt
It is the changing magnetic flux that creates the circular
electric field E shown in Figure (1b).
B = µoni

positive
path
E
i

a) side view of coil and magnetic field

Bup
increasing
area A E
of coil
right hand rule
for positive path
Bup
increasing

d ΦB
E⋅d = – (Faraday's Law)
dt
b) top view showing the electric field
Figure 2 Sign conventions
surrounding the increasing magnetic flux
We start by defining up, out of the paper, as the positive
Figure 1 direction. Then use the right hand rule to define a
When we turn up the current in a solenoid, we increase positively oriented path. As a result, counter clockwise
the magnetic field and therefore the magnetic flux up is positive, clockwise is negative. With these
through the coil. This increasing magnetic flux is the conventions, dΦ B / dt is positive for an increasing
source of the circular electric field seen in the top view. upward directed magnetic flux. In calculating the line
integral E ⋅ d , we go around in a positive direction,
counter clockwise. Everything is positive except the –
sign in Faraday's law, thus the electric field goes
around in a negative direction, clockwise as shown.
 31-3

In Figure (2) we have shown the top view of the Direction of the Electric Field
solenoid in Figure (1) and added in the circular electric In Figure (2) and in the above exercise, we saw that an
field we would get if we had an increasing magnetic increasing magnetic flux in the coil created a clockwise
flux up through the solenoid. We have also drawn a circular electric field both inside and outside the wire as
circular path of radius r around the solenoid as shown. shown in Figure (3). In particular we have a circular
If we calculate the line integral E⋅d for this closed electric field at the wire, and this circular electric field
path, we get by Faraday’s law will act on the charges carrying the current in the wire.
dΦB
E⋅d = – (30-17) To maintain our sign conventions, think of the current
dt in the wire as being carried by the flow of positive
di charge. The up directed magnetic field of Figure (3)
E×2πr = – µ0nA (3) will be produced by a current flowing counterclock-
dt
wise as shown (right hand rule). In order to have an
where the integral E ⋅ d is simply E times the cir- increasing flux, this counterclockwise current must be
cumference of the circle, and we used Equation (2) for increasing.
dΦB /dt.
We saw that the electric field is clockwise, opposite to
The minus sign in Equation (3) tells us that if we use a the direction of the current. We are turning up the
positive path as given by the right hand rule, and we are current to increase the magnetic field, and the electric
increasing the flux up through this path, then E⋅d field is opposing the increase.
must be negative. I.e., the electric field must go
If we already have a current in a solenoid, already have
clockwise, opposite to the positive path. (Do not worry
an established B field and try to decrease it, di/dt is
too much about signs in this discussion. We will
negative for this operation, and we get an extra minus
shortly find a simple, easily remembered, rule that tells
sign in Equation (3) that reverses the direction of E. As
us which way the electric field points.)

Equation (3) tells us that the strength E of the circular E

i
field is proportional to the rate of change of current i in
the solenoid, and drops off as 1/r if we are outside the
solenoid. In the following exercise, you are to show
that we also have a circular field inside the solenoid, a B up
increasing
field that decreases linearly to zero at the center.

Exercise 1
Use Faraday’s law to calculate the electric field inside increasing current
the solenoid. Note that for a circular path of radius r i creates the increasing
inside the solenoid, the flux ΦB through the path is magnetic flux
proportional to the area of the path and not the area A Figure 3
of the solenoid. If the sign conventions described in Fig. 2 seemed too
arbitrary, here is a physical way to determine the
The calculation of the circular electric field inside and direction of E . The rule is that the electric field E
outside a solenoid, when i is changing, is a good opposes any change in the current i. In this case, to
example of the use of both Ampere’s law to calculate B create an increasing upward directed magnetic flux,
and Faraday’s law to calculate E . It should be saved in the current i must be flowing counter clockwise as
your collection of good examples. shown, and be increasing. To oppose this increase, the
electric field must be clockwise.
31-4 Inductors and Magnetic Moment

a result we get a counterclockwise electric field that Induced Voltage

exerts a force in the direction of i. Thus when we try to We have just seen that the changing magnetic flux in a
decrease the current, the electric field tries to maintain solenoid creates an electric field that acts on the current
it. in the solenoid to oppose the change in the current.
From Equation (3), we see that the formula for the line
There is a general rule for determining the direction of integral of this electric field around one loop of the coil
the electric field. The electric field produced by the is given by
changing magnetic flux always opposes the change.
di
If you have a counterclockwise current and increase it, E ⋅ d = – µ 0nA (4)
you will get a clockwise electric field that opposes the dt
increase. If you have a counterclockwise current and where the path is at the wire as shown in Figure (4). The
decrease it you get a counterclockwise electric field n in Equation (4), which comes from the formula for the
that opposes the decrease. If you have a clockwise magnetic field of a solenoid, is the number of turns per
current and try to increase it, you get a counter clock- unit length in the solenoid.
wise electric field that opposes the increase, etc. There
are many possibilities, but one rule—the electric field In our discussion of the betatron, we saw that the
always opposes the change. circular electric field accelerated electrons as they went
around the evacuated donut. Each time the electrons
went around once, they gained an amount of kinetic
energy which, in electron volts, was equal to E⋅d .
In our discussion of the electron gun, we saw that using
a battery of voltage Vacc to accelerate the electrons,
E positively oriented
path inside wire for produced electrons whose kinetic energy, in electron
calculating E ⋅ d volts, was equal to Vacc . In other words, the circular
current i
in wire
electric field can act like a battery of voltage
Vacc = E⋅d .
Bup
increasing
When acting on the electrons in one loop of wire, the
circular electric field produces a voltage change ∆V1
given by
change in electric
∆V1 = E⋅d voltage in one (5a)
turn of the coil

Figure 4 If we have a coil with N turns as shown in Figure (5),

The electric field penetrates the wire, opposing the then the change in voltage ∆VN across all N turns is N
change in the current i. The line integral E⋅d times as great, and we have
around the coil is just equal to the change in voltage
∆V 1 around each turn of the coil. change in electric
∆VN = N E⋅d voltage in N (5b)
turns of the coil
 31-5

Using Equation 4 for the E⋅d for a solenoid, we see Inductance

that the voltage change ∆VN across the entire solenoid If you take a piece of insulated wire, tangle it up in any
has a magnitude way you want, and run a current through it, you will get
di an induced voltage Vinduced that is proportional to the
∆VN = N E⋅d = µ0NnA (6) rate of change of current di/dt, and directed in a way
dt
such that it opposes the change in the current. If we
where N is the total number of turns, n = N/h is the designate the proportionality constant by the letter L,
number of turns per unit length, A is the cross-sectional then the relationship between Vinduced and di/dt can be
area of the solenoid, and i the current through it. written
To get the correct sign of ∆VN , to see whether we have inducedvoltages
a voltage rise or a voltage drop, we will use the rule that di
Vinduced = L are proportional
the circular electric field opposes any change in the dt to di/dt
(7)
current. This rule is much easier to use than trying to
keep track of all the minus signs in the equations. The constant L is called the inductance of the coil or
tangle of wire. In the MKS system, inductance has the
In summary, Equation (6) is telling us that if you try to dimension of volt seconds/ampere, which is called a
change the amount of current flowing in a solenoid, if henry.
di dt is not zero, then a voltage will appear across the
ends of the solenoid. The voltage has a magnitude Comparing Equations (6) and (7), we immediately
proportional to the rate di dt that we are trying to obtain the formula for the inductance of a solenoid
change the current, and a direction that opposes the 2
change. It is traditional to call this voltage ∆VN the µ0N A inductanceof
induced voltage. One says that the changing magnetic L = µ0 NnA = a solenoid
(8)
h
flux in the coil induces a voltage. Such a coil of wire
is often called an inductor. where N is the number of turns in the solenoid, A the
cross-sectional area and h the length. In the middle
E
i term, n = N/h is the number of turns per unit length.
i
coil of
length h
∆VΝ
(voltage h
across n = N/h is the
coil) number of turns
per unit length

N turns, area A
Figure 5
Our standard coil with N turns, area A and length h.
If you try to increase the current i in the coil, you get
an opposing voltage.
31-6 Inductors and Magnetic Moment

Example 1 The toroidal Inductor With

The simplest solenoid we can use is a toroidal one, like
that shown in Figure (6), where the magnetic field is µ0N2 A
L =
completely confined to the region inside the coil. h
Essentially, the toroid is an ideal solenoid (no end we get
effects) of length h = 2πR. To develop an intuitive
–6 2 –4
feeling for inductance and the size of a henry, let us 1.26×10 × 696 × 5.31 ×10
calculate the inductance of the toroidal solenoid shown L =
1.35
in the photograph of Figure (6b). This solenoid has 696
turns and a radius of R = 21.5 cm. Each coil has a radius = 2.40 ×10 – 4 henry
of r = 1.3 cm. Thus we have
We see that even a fairly big solenoid like the one
N = 696 turns shown in Figure (6) has a small inductance at least
when measured in henrys. At the end of the chapter we
R = 21.5cm
will see that inserting an iron core into a solenoid
h = 2πR = 2π *.215 = 1.35m greatly increases the inductance. Inductances as large
r = 1.3cm or larger than one henry are easily obtained with iron
A = πr2 = π×.013 2 = 5.31 × 10 – 4 m2 core inductors.
µ0 = 1.26 × 10 – 6 henry/m

2r

h = 2πR

Figure 6b
Photograph of the toroidal solenoid used in various
experiments. Although the coil looks big, the
R inductance is only 2.40× 10– 4 henry. (If you put iron
inside the coil, you could greatly increase the
inductance, but you would not be able to calculate
its value.)

Figure 6a
A toroid is an ideal solenoid of length h = 2π R .
 31-7

INDUCTOR AS A CIRCUIT i
ELEMENT
Because a changing electric current in a coil of wire VR = i R
a) R
produces a voltage rise, small coils are often used as
circuit elements. Such a device is called an inductor,
and the symbol used in circuit diagrams is a sketch of
a solenoid and usually designated by the
i
symbol L. The voltage rise across the three circuit
elements we have considered so far are
VR = iR resistor (27-8) + VC = Q
b) C
– C
Q
VC = capacitor (27-31)
C
i (increasing)
di
VL = L inductor (7)
dt
As shown in Figure (7) the direction of the rise is c) L VL = L di
dt
opposite to the current in a resistor, toward the positive
charge in a capacitor, and in a direction to oppose a
change in the current i in an inductor. In (c), we are
Figure 7
showing the direction of the voltage rise for an increas- The resistor R, capacitor C and
ing current. The voltage in the inductor is opposing an inductor L as circuit elements.
increase in the current, just as the voltage in the resistor
(a) opposes the current i itself.
31-8 Inductors and Magnetic Moment

The LR Circuit When we open the switch of Figure (9), the battery and
We will begin our discussion of the inductor as a circuit resistor R1 are immediately disconnected from the
element with the LR circuit shown in Figure (8). circuit, and we have the simple LR circuit shown in
Although this circuit is fairly easy to analyze, it is a bit Figure (8). Everything changes instantly except the
tricky to get the current i started. One way to start the current i in the inductor. The inductor instantly sets
current is shown in Figure (9) where we have a battery up a voltage VL to oppose any change in the current.
and another resistor R1 attached as shown.
Figure (10) is a recording of the voltage VL across the
When the switch of Figure (9) has been closed for a inductor, where the switch in Figure (9) is opened at
while, we have a constant current i 0 that flows down time t = 0. Before t = 0, we have a constant current i 0
out of the battery, through the resistor R1, around up and no voltage VL . When the switch is opened the
through the inductor and back to the battery. Because voltage jumps up to VL = V0 and then decays expo-
the current is constant, di 0 /dt = 0 and there is no nentially just as in the RC circuit. What we want to do
voltage across the inductor. When we have constant is apply Kirchoff’s law to Figure (8) and see if we can
“DC” currents, inductors act like short circuits. That is determine the time constant for this exponential decay.
why the current, given the choice of going up through
switch
the inductor L or the resistor R, all goes up through L. i0
(To say this another way, since the voltage across L is
zero, the voltage VR across R must also be zero, and the – Vb
current i R = VR/R = 0.) Since R1 is the only thing +
L R
that limits the current in Figure (9), i 0 is given by R1
VB
i0 = (9) i0 i0
R1
Figure 9
Equation (9) tells us that we have a serious problem if To get a current started in an LR circuit, we begin with
we forget to include the current limiting resistor R1. the extra battery and resistor attached as shown. With
the switch closed, in the steady state all the current i0
i flows up through the inductor because it (theoretically)
has no resistance. When the switch is opened, the
battery is disconnected, and we are left with the RL
L di L R iR circuit starting with an initial current i0 .
dt

Figure 8
The LR circuit. If we have a decreasing current, the
voltage in the inductor opposes the decrease and
creates a voltage that continues to push the current
through the resistor. But, to label the voltages for
Kirchoff's law, it is easier to work with positive
quantities. I.e. we label the circuit as if both i and di/dt
were positive. With a positive di/dt, the voltage on the
inductor opposes the current, as shown.
 31-9

If we walk around the circuit of Figure (8) in the Differentiating Equation (11) to get di/dt, we have
direction of i, and add up the voltage rises we encoun-
ter, and set the sum equal to zero (Kirchoff’s law), we di
= – αi 0 e –αt (12)
get dt

di and substituting (11) and (12) in Equation (10) gives

– iR – L = 0
dt R –αt
–αt
– αi 0 e + i e = 0 (13)
di R L 0
+ i = 0 (10)
dt L
In Equation (13), i 0 and e-α t cancel and we get
Equation (10) is a simple first order differential equa- α = R/L
tion for the current i. We guess from our experimental Equation (11) for i becomes
results in Figure (10) that i should be given by an
exponential decay of the form – t ( L R)
i = i0 e (
– R L
L)t R
= i0 e
i = i 0 e –αt (11) (14)
–t
≡ i0 e T

We see from Equation (14) that the time constant T for

the decay is

time constant
L
T = = for the decay (15)
R
of an LR circuit

Everything we said about exponential decays and time

constants for RC circuits at the end of Chapter (27)
applies to the LR circuit, except that the time constant
is now L/R rather than RC.

Exercise 2
Figure 10
The LR circuit that produced the experimental results
Experimental recording of the voltage in an RL
circuit. We see that once the switch of Fig. 9 is shown in Figure (10) had a resistor whose resistance R
opened, the voltage across the inductor jumps from was 15 ohms. Quickly estimate the inductance L. (You
zero to V0 = i0R . This voltage on the inductor is should be able to make this estimate accurate to within
trying to maintain the current now that the battery about 10% simply by sketching a straight line on the
is disconnected. The voltage and the current then graph of Figure 10.) Compare your result with the
die with an exponential decay. (For this experiment, inductance of the toroidal solenoid discussed in Figure
we used the toroidal inductor of Figure 6, with (6) on page 6.
R = 15Ω , R1 = 4Ω , and Vb = 2.5 volts.)
31-10 Inductors and Magnetic Moment

THE LC CIRCUIT Finally use Equation (17) i = dQ/dt and we get the
The next circuit we wish to look at is the LC circuit second order differential equation
shown in Figure (11). All we have done is replace the
resistor R in Figure (8) with a capacitor C as shown. It d 2i 1
2
+ i = 0 (18)
does not seem like much of a change, but the behavior dt LC
of the circuit is very different. The exponential decays
we saw in our LR and RC circuits occur because we are The fact that we get a second order differential equation
losing energy in the resistor R. In the LC circuit we (with a second derivative of i) instead of the first order
have no resistor, no energy loss, and we will not get an differential equations we got for LR and RC circuits,
exponential decay. To see what we should get, we will shows that we have a very different kind of problem. If
apply Kirchoff’s law to the LC circuit and see if we can we try an exponential decay in Equation (18), it will not
guess the solution to the resulting differential equation. work.

Walking clockwise around the circuit in Figure (11) Exercise 3

and setting the sum of the voltage rises to zero, we get Try the solution i = i 0e– α t in Equation (18) and see
Q di what goes wrong.
– –L = 0
C dt
We have previously seen a second order differential
di Q (16) equation in just the form of Equation (18) in our
+ = 0
dt LC discussion of simple harmonic motion. We expect a
sinusoidal solution of the form
The problem we have with Equation (16) is that we
have two variables, i and Q, and one equation. But we i = i 0 sin ωt (19)
had this problem before in our analysis of the RC
In order to try this guess, Equation (19), we differenti-
circuit, and solved it by noticing that the charge Q on the
ate twice to get
capacitor is related to the current i flowing into the
capacitor by di
= ωi 0 cos ωt
dt
dQ
i = (17)
dt d 2i
2
= – ω 2 i 0 sin ωt (20)
If we differentiate Equation (16) once with respect to dt
time to get and substitute Equation (20) into (18) to get
d2i 1 dQ i0
2
+ = 0 – ω 2 i 0 sin ωt + sin ωt = 0 (21)
dt LC dt LC
The quantity i 0 sin ωt cancels from Equation (21) and
we get
i
1 1
ω2 = ; ω = (22)
LC LC
L di L C
Q
dt C

We see that an oscillating current is a solution to

Kirchoff’s law, and that the frequency ω of oscillation
Figure 11
The LC circuit. This is the same as the LR is determined by the values of L and C.
circuit of Figure (8), except that the resistor
has been replaced by a capacitor.
 31-11

Exercise 4 i
In Figure (12) we have an LC circuit consisting of a
toroidal coil shown in Figure (6) (on page 31-6), and the
parallel plate capacitor made of two aluminum plates L C
with small glass spacers. The voltage in Figure (12c) is
oscillating at the natural frequency of the circuit.

a) What is the capacitance of the capacitor?

b) The aluminum plates have a radius of 11 cm. a) The LC circuit

Assuming that we can use the parallel plate capacitor
formula
ε0 AC
C =
d

where AC is the area of the plates, estimate the thick-

ness d of the glass spacers used in this experiment.

(The measured value was 1.56 millimeters. You should

get an answer closer to 1 mm. Errors could arise from
fringing fields, effect of the glass, and non-uniformity of
the surface of the plates.)

b) Inductor and capacitor used in the experiment

c) Oscillating voltage at the resonant frequency.

Figure 12
Oscillating current in an LC circuit consisting of the
toroidal inductor of Fig. 6 and a parallel plate
capacitor. We will discuss shortly how we got the
current oscillating and measured the voltage.
31-12 Inductors and Magnetic Moment

Intuitive Picture of the LC Oscillation Finally, in (f), enough positive charge has built up on
The rather striking behavior of the LC circuit deserves the bottom of the capacitor to stop the flow of the
an attempt at an intuitive explanation. The key to current. In (g) the current reverses and the capacitor
understanding why the current oscillates lies in under- begins to discharge. The inductor supplies the inertia
standing the behavior of the inductor. As we have to keep this reversed current going until the capacitor is
mentioned, the voltage rise on an inductor is always in charged the other way in (i). But this is the same picture
a direction to oppose a change in current. The closest as (a), and the cycle begins again.
analogy is the concept of inertia. If you have a massive
object, a large force is required to accelerate it. But This intuitive picture allows you to make a rough
once you have the massive object moving, a large force estimate of how the frequency of the oscillation should
is required to stop it. depend upon the size of the inductance L and capaci-
tance C. If the inductance L is large, the current has
An inductor effectively supplies inertia to the current more inertia, it will charge up the capacitor more, and
flowing through it. If you have a large inductor, a lot should take longer. If the capacitance C is larger, it
of work is required to get the current started. But once should take longer to fill up. In other words, the period
the current is established, a lot of effort is required to should be longer, the frequency ω lower, if either L or
stop it. In our LR circuit of Figure (10), once we got a C are increased. This is consistent with the result
current going through the inductor L, the current con- ω = 1/ LC we saw in Equation (22).
tinued to flow, even though there was no battery in the
circuit, because of the inertia supplied to the current by Before leaving Figure (13) go back over the individual
the inductor. sketches and check two things. First, verify that
Kirchoff’s law works for each stage; i.e., that the sum
Let us now see why an LC circuit oscillates. One cycle of the voltage rises around the circuit is zero for each
of an oscillation is shown in Figure (13) where we stage. Then note that whenever there is a voltage VLon
begin in (a) with a current flowing up through the the inductor, the direction of VL always opposes the
inductor and over to the capacitor. The capacitor change in current.
already has some positive charge on the upper plate and i=0
the current is supplying more. The capacitor voltage i + + – –
a) VL VC f) VL + + VC
VC is opposing the flow of the current, but the inertia – –

supplied to the current by the inductor keeps the current

flowing. i=0
– –
In the next stage, (b), so much charge has built up in the b) VL + +
– –
VC g) VL + + VC
i
capacitor, VC has become so large, that the current
stops flowing. Now we have a charged up capacitor
which in (c) begins to discharge. The current starts to
i
flow back down through tin inductor. The current c) VL + + VC h) L C
– –
continues to flow out of the capacitor until we reach (d) i
where the capacitor is finally discharged. di = 0; Q = 0
dt
The important point in (d) is that, although the capacitor i i + +
d) L C i) VL VC
is empty, we still have a current and the inductor gives – –
the current inertia. The current will continue to flow di = 0; Q = 0
even though it is no longer being pushed by the dt
capacitor. Now in (e), the continuing current starts to i – –
charge the capacitor up the other way. The capacitor e) VL + + VC
voltage is trying to slow the current down but the
inductor voltage keeps it going. Figure 13
The various stages in the oscillation of
the electric charge in an LC circuit.
 31-13

The LC Circuit Experiment off the spring or some form of dampening or energy
The oscillation of the LC circuit in Figure (13) is a loss comes into play.
resonance phenomena and the frequency
It is clear from Figure (14) how to drive the motion of
ω 0 = 1/ LC is the resonance frequency of the
circuit. If we drive the circuit, force the current to a mass on a spring; just oscillate our hand up and down.
oscillate, it will do so at any frequency, but the response But how do we drive the LC circuit? It turns out that for
is biggest when we drive the circuit at the resonant the parallel plate capacitor and air core toroidal induc-
frequency ω 0. tor we are using, the resonance is so delicate that if we
insert something into the circuit to drive it, we kill the
There turns out to be a very close analogy between the resonance. We need a way to drive it from the outside,
LC circuit and a mass hanging on a spring as shown in and an effective way to do that is shown in Figure (15).
Figure (14). The amplitude of the current in the circuit
is analogous to the amplitude of the motion of the mass.
If we oscillate the upper end of the spring at a low oscillator
frequency ω much less than the resonant frequency ω 0,
i = i 0 sin ωt
the mass just moves up and down with our hand. If ω
is much higher than ω 0, the mass vibrates at a small
amplitude and its motion is out of phase with the
motion of our hand. I.e., when our hand comes down,
the spring comes up, and vice versa. But when we L C
oscillate our hand at the resonant frequency, the ampli-
tude of vibration increases until either the mass jumps
scope
Run a wire from an oscillator around the coil and
back to the oscillator. Do the same for the scope.
ω

B
m
magnetic field created
by the current from the
oscillator
Figure 14 Figure 15
To get the mass on the end of a spring Driving the LC circuit. The turns of wire from the
oscillating at some frequency ω you move your oscillator produce an oscillating magnetic field inside
hand up and down at the frequency ω . If ω is the coil. This in turn produces an electric field at the
the resonant frequency of the mass and spring coil wires which also oscillates and drives the current
system, the oscillations become quite large. in the coil. (A second wire wrapped around the coil is
used to detect the voltage. The alternating magnetic
field in the coil produces a voltage in the scope wire.)
31-14 Inductors and Magnetic Moment

In that figure we have taken a wire lead, wrapped it which drives the current in the LC circuit. We can
around the toroidal coil a couple of times, and plugged change the driving frequency simply by adjusting the
the ends into an oscillator as shown. (Some oscillators oscillator.
might not behave very well if you short them out this
way. You may have to include a series resistor with the To detect the oscillating current in the coil, we wrap
wire that goes to the oscillator.) When we turn on the another wire around the coil, and plug that into an
oscillator, we get a current iosc = i 0sinωt in the wire, oscilloscope. The changing magnetic flux in the coil
where the frequency ω is determined by the oscillator induces a voltage in the wire, a voltage that is detected
setting. by the scope.

The important part of this setup is shown in Figure In Figure (16a) we carried out the experiment shown in
(15b) where the wire lead wraps a few times around the Figure (15), and recorded the amplitude VC of the
toroid. Since the wire lead itself forms a small coil and capacitor voltage as we changed the frequency ω on the
since it carries a current iosc, it will create a magnetic oscillator. We see that the amplitude is very small until
field Bosc as shown. Part of the field Bosc will lie inside we get to a narrow band of frequencies centered on ω 0,
the toroid and create magnetic flux Φosc down the in what is a typical resonance curve. The height of the
toroid. Since the current producing Bosc is oscillating peak at ω = ω 0 is limited by residual resistance in the
at a frequency ω, the field and the flux will also oscillate LC circuit. Theory predicts that if there were no
at the same frequency. As a result we have an oscillat- resistance, the amplitude at ω = ω 0 would go to
ing magnetic flux in the toroid, which by Faraday’s law infinity, but the wires in the toroid would melt first. In
creates an electric field of magnitude general, however, the less resistance in the circuit, the
E ⋅ d = –d d Φosc dt around the turns of the sole- narrower the peak in Figure (16a), and the sharper the
noid. This electric field induces a voltage in the toroid resonance.
torroidal capacitor plate
0.75 amplitude amplitude N turns radius RT area Ac
in volts peaks at
ω0 = 3.6 106 rad/sec

RT
0.50 d

capacitor
0.25
cross-sectional plate
area AT of the turns separation

2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.2 4.4 106
ω0
resonant radians/sec
frequency

Figure 16a
As we tune the oscillator frequency through
the resonant frequency, the amplitude of the
LC voltage goes through a peak. Figure 16b
The LC apparatus.
 31-15

MEASURING THE SPEED OF LIGHT Finally, recall in our early discussion of magnetism,
The main reason we have focused on the LC resonance that µ 0 ε 0 was related to the speed of light c by
experiment shown in Figure (15) is that this apparatus 1
c2 = (27-18)
can be used to measure the speed of light. We will first µ 0ε 0
show how, and then discuss the philosophical implica-
Using Equation (25) in (27-18), and taking the square
tions of such a measurement.
root gives
The calculation is straightforward but a bit messy. We
start with Equation (22) for the resonance frequency 1 ATAC
ω0 c = = ω 0N (26)
µ0ε 0 2πR Td
1
ω0 =
LC (22)
and then use Equation (8) for the inductance L of a Exercise 5
solenoid Show that c in Equation (26) has the dimensions of a
velocity. (Radians are really dimensionless.)
L = µ 0 N 2 A/h (8)
and Equation (27-32) for the capacitance of a parallel At first sight Equation (26) appears complex. But look
plate capacitor at the quantities involved.

C = ε 0 AC/d (27-32) ω0 = the measured resonant frequency

N = the number of turns in the solenoid
For the apparatus shown in Figure (16b), the length of
the toroidal solenoid is h = 2π R T, and the cross- AT = cross–sectional area of the toroid
sectional area is A = A T , so that Equation (8) be- AC = area of capacitor plates
comes
R T = radius of toroid
L toroid = µ 0 N 2 A T /2πR T (23)
d = separation of capacitor plates
For the capacitor, A C is the area of the plates, d their
separation, and we can use Equation (27-32) as it Although it is a lot of stuff, everything can be counted,
stands. If we square Equation (22) to remove the square measured with a ruler, or in the case of ω 0, determined
root from the oscilloscope trace. And the result is the speed
1 of light c. We have determined the speed of light from
ω 02 = (22a)
LC a table top experiment that does not involve light.
and use Equations (23) and (27-32) for L and C, we get
Exercise 6
2 2 π RT d
ω0 = × The resonant curve in Figure (16a) was measured using
µ 0 N AT ε0AC
2
the apparatus shown in Figure (16b). For an inductor,
we used the toroid described in Figure (6). The parallel
1 2 π R Td (24) plates have a radius of 11 cm, and a separation d =
= × 1.56mm. Use the experimental results of Figure (16a),
µ 0ε 0 2
N ATAC along with the measured parameters of the toroid and
parallel plates to predict the speed of light. (The result
The important point is that the product µ 0 ε 0 appears in is about 20% low due to problems determining the
Equation (24), and we can solve for 1/µ 0 ε 0 to get capacitance, as we discussed in Exercise 3.)

1 = ω 0 N AT AC
2 2
µ 0 ε0 2π R T d (25)
31-16 Inductors and Magnetic Moment

In our initial discussion of the special theory of relativ- Exercise 7

ity in Chapter 1, we pointed out that according to In Figure (17a) we have an LRC series circuit driven by
Maxwell’s theory of light, the speed of light c could be a sinusoidal oscillator at a frequency ω radians/sec.
predicted from a table top experiment that did not The voltage V R is given by the equation
involve light. This theory, developed in 1860, pre-
dicted that light should travel at a speed c = 1/ µ 0 ε 0 , V R = V R0 cos (ωt)
and Maxwell knew that the product µ 0 ε 0 could be
as shown in the upper sketch of Figure (17b).
determined from an experiment like the one we just
described. (Different notation was used in 1860, but Knowing V R , find the formulas and sketch the voltages
the ideas were the same.) for VL and V C . Determine the formulas for the ampli-
tudes V L0 and V C0 in terms of V R 0 and ω .
This raised the fundamental question: if you went out
and actually measured the speed of a pulse of light as
it passed by, would you get the predicted answer
1/ µ 0 ε 0 ? If you did, that would be evidence that you
were at rest. If you did not, then you could use the R VR
difference between the observed speed of the pulse and Figure 17a
1/ µ 0 ε 0 as a measurement of your speed through An LRC circuit driven
by a sinusoidal
space. This was the basis for the series of experiments oscillator. The voltage
performed by Michaelson and Morley to detect the VR across the resistor is L VL
motion of the earth. It was the basis for the rather firm shown in Figure (17b).
conviction during the last half of the 19th century that
the principle of relativity was wrong.
C VC
It was not until 1905 that Einstein resolved the problem
by assuming that anyone who measured the speed of a
pulse of light moving past them would get the answer VR
c = 1/ µ 0 ε 0 = 3×10 8 m/s, no matter how they were
moving. And if everyone always got the same answer
for c, then a measurement of the speed of light could
not be used as a way of detecting one’s own motion and
violating the principle of relativity. The importance of
the LC resonance experiment, of the determination of VR = VR0cos(ωt)
the speed of light without looking at light, is that it
Figure 17b VL
focuses attention on the fundamental questions that Knowing VR , find
lead to Einstein’s special theory of relativity. the formulas and
sketch the voltages
In the next chapter we will discuss Maxwell’s equa- for VL and VC .
tions which are the grand finale of electricity theory. It
was the solution of these equations that led Maxwell to
his theory of light and all the interesting problems that VL = VL0 ...
were raised concerning the principle of relativity.
VC

VC = VC0...
 31-17

The second half of this chapter which discusses the

concept of magnetic moment, provides additional labo-
ratory oriented applications of Faraday’s law and the
Lorentz force law. This topic contains essential back-
ground material for our later discussion of the behav-
ior of atoms and elementary particles in a magnetic
field, but is not required for the discussion of Maxwell’s
equations in the next chapter. You may wish to read
through the magnetic moment discussion to get the
general idea now, and worry about the details when
you need them later.
31-18 Inductors and Magnetic Moment

MAGNETIC MOMENT In order to calculate the force exerted by B on i, we will

We will see, using the Lorentz force law, that when a use our model of a current as consisting of rods of
current loop (a loop of wire with a current flowing in it) charge moving past each other as shown in Figure
is placed in a magnetic field, the field can exert a torque (18b). The rods have equal and opposite charge
on the loop. This has an immediate practical applica- densities Q , and the positive rod is moving at a speed
tion in the design of electric motors. But it also has an v to represent a positive current. The current i is the
impact on an atomic scale. For example, iron atoms act amount of charge per second carried past any cross-
like current loops that can be aligned by a magnetic sectional area of the wire. This is the amount of charge
field. This alignment itself produces a magnetic field per meter, Q , times the number of meters per second,
and helps explain the magnetic properties of iron. On v, passing the cross-sectional area. Thus
a still smaller scale elementary particles like the elec-
tron, proton, and neutron behave somewhat like a Q coulombs meter
i = ×v
current loop in that a magnetic field can exert a torque meter second
on them. The phenomena related to this torque, al-
Q coulombs (27)
though occurring on a subatomic scale, are surprisingly = v
well described by the so called “classical” theory we second
will discuss here.
In Figure (18b) we see that the downward magnetic
field B acts on the moving positive charges to produce
Magnetic Force on a Current a force FQ of magnitude
Before we consider a current loop, we will begin with
a derivation of the force exerted by a straight wire FQ = Q v × B = QvB
carrying a current i as shown in Figure (18a). In that which points toward the top of the page. The force f
figure we have a positive current i flowing to the right on a unit length of the wire is equal to the force on one
and a uniform magnetic field B directed down into the charge Q times the number of charges per unit length,
paper. which is 1 . Thus
x x x x x x x x
force on a
1 QvB
x x x x
i x x x x B(down) unit length ≡ f = FQ × = (28)
of wire
a) current in a magnetic field

FB
FB
x x x x x x x x
v
+ + + + + + + x x x x x x x x
– – – – – – –
x x x x x x x x B(down) i
x x x x x x x x
B(down)
b) model of the neutral current as two rods Figure 19
of charge, the positive rod moving in the Sideways magnetic force on a current in a magnetic
direction of the current i field. The force per unit length f is related to the
charge per unit length λ by f = λ v × B . Since λ v is
Figure 18
the current i , we get f = i × B .
Magnetic force on the moving charges when
a current i is placed in a magnetic field.
 31-19

Using Equation (27) to replace Qv/ by the current i in The resulting force per unit length on B2 is
(28), and noting that f points in the direction of i ×B ,
as shown in Figure (19), we get f = i 2 ×B1

which is directed in toward i1 and has a magnitude

force per unit length
f = i×B exerted by a magnetic (29) µ 0 i1 i 2
fieldB on a current i f = i2B1 = (30)
2πr
where i is a vector of magnitude i pointing in the
Equation (30) is used in the MKS definition of the
direction of the positive current.
ampere and the coulomb. In 1946 the following
definition of the ampere was adopted:
Example 2
Calculate the magnetic force between two straight The ampere is the constant current which, if main-
parallel wires separated by a distance r, carrying paral- tained in two straight parallel conductors of infinite
lel positive currents i1 and i 2 as shown in Figure (21). length, of negligible circular cross section, and placed
1 meter apart in a vacuum, would produce on each of
Solution
these conductors a force equal to 2 × 10 –7 newtons per
The current i1 produces a magnetic field B1 , which meter of length.
acts on i2 as shown in Figure (20) (and vice versa).
Since B1 is the field of a straight wire, it has a Applying this definition to Equation (30), we set
magnitude given by Equation (28-18) as i 1 = i 2 = 1 to represent one ampere currents, r = 1 to
represent the one meter separation, and f = 2 × 10 –7
µ 0 i1 as the force per meter of length. We get
B1 = (29-18)
2πr
µ0
2 × 10 –7 =
2π
i1 From this we see that µ0 is now a defined constant with
r
the exact value

i2
µ0 = 4 π × 10 –7 by definition (31)
a) top view

With the above definition of the ampere, the coulomb

B1 is officially defined by the amount of charge carried by
FB a one ampere current, per second, past a cross-sectional
area of a wire .
i1 i2
(up) (up) Looking back over our derivation of the formula
f = i × B, and then the above MKS definitions, we
see that it is the magnetic force law F = Q v × B which
b) end view showing the magnetic field of current now underlies the official definitions of charge and
i1 exerting a magnetic force on current i 2
current.
The force between parallel currents is attractive.

Figure 20
Force between two currents.
31-20 Inductors and Magnetic Moment

Torque on a Current Loop With this convention, the loop area A points toward the
In an easily performed experiment, we place a square upper left part of the page in Figure (21b) as shown.
loop of wire of sides ( ) and (w) as shown in Figure And we see that the torque caused by the magnetic
(21a), into a uniform magnetic field as shown in (21b). forces, is trying to orient the loop so that the loop
The loop is allowed to rotate around the axis and is now area a is parallel to B . This is a key result we will use
orientated at an angle θ as seen in (21c). often.

If we now turn on a current i, we get an upward To calculate the magnitude of the magnetic torque, we
magnetic force proportional to i × B in the section from note that the magnitude of the force on side (1)-(2) or
point (1) to point (2), and a downward magnetic force side (3)-(4) is the force per unit length f = i × B times
proportional to i × B in the section from point (3) to the length of the side
point (4). These two forces exert a torque about the axis
of the loop, a torque that is trying to increase the angle F1,2 = F3,4 = f = iB
θ . (This torque is what turns the armature of an electric
motor.) When the loop is orientated at an angle θ as shown, then
the lever arm for these forces is
Following our earlier right hand conventions, we will
define the area a of the loop as a vector whose
w
magnitude is the area (a = w) of the loop, and whose lever arm = sin θ
direction is given by a right hand rule for the current in 2
the loop. Curl the fingers of your right hand in the
Since both forces are trying to turn the same way, the
direction of the positive current i and your thumb points
total torque is twice the torque produced by one force,
in the direction of a as shown in Figure (22).
and we have
2 times
w
torque = 2 × sinθ × iB leverarm
2 times force

end of
w axis loop
torque = iB w sin θ (32)

i
a) a wire loop carrying a current i, free to turn on the axis F1,2 = ixB
end view
of loop
F1,2
F1,2 = ixB
(2)
θ
w i
axis (1)
B
(3)
i
(4)
B
F3,4 ( w sin θ)
F3,4 = ixB 2
F3,4
w = width of loop

b) magnetic force acting on horizontal loop c) magnetic force acting on tilted loop
Figure 21
Analysis of the forces on a current loop in a magnetic field.
 31-21

The final step is to convert Equation (32) into a vector Magnetic Moment
equation. First recall that the vector torque τ is defined When you put a current loop in a magnetic field, there
as is no net force on the loop ( F1,2 = - F3,4 in Figure 19b),
but we do get a torque. Thus magnetic fields do not
τ = r×F
accelerate current loops, but they do turn them. In the
where r and F1,2 are shown in Figure (23). In the figure study of the behavior of current loops, it is the torque
we see that r × F and therefore τ points up out of the that is important, and the torque is given by the simple
paper. formula of Equation (33).

Next we note that in Equation (32), θ is the angle This result can be written in an even more compact
between the magnetic field B and the loop area a . In form if we define the magnetic moment µ of a current
addition, the vector cross product a × B has a magni- loop as the current i times the vector area a of the loop
tude
definition of
µ ≡ ia (34)
magnetic moment
a × B = aBsin θ = w Bsin θ
With this definition, the formula for the torque on a
and points up, in the same direction as the torque τ . current loop reduces to
Thus Equation (32) immediately converts to the vector
equation
τ = µ×B (35)
τ = ia×B (33)
Although we derived Equations (34) and (35) for a
square loop, they also apply to other shapes such as
where i is the current in the loop, and a is the vector area round loops.
defined by the right hand convention of Figure (22).
F1,2
a
i
end view
of loop
r

Figure 22 τ1,2 = r × F1,2

Right hand convention for the loop area A.
τ1,2 points up,
out of the paper
Figure 23
The torque τ 1,2 exerted by the force F1,2 acting on
the side of the current loop. The vector r is the
lever arm of F1,2 about the axis of the coil. You
can see that r × F1,2 points up out of the paper.
31-22 Inductors and Magnetic Moment

Magnetic Energy
In Figure (24) we start with a current loop with its
magnetic moment µ aligned with the magnetic field as (a) θ=0
shown in (24a). We saw in Figure (21b) that this is the
orientation towards which the magnetic force is trying
to turn the loop. µ

If we grab the loop and rotate it around as shown in B

(24b) until µ is finally orientated opposite B as in (24c)
we have to do work on the loop. We can calculate the
amount of work we do rotating the loop from an angle
θ = 0 to θ = π using the angular analogy for the formula
for work. The linear formula for work is
(b) τ=µxB
x2 θ τ = µ B sin θ
W = Fx dx (10-19) µ
x1
B
Replacing the linear force Fx by the angular force τ ,
and the linear distances dx, x1 ,and x2 by the angular
distances d θ , 0 and π , we get
π µ
W = τdθ (36)
0
(c) θ=π
If we let go of the loop, the magnetic force will try to
reorient the loop back in the θ = 0 position shown in
Figure (24a). We can think of the loop as falling back
down to the θ = 0 position releasing all the energy we
stored in it by the work we did. In the θ = π position of B
Figure (24c) the current loop has a potential energy
equal to the work we did in rotating the loop from
θ = 0 to θ = π. Figure 24
The resting, low energy position of a current loop is
with µ parallel to B as shown in (a). To turn the
loop the other way, we have to do work against the
restoring torque τ = µ × B as shown in (b). The
total work we do to get the loop into its high energy
position (c) is 2 µ B . We can think of this as
magnetic potential energy that would be released if
we let the loop flip back down again. We choose the
zero of this potential energy half way between so
that the magnetic potential energy ranges from
+ µ B in (c) to – µ B in (a).
 31-23

We can calculate the magnetic potential energy by

evaluating the integral in Equation (36). From Equa-
tion (35) we have

τ = µ × B = µBsin θ

so that
π

W = µBsin θdθ
0
π
(36a)
= – µBcos θ = 2 µB
0

Thus the current loop in the θ = π position of (24c) has

an energy 2µΒ greater than the energy in the θ = 0
position of (24a).

It is very reasonable to define the zero of magnetic

potential energy for the position θ = π/2, half way
between the low and high energy positions. Then the
magnetic potential energy is + µB in the high energy
position and – µB in the low energy position. We
immediately guess that a more general formula for
magnetic potential energy of the current loop is

magnetic potential
energy of a Emag = – µ⋅B
current loop
(37)

This gives Emag = + µB when the loop is in the high

energy position with µ opposite B, and Emag = - µB
in the low energy position where µ and B are parallel.
At an arbitrary angle θ , Equation (37) gives
Emag = - µBcos θ , a result you can obtain from equa-
tion (36a) if you integrate from θ = 0 to θ = θ , and
adjust the zero of potential energy to be at θ = π/2.
31-24 Inductors and Magnetic Moment

Summary of Magnetic Moment Since the area of the loop is π r2, we get as the formula
Equations for the magnetic moment
Since we will be using the magnetic moment equations
v
in later discussions, it will be convenient to summarize µ = iA = q πr 2
them in one place. They are a short set of surprisingly 2πr *
compact equations. magnetic momentof
qv r a charge q traveling
A = area of current loop µ = at a speed v in a (38)
2
A
circularorbit of radius r
µ ≡ iA (34)
We can make a further refinement of Equation (38) by
(35) noting that the angular momentum J (we have already
τ = µ×B i B
used L for inductance) of a particle of mass m traveling
(37) at a speed v in a circle of radius r has a magnitude
Emag = –µ⋅ B
J = mv r

Charge q in a Circular Orbit More importantly, J points perpendicular to the plane

Most applications of the concept of magnetic moment of the orbit in a right handed sense as shown in Figure
are to atoms and elementary particles. In the case of (26a). This is the same direction as the magnetic
atoms, we can often picture the magnetic moment as moment µ seen in (26b), thus if we write Equation (38)
resulting from an electron traveling in a circular orbit in the form
like that shown in Figure (25). In that figure we show
q
a charge q traveling at a speed v in a circular orbit of µ = mv r (38a)
radius r. Since charge is being carried around this loop, 2m
this is a current loop, where the current i is the amount and use J for mvr, we can write (38a) as the vector
of charge per second being carried past a point on the equation
orbit. In one second the charge q goes around v/2πr
times, therefore relationbetween the angular
q momentum J and magnetic
µ = J
2m moment µ for a particle (39)
v meter / sec traveling in a circularorbit
i = q coulombs
2πr meter
Equation (39) is as far as we want to go in developing
v magnetic moment formulas using strictly classical
i = q
2πr physics. We will come back to these equations when
we study the behavior of atoms in a magnetic field.
v
q

Figure 25
A charged particle in a circular orbit acts like a current
loop. Its magnetic moment turns out to be µ = qvr/2.
 31-25

(angular momentum)
J

J = mvr

m r
v

(a) angular momentum of a particle in

a circular orbit

(magnetic moment)
µ

µ = q (mvr)
2m

q r
v

(b) magnetic moment of a charged particle

in a circular orbit

Figure 26
Comparing the magnetic moment µ and angular
momentum J of a particle in a circular orbit, we see
that q
µ= J
2m
31-26 Inductors and Magnetic Moment

IRON MAGNETS Our purpose in this discussion of iron magnets is not to

In iron and many other elements, the atoms have a net go over the details of how magnetic moments are
magnetic moment due to the motion of the electrons aligned, what keeps them aligned or what disrupts the
about the nucleus. The classical picture is a small alignment. We will consider only the more fundamen-
current loop consisting of a charged particle moving in tal question – what is the effect of lining up the
a circular orbit as shown previously in Figures (25) and magnetic moments in a sample of matter. What
(26). happens if we line them all up as shown in Figure (27)?

If a material where the atoms have a net magnetic A current loop has its own magnetic field which we saw
moment is placed in an external magnetic field Bext the in our original discussion of magnetic field patterns and
torque exerted by the magnetic field tends to line up the which we have reproduced here in Figure (28). This is
magnetic moments parallel to Bext as illustrated in a fairly complex field shape. (Out from the loop at
Figure (27). This picture, where we show all the atomic distances of several loop radii, the field has the shape of
magnetic moments aligned with Bext is an exaggera- what is called a “dipole” magnetic field. In certain
tion. In most cases the thermal motion of the atoms regions earth’s magnetic field has this dipole magnetic
seriously disrupts the alignment. Only at temperatures field shape.)
of the order of one degree above absolute zero and in
external fields of the order of one tesla do we get a
nearly complete alignment.

Iron and a few other elements are an exception. A small

external field, of the order of 10 gauss (.001 tesla) or
less, can align the magnetic moments at room tempera-
ture. This happens because neighboring atoms interact
with each other to preserve the alignment in an effect
called ferromagnetism. The theory of how this inter-
action takes place, and why it suddenly disappears at a
certain temperature (at 1043 K for iron) has been and
still is one of the challenging problems of theoretical
Bext Bext
physics. (The problem was solved by Lars Onsager for
a two-dimensional array of iron atoms, but no one has
yet succeeded in working out the theory for a three-
Figure 27
dimensional array.)
Ferromagnetism. When you apply an external
The behavior of iron or other ferromagnetic materials magnetic field to a piece of "magnetically soft"
iron (like a nail), the external magnetic field
depends very much on how the substance was physi- aligns all of the magnetic moments of the iron
cally prepared, i.e., on how it was cooled from the atoms inside the iron. The magnetic field of the
molten mixture, what impurities are present, etc. In one current loops can be enormous compared to the
extreme, it takes a fairly strong external field to align external field lining them up. As a result a
the magnetic moments, but once aligned they stay small external field produced say by a coil of
there. This preparation, called magnetically hard iron, wire, can create a strong field in the iron and we
have an electromagnet. This phenomena is
is used for permanent magnets. In the other extreme, called ferromagnetism.
a small external field of a few gauss causes a major
alignment which disappears when the external field is
removed. This preparation called magnetically soft
iron is used for electromagnets.
 31-27

When you have a large collection of aligned current The cancellation is complete everywhere except at the
loops as shown in Figure (27), the magnetic fields of outside surface. At this surface we essentially have a
each of the current loops add together to produce the single large current loop with a current i equal to the
magnetic field of the magnet. The magnetic field of a current i in each of the little loops. It was in this way that
single current loop, shown in Figure (28), is bad Ampere saw that the magnetic field produced by all the
enough. What kind of a mess do we get if we add up small current loops packed together must be the same
the fields of thousands, billions, 6 × 1023 of these as the magnetic field of one big loop. What an
current loops? The calculation seems impossibly enormous simplification!
difficult.
B
Ampere discovered a simple, elegant way to solve the
problem. Instead of adding up the magnetic fields of
each current loop, he first added up the currents using
a diagram like that shown in Figure (29).
iron bar effective current
We can think of Figure (29) as the top view of the magnet i around the
aligned current loops of Figure (27). If you look at surface
i
Figure (29) for a while, you see that all the currents i i
inside the large circle lie next to, or very close to, an
equal current flowing in the opposite direction. We can
say that these currents inside the big circle cancel each current loop of
other. They do not carry a net charge, and therefore do individual atom
(greatly enlarged)
not produce a net magnetic field.
enlarged
cross-section
of iron bar
magnet
i
Figure 29
In an iron bar magnet, the iron atoms are
permanently aligned. In the cross-sectional view
we are looking down on the aligned current
loops of the atoms. Inside the iron, we picture
the currents as cancelling, leaving a net current
i (the same as the current in each loop) going
Figure 28 around the surface of the nail. This picture of a
Magnetic field of a current loop. surface current replacing the actual current
loops was proposed by Ampere, and the surface
current is known as an Amperian current.
31-28 Inductors and Magnetic Moment

Now let us return to Figure (27), redrawn in (30a), Although a bar magnet and a solenoid have the same
where we had a collection of aligned current loops. We field shape, the strength of the field in a bar magnet is
can think of this as a model of a magnetized iron rod usually far stronger. If the majority of the magnetic
where all the iron atom magnetic moments are aligned. moments in an iron bar are aligned, we get a field of the
From Figure (29) we see that one horizontal layer of order of one tesla inside the bar. To obtain comparable
these current loops can be replaced by one large loop field strengths inside a solenoid made using copper
carrying a current i that goes all the way around the iron wire, we would have to use currents so strong that the
rod. This is shown again at the top of Figure (30). copper wire would soon heat up and perhaps melt due
to electrical resistance in the copper.
Now our rod consists of a number of layers of small
loops shown in (30a). If each of these layers is replaced
by a single large loop, we end up with the stack of large The Electromagnet
loops shown in (30b). But this stack of large loops is If we insert a magnetically soft iron rod into the core of
just the same current distribution we get in a sole- a solenoid as shown in Figure (32), we have an electro-
noid! Thus we get the remarkable result that the vector magnet. It only takes a small external field to align a
sum of the magnetic fields of all the current loops in majority of the magnetic moments in magnetically soft
Figure (30a) is just the simple field of a solenoid. This iron. And when the moments are aligned, we an get
is why a bar magnet and a solenoid of the same size fields approaching one tesla, 104 gauss, as a result. This
have the same field shape, as seen in Figure (31). is the principle of an electromagnet where a weak field
produced by a small current in the windings produces
a strong field in the iron.
i
i
i i
i
i i i
i

i
amperian wire
current current

a) small current loops b) large Amperian bar magnet solenoid

currents around
Figure 30 surface Figure 31
Ampere's picture of replacing Comparison of the magnetic fields of an iron magnet
small current loop throughout and a solenoid. The fields are essentially the same
the substance by large ones on because the Amperian currents in a bar magnet are
the surface. essentially the same as the current in the coils of a
solenoid.
 31-29

Figure (33) is a graph showing the strength of the The Iron Core Inductor
magnetic field inside the iron core of an electromagnet When the external field is less than 2 gauss in Figure
as a function of the strength of the external magnetic (33), we have a more or less linear relationship shown
field produced by the windings of the solenoid. In this by the dotted line between the external field and the
case a toroidal solenoid was used, the iron core is an field in the iron. In this region of the curve, for Bext <
iron ring inside the toroid, and the results in Figure (33) 2 gauss, the iron is essentially acting as a magnetic field
are for one particular sample of iron. We can get amplifier. For this sample, a 2 gauss external field
different results for different samples of iron prepared produces a 50,000 gauss magnetic field in the iron, an
in different ways. amplification by a factor of 25,000.
The vertical axis in Figure (33) shows the percentage of If we amplify the magnetic field in our solenoid 25,000
the maximum field Bmax we can get in the iron. Bmax times, we are also amplifying the magnetic flux ΦB by
is the “saturated” field we get when all the iron atoms the same factor. If we have a varying current in the
magnetic moments are aligned and has a typical value solenoid, but keep Bext under 2 gauss, we will get a
of about 1 tesla. We see that a very small external field varying magnetic field in the iron and a varying mag-
of 2 gauss brings the magnetic field up to 50% of its netic flux ΦB that is roughly proportional to the current
saturated value. Getting the other 50% is much harder. i in the solenoid. The difference that the iron makes is
We can more or less turn on the electromagnet using a that the flux ΦB, and the rate of change of flux dΦB /dt
2 gauss external field, and that not much is to be gained will be 25,000 times larger. And so will the induced
by using a stronger external field. voltage in the turns of the solenoid. This means that the
inductance of the solenoid is also increased by 25,000
iron
times. If we inserted an iron ring into our air core
core
solenoid shown in Figure (6), and the iron had the same
magnetic properties as the iron sample studied in
Bexternal Figure (33), the inductance of our toroidal solenoid
would increase 25,000 times from 1.8 × 10-4 henry up
i
to about 4 henrys.

Bmax 1.2 tesla

= 12000 gauss

75

Need student project

50
data for this.
25
Figure 32
In an electromagnet, some turns of wire are wrapped Bexternal
around an iron bar. When a current i is turned on, the 2 4 6 8 10 12 14 gauss (10 – 4 tesla)
magnetic field of the turns of wire provide the external
field to align the iron atoms. When the current i is Figure 33
shut off, and the external field disappears, the iron Example of a magnetization curve for
atoms return to a random alignment and the magnetically soft iron. The impressive
electromagnet shuts off. Whether the iron atoms feature is that an external field of only
remain aligned or not, whether we have a permanent a few gauss can produce fields in
magnet or not, depends upon the alloys (impurities) in excess of xxxxx gauss inside the iron.
the iron and the way the iron was cooled after casting.
31-30 Inductors and Magnetic Moment

We can easily get large inductances from iron core The introduction in the early 1970s of superconducting
inductors, but there are certain disadvantages. The wire that could carry huge currents yet had zero electri-
curve in Figure (33) is not strictly linear, therefore the cal resistance revolutionized the design of strong field
inductance has some dependence on the strength of the magnets. Magnets made from superconducting wire,
current in the coil. When we use an AC current in the called superconducting magnets are routinely designed
solenoid, the iron atoms have to flip back and forth to to create magnetic fields of strengths up to 5 tesla. Such
keep their magnetic moments aligned with the AC magnets will be used in the superconducting
external field. There is always some energy dissipated supercollider discussed earlier, and are now found in
in the process and the iron can get hot. And if we try to the magnetic resonance imaging devices in most large
go to too high a frequency, the iron atoms may not be hospitals. The major problem with the superconduct-
able to flip fast enough, the magnetic field in the iron ing magnets is that the superconducting wire has to be
will no longer be able to follow the external field, and cooled by liquid helium to keep the wire in its supercon-
the amplification is lost. None of these problems is ducting state. And helium is a rare substance (at least
present with a air core inductor that has no iron. on earth) that is difficult to liquefy and hard to maintain
as a liquid.
Superconducting Magnets In the late 1980s substances were discovered that are
The fact that iron saturates, the fact that we can do no superconducting when immersed in liquid nitrogen, an
better than aligning all the iron atoms current loops, inexpensive substance to create and maintain. So far
places a fundamental limit on the usefulness of electro- we have not been able to make wires out of these “high
magnets for producing strong magnetic fields. Instead temperature” superconductors that can carry the huge
it is necessary to return to air core solenoids or other currents needed for big superconducting magnets. But
arrangements of coils of wire, and simply use huge this seems to be an engineering problem that when
currents. solved, may have a revolutionary effect not just on the
design and use of superconducting magnets, but on
The problem with using copper wire for coils that
technology in general.
produce magnetic fields stronger than 1 tesla is that
such strong currents are required that even the small
resistance in copper produces enough heat to melt the
wire. The only solutions for copper are to use an
elaborate cooling system to keep the copper from
heating, or do the experiment so fast that either the
copper does not have time to heat, or you do not mind
if it melts.
 31-31

APPENDIX
THE LC CIRCUIT
AND FOURIER ANALYSIS
The special feature of an LC circuit, like the one shown
in Figure (A1), is it’s resonance at an angular frequency
ω 0 = 1 LC . If you drive the circuit with an oscillator R
that puts out a sine wave voltage V = V0 sin ωt, the
circuit will respond with a large voltage output when V L C
the driving frequency ω equals the circuit’s resonant
frequency ω 0 . We saw this resonance in our discus-
sion of the LC circuit shown in Figure (12) (p 31-11). scope oscillator
Figure A1
In Chapter 16, in our discussion of Fourier analysis, we The LC circuit. For this experiment we used a fairly
saw that a square wave of frequency ω 0 can be large commercial 9.1 × 10– 3 henry inductor. This
constructed by adding up a series of harmonic sine large inductance made the circuit more stable and less
waves. The first harmonic, of the form A 1 sin ω1t , has noisy than when we tried to do the experiment with the
the same frequency as the square wave. For a square toroidal inductor of Figure (12). This inductor turned
wave the second and all even harmonics are missing. out to have an internal capacitance of 9.9 × 10– 10
farads (990 picofarads) due to the coil windings
The third harmonic is of the form A 3 sin ω3t . That is, themselves. We used this internal capacitance for the
the third harmonic’s frequency is three times the fre- capacitor C of the circuit. (The internal capacitance
quency ω1 of the first harmonic. The fifth harmonic is was determined by measuring the resonant frequency
of the form A 5 sin ω5t . The amplitudes A 1 , A 3 , ω 0 = 1 LC and solving for C.)
A 5 ,... of the harmonic sine waves present in the square By using a large inductor L, we can attach the scope
wave, which are shown by the vertical bars in Figure directly across the inductor, as shown, without the
(A2), were determined by Fourier analysis. (For a scope having a serious effect on the circuit. The
square wave, there is the simple relationship A 3= A 1 3 , resistance R, with R = 150KΩ , partially isolated the LC
A 5= A 1 5 , etc.) circuit from the oscillator. This allowed the oscillator
to gently drive the circuit without putting out much
The point of this lab is to demonstrate the physical current and without distorting the shape of the square
wave. (The oscillator could not maintain a square
reality of the harmonics in a square wave. We have
wave when we used the LC circuit shown in Figure
seen that an LC circuit can be driven to a large ampli- (12).)
tude resonance only when the driving frequency is
If one wants to try values of C other than the internal
equal to or close to the resonant frequency ω 0 = 1 LC . capacitance of the coil, one can add an external
To put it another way, we can use the LC circuit to capacitor in parallel with L and C of Figure (A1). If
detect the presence of a sine wave of frequency ω 0 in you use an external capacitor more than about 10 times
the driving signal. If that frequency is present in the the internal capacitance, the internal capacitance of
driving signal, the circuit will resonate. If it is not the inductor can be neglected.
present, the circuit will not resonate.

Our experiment is to drive an LC circuit with a square

wave, and see if the various harmonics in the square
wave can each cause a resonance in the circuit. For
example, if we adjust the frequency ω1 of the square
wave to equal ω 0 , then we expect the first harmonic
A 1 sin ω1t to drive the circuit in resonance.
31-32 Inductors and Magnetic Moment

If we then lower the frequency ω1 of the square wave

so that 3ω1 = ω 0 , then we expect that the third har-
monic A 3 sin ω3t should drive the circuit in resonance.
We should get another resonance when 5ω1 = ω 0 , and
another at 7ω1 = ω 0 , etc. When the LC circuit is
driven by a square wave, there should be a whole series
of resonances, where in each case one of the harmonics
has the right frequency to drive the circuit. These
resonances provide direct experimental evidence that
the various harmonics are physically present in the
square wave, that they have energy that can drive the A1
resonance.
A3
In Figures A3 and A4 we look at the shape of the first A5
few harmonics in the square wave of Figure (A2), and
then watch as a square wave emerges as the harmonics
are added together. (This is mostly a review of what we Figure A2
Fourier analysis of a square wave. The top part of the
did back in Chapter 16.) After that, we study the
MacScope output shows an experimental square wave.
resonances that occur when ω1 = ω 0 , 3ω1 = ω 0 , We selected one cycle of the wave, chose Fourier
5ω1 = ω 0 , etc. Finally we drop the square wave fre- analysis, and see that the wave consists of a series of
quency to about 23ω1 = ω 0 , and watch the LC circuit odd harmonics. You can see the progression of
ring like a bell repeatedly struck by a hammer.Figure amplitudes with A 3 = A 1 3 , A 5 = A 1 5 , etc. In Figure
A1 A3 we show the harmonics A 1 sin ω 1 t , A 3 sin ω 3 t ,and
A 5 sin ω 5 t . In Figure A4 we add together the
harmonics to create a square wave.
 31-33

Harmonic 1 selected.

Harmonic 1

Harmonics 1 and 3 selected.

Harmonic 3

Harmonics 1,3 and 5 selected.

Harmonics 1,3,5 and 7 selected.

Harmonic 5

Figure A3
Displaying selected harmonics. Note that when you
select a harmonic, you not only see the shape of the Harmonics 1,3,5,7,9 and 11 selected.
harmonic, but also see the harmonic’s frequency
displayed. (We highlighted this display with small Figure A4
rectangles.) You can see, for example, that the third Adding up harmonics to create a square wave.
harmonic is 159.5 kHz, 3 times the fundamental The more harmonics we add, the closer we get.
frequency of 53.19 kHz.
31-34 Inductors and Magnetic Moment

Figure A5 – Resonance at ω1 = ω0
We get the biggest resonance when the frequency ω 1 of the square wave is equal to the resonant
frequency ω 0 of the circuit. We displayed the first harmonic by clicking on the bar over the 1 in
the Fourier analysis window, and see that the frequency of the first harmonic is 52.91 kilohertz.

Figure A6 – Resonance at 3ω 1 = ω0
Lowering the frequency of the square wave to 17.54 kilohertz, we get another resonance
shown above. In the Fourier analysis window, we selected the third harmonic, and see
that the frequency of this harmonic is 52.63 kilohertz. To within experimental accuracy,
this is equal to the LC circuit’s resonant frequency of 52.91 kilohertz.
 31-35

Figure A7 – Resonance at 5ω 1 = ω0
Lowering the frequency of the square wave to 10.25 kilohertz, we get another resonance. In
the Fourier analysis window, we selected the fifth harmonic, and see that the frequency of
this harmonic is 52.63 kilohertz. To within experimental accuracy, this is again equal to
the LC circuit’s resonant frequency of 52.91 kilohertz.

Figure A8 – Resonance at 11ω 1 = ω0

Skipping two resonances and lowering the frequency of the square wave to 4.712 kilohertz, we
get a sixth resonance. In the Fourier analysis window, we selected the eleventh harmonic,
and see that the frequency of this harmonic is 52.63 kilohertz. To within experimental
accuracy, this is again equal to the LC circuit’s resonant frequency of 52.91 kilohertz.
31-36 Inductors and Magnetic Moment

Figure A9 – Ringing like a bell

Dropping the square wave frequency even further, we see that every time the voltage of the
square wave changes, the circuit responds like a bell struck by a hammer. This setup can
be used as the starting point for the study of damped resonant (LRC) circuits.
 32-1

Chapter 32
Maxwell's Equations

CHAPTER 32 MAXWELL'S EQUATIONS

In 1860 James Clerk Maxwell summarized the entire In electricity theory we have two vector fields E and B ,
content of the theory of electricity and magnetism in a and two equations are needed to define each field.
few short equations. In this chapter we will review Therefore the total number of equations required must
these equations and investigate some of the predictions be four.
one can make when the entire theory is available.
How many of the required equations have we discussed
What does a complete theory of electricity and magne- so far? We have Gauss’ law for the divergent part of
tism involve? We have to fully specify the electric field E , and Faraday’s law for the solenoidal part. It
E , the magnetic field B , and describe what effect the appears that we already have a complete theory of the
fields have when they interact with matter. electric field, and we do. Gauss’ law and Faraday’s
law are two of the four equations needed.
The interaction is described by the Lorentz force law
F = qE + q v × B (28-18) For magnetism, we have Ampere’s law that defines the
solenoidal part of B . But we have not written an
which tells us the force exerted on a charge q by the E equation involving the surface integral of B . We are
and B fields. As long as we stay away from the atomic missing a Gauss’ law type equation for the magnetic
world where quantum mechanics dominates, then the field. It would appear that the missing Gauss’ law for
Lorentz force law combined with Newton’s second law B , plus Ampere’s law make up the remaining two
fully explains the behavior of charges in the presence equations.
of electric and magnetic fields, whatever the origin of
the fields may be. This is not quite correct. The missing Gauss’ law is one
of the needed equations for B , and it is easily written
To handle the electric and magnetic fields, recall our down because there are no known sources for a diver-
discussion in Chapter 30 (on two kinds of fields) where gent B field. But Ampere’s law, in the form we have
we saw that any vector field can be separated into two been using
parts; a divergent part like the electric field of static
charges, and a solenoidal part like the electric field in B⋅dl = µ0 i (29-18)
a betatron or inductor. To completely specify a vector has a logical flaw that was discovered by Maxwell.
field, we need two equations – one involving a surface When Maxwell corrected this flaw by adding another
integral or its equivalent to define the divergent part of source term to the right side of Equation (29-18), he
the field, and another involving a line integral or its then had the complete, correct set of four equations for
equivalent defining the solenoidal part. E and B .
32-2 Maxwell's Equations

All Maxwell did was to add one term to the four GAUSS’ LAW FOR
equations for E and B , and yet the entire set of MAGNETIC FIELDS
equations are named after him. The reason for this is
Let us review a calculation we have done several times
that with the correct set of equations, Maxwell was able
now—the use of Gauss’ law to calculate the electric
to obtain solutions of the four equations, predictions of
field of a point particle. Our latest form of the law is
these equations that could not be obtained until
Ampere’s law had been corrected. The most famous of Q in
E ⋅ dA = (29-5)
these predictions was that a certain structure of electric closed ε0
surface
and magnetic fields could travel through empty space
at a speed v = 1/ µ0ε0 . Since Maxwell knew that where Qin is the total amount of electric charge inside
1/ µ0ε0 was close to the observed speed 3 × 10 8 m/s the surface.
for light, he proposed that this structure of electric and
magnetic fields was light itself. In Figure (1), we have a point charge Q and have
constructed a closed spherical surface of radius r cen-
In this chapter, we will first describe the missing Gauss' tered on the charge. For this surface, E is everywhere
law for magnetic fields, then correct Ampere’s law to perpendicular to the surface or parallel to every surface
get the complete set of Maxwell’s four equations. We element dA , thus E ⋅ dA = E dA . Since E is of
will then solve these equations for a structure of electric constant magnitude, we get
and magnetic fields that moves through empty space at
a speed v = 1/ µ0ε0 . We will see that this structure
E ⋅ dA = E dA
explains various properties of light waves, radio waves,
and other components of the electromagnetic spec- closed closed
surface surface
trum. We will find, for example, that we can detect
radio waves by using the same equipment and proce- = E 4πr2
dures we have used in earlier chapters to detect and (1)
map electric and magnetic fields. Qin
=
ε0
where Qin = Q .

Q r

Figure 1 E
Field of point charge.
 32-3

The solution of Equation (1) gives E = Q/4π ε0 r2 as In the last 20 years there has been an extensive search
the strength of the electric field of a point charge. A for evidence for the decay of protons or the existence of
similar calculation using a cylindrical surface gave us magnetic monopoles. So far we have found no evi-
the electric field of a charged rod. By being clever, or dence for either. (You do not have to wait 10 32 years
working very hard, one can use Gauss’ law in the form to see if protons decay; instead you can see if one out
of Equation (29-5) to solve for the electric field of any of 10 32 protons decays in one year.)
static distribution of electric charge.
The failure to find the magnetic monopole, the fact that
But the simple example of the field of a point charge no one has yet seen a magnetic field with the shape
illustrates the point we wish to make. Gauss’ law shown in Figure (2), can be stated mathematically by
determines the diverging kind of field we get from a writing a form of Gauss’ law for magnetic fields with
point source. Electric fields have point sources, namely the magnetic charge Qin set to zero
electric charges, and it is these sources in the form of
Qin that appear on the right hand side of Equation closed
B ⋅ dA = 0 (2)
(29-5). surface

Figure (2) shows a magnetic field emerging from a When reading Equation (2), interpret the zero on the
point source of magnetism. Such a point source of right side of Equation (2) as a statement that the
magnetism is given the name magnetic monopole and divergent part of the magnetic field has no source
magnetic monopoles are predicted to exist by various term. This is in contrast to Gauss’ law for electric
recent theories of elementary particles. These theories fields, where Qin /ε 0 is the source term.
are designed to unify three of the four basic interactions
– the electrical, the weak, and the nuclear interactions.
(They are called “Grand Unified Theories” or “GUT”
theories. Gravity raises problems that are not handled
by GUT theories.) These theories also predict that the
proton should decay with a half life of 10 32 years.

B
Figure 2
Magnetic field produced by a point source.
32-4 Maxwell's Equations

MAXWELL’S CORRECTION path on the wire so that B and sections d of the path
TO AMPERE’S LAW are everywhere parallel. Thus B ⋅ d = B d , and
since B is constant along the path, we have
As we mentioned in the introduction, Maxwell de-
tected a logical flaw in Ampere’s law which, when B ⋅ d = B d = B×2πr = µ0 i
corrected, gave him the complete set of equations for
the electric and magnetic fields. With the complete set which gives our old formula B = µ0 i/2π r for the
of equations, Maxwell was able to obtain a theory of magnetic field of a wire.
light. No theory of light could be obtained without the
correction. To see the flaw with Ampere’s law, consider a circuit
where a capacitor is being charged up by a current i as
Ampere’s law, Equation (29-18), uses the line integral shown in Figure (4). When a capacitor becomes
to detect the solenoidal component of the magnetic charged, one plate becomes positively charged and the
field. We had other negatively charged as shown. We can think of the
capacitor being charged because a positive current is
Ampere's
B ⋅ d = µ0 ienclosed Law
(29-18) flowing into the left plate, making that plate positive,
and a positive current is flowing out of the right plate,
where i enclosed is the total current encircled by the making that plate negative.
closed path used to evaluate B ⋅ d . We can say that
µ0 ienclosed is the source term for this equation, in Figure (4) looks somewhat peculiar in that the current
analogy to Qin /ε 0 being the source term for Gauss’ i almost appears to be flowing through the capacitor.
law. We have a current i on the left, which continues on the
right, with a break between the capacitor plates. To
Before we discuss Maxwell’s correction, let us review emphasize the peculiar nature of this discontinuity in
the use of Equation (29-18) to calculate the magnetic the current, imagine that the wires leading to the
field of a straight current i as shown in Figure (3). In capacitor are huge wires, and that the capacitor plates
(3a) we see the wire carrying the current, and in (3b) we are just the ends of the wires as shown in Figure (5).
show the circular magnetic field produced by the
current. To apply Equation (29-18) we draw a closed Now let us apply Ampere’s law to the situation shown
circular path of radius r around the wire, centering the in Figure (5). We have drawn three paths, Path (1)
around the wire leading into the positive plate of the
capacitor, Path (2) around the wire leading out of the
negative plate, and Path (3) around the gap between the
i plates. Applying Ampere’s law we have
r
(a)
+ –
+ –
B
i + – i
+ –
+ –
r (b) + –

Figure 4
Charging up a capacitor.

Figure 3
Using Ampere's law.
 32-5

Maxwell’s solution was that even inside the gap at the

path 1 goes
B ⋅ d = µ0 i around a current i
(3) capacitor there was a source for B ⋅ d , and that the
path 1
strength of the source was still µ0 i. What actually
path 2 goes
exists inside the gap is the electric field E due to the +
B ⋅ d = µ0 i around a current i
(4) and – charge accumulating on the capacitor plates as
path 2
shown in Figure (6). Perhaps this electric field can
path 3 does not go somehow replace the missing current in the gap.
B⋅d = 0 around any current
(5)
path 3
The capacitor plates or rod ends in Figure (6) have a
When we write out Ampere’s law this way, the charge density σ = Q/A where Q is the present
discontinuity in the current at the capacitor plates charge on the capacitor and A is the area of the plates.
looks a bit more disturbing. In one of our early Gauss’ law calculations we saw that
a charge density on a conducting surface produces an
For greater emphasis of the problem, imagine that the electric field of strength E = σ/ε0, thus E between
gap in Figure (5) is very narrow, like Figure (5a) only the plates is related to the charge Q on them by
worse. Assume we have a 1 mm diameter wire and the
gap is only 10 atomic diameters. Then according to σ Q
Ampere’s law, B ⋅ d should still be zero if it is E = = ; Q = ε0 EA (6)
ε0 ε0 A
correctly centered on the gap. But can we possibly
center a path on a gap that is only 10 atomic diameters The current flowing into the capacitor plates is related
wide? And even if we could, would B ⋅ d be zero to the charge Q that has accumulated by
for this path, and have the full value µ0 i for the path 10
atomic diameters away? No, we simply cannot have dQ
i =
such a discontinuity in the magnetic field and there dt (7)
must be something wrong with Ampere’s law. This Using Equation (6) in Equation (7), we get
was the problem recognized by Maxwell.
Path 1 Path 3 Path 2 d
i = ε0 EA (8)
dt
capacitor
i + – i plate of
+ – area A
i + – i
+ – + –
i + –
i i E i
+ –
+ – i + E – i
+ –
i + E – i
Figure 5
+Q + – –Q
Current flows through Paths (1)
& (2), but not through Path (3).
Figure 6
An electric field E exists between the plates.

i i
i i
i i

Figure 5a
Very narrow gap
32-6 Maxwell's Equations

Noting that the flux ΦE of electric field between the I.e., for Paths (1) and (2), there is no electric flux
plates is through the path and the source of B ⋅ d is the
current. For Path (3), no current flows through the path
ΦE = E ⋅ dA = E A⊥ = E A (9) and the source of B ⋅ d is the changing electric
flux. But, because µ0 i in Equation (12) has the same
and multiplying through by µ0, we can write Equation magnitude µ0ε0 dΦE dt in Equation (13), the term
(8) in the form B ⋅ d has the same value for Path (3) as (1) and (2),
dΦE and there is no discontinuity in the magnetic field.
µ0 i = µ0ε0 (10)
dt
Example: Magnetic Field
We get the somewhat surprising result that µ0 ε 0 times between the Capacitor Plates
the rate of change of electric flux inside the capacitor As an example of the use of the new term in the
has the same magnitude as µ0 i , where i is the current corrected Ampere’s law, let us calculate the magnetic
in the wire leading to the capacitor. Maxwell proposed field in the region between the capacitor plates. To do
that µ0ε0 dΦE dt played the same role, inside the this we draw a centered circular path of radius r smaller
capacitor, as a source term for B ⋅ dl, that µ0 i did than the capacitor radius R as shown in Figure (8).
outside in the wire. There is no current through this path, but there is an
electric flux ΦE r = EA r = Eπ r2 through the
As a result, Maxwell proposed that Ampere’s law be path. Thus we set i = 0 in Ampere’s corrected law, and
corrected to read replace ΦE by the flux ΦE r through our path to get
dΦE corrected dΦE r
B ⋅ d = µ 0 i + µ 0 ε0 Ampere's (11) B ⋅ d = µ0 ε0 (14)
dt law
dt
Equation (14) tells us that because we have an in-
Applying Equation (11) to the three paths shown in creasing electric field between the plates, and thus an
Figure (7), we have increasing electric flux through our path, there must be
B ⋅ d = µ0 i ( ΦE = 0 ) (12) a magnetic field around the path.
paths 1&2

dΦE
B ⋅ d = µ0ε0 (i = 0) (13)
paths 3 dt

1 3 2

i E i
i i i R i
E E
r
i E i

Figure 8
Figure 7 Calculating B in the region between the plates.
Path (2) surrounds a changing electric flux. Inside the
gap, µ0 ε0 (dΦ E/dt) replaces µ0 i as the source of B .
 32-7

Due to the cylindrical symmetry of the problem, the Finally using Equation (15) and (17) in (14) we get
only possible shape for the magnetic field inside the 2
capacitor is circular, just like the field outside. This r
B×2πr = µ0i
circular field and our path are shown in the end view, R2
Figure (9). Since B and d are parallel for all the steps
around the circular path, we have B ⋅ d = B d . And magnetic
µ0i r
since B is constant in magnitude along the path, we get B = field between (18)
2π R 2
capacitor plates
B ⋅ d = B d = B× 2πr (15)
Figure (10) is a graph of the magnitude of B both
To evaluate the right hand side of Equation (14), note inside and outside the plates. They match up at r =
that the flux through our path ΦE r is equal to the total R, and the field strength decreases linearly to zero
flux ΦE total times the ratio of the area π r2 of our path inside the plates.
to the total area π R2 of the capacitor plates
Exercise 1
ΦE r = ΦE total π r
2
(16) Calculate the magnetic field inside the copper wires
π R2
that lead to the capacitor plates of Figure (5). Use
so that the right hand side becomes Ampere’s law and a circular path of radius r inside the
dΦE r dΦ total r2 copper as shown in Figure (11). Assuming that there is
µ 0 ε0 = µ 0 ε0 E a uniform current density in the wire, you should get
dt dt R2 Equation (18) as an answer. Thus the magnetic field is
2
= µ0 i r (17) continuous as we go out from the copper to between the
R2 capacitor plates.
where in the last step we used Equation (10) to replace
µ0 ε0 dΦE total /dt by a term of the same magnitude,
namely µ0 i.
circular R
i
Magnetic r
Field
B
circular

;;;
Path of B Figure 11
radius r

;
r
B(r)

Figure 9
End view of capacitor plate.
R r
Figure 10
Magnetic fields inside and outside the gap.
32-8 Maxwell's Equations

MAXWELL’S EQUATIONS The best way to give these equations meaning is to

know inside out at least one specific example that
Now that we have corrected Ampere’s law, we are
illustrates the use of each term in the equations. For
ready to write the four equations that completely Gauss’ law, we have emphasized the calculation of the
govern the behavior of classical electric and magnetic electric field of a point and a line charge. We have the
fields. They are nonexistence of the divergent magnetic field in Figure
(2) to illustrate Gauss’ law for magnetic fields. We have
used Ampere’s law to calculate the magnetic field of a
Q in Gauss'
(a) E ⋅ dA = wire and a solenoid. The new term in Ampere’s law was
closed ε0 Law used to calculate the magnetic field inside a parallel
surface
plate capacitor that is being charged up.
No
(b) closed
B ⋅ dA = 0 Monopole
Faraday’s law has numerous applications including the
surface air cart speed meter, the betatron, the AC voltage
generator, and the inductance of a solenoid. Perhaps
dΦE the most important concept with Faraday’s law is that
Ampere's
(c) B⋅d = µ0 i + µ0ε0 Law
E⋅dl is the voltage rise created by solenoidal electric
dt
fields, which for circuits can be read directly by a
dΦB voltmeter. This lead to the interpretation of a loop of wire
Faraday's
(d) E⋅d = – Law
with a voltmeter attached as an E⋅dl meter. We used
dt an E⋅dl meter in the design of the air cart speed
(19) detector and experiment where we mapped the mag-
netic field of a Helmholtz coil.
The only other thing you need for the classical theory Then there is the Lorentz force law with the formulas for
of electromagnetism is the Lorentz force law and the electric and magnetic force on a charged particle.
Newton’s second law to calculate the effect of electric As an example of an electric force we calculated the
and magnetic fields on charged particles. trajectory of an electron beam between charged plates,
and for a magnetic force we studied the circular motion
Lorentz of electrons in a uniform magnetic field.
F = qE + qv ×B Force
Law The assignment of this exercise is to write out Maxwell’s
equations one by one, and with each equation write
(20) down a fully worked out example of the use of each term.
Do this neatly, and save it for later reference. This is what
This is a complete formal summary of everything we turns the hen scratchings shown on the previous page
have learned in the past ten chapters. into a meaningful theory. When you buy a T-shirt with
Maxwell’s equations on it, you will be able to wear it with
confidence.
Exercise 2
This is one of the most important exercises in the text.
The four Maxwell’s equations and the Lorentz force law We have just crossed what you might call a conti-
represent an elegant summary of many ideas. But nental divide in our study of the theory of electricity
these equations are nothing but hen scratchings on a
and magnetism. We spent the last ten chapters
piece of paper if you do not have a clear idea of how
each term is used.
building up to Maxwell’s equations. Now we de-
scend into applications of the theory. We will focus
on applications and discussions that would not have
made sense until we had the complete set of equa-
tions—discussions on the symmetry of the equations
and applications like Maxwell’s theory of light.
 32-9

SYMMETRY OF Equations (22) immediately demonstrate the lack of

MAXWELL’S EQUATIONS symmetry caused by the absence of magnetic mono-
poles, and so does the Lorentz force law of Equation
Maxwell’s Equations (19 a, b, c, and d), display consid-
(20). If the magnetic monopole is discovered, and we
erable symmetry, and a special lack of symmetry. But
assign to it the “magnetic charge” Q B , then for ex-
the symmetry or lack of it is clouded by our choice of
ample Equation (22b) would become
the MKS units with its historical constants µ0 and ε0
that appear, somewhat randomly, either in the numera- B ⋅ dA = Q B (22b')
tor or denominator at various places. closed
surface

For this section, let us use a special set of units where

If we have magnetic monopoles, a magnetic field
the constants µ0 and ε0 have the value 1
should exert a force FB = Q BB and perhaps an electric
in a special field should exert a force something like FE = Q B v × E .
µ0 = 1 ; ε 0 = 1 set of units
(21)
Aside from Equation (22b), the other glaring asymme-
Because the speed of light c is related to µ0 and ε0 by try is the presence of an electric current i in Ampere’s
c = 1/ µ0 ε0 , we are now using a set of units where law (22c) but no current term in Faraday’s law (22d).
the speed of light is 1. If we set µ0 = ε0 = 1 in If, however, we have magnetic monopoles we can also
Equations (19) we get have a current iB of magnetic monopoles, and this
asymmetry can be removed.
closed
E ⋅ dA = Q in (22a)
surface
Exercise 3
Assume that the magnetic monopole has been dis-
closed
B ⋅ dA = 0 (22b) covered, and that we now have magnetic charge QB
surface and a current iB of magnetic charge. Correct Maxwell’s
Equations (22) and the Lorentz force law (20) to include
the magnetic monopole. For each new term you add to
dΦE
B⋅d = i + (22c) these equations, provide a worked-out example of its
dt use.
dΦB In this exercise, use symmetry to guess what terms
E⋅d = – (22d) should be added. If you want to go beyond what we are
dt
asking for in this exercise, you can start with the formula
Stripping out µ0 and ε 0 gives a clearer picture of what FB = QB B for the magnetic force on a magnetic charge,
Maxwell’s equations are trying to say. Equation (22a) and with the kind of thought experiments we used in the
tells us that electric charge is the source of divergent chapter on magnetism, derive the formula for the elec-
electric fields. Equation (22b) says that we haven’t tric force on a magnetic charge QB . You will also end
found any source for divergent magnetic fields. Equa- up with a derivation of the correction to Faraday’s law
tion (22c) tells us that an electric current or a changing caused by a current of magnetic charge. (This is more
electric flux is a source for solenoidal magnetic fields, of a project than an exercise.)
and (22d) tells us that a changing magnetic flux creates
a solenoidal electric field.
32-10 Maxwell's Equations

MAXWELL’S EQUATIONS Equations (24a, b) suggest a coupling between electric

IN EMPTY SPACE and magnetic fields. Let us first discuss this coupling
in a qualitative, somewhat sloppy way, and then work
In the remainder of this chapter we will discuss the
out explicit examples to see precisely what is happen-
behavior of electric and magnetic fields in empty space
ing.
where there are no charges or currents. A few chapters
ago, there would not have been much point in such a Roughly speaking, Equation (24a) tells us that a chang-
discussion, for electric fields were produced by charges, ing electric flux or field creates a magnetic field, and
magnetic fields by currents, and without charges and (24b) tells us that a changing magnetic field creates an
currents, we had no fields. But with Faraday’s law, we electric field. These fields interact, and in some sense
see that a changing magnetic flux dΦB dt acts as the support each other.
source of a solenoidal electric field. And with the
correction to Ampere’s law, we see that a changing If we were experts in integral and differential equa-
electric flux is a source of solenoidal magnetic fields. tions, we would look at Equations (24) and say, “Oh,
Even without charges and currents we have sources for yes, this is just one form of the standard wave equation.
both electric and magnetic fields. The solution is a wave of electric and magnetic fields
traveling through space.” Maxwell was able to do this,
First note that if we have no electric charge (or mag- and solve Equations (24) for both the structure and the
netic monopoles), then we have no sources for either a speed of the wave. The speed turns out to be c, and he
divergent electric or divergent magnetic field. In guessed that the wave was light.
empty space diverging fields do not play an important
role and we can focus our attention on the equations for Because the reader is not expected to be an expert in
the solenoidal magnetic and solenoidal electric field, integral and differential equations, we will go slower,
namely Ampere’s and Faraday’s laws. working out specific examples to see what kind of
structures and behavior we do get from Equations (24).
Setting i = 0 in Equation (19c), the Equations (19c) and We are just beginning to touch upon the enormous
(19d) for the solenoidal fields in empty space become subject of electromagnetic radiation.
dΦE
B⋅d = µ0ε0 (23a)
dt A Radiated Electromagnetic Pulse
dΦB We will solve Equations (24) the same way we have
E⋅d = – (23b) been solving all equations involving derivatives or
dt integrals—by guessing and checking. The rules of the
We can make these equations look better if we write game are as follows. Guess a solution, then apply
µ0ε0 as 1/c 2 , where c = 3 × 10 8 m/s as determined in Equations (24) to your guess in every possible way you
our LC circuit experiment. Then Equations (23) be- can think of. If you cannot find an inconsistency, your
come guess may be correct.

In order to guess a solution, we want to pick an example

1 dΦE that we know as much as possible about and use every
B⋅d = 2 Maxwell's (24a)
c dt equations
insight we can to improve our chances of getting the
in empty right answer. Since we are already familiar with the
dΦB fields associated with a current in a wire, we will focus
E⋅d = – space (24b)
dt on that situation. Explicitly, we will consider what
happens, what kind of fields we get, when we first turn
on a current in a wire. We will see that a structure of
magnetic and electric fields travels out from the wire,
in what will be an example of a radiated electromag-
netic pulse or wave.
 32-11

A Thought Experiment The pure, expanding magnetic field shown in Figure

Let us picture a very long, straight, copper wire with no (13) seems like a good guess. But it is wrong, as we can
current in it. At time t = 0 we start an upward directed see if we apply Ampere’s law to Path (a) which has not
current i everywhere in the wire as shown in Figure yet been reached by the growing magnetic field. For
(12). This is the tricky part of the experiment, having this path that lies outside the magnetic field,
the current i start everywhere at the same time. If we B ⋅ d = 0 , and the corrected Ampere’s law, Equa-
closed a switch, the motion of charge would begin at the tion (19c), gives
switch and advance down the wire. To avoid this,
imagine that we have many observers with synchro- dΦE
B ⋅ d = µ0 i + µ0 ε0 =0 (25)
nized watches, and they all reach into the wire and start dt
the positive charge moving at t = 0. However you want
to picture it, just make sure that there is no current in the In our picture of Figure (13) we have no electric field,
wire before t = 0, and that we have a uniform current i therefore ΦE = 0 and Equation (25) implies that µ0i is
afterward. zero, or the current i through Path (a) is zero. But the
current is not zero and we thus have an inconsistency.
In our previous discussions, we saw that a current i in The growing magnetic field of Figure (13) is not a
a straight wire produced a circular magnetic field of solution of Maxwell’s equations. (This is how we play
magnitude B = µ0i/2πr everywhere outside the the game. Guess and try, and this time we failed.)
wire. This cannot be the solution we need because it
implies that as soon as the current is turned on, we have Equation (25) gives us a hint of what is wrong with our
a magnetic field throughout all of space. The existence guess. It says that
of the magnetic field carries the information that we
have turned on the current. Thus the instantaneous dΦE i
= - (25a)
spread of the field throughout space carries this infor- dt ε0
mation faster than the speed of light and violates the
principle of causality. As we saw in Chapter 1, we thus if we have a current i and have the growing
could get answers to questions that have not yet been magnetic field shown in Figure (13) we must also have
asked. a changing electric flux ΦE through Path (a). Some-
where there must be an electric field E to produce the
Using our knowledge of special relativity as a guide,
changing flux ΦE , a field that points either up or down,
we suspect that the solution B = µ0i/2πr every-
passing through the circular path of Figure (13).
where in space, instantaneously, is not a good guess. A
more reasonable guess is that the magnetic field grows
at some speed v out from the wire. Inside the growing v path (a)
front, the field may be somewhat like its final form
B = µ0i/2πr , but outside we will assume B = 0. v
v
B B=0
i
i up at t = 0
Figure 12 v
A current i is started v
all along the wire at
time t = 0.
v
Figure 13
As a guess, we will assume that the magnetic
field expands at a speed v out from the wire,
when the current is turned on.
32-12 Maxwell's Equations

In our earlier discussion of inductance and induced In Figure (15), we have sketched a field structure
voltage, we saw that a changing current creates an consisting of a circular magnetic field and a downward
electric field that opposed the change. This is what electric field that started out at the wire and is expand-
gives an effective inertia to the current in an inductor. ing radially outward at a speed v as shown. This
Thus when we suddenly turn on the upward directed structure has not yet expanded out to our Path (1), so
current as shown in Figure (12), we expect that we that the line integral B ⋅ d is still zero and Ampere’s
should have a downward directed electric field as law still requires that
indicated in Figure (14), opposing our trying to start the dΦE Path 1 of
current. 0 = µ0 i + µ0 ε0
dt Figure 15
Initially the downward directed electric field should be
inside the wire where it can act on the current carrying dΦE i
= – (26)
charges. But our growing circular magnetic field dt ε0
shown in Figure (13) must also have started inside the which is the same as Equation (25).
wire. Since a growing magnetic field alone is not a
solution of Maxwell’s equations and since there must Looking at Figure (15), we see that the downward
be an associated electric field, let us propose that both electric field gives us a negative flux ΦE through our
the circular magnetic field of Figure (13), and the path. (We chose the direction of the path so that by the
downward electric field of Figure (14) grow together as right hand convention, the current i is positive.) And as
shown in Figure (15). the field structure expands, we have more negative flux
through the path. This increasing negative flux is just
what is required by Equation (26).

i
i up, increasing

B
Figure 15 path (1)
As a second guess,
Figure 14
we will assume
When a current starts that there is a v v
E E downward directed
up, it is opposed by an
electric field. electric field
associated with the
expanding
magnetic field.
Again, Path (1) is
out where the
fields have not yet
arrived.
E E
 32-13

What happens when the field structure gets to and The problems associated with the electric field dying
passes our path? The situation suddenly changes. Now away can be avoided if we turn on the current at time
we have a magnetic field at the path, so that B ⋅ d t = 0, and then shut it off a very short time later. In that
is no longer zero. And now the expanding front is case we should expect to see an expanding cylindrical
outside our path so that the expansion no longer con- shell of electric and magnetic fields as shown in Figure
tributes to – dΦE dt . The sudden appearance of (16). The front of the shell started out when the current
B ⋅ d is precisely compensated by the sudden loss was turned on, and the back should start out when the
of the dΦE /dt due to expansion of the field structure. current is shut off. We will guess that the front and back
should both travel radially outward at a speed v as
The alert student, who calculates E ⋅ d for some shown.
paths inside the field structure of Figure (15) will
discover that we have not yet found a completely
satisfactory solution to Maxwell’s equations. The
electric fields in close to the wire eventually die away,
and only when they have gone do we get a static
magnetic field given by B ⋅ d = µ0 i + 0 .

Figure 16
Electromagnetic
pulse produced
by turning the
current on and v v
then quickly off. v v
We will see that
this structure
agrees with
Maxwell's
equations.

E E
32-14 Maxwell's Equations

Speed of an Electromagnetic Pulse In order to apply Maxwell’s equations to the fields in

Let us use Figure (16), redrawn as Figure (17a), as our Figure (17a), we will focus our attention on a small
best guess for the structure of an electromagnetic pulse. piece of the shell on the right side that is moving to the
The first step is to check that this field structure obeys right at a speed v. For this analysis, we will use the two
Maxwell’s equations. If it does, then we will see if we paths labeled Path (1) and Path (2). Path (1) has a side
can solve for the speed v of the wave front. parallel to the electric field, and will be used for
Equation 23b. Path (2) has a side parallel to the mag-
In Figure (17a), where we have shut the current off, netic field, and will be used for Equation 23a.
there is no net charge or current and all we need to
consider is the expanding shell of electric and magnetic Analysis of Path 1
fields moving through space. We have no divergent In Figure (17b), we have a close up view of Path (1).
fields, no current, and the equations for E and B The path was chosen so that only the left edge of length
become h was in the electric field, so that
dΦE
B ⋅ d = µ0ε0 (23a)
dt E ⋅ d = Eh (27)

dΦB In order to make E ⋅ d positive on this left edge, we

E⋅d = – (23b) went around Path (1) in a counterclockwise direction.
dt
By the right hand convention, any vector up through
which we wrote down earlier as Maxwell’s equation this path is positive, therefore the downward directed
for empty space. magnetic field is going through Path (1) in a negative
direction. (We will be very careful about signs in this
path (1) discussion.)
B

v
path 1
path (2)

h
v v
v v
B into paper
v

E E E
Figure 17a Figure 17b
In order to analyze the electromagnetic pulse Side view showing path (1). An increasing (negative)
produced by turning the current on and off, we magnetic flux flows down through Path (1).
introduce the two paths shown above. Path (1)
has one side parallel to the electric field, while
Path (2) has a side parallel to the magnetic field.
 32-15

In Figure (18), we are looking at Path (1), first at a time We now have a formula for E ⋅ d (Equation 27) and
t (18a) where the expanding front has reached a for dΦE /dt (Equation 29) which we can substitute into
position x as shown, then at a time t + ∆t where the front Faraday’s law (23b) to get
has reached x + ∆x. Since the front is moving at an dΦB
assumed speed v , we have E⋅d = –
dt
v = ∆x
∆t Eh = – ( –Bvh) = +Bhv
At time t + ∆t , there is additional magnetic flux through The factor of h cancels and we are left with
Path (1). The amount of additional magnetic flux ∆ΦB
is equal to the strength B of the field times the additional E = Bv
from
(30)
area ( h∆x ). Since B points down through Path (1), in Faraday's law
a negative direction, the additional flux is negative and which is a surprisingly simple relationship between the
we have strengths of the electric and magnetic fields.
∆Φ B = –B(∆A⊥) = –B(h∆x) (28)
Dividing Equation (28) through by ∆t, and taking the
limit that ∆t goes to dt, gives
∆ΦB ∆x
= –Bh
∆t ∆t
dΦB dx
= –Bh = –Bhv (29)
dt dt

x
B(down)

a) path 1
v

at time t

additional area = h∆x

x ∆x

b) path 1
h v

at time t+∆t
Figure 18
As the front expands, there is more
magnetic flux down through Path (1).
32-16 Maxwell's Equations

Analysis of Path 2 In Figure (19), we show the expanding front at time t

Path (2), shown in Figure (17c), is chosen to have one (19a) and at time t + ∆t (19b). The increase in electric
side in and parallel to the magnetic field. We have gone flux ∆ΦE is (E) times the increased area ( h∆x )
around clockwise so that B and d point in the same
direction. Integrating B around the path gives ∆ΦE = E h∆x

B ⋅d = Bh (31) Dividing through by ∆t, and taking the limit that ∆t goes
to dt, gives
Combining Equation 31 with Ampere's law
∆ΦE ∆x
dΦ E = Eh
B ⋅d = µ0ε0 ∆t ∆t
dt dΦE dx
gives = Eh = Ehv (33)
dt dt
dΦ E Using Equation 33 in 32, and then cancelling h, gives
Bh = µ0 ε 0 (32)
dt Bh = µ 0 ε 0 E hv
To evaluate dΦ E dt , we first note that for a clockwise (34)
From
path, the positive direction is down into the paper in B = µ 0ε 0 E v Ampere's law
Figure (17c). This is the same direction as the electric
field, thus we have a positive electric flux through path which is another simple relationship between E and B.
(2).
path (1)
B
B

v
path (2)
path 2

v v
v v h

E into
v paper

E E

Figure 17a (repeated) Figure 17c

We will now turn our attention to path 2 which An increasing (negative) electric flux
has one side parallel to the magnetic field. flows down through Path (2).
 32-17

If we divide Equation (30) B v = E , by Equation (34) Exercise 4

B = µ0 ε 0 Ev, both E and B cancel giving Construct paths like (1) and (2) of Figure (17), but which
Bv E include the back side, rather than the front side, of the
= electromagnetic pulse. Repeat the kind of steps used
B µ0ε0 E v
to derive Equation (35) to show that the back of the pulse
1 also travels outward at a speed v = 1/ µ0ε0 . As a
v2 = result the pulse maintains its thickness as it expands out
µ0ε0
through space.

1 speed of Exercise 5
v =
µ0ε0 light!!! (35) After a class in which we discussed the electromagnetic
pulse shown in Figure (20a), a student said she thought
that the electric field would get ahead of the magnetic
Thus the electromagnetic pulse of Figures (16) and (17) field as shown in Figure (20b). Use Maxwell's equations
expands outward at the speed 1/ µ0ε0 which we have to show that this does not happen.
seen is 3 × 10 8 meters per second. Maxwell recog-
nized that this was the speed of light and recognized
that the electromagnetic pulse must be closely related B
to light itself.

Using v = 1/ µ0ε0 = c in Equation (34) we get

v
B =E
c (36)
as the relative strength of the electric and magnetic
fields in an electromagnetic pulse, or as we shall see,
any light wave. If we had used a reasonable set of units E
Figure 20a
where c = 1 (like feet and nanoseconds), then E and B
The radiated electromagnetic pulse we
would have equal strengths in a light wave. saw in Figures (16) and (17).
x
E(down)
path 2 Path 1
a) v

at time t
B

v
additional area = h∆x
x ∆x

path 2 Figure 20b

E
b) h v
The student guessed that the electric fields would
get out ahead of the magnetic field. Use Path (1)
at time t+∆t to show that this does not happen.
Figure 19
As the front expands, there is more
electric flux down through Path (2).
32-18 Maxwell's Equations

ELECTROMAGNETIC WAVES If we used a sinusoidally oscillating current in the wire

The single electromagnetic pulse shown in Figure (17) of Figure (22), then the series of electromagnetic pulses
is an example of an electromagnetic wave. We usually would blend together to form the sinusoidally varying
think of a wave as some kind of oscillating sinusoidal electric and magnetic fields structure shown in Figure
thing, but as we saw in our discussion of waves on a (23b). This is the wave structure one usually associates
Slinky in Chapter 1, the simplest form of a wave is a with an electromagnetic wave.
single pulse like that shown in Figure (21). The basic When you think of an electromagnetic wave, picture
feature of the Slinky wave pulse was that it maintained the fields shown in Figure (23), moving more or less as
its shape while it moved down the Slinky at the wave a rigid object past you at a speed c. The distance λ
speed v . Now we see that the electromagnetic pulse between crests is called the wavelength of the wave.
maintains its structure of E and B fields while it moves The time T it takes one wavelength or cycle to pass you
at a speed v = c through space. is
We made a more or less sinusoidal wave on the λ meter
= λ second
cycle
Slinky by shaking one end up and down to produce T second = meter (37)
a series of alternate up and down pulses that traveled cycle c c cycle
second
together down the Slinky. Similarly, if we use an
alternating current in the wire of Figure (17), we will T is called the period of the wave. The frequency of
get a series of electromagnetic pulses that travel out the wave, the number of wavelengths or full cycles of
from the wire. This series of pulses will more the wave that pass you per second is
closely resemble what we usually think of as an
c meter
electromagnetic wave. f
cycle
= second = λ cycle (38)
second λ meter c second
Figure (22a) is a graph of a rather jerky alternating cycle
current where we turn on an upward directed current of In Equations (37) and (38) we gave λ the dimensions
magnitude i 0, then shut off the current for a while, then meters/cycle, T of seconds/cycle and f of cycles/sec-
turn on a downward directed current i 0, etc. This series ond so that we can use the dimensions to remember the
of current pulse produces the series of electromagnetic
pulses shown in Figure (22b). Far out from the wire i
where we can neglect the curvature of the magnetic i 0 up
a) Graph of current
field, we see a series of pulses shown in the close-up –t
pulses in wire
view, Figure (23a). This series of flat or non-curved –i 0 down
pulses is called a plane wave of electromagnetic radia-
tion. B
c c

E E

Figure 21
Slinky wave pulse. b) Resulting electric and magnetic fields
Figure 22
Fields produced by a series of current pulses.
 32-19

formulas T = λ /c, f = c/ λ . (It is now common to use You can remember where the 2π goes by giving it the
“hertz” or “Hz” for the dimensions of frequency. This dimensions 2π radians/cycle. (Think of a full circle or
is a classic example of ruining simple dimensional full cycle as having 2π radians.) We will indiscrimi-
analysis by using people’s names.) Finally, the angular nately use the word frequency to describe either
frequency ω radians per second is defined as f cycles/second or ω radians/second, whichever is
radians radians cycles more appropriate. If, however, we say that something
ω = 2π ×f has a frequency of so many hertz, as in 60 Hz, we will
sec ond cycle second always mean cycles/second.
radians (39)
= 2πf
second

Fields move as a fixed unit

at the speed of light.

E c c
Electric
B field
Magnetic
field

a) Electric and magnetic fields produced

by abruptly switching the antenna current.

λ
One wavelength l = the distance between similar crests

c c

b) Electric and magnetic fields produced

by smoothly switching the antenna current.

Figure 23
Structure of electric and magnetic fields in light and radio waves.
32-20 Maxwell's Equations

ELECTROMAGNETIC SPECTRUM c meters

λ meters = second = 3 × 10 8m/s
We have seen by direct calculation that the electromag- cycle cycles 103 × 10 6c/s
netic pulse of Figure (17), and the series of pulses in f
sec ond
Figure (22) are a solution of Maxwell’s equations. It is
= 297 meters
not much of an extension of our work to show that the
sinusoidal wave structure of Figure (23b) is also a Such waves were discovered within 10 years of
solution. The fact that all of these structures move at a Maxwell’s theory, and were called radio waves. The
speed c = 1/ µ0ε0 = 3 × 10 8 m/s is what suggested frequency 1030 kc is the frequency of radio station
to Maxwell that these electromagnetic waves were WBZ in Boston, Mass.
light, that he had discovered the theory of light.

But there is nothing in Maxwell’s equations that re- Components of the

stricts our sinusoidal solution in Figure (23b) to certain Electromagnetic Spectrum
values or ranges of frequency or wavelength. One Figure (24) shows the complete electromagnetic spec-
hundred years before Maxwell it was known from trum as we know it today. We have labeled various
interference experiments (which we will discuss in the components that may be familiar to the reader. These
next chapter) that light had a wave nature and that the components, and the corresponding range of wave-
wavelengths of light ranged from about 6 × 10 –5 cm in lengths are as follows:
the red part of the spectrum down to 4 × 10 –5 cm in the
blue part. With the discovery of Maxwell’s theory of Radio Waves 106 m to .05 mm
light, it became clear that there must be a complete AM Band 500 m to 190 m
spectrum of electromagnetic radiation, from very long Short Wave 60 m to 15 m
down to very short wavelengths, and that visible light
TV VHF Band 10 m to 1 m
was just a tiny piece of this spectrum.
TV UHF Band 1 m to 10 cm
More importantly, Maxwell’s theory provided the clue Microwaves 10 cm to .05 mm
as to how you might be able to create electromagnetic
Infrared Light .05 mm to 6 ×10 – 5 cm
waves at other frequencies. We have seen that an
oscillating current in a wire produces an electromag- Visible Light 6 ×10 – 5 to 4 ×10 – 5cm
netic wave whose frequency is the same as that of the Ultraviolet Light 4 ×10 – 5 cm to 10 – 6cm
current. If, for example, the frequency of the current is X Rays 10 –6 cm to 10 –9 cm
1030 kc (1030 kilocycles) = 1.03 × 10 6 cycles/sec,
γ Rays 10 –9 cm and shorter
then the electromagnetic wave produced should have a
wavelength

visible ultraviolet
radio, television, radar light rays gamma rays

wavelength, cm

106 5 4 3 2 10 1 10 -1 -2 -3 -4 -5 -6 -7 -8 -9 -10 -11 -12

infrared rays X-rays

Figure 24
The electromagnetic spectrum extends from long wavelength radio waves down to short wavelength
X rays and gamma rays. The visible part of the spectrum is indicated by the small box.
 32-21

In each of these ranges, the most efficient way to emit meters, while those for UHF are of the order of a foot.
or detect the radiation is to use antennas whose size is Those with separate VHF and UHF television antennas
comparable to the wavelength of the radiation. For will be familiar with the fact that the UHF antenna,
radio waves the antennas are generally some kind of a which detects the shorter wavelengths, is smaller in
structure made from wire. In the infrared and the size.
visible region, radiation is generally emitted by mol-
ecules and atoms. The short wavelength x rays and γ Adjusting the rabbit ears antenna on a television set
rays generally come from atomic nuclei or subatomic provides practical experience with the problems of
particles. detecting an electromagnetic wave. As the TV signal
strikes the antenna, the electric field in the wave acts on
The longest wavelength radio waves that have been the electrons in the TV antenna wire. If the wire is
studied are the so-called “whistlers”, radio waves with parallel to the electric field, the electrons are pushed
an audio frequency, that are produced by lightening along in the wire producing a voltage that is detected by
bolts and reflected back and forth around the earth by the television set. If the wire is perpendicular, the
charged particles trapped in the earth’s magnetic field. electrons will not be pushed up and down and no
On a shorter scale of distance are the long wavelength voltage will be produced.
radio waves which penetrate the ocean and are used for
communications with submarines. The radio station in The length of the wire is also important. If the antenna
Cutler, Maine, shown in Figure (25), has twenty-six were one half wavelength, then the electric field at one
towers over 1000 feet tall to support the antenna to end would be pushing in the opposite direction from the
produce such waves. This station, operated by the field at the other end, the integral E ⋅ dl down the
United States Navy, is the world’s most powerful. antenna would be zero, and you would get no net
voltage or signal. You want the antenna long enough
As we go to shorter wavelengths and smaller antennas, to get a big voltage, but not so long that the electric field
we get to the broadcast band, short wave radio, then to in one part of the antenna works against the field in
the VHF and UHF television frequencies. (FM radio another part. One quarter wavelength is generally the
is tucked into the VHF band next to Channel 6). The optimum antenna length.
wavelengths for VHF television are of the order of

Figure 25
The worlds largest radio station at Cutler, Maine. This structure, with 75 miles of antenna wire and 26 towers
over 1000 ft high, generates long wavelength low frequency, radio waves for communications with submarines.
32-22 Maxwell's Equations

The microwave region, now familiar from microwave All objects emit blackbody radiation. You, yourself,
communications and particularly microwave ovens, are like a small star emitting infrared radiation at a
lies between the television frequencies and infrared wavelength corresponding to a temperature of 300K.
radiation. The fact that you heat food in a microwave In an infrared photograph taken at night, you would
oven emphasizes the fact that electromagnetic radia- show up distinctly due to this radiation. Infrared
tion carries energy. One can derive that the energy photographs are now taken of houses at night to show
density in an electromagnetic wave is given by the up hot spots and heat leaks in the house.
formula
Perhaps the most famous example of blackbody radia-
tion is the 3K cosmic background radiation which is the
energy density in an 2 remnant of the big bang which created the universe.
ε0E B2
electromagnetic = + (37) We will say much more about this radiation in Chapter
2 2µ0 34.
wave

We have already seen the first term ε0E2 /2, when we UV, X Rays, and Gamma Rays
calculated the energy stored in a capacitor (see When we get to wavelengths shorter than the visible
Equation 27-36 on page 27-19). If we had calculated spectrum, and even in the visible spectrum, we begin to
the energy to start a current in an inductor, we would run into problems with Maxwell’s theory of light.
have gotten the formula B2 /2µ0 for the energy These problems were first clearly displayed by Max
density in that device. Equation (37) tells us the Planck who in 1900 developed a theory that explained
amount of energy is associated with electric and the blackbody spectrum of radiation. The problem
magnetic fields whenever we find them. with Planck’s theory of blackbody radiation is that it
could not be derived from Maxwell’s theory of light
and Newtonian mechanics. His theory involved arbi-
Blackbody Radiation trary assumptions that would not be understood for
Atoms and molecules emit radiation in the infrared, another 23 years, until after the development of quan-
visible and ultraviolet part of the spectrum. One of the tum mechanics.
main sources of radiation in this part of the spectrum is
the so-called blackbody radiation emitted by objects Despite the failure of Newton’s and Maxwell’s theo-
due to the thermal motion of their atoms and molecules. ries to explain all the details, the electromagnetic
spectrum continues right on up into the shorter wave-
If you heat an iron poker in a fire, the poker first gets lengths of ultraviolet (UV) light, then to x rays and
warm, then begins to glow a dull red, then a bright red finally to γ rays. Ultraviolet light is most familiar from
or even, orange. At higher temperatures the poker the effect it has on us, causing tanning, sunburns, and
becomes white, like the filaments in an electric light skin cancer depending on the intensity and duration of
bulb. At still higher temperatures, if the poker did not the dose. The ozone layer in the upper atmosphere, as
melt, it would become bluish. The name blackbody long as it lasts, is important because it filters out much
radiation is related to the fact that an initially cold, of the ultraviolet light emitted by the sun.
black object emits these colors of light when heated.
X rays are famous for their ability to penetrate flesh and
There is a well studied relationship between the tem- produce photographs of bones. These rays are usually
perature of an object and the predominant frequency of emitted by the tightly bound electrons on the inside of
the blackbody radiation it emits. Basically, the higher large atoms, and also by nuclear reactions. The highest
the temperature, the higher the frequency. Astrono- frequency radiation, γ rays, are emitted by the smallest
mers use this relationship to determine the temperature objects—nuclei and elementary particles.
of stars from their color. The infrared stars are quite
cool, our yellow sun has about the same temperature as
the yellow filament in an incandescent lamp, and the
blue stars are the hottest.
 32-23

POLARIZATION Because E must lie in the plane perpendicular to the

One of the immediate tests of our picture of a light or direction of motion of an electromagnetic wave, E has
radio wave, shown in Figure (23), is the phenomena of only two independent components, which we can
polarization. We mentioned that the reason that you call the vertical and horizontal polarizations, or the x
had to adjust the angle of the wires on a rabbit ears and y polarizations as shown in Figures (27a) and (27b)
antenna was that the electric field of the television respectively. If we happen to encounter an electromag-
signal had to have a significant component parallel to netic wave where E is neither vertical or horizontal,
the wires in order to push the electrons up and down the but at some angle θ, we can decompose E into its x and
wire. Or, in the terminology of the last few chapters, we y components as shown in (27c). Thus we can consider
needed the parallel component of E so that the voltage a wave polarized at an arbitrary angle θ as a mixture of
the two independent polarizations.
V = E ⋅ dl would be large enough to be detected by
the television circuitry. (In this case, the line integral
E ⋅ dl is along the antenna wire.) Polarization is a Electric
E
Magnetic
field field
phenomena that results from the fact that the electric
B
field E in an electromagnetic wave can have various
orientations as the wave moves through space.

Although we have derived the structure of an electro- c c

magnetic wave for the specific case of a wave produced a) Vertically polarized electromagnetic wave.

by an alternating current in a long, straight wire, some

Magnetic Electric
of the general features of electromagnetic waves are field B field
clearly present in our solution. The general features E
that are present in all electromagnetic waves are:

1) All electromagnetic waves are a structure consist-

c c
ing of an electric field E and a magnetic field B . b) Horizontally polarized electromagnetic wave.

2) E and B are at right angles to each other as Figure 26

shown in Figure (23). Two possible polarizations of
an electromagnetic wave.
3) The wave travels in a direction perpendicular to
the plane of E and B .

4) The speed of the wave is c = 3 × 10 8 m/s . E

5) The relative strengths of E and B are given by

Equation (36) as B = E/c.

Even with these restrictions, and even if we consider a) Vertical Polarization

only flat or plane electromagnetic waves, there are still Figure 27
various possible orientations of the electric field as We define the direction
of polarization of an E
shown in Figure (26). In Figure (26a) we see a plane electromagnetic wave as
wave with a vertical electrical field. This would be the direction of the b) Horizontal Polarization

called a vertically polarized wave. In Figure (26b), electric field.

where the electric field is horizontal, we have a hori- Ey E
zontally polarized wave. By convention we say that the θ
Ex
direction of polarization is the direction of the electric
field in an electromagnetic wave. c) Mixture
32-24 Maxwell's Equations

Polarizers If you then rotate the wires 90º, so that the E field in the
A polarizer is a device that lets only one of the two wave is perpendicular to the wires, the electric field can
possible polarizations of an electromagnetic wave pass no longer move electrons along the wires and the wires
through. If we are working with microwaves whose have no effect. The wave passes through without
wavelength is of the order of a few centimeters, a frame attenuation.
strung with parallel copper wires, as seen in Figure
(28), makes an excellent polarizer. If a vertically If you do not happen to know the direction of po-
polarized wave strikes this vertical array of wires, the larization of the microwave, put the polarizer in the
electric field E in the wave will be parallel to the wires. beam and rotate it. For one orientation the microwave
This parallel E field will cause electrons to move up beam will be completely blocked. Rotate the polarizer
and down in the wires, taking energy out of the incident by 90° and you will get a maximum transmission.
wave. As a result the vertically polarized wave cannot
get through. (One can observe that the wave is actually
reflected by the parallel wires.)

Figure 28
Microwave polarizer, made from an array of copper wires. The microwave
transmitter is seen on the other side of the wires, the detector is on this side.
When the wires are parallel to the transmitted electric field, no signal is
detected. Rotate the wires 90 degrees, and the full signal is detected.
 32-25

Light Polarizers Since light from the sun or from standard electric light
We can picture light from the sun as a mixture of light bulbs consists of many randomly polarized waves, a
waves with randomly oriented polarizations. (The E single sheet of Polaroid removes half of the waves no
fields are, of course, always in the plane perpendicular matter how we orient the Polaroid (as long as the sheet
to the direction of motion of the light wave. Only the of Polaroid is perpendicular to the direction of motion
angle in that plane is random.) A polarizer made of an of the light beam). But once the light has gone through
array of copper wires like that shown in Figure (28), one sheet of Polaroid, all the surviving light waves have
will not work for light because the wavelength of light the same polarization. If we place a second sheet of
is so short λ ≈ 5 × 10- 5 cm that the light passes right Polaroid over the first, all the light will be absorbed if
between the wires. For such a polarizer to be effective, the long molecules in the second sheet are perpendicu-
the spacing between the wires would have to be of the lar to the long molecules in the first sheet. If the long
order of a wavelength of light or less. molecules in the second sheet are parallel to those in the
first, most of the waves that make it through the first,
A polarizer for light can be constructed by imbedding make it through the second also. This effect is seen
long-chain molecules in a flexible plastic sheet, and clearly in Figure (29).
then stretching the sheet so that the molecules are
aligned parallel to each other. The molecules act like
the wires in our copper wire array, but have a spacing
of the order of the wavelength of light. As a result the
molecules block light waves whose electric field is
parallel to them, while allowing waves with a perpen-
dicular electric field to pass. (The commercial name for
such a sheet of plastic is Polaroid.)

Figure 29
Light polarizers. Two sheets of polaroid are placed on
top of a drawing. On the left, the axes of the sheets are
parallel, so that nearly half the light passes through. On
the right, the axes are perpendicular, so that no light
passes through. (Photo from Halliday & Resnick)
32-26 Maxwell's Equations

Magnetic Field Detector In a typical scene we see a car with a metal loop
So far, our discussion of electromagnetic radiation has mounted on top as shown in Figure (30a). It is chasing
focused primarily on detecting the electric field in the another car with a hidden transmitter, or looking for a
wave. The rabbit ear antenna wire had to be partially clandestine enemy transmitter.
parallel to the electric field so that E⋅ d and there-
fore the voltage on the antenna would not be zero. In If the transmitter is a radio antenna with a vertical
our discussion of polarization, we aligned the parallel transmitting wire as shown in Figure (30b), the mag-
array of wires or molecules parallel to the electric field netic field of the radiated wave will be concentric
when we wanted the radiation to be reflected or ab- circles as shown. Objects on the ground, the ground
sorbed. itself, and nearby buildings and hills can distort this
picture, but for now we will neglect the distortions.
It is also fairly easy to detect the magnetic field in a
radio wave by using one of our E⋅ d meters to detect circular magnetic field
a changing magnetic flux (an application of Faraday’s radiated by the antenna
law). This is the principle behind the radio direction
finders featured in a few World War II spy pictures.
B v
loop antenna

detector loop
vertical antenna on car

Figure 30a Figure 30b

Car with radio direction finder Car driving toward radio transmitter.
loop antenna mounted on top.

Figure 31
In a January 1998 National Geographic
article on Amelia Earhardt, there
appeared a picture of a vintage Electra
airplane similar to the one flown by
Earhardt on her last trip in 1938. On the
top of the plane, you can see the kind of
radio direction finder we have been
discussing. (The plane is being flown by
Linda Finch.)
 32-27

In Figures (32a) and (32b), we show the magnetic field The most sensitive way to use this radio direction finder
of the radio wave as it passes the detector loop mounted is to get a zero or “null” reading on the voltmeter. Only
on the car. A voltmeter is attached to the loop as shown when the loop is oriented as in Figure (32a), with its
in Figure (33). In (32a), the plane of the loop is parallel plane perpendicular to the direction of motion of the
to B, the magnetic flux ΦB through the loop is zero, and radio wave, will we get a null reading. At any other
Faraday’s law gives orientation some magnetic flux will pass through the
dΦ B loop and we get some voltage.
V = E⋅d = = 0
dt Spy pictures, set in more modern times, do not show
antenna loops like that in Figure (30) because modern
In this orientation there is no voltage reading on the
radio direction finders use so-called “ferrite” antennas
voltmeter attached to the loop.
that detect the electric field in the radio wave. We get
In the orientation of Figure (32b), the magnetic field a voltage on a ferrite antenna when the electric field in
passes through the loop and we get a maximum amount the radio wave has a component along the ferrite rod,
of magnetic flux ΦB. As the radio wave passes by the just as it needed a component along the wires of a rabbit
loop, this flux alternates signs at the frequency of the ears antenna. Again these direction finders are most
wave, therefore the rate of change of flux dΦB /dt is at accurate when detecting a null or zero voltage. This
a maximum. In this orientation we get a maximum occurs only when the rod is parallel to the direction of
voltmeter reading. motion of the radio wave, i.e. points toward the station.
(This effect is very obvious in a small portable radio.
You will notice that the reception disappears and you
get a null detection, for some orientations of the radio.)
Magnetic Electric
field B field
E c

metal loop

c
a) Loop oriented so that no
magnetic flux goes through it E⋅ d meter

B
E c V
voltmeter

Figure 33
c We can think of a wire loop connected
b) Loop oriented so that
to a voltmeter as an E⋅d meter. Any
magnetic flux goes through it changing magnetic flux through the
Figure 32 loop induces a voltage around the loop.
Electromagnetic field impinging upon a loop antenna. This voltage is read by the voltmeter.
In (a), the magnetic field is parallel to the plane of the
loop, and therefore no magnetic flux goes through the
loop. In (b), the magnetic flux goes through the loop.
As the wave passes by, the amount of flux changes,
inducing a voltage in the loop antenna.
32-28 Maxwell's Equations

RADIATED ELECTRIC FIELDS At time t = t1, we stop moving the positively charged
One of the best computer simulations of physical rod. The information that the charged rod has stopped
phenomena is the series of short films about the electric moving cannot travel faster than the speed of light, thus
fields produced by moving and accelerated charges. the displaced radial field next to the rod cannot be any
We will describe a few of the frames from these films, farther out than a distance c t- t1 as shown in Figure
but nothing replaces watching them. (34c). The effect of starting, then stopping the positive
rod is an outward traveling kink in the electric field
Two basic ideas underlie these films. One is Gauss’ lines. It is as if we had ropes attached to the positive rod,
law which requires that electric field lines not break, and jerking the rod produced an outward traveling kink
do not end, in empty space. The other is that distur- or wave on the ropes.
bances on an electric field line travel outward at the
speed of light. No disturbance, no change in the electric In Figure (34d), we have added in a stationary nega-
field structure, can travel faster than the speed of light tively charged rod and the inward directed electric field
without violating causality. (You could get answers to produced by that rod. The charge density on the
questions that have not yet been asked.) negative rod is opposite that of the positive rod, so that
there is no net charge on the two rods. When we
As an introduction to the computer simulations of combine these rods, all we have left is a positive
radiation, let us see how a simple application of these upward directed current during the time interval t = 0 to
two basic ideas leads to the picture of the electromag- t = t1. We have a short current pulse, and the electric
netic pulse shown back in Figure (16). In Figure (34a) field produced by the current pulse must be the vector
we show the electric field of a stationary, positively sum of the electric fields of the two rods.
charged rod. The electric field lines go radially out-
ward to infinity. (It’s a long rod, and it has been at rest In Figure (34e), we add up the two electric fields. In the
for a long time.) region r > ct beyond the kink, the positive and negative
fields must cancel exactly. In the region r < t - t1 we
At time t = 0 we start moving the entire rod upward at should also have nearly complete cancellation. Thus
a speed v. By Gauss’ law the electric field lines must all we are left with are the fields E + and E – inside the
stay attached to the charges Q in the rod, so that the ends kink as shown in Figure (34f). Since electric field lines
of the electric field lines have to start moving up with cannot end in empty space, E + and E – must add up to
the rod. produce the downward directed E net shown in Figure
(34g). Note that this downward directed electric field
No information about our moving the rod can travel pulse was produced by an upward directed current
outward from the rod faster than the speed of light. If pulse. As we have seen before, this induced electric
the time is now t > 0, then beyond a distance ct, the field opposes the change in current.
electric field lines must still be radially outward as in
Figure (34b). To keep the field lines radial beyond In Figure (34h) we added the expanding magnetic field
r = ct, and keep them attached to the charges +Q in the pulse that should be associated with the current pulse.
rod, there must be some kind of expanding kink in the What we see is an expanding electromagnetic pulse
lines as indicated. that has the structure shown in Figure (16). Simple
arguments based on Gauss’ law and causality gave us
most of the results we worked so hard to get earlier.
What we did get earlier, however, when we applied
Ampere’s and Faraday’s law to this field structure, was
the explicit prediction that the pulse expands at the
speed 1/ µ0 ε0 = 3 × 108 m/s.
 32-29

+
+ e) When we add up the
+
E– electric fields of the
+ –
+ a) Electric field of a + positive and negative
stationary, positively rods, the fields cancel
t<0 +
+ charged rod. E+ everywhere except at
+ the outward going
+ pulse.
+

+ ct
E–
v ++ b) At time t = 0, we E+
+
start moving the entire f) At the pulse, the
+ rod upward at a speed vector sum of E+ and
+ v. The ends of the field Enet E+ E– is a downward
t>0 + lines must stay directed field Enet as
+ attached to the charges shown.
+ in the rod. E–
+

+ c(t–t1)
+ g) Thus a short
+ upward directed
c) We stop moving the
+
rod at time t = t1 . No c current pulse
t > t1 + produces a downward
+ information about our
Enet directed electric field
+ having moved the rod
that travels outward
+ can travel out faster
from the wire at a
+ than the speed of light.
speed c.
+
ct
B
+
– d) We have added in
+ h) Add in the magnetic
–
+ the electric field of a
– field of the current
+ line of stationary c
–
+
pulse, and we have the
– negative charges. As a
+ electromagnetic wave
– result, the net charge Enet
+ structure seen in
– on the rod is zero and
+ Figure (16).
– we have only a current
+
–
+
pulse that lasted from
– t = 0 to t = t1 .

Figure 34
Using the fact that electric field lines cannot break in empty space (Gauss' Law), and the idea that kinks
in the field lines travel at the speed of light, we can guess the structure of an electromagnetic pulse.
32-30 Maxwell's Equations

Field of a Point Charge

The computer simulations show the electric field of a
point charge under varying situations. In the first, we
see the electric field of a point charge at rest, as shown
in Figure (35a). Then we see a charge moving at
constant velocity v. As the speed of the charge ap-
proaches c, the electric field scrunches up as shown in
Figure (35b). Figure 35a
Electric field of a stationary charge.
The next film segment shows what happens when we
have a moving charge that stops. If the charge stopped
at time t = 0, then at a distance r = ct or greater, we must
have the electric field of a moving charge, because no
information that the charge has stopped can reach
beyond this distance. In close we have the electric field
of a static charge. The expanding kink that connects the
two regions is the electromagnetic wave. The result is
v
shown in Figure (35c). The final film segment shows
the electric field of an oscillating charge. Figure (36)
shows one frame of the film. This still picture does a
serious injustice to the animated film. There is no
substitute, or words to explain, what you see and feel Figure 35b
when you watch this film. Electric field of a moving charge. If the charge has
been moving at constant speed for a long time, the field
is radial, but squeezed up at the top and bottom.

Figure 35c
Field of a charge that stopped. Assume that the
charge stopped t seconds ago. Inside a circle of
radius ct, we have the field of a stationary charge.
Outside, where there is no information that the
charge has stopped, we still have the field of a
moving charge. The kink that connects the two
fields is the electromagnetic radiation.
 32-31

Figure 35c (enlarged)

Electric field of a charge that stopped. The
dotted lines show the field structure we
would have seen had the charge not stopped.

Figure 36
Electric field of an oscillating charge.
32-32 Maxwell's Equations

Exercise 6
Assume that we have a supply of ping pong balls and E
cardboard tubes shown in Figures (37). By looking at
the fields outside these objects decide what could be
inside producing the fields. Explicitly do the following
for each case.
Ball
i) Write down the Maxwell equation which you used to
decide what is inside the ball or tube, and explain how
you used the equation.

ii) If more than one kind of source could produce the

field shown, describe both (or all) sources and show the
appropriate Maxwell equations. Figure 37a
Electric field emerging from ping pong ball.
iii) If the field is impossible, explain why, using a
Maxwell equation to back up your explanation.

In each case, we have indicated whether the source is

in a ball or tube. Magnetic fields are dashed lines,
electric fields are solid lines, and the balls and tubes are B
surrounded by empty space.

Tube
(end view)

Figure 37b
Magnetic field emerging from ping pong ball.

Ball

Figure 37c
Electric field emerging above ping pong ball.
 32-33

x x
x x

x x
x x

x x
x

x
x
x

x
x

x
x

x
E

x
x x
x
x = electric field into paper x
Tube
x x
x x
(end view) x x x x
x x x x
x x Tube x x
(end view)
x x x x
x x x x
x x
x x
Figure 37d
x x
x x
Electric field around tube.

x
x

x
x

x
x

x
x x

x x
x

x x

x x

x
x x
Figure 37f
For this example, explain what is happening to the
fields, what is in the tube, and what happened inside.

Tube E
(end view)

B
Tube
Figure 37e (end view) E
Electric field passing through tube.

Figure 37g
There is only ONE object inside this tube.
What is it? What is it doing?
Chapter 33
Light Waves CHAPTER 33 LIGHT WAVES

Ripples produced by rain drops. (Bill Jack Rodgers, Los Alamos Scientific Laboratory)

In the examples of wave motion we studied back in There are general rules of wave motion that transcend
Chapter 15, like waves on a rope and sound in a gas, the nature of the medium or type of wave. One is the
we could picture the wave motion as a consequence of principle of superposition that we used extensively in
the mechanical behavior particles in the rope or Chapter 15. It is the idea that as waves move through
molecules in the gas. We used Newton's laws to predict each other, they produce an overall wave whose
the speed of a rope wave and could have done the same amplitude is the sum of the amplitudes of the individual
for a sound. When we discuss light waves, we go waves. The other is a concept we will use extensively
beyond the Newtonian behavior. Waves on a rope, on in this chapter called the Huygens principle, named
water or in a gas are mechanical undulations of an after its discoverer Christian Huygens, a contempo-
explicit medium. Light waves travel through empty rary of Isaac Newton.
space; there is nothing to undulate, nothing to which
we can apply Newton's laws. Yet, in many ways, the We will see that a straightforward application of the
behavior of light waves, water waves, sound waves, principle of superposition and the Huygens principle
and even the waves of quantum theory, are remarkably allows us to make detailed predictions that even can be
similar. used as a test of the wave nature of the phenomena we
are studying.
 33-2 Light Waves

SUPERPOSITION OF When we extend our study of wave motion to two and

CIRCULAR WAVE PATTERNS three dimensions, the principle of superposition works
the same way, but now we have to add patterns rather
When we studied the interaction of waves on a rope, it
than just heights along a line. If, for example, we are
was a relatively simple process of adding up the
studying wave motion on the surface of water, and two
individual waves to see what the resultant wave would
wave patterns move through each other, the resulting
be. For example, in Figure (15-6) reproduced here, we
wave is the sum of the heights of the individual waves
see that when a crest and a trough run into each other,
at every point on the surface. We do the same addition
for an instant they add up to produce a flat rope. At this
as we did for one dimensional waves, but at many more
instant the crest and the trough cancel each other. In
points.
contrast two crests add to produce a big crest, and two
troughs add to produce a deeper trough. A relatively simple, but important example of the
superposition of wave patterns is the pattern we get
a) when concentric circular waves from two nearby
sources run into each other. The pattern is easy to set
up in a ripple tank using two oscillating plungers.

Figure (1a) shows the circular wave pattern produced

b)
by a single oscillating plunger. From this picture we
can easily see the circular waves emerging from the
plunger. The only difficulty is distinguishing crests
from troughs. We will handle this by using a solid line
c)
to represent the crest of a wave and a dashed line for a
trough, as illustrated in Figure (1b).
d)

Figure 15-6
When a crest meets a trough, there is
a short time when the waves cancel.

Figure 1a,b
Circular wave pattern
produced in a ripple tank
by a plunger. The pattern
consists of alternate
crests and troughs. To
diagram the circular
wave pattern, we will use
solid lines for crests and
dashed lines for troughs.
a) b)
 33-3

In Figure (2a), we see the wave pattern produced by two Between the lines of nodes we get beams of waves. In
plungers oscillating side by side. Each plunger sends each beam, crests from one plunger meet crests from
out a circular set of waves like that seen in Figure (1). the other producing a higher crest. And troughs from
When the two sets of circular waves cross each other, one set meet troughs from the other producing deeper
we get cancellation where crests from one set meet troughs. In our drawing of circles, Figures (2b) and
troughs from the other set (where a solid line from one (2c), we get beams of waves along the lines where solid
set of circles meets a dashed line from the other set of circles cross solid circles and dashed circles cross
circles in Figure (2b). This cancellation occurs along dashed circles.
lines called lines of nodes which are clearly seen in
Figure (2a).

d es
no
e of
lin

nodes
line of

line of n
odes

line
of
no
de
s

a) b)
Figure 2a,b
Ripple tank photograph of an interference pattern.
When two sets of circular waves move through each
other, there are lines along which crests from one beam
set always meet troughs from the other set. These
are called lines of nodes. Between the lines of node
nodes, we get beams of waves. The resulting pattern
is called an interference pattern. beam

node

beam
Figure 2c
We get beams of waves where crests meet crests
and troughs meet troughs. The lines of nodes are
where crests meet troughs and the waves cancel.
33-4 Light Waves

HUYGENS PRINCIPLE consider the relatively smooth wave front shown in

When sunlight streams in through an open kitchen Figure (4). To predict the position of the wave front a
door, we see a distinct shadow on the floor. The short time later, we treat each point on the front of the
shadow can be explained by assuming that the light wave as a source of circular waves. We can see the
beams travel in straight lines from the sun through the effect by drawing a series of circles at closely spaced
doorway. The whole subject of geometrical optics and points along the wave. The circular waves add up to
lens design is based on the assumption that light travels produce a new wave front farther out. While you can
in straight lines (except at the interface of two media of use the same construction to figure out what is happen-
different indices of refraction). ing throughout the wave, it is much easier to see what
is happening at the front.
In Figure (3) we see what happens when a wave
impinges upon a slit whose width is comparable to the Exercise 1
wavelength of the waves. Instead of there being a At some instant of time, the front of a wave has a sharp,
shadow of the slit, we see that the emerging wave right angle corner. Use Huygens principle to find the
comes out in all directions. The wave pattern on the shape of the wave front at some later instant of time.
right side of the slit is essentially identical to the wave (Draw a right angle corner and use the kind of construc-
pattern produced by the oscillating plunger in Figure tion shown in Figure (4).)
(1a). We can explain Figure (3) by saying that the small
v wave
piece of wave front that gets through the slit acts as a
source of waves in much the same way that the oscil-
lating plunger acted as a source of waves.

Christian Huygens noted this phenomena and from it

developed a general principle of wave motion. His idea
was that as a wave pattern evolved, each point of a wave
front acts as the source of a new circular or spherical
wave. To see how this principle can be applied,

front edge
of wave

Figure 3 Figure 4
A wave emerging from a narrow slit spreads out in all Huygen’s construction. The future position of a
directions, just as if the wave in the slit were a plunger. wavefront can accurately be predicted by assuming that
each point on the wavefront is a source of a new wave.
 33-5

By using the construction of Figure (4) to predict the

future shape of a wave front, we see that if we use a slit
to block all but a small section of the wave front, as
illustrated in Figure (5), then the remaining piece of
wave front will act as a source of circular waves
emerging from the other side. This is what we saw in
Figure (3). Thus the Huygens construction allows us to
see not only how a smooth wave travels forward intact,
but also why circular waves emerge from a narrow slit
as we saw in Figure (3).

The Huygens construction also provides a picture of

Figure 6
what happens as waves go through progressively wider When the slit is about 2 wavelengths wide,
slits. If the slit is wider than a wavelength then we have the wave in the slit acts as 2 point sources.
more sources in the slit and the waves from the sources
begin to interfere with each other. In Figures (6, 7, 8)
we see the wave patterns for increasingly wide slits and
the corresponding Huygens constructions. For the
wider slits, more of the wave goes through the center
intact, but there is always a circular wave coming out
at the edges. For the slit of Figure (8), the circular waves
at the edges are relatively unimportant, and the edges of
the slit cast a shadow. This is beginning to resemble our
example of sunlight coming through the kitchen door-
way. The name diffraction is used to describe the
spreading of the waves that we see at the edges of the Figure 7
slits in Figures (5) through (8). As the slit is widened, more of the wave comes
through intact. In the center we are beginning to
get a beam of waves, yet at the edges, the wavefront
continues to act as a source of circular waves.

Figure 5
The small piece of wave in a narrow
slit acting as a single point source.

Figure 8
When the slit is wide compared to a wavelength, we
get a distinct beam of waves. Yet no matter how wide
the slit, there are still circular waves at the edges.
33-6 Light Waves

TWO SLIT Sending a wave through two slits and observing the
INTERFERENCE PATTERN resulting wave pattern is a convenient way to analyze
various kinds of wave motion. But in most cases we do
If a single narrow slit can produce the same wave
not see the full interference pattern as we do for these
pattern as an oscillating plunger, as we saw in Figure
ripple tank photographs. Instead, we observe only
(3), then we should expect that two slits next to each
where the waves strike some object, and from this
other should produce an interference pattern similar to
deduce the nature of the waves.
the one produced by two oscillating plungers seen in
Figure (2). That this is indeed correct is demonstrated To illustrate what we mean, imagine a harbor with a sea
in Figure (9). On the left we have repeated the wave wall and two narrow entrances in the wall as shown in
pattern of 2 plungers. On the right we have a wave Figure (10). Waves coming in from the ocean emerge
impinging upon two narrow slits. We see that both as circular waves from each entrance and produce a two
have the same structure of lines of nodes, with beam of slit interference pattern in the harbor. Opposite the sea
waves coming out between the lines of nodes. Because wall is a beach as shown.
the patterns are the same, we can use the same analysis
for both situations. If we are at point A on the beach directly across from
the center of the two entrances, we are standing in the
center beam of waves in the interference pattern. Here

Figure 9
The wave pattern emerging from 2 slits is similar to the wave pattern produced by two plungers.
 33-7

large waves wash up on the beach. Walking north striking the shore eroded the beach. As you walk along
along the beach we cross the first line of nodes at point the beach you notice a series of indentations, at points
B. Here the water is calm. Going farther up to point C A, C, C', etc. where the beach was eroded. The sand
we are again in the center of a beam of waves. We will was not eroded at points B, B', D and D'. If someone
call this the first maximum above the central maxi- asked what the ocean waves were like during the storm,
mum. Farther up we cross the second line of nodes at could you tell them?
point D and encounter the second minimum in the
height of waves striking the shore. By measuring the distance between the maximum
erosions and knowing the geometry of the harbor, you
Going south from point A we encounter the same can determine the wavelength of the ocean waves that
alternate series of maxima and minima at points B', C', struck the sea wall during the storm. Similar calcula-
D', etc. If we graphed the amplitude of the waves tions can be made to determine the wavelength of any
striking the shore, we would get the pattern shown at kind of wave striking two narrow slits producing an
the right side of Figure (10). interference pattern on the other side. We do not have
to see the actual wave pattern, we only have to note the
Now suppose that we walk along the beach on a calm location of the maxima and minima of the waves
day where there are no waves, but on the previous day striking an object like the shore in Figure (10).
there had been a storm. During the storm, the waves

N E
second maximum

C first maximum

B
north entrance
central
A
maximum

south entrance
B'

C' first maximum

D'

E' second maximum

Figure 10
Hypothetical harbor with two entrances through the sea wall. If ocean waves
are coming straight in toward the sea wall, there will be a 2 slit interference
pattern inside the harbor, with a series of maxima and minima along the beach.
33-8 Light Waves

We begin our analysis of the two slit wave pattern by

drawing a series of circles to represent the wave crests
and troughs emerging from the two slits. The results,
first
which are shown in Figure (11), are essentially the 9
10
maximum
8 11
same as our analysis of the two plunger interference 5
6
7
10
9
pattern in Figure (2). The maxima occur where crests 3
4
8
2
meet crests and troughs meet troughs. The minima or 1
6
7

lines of nodes are where crests meet troughs. 5

4 central
3
2
maximum
Exercise 2 1

On Figure (11), sketch the lines along which crests meet

troughs, i.e., where solid and dashed circles intersect.
This should be where the lines of nodes are located.
Figure 12
One more wave fits in the path from the bottom slit to
the first maxima, than in the path from the top slit.
The First Maxima
The central maximum is straight across from the center from each slit. We see that at first maximum, the 12th
of the two slits (if the incoming waves are parallel to the crest from the lower slit has run into the 11th crest from
slits as in Figure 11). To figure out where the first the upper slit, producing a maximum crest. The
maximum is located, consider the sketch in Figure (12). distance from the lower slit to the first maximum is
We have reduced the complexity of the sketch by exactly one wavelength longer than the distance from
drawing only the solid circles representing wave crests. the upper slit to the first maximum. This is what
In addition we have numbered the crests emerging determines the location of the first maximum.

In Figure (13) we have repeated the sketch

of Figure (12), but now focus our attention
on the difference in the length of the two
first paths from the slits to the first maximum.
maximum Since an extra wavelength λ fits into the
lower path, the path length difference is λ as
shown. The bottom path, with λ removed,
and the upper path, both shown as dashed
lines in Figure (13), are thus the same length
central
maximum and therefore form 2 sides of an isosceles
triangle.

Let us denote by θ 1 the angle from the

center of the two slits up to the first maxi-
first mum. Since this line bisects the isosceles
maximum triangle formed by the two dashed lines, it is
perpendicular to the base of the isosceles
triangle which is the line from the center of
the upper slit down to the point (a) on the
Figure 11
lower path. As a result, the base of the
Analysis of the two slit wave pattern, assuming isosceles triangle makes the same angle θ 1
that circular waves emerge from each slit and with the plane of the slits as the line to the
interfere with each other. The maxima are where first maximum does with the horizontal line
crests from one slit meet crests from the other. to the central maximum. (Picture rotating
Cancellation occurs where crests meet troughs.
 33-9

the isosceles triangle up around its base. If you rotate An easy way to remember this derivation is to note that
the isosceles triangle by an angle θ 1 , its base will rotate the two triangles in Figures (13) and (14), drawn
by the same angle θ 1 , thus the 2 angles labeled θ 1 in separately in Figure (15), are similar triangles. Thus the
Figure (13) are the same.) ratios of the small sides to the hypotenuses must be
equal, giving
Our approximation in this analysis is that the separation
d between the slits is very small compared to the λ = Ymax
(3′)
distance over to where we are viewing the first maxi- d 2
D2 +Ymax
mum. If this is true, then the two paths to the first
maximum are essentially parallel and the small bold The importance of Equation 3 is that it allows us to
triangle in Figure (13) is very nearly a right triangle. calculate the wavelength of a wave by observing the
Assuming that this is a right triangle, we immediately distance Ymax between maxima of the interference
get pattern. For example, in our problem of determining
the character of the waves eroding the beach in Figure
sin θ1 = λ
angle to first
maximum (1) (10), we could use a map to determine the distance D
d
from the breakwater to the shore and the distance d
between entrances through the breakwater. Then
In Figure (14) we have another right triangle involving
pacing off the distance Ymax between erosions on the
the angle θ 1 . If the distance from the slits to where we
beach, we could use Equation 3 to determine what the
are viewing the maxima is D, and if we designate by
wavelength of the waves were during the storm.
Ymax the distance from the central to the first maxi-
mum, then the hypotenuse of this right triangle is given
by the Pythagorean theorem as D2 +Ymax 2
. From this first
maximum
triangle we have 2
2+ Ymax
sin θ1 =
Ymax
2 2
(2) √D Ymax
D +Ymax
θ1 central
maximum
Equating the two formulas for sin θ 1 and solving for D
λ gives

λ = Ymax d Figure 14
2
2
(3)
D +Ymax The angle θ 1 up to the first maxima is the same as
the angle in the small triangle of Figure (13).

isosceles triangle first

maximum

2
θ1 2+ Ymax
θ1 central d θ1 √D Ymax
d maximum θ1
(a) λ
λ Figure 15
D

Since the two triangles are similar,

Figure 13 we have λ /d = ymax D2 + Ymax 2

The path length difference to the first

maximum is one wavelength λ .
33-10 Light Waves

Exercise 3 TWO SLIT PATTERN FOR LIGHT

Repeat the derivation that led to Equation 3 except do Christian Huygens discovered his principle of wave
the calculation in terms of the distance Ymin from the motion in 1678, and developed a wave theory of light
central maximum to the first minimum. (Now the path that competed with Newton's particle theory of light. It
length difference is λ /2 .) was not until 1801, over 120 years later, that Thomas
Young first demonstrated the wave nature of light
Exercise 4 using a two slit interference experiment. Why did it
We will see that Equation 3 has an applicability that goes take so long to do this demonstration?
far beyond the analysis of two slit interference patterns.
You will need this formula several times later in this Two major problems arise when you try to test for the
course, and quite likely in other research work. Rather wave nature of light. One is the fact that the wavelength
than memorizing the formula, it is much better to memo- of light is very short, on the order of one hundred
rize the derivation. The best way to do this is to treat the thousand times shorter than the wavelengths of the
derivation as a clean desk problem. Some time, a day water waves we observe in the ripple tank photographs.
or so after you have read this section, clean off your A more serious problem is that individual atoms in the
desk, take out a blank sheet of paper, and derive sun or a light bulb emit short bursts of light that are not
Equation 3. The first time you try it, you may have
coordinated with each other. The result is a chopped
forgotten some steps. If that happens, review the
up, incoherent beam of light that may also include a
derivation and try to do the clean desk problem a day or
so later. It is worth the effort because the derivation
mixture of frequencies.
summarizes all the formulas used in this chapter.
In our analogy of a sea wall with two entrances, it is
likely that a real storm would produce a mixture of
waves of different wavelengths heading in different
directions. Many different interference patterns would
be superimposed on the inside of the sea wall, different
maxima and minima would overlap at the beach and
the beach would be more or less uniformly eroded.
Walking along the beach the next day, you would not
find enough evidence to prove that the damage was
done by ocean waves, let alone trying to determine the
wavelength of the waves.

The invention of the laser by Charles Townes in 1960

eliminated the experimental problems. The laser emits
a continuous coherent beam of light that more closely
resembles the orderly ripple tank waves approaching
the slits in Figure (10) than the confused wave motion
seen in a storm. If you send a laser beam through two
closely spaced slits, you cannot help but see a two slit
interference pattern.
 33-11

2
Even in a demonstration lecture, the two slit pattern neglect the Ymax in the square root and we get the
produced by a laser beam can be used to measure the simpler formula
wavelength of the light in the beam. In Figure (16) we
placed a two slit mask next to a millimeter scale on the λ ≈ Ymax d if D > > Ymax (3a)
D
top of an overhead projector and projected the image on
a large screen. You can see that the spacing between the Putting in the numbers obtained from Figures (16) and
two slits is about 1/3 of a millimeter. In Figure (17) we (17), we get
aimed the red beam of a common helium neon laser
λ = 2.3 cm × .3 × 10 m = 7 × 10 – 5cm (4)
–3
through the two slits of Figure (16), onto a screen 10 10m
meters from the slits. The resulting two slit pattern
consisting of the alternate maxima and minima are While this demonstration experiment gives fairly ap-
easily seen by the class. Marking the separation of two proximate results, accurate to about one significant
maxima on a piece of paper and measuring the distance figure, it may be somewhat surprising that a piece of
we found that the separation Ymax between maxima apparatus as crude as the two slits seen in Figure (16)
was about 2.3 cm. even allows us to measure something as small as
7 × 10 – 5cm .
In using Equation 3, λ = Ymaxd/ D2 + Ymax
2
to calculate
the wavelength λ , we note that the 10 meter distance
D is much greater than the 2.3 cm Ymax . Thus we can

Shows that d = .3mm

Figure 16
The two slits and a plastic ruler are placed on an overhead
projector and projected onto a screen 10 meters away. This
is a photograph of the screen.

Shows that Ymax = 2.3 mm for D = 10 m.

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 cm
Figure 17
The 2 slit laser pattern is then projected on the screen. Below
is a centimeter scale, showing that the maxima are about 2.3
centimeters apart.
33-12 Light Waves

THE DIFFRACTION GRATING had excellent control over the slit width and spacing.
The crudeness of our measurement of the wavelength For these images, the slits are 50 microns (50µ) wide
of the laser light in our two slit experiment could be and spaced 150 microns apart on centers.
improved somewhat by a more accurate measurement The photographs of the interference patterns produced
of the separation of the two slits, but the improvement by the slits of Figure (18) are all enlarged to the same
would not be great. There is, however, a simple way to scale. The important point to notice is that while the
make far more accurate measurements of the wave- maxima become sharper as we increase the number of
length of a beam of light. The trick is simply add more slits, the spacing between maxima remains the same.
slits. Adding more identical slits sharpens the maxima but
To see why adding more slits gives more accurate does not change their spacing! As a result the two slit
results, we show in Figure (18) the wave patterns we get formula, Equation 3, can be applied to any number of
when the laser beam is sent through two slits, three slits, slits as long as the spacing d between slits remains
four slits, five slits, and seven slits. We created the slits constant.
using a Macintosh computer using the Adobe Photoshop If there are many slits, the device is called a diffraction
program and a Linatronic printer to produce the film grating and Equation 3, which we repeat below, is
images of the slits. The Linatronic printer can draw known as the diffraction grating formula.
precise lines one micron wide ( 10 – 6 meters); thus we

w=50µ d=150µ 3 slits 4 slits 5 slits 7 slits

1 slit
1 slit

2 slits
2 slits

3 slits
3 slits

4 slits
4 slits

5 slits
5 slits

7 slits
7 slits
Figure 18
Interference patterns for various slit structures. If we keep the spacing between slits the same, then there is
no change in the location of the maxima, no matter how many slits the laser beam passes through. Thus an
analysis of the location of the maxima for 2 slits applies to any number of slits. Also note that the single slit
pattern acts as an envelope for the multiple slit patterns.
 33-13

produced and very precise wavelength measurements

d diffraction can be made. It is possible to make inexpensive plastic
λ = Ymax grating (3 repeat) replicas of fine diffraction gratings for use in all kinds
22
D +Ymax formula
of laboratory work, or even for making jewelry. It turns
out that compact disks (CDs) also make superb diffrac-
Exercise 5
tion gratings. We will not tell you the spacing of the
lines on a CD for it is a nice project to figure that out for
In Figure (18) the separation of the slits is 150 microns
and the separation of 10 maxima is 26.4 cm. The screen yourself. (All you need is a common helium neon laser.
is a distance of 6.00 meters from the slits. From this The wavelength of the laser beam can be gotten from
determine the wavelength of the light in the laser beam Exercise 6.)

(a) using the exact formula, Equation 3.

Exercise 6
(b) using the approximate formula, Equation 3a. In Figure (19), a laser beam is sent through a smoke
filled box with a diffraction grating at the center of the
How many significant figures are meaningful in your box as shown in the sketch (19a). The smoke allows you
result? To this accuracy, did it make any difference to see and photograph the central laser beam and two
whether you used the exact Equation 3 or the approxi- maxima on each side. You also see maxima reflected
mate Equation 3a. from the back side of the grating. (When you shine a
laser beam on a CD you get only the reflected maxima,
Figure (18) demonstrates that the more slits you use, the no light goes through the record.)
sharper the maxima and the more accurately you can
determine the wavelength of the light passing through The grating used in Figure (19) had 15,000 lines per
the slits. In the latter part of the 1800s, the diffraction inch (1 inch = 2.54 cm). From this information and the
photograph of Figure (19b), determine the wavelength
grating was recognized as an excellent tool for scien-
of the laser beam used. Try both the exact Equation 3
tific research, and a great effort was put into producing
and the approximate Equation 3a. Explain why Equa-
gratings with as many closely spaced lines as possible. tion 3a does not work well for this case.
Fine ruling machines were developed that produced
gratings on the order of 6000 lines or slits per centime-
ter. With so many lines, very sharp maxima are

smoke filled box with glass top

second maximum
reflected
maxima
first maximum

central maximum
laser
beam
white
screen
grating

a)
Figure 19
Laser beam passing through a diffraction grating. The beam
is made visible by placing the grating in a smoke filled box.
Because the lines are so close together, the maxima are widely
separated. You can also see reflected maxima on the back side. b)
33-14 Light Waves

More About Diffraction Gratings When we get down to slit # 501, just over half way
The results of Figure (18) demonstrated that the maxima down, the path length difference is
got sharper but remained in the same place as we added 500λ – 500λ/1000 = 500λ – λ/2 . In other words,
slits. Let us now see why this happens. the waves from slit 1 and slit 501 are precisely one half
a wavelength out of phase, crests exactly meet troughs,
The maxima of a diffraction grating occur at those and there is precise cancellation. A similar argument
points on the screen where the waves from every slit shows that waves from slits #2 and #502 are λ/2 out of
add up constructively. This can happen only when the phase and cancel exactly. The same goes for the pairs
path length difference between neighboring maxima is #3 and #503, #4 and #504, all the way down to 500 and
0 (central maximum), λ (first maxima), 2λ (second 1000. In other words, the waves all cancel in pairs and
maxima), etc. In Figure (20) we are looking at a small we have a minimum, complete cancellation at the point
section of a diffraction grating where we have drawn in just below the first maximum where the path length
the paths to the first maxima. The path length differ- difference is λ – λ/1000 instead of λ . With two slits
ences between neighboring slits are all λ and the angle we got complete cancellation half way between maxima.
θ 1 to the first maxima is given by θ 1 = λ/d , the same With 1000 slits, we only have to go approximately
results we had for the two slit problem in Figure (13). 1/1000 the way toward the next maxima before we get
This angle does not depend upon the number of slits, complete cancellation. The maxima are roughly 500
thus the position of the maxima do not change when we times sharper. You can see that with n slits, the maxima
add slits as in Figure (18). will be about n/2 times sharper than for the two slit
example.
To see why the maxima become narrower as we add
slits, let us consider the example of a 1000 slit grating
illustrated in Figure (21). We have numbered the slits
from 1 to 1000, and are showing the paths to a point just 1
below the first maximum where the path length differ- 2 (λ − λ/1000)
ence between neighboring slits is ( λ – λ/1000 ) in-
(2λ − 2λ/1000)
stead of λ . 3

4 (3λ − 3λ/1000)
On the figure we are indicating, not the path length
difference between neighboring slits, but instead, the
path length difference between the first slit and the
others. This difference is ( λ – λ/1000 ) for slit #2,
500
( 2λ – 2λ/1000 ) for slit #3, ( 3λ – 3λ/1000 ) for slit
#4, etc. 501 (500λ − 500λ/1000)
= (500λ − λ/2)
502
paths
to
θ1 first
d maxima
998
λ
999
λ 1000
θ1
Figure 21
In a thousand slit grating, we get complete cancellation
when the path length difference between neighboring
Figure 20
slits is reduced from λ to λ – λ / 1000 .
When the path length difference between
neighboring paths is λ , then the waves from all
slits add constructively and we get the first
maxima.
 33-15

The maxima will also be much more intense because The Visible Spectrum
the light is coming in from more slits. If we have n slits, Thus far we have been using a laser beam to study the
the amplitude of the wave at the center of a maxima will operation of a diffraction grating. Now we will reverse
be n times as great as the amplitude from a single slit. the process and use diffraction gratings to study the
It turns out that the amount of energy in a wave, the nature of beams of light.
intensity, or, for light, the brightness, is proportional to
the square of the amplitude of the wave. Thus the If you send a beam of white light through a diffraction
brightness at the center of the maxima for an n slit grating, you get a series of maxima. In all but the central
grating is n 2 times as bright at the brightness we would maxima light is spread out into a rainbow of colors
have for a single slit. The maxima for the 1000 slit illustrated in Figure (22). In each maxima the red light
grating illustrated in Figure (21) would be one million is bent the most, and blue the least. As we saw from
times brighter than if we let light go through only one Equation 1, sin θ1 = λ /d, the longer the wavelength
of the slits. (To see how the total energy works out, the greater the angle the wave is bent or diffracted.
consider the following argument. Compared to one Thus red light has the longest wavelength and blue the
slit, when you have n slits, you have n times as much shortest in the mixture of wavelengths that make up
light energy that is compressed into a maxima that is white light.
only 1/n as wide. You get one factor of n in brightness
The longest wavelength that the human eye can see is
due to the compression, and the other factor of n due to
about 7.0 × 10 – 5 cm , a deep red light, and the shortest
there being n slits.)
is about 4.0 × 10 – 5 cm , a deep purple. All other
visible wavelengths, the entire spectrum of visible
light, lies in the range between 4.0 × 10 – 5 cm to
7.0 × 10 – 5 cm . Yellow light, for example, has a
wavelength around 5.7 × 10 – 5 cm , and green light is
near 5.0 × 10 – 5 cm .
red
yellow
green As we saw in Chapter 32, visible light is just a part of
m blue
mu
maxi the complete electromagnetic spectrum. A surpris-
o nd red ingly small part. As radio, television, microwave
sec yellow
ovens, infra red sensors, ultraviolet sunscreens, x ray
imum green
max photographs, and γ ray bursts in the sky have entered
first blue
our experience of the world, we have become familiar
white white with a much greater range of the electromagnetic
light light
spectrum. As indicated in Figure (23), AM radio waves
have wavelengths in the range of 10 to 100 meters,
Figure 22 VHF television a few meters, VHF from around 10 cm
When white light passes through a diffraction grating, to a meter, microwaves from around a millimeter to 10
the maxima for different colors emerge at different cm, infra red from less than a millimeter down to visible
angles. Since red light has the longest wavelength of
the visible colors, it emerges at the greatest angle.
red light at 7.0 × 10 – 5 cm . At shorter wavelengths

AM radio FM,TV microwaves infra red ultra violet x rays γ rays

red yellow green blue

Figure 23
Visible light is a tiny piece of the electromagnetic spectrum.
33-16 Light Waves

than deep blue we have ultraviolet, then x rays, and the something about the age of the star and the environ-
very shortest wavelengths are called γ (gamma) rays. ment in which it was formed.
To study the electromagnetic spectrum, different de- Our main reason for studying the spectrum of light
vices are used at different wavelengths. In Chapter 32 emitted by atoms will be to learn something about the
we used a loop of wire and Faraday's law to detect the atoms themselves. Since Rutherford's discovery of the
magnetic fields of a radio wave. This required the use atomic nucleus in 1912, it has been known that atoms
of an oscilloscope that could display radio wave fre- consist of a positively charged nucleus surrounded by
quencies, typically of the order of a megacycle for AM negatively charged electrons. If we apply Newtonian
radio. For visible light the frequencies are too high, the mechanics to predict the motion of the electrons, and
wavelengths too short for light to be studied by similar Maxwell's equations to predict the kind of electromag-
techniques. Instead the diffraction grating will be our netic radiation the moving electrons should radiate, we
main tool for studying the electromagnetic waves in the get the wrong answer. There is no way that we can
visible spectrum. explain the spectrum of light emitted by atoms from
Maxwell's equations and Newtonian mechanics. The
Exercise 7 existence of detailed atomic spectra is a clue that
What are the lowest and highest frequencies of the
something is wrong with this classical picture of the
waves in the visible spectrum? What is the color of the atom. It is also the evidence upon which to test new
lowest frequency? What is the color of the highest? theories.
What is the frequency of yellow light?
We do not have to study many kinds of atoms to find
something wrong with the predictions of classical
Atomic Spectra theory. The simplest of all atoms, the hydrogen atom
Our main application of the diffraction grating will be consisting of one proton for a nucleus, surrounded by
to study the spectrum of light emitted by atoms. It has one electron, is all we need. Heated hydrogen gas emits
long been known that if you have a gas of a particular a distinct, orderly, spectrum of light that provides the
kind of atom, like nitrogen, oxygen, helium, or hydro- essential clues of what is going on inside a hydrogen
gen, a special kind of light is emitted. You do not get atom. In this chapter we will focus on using a diffrac-
the continuous blend of wavelengths seen in white tion grating to learn what the spectrum of hydrogen is.
light. Instead the light consists of a mixture of distinct In the following chapters we use the hydrogen spec-
wavelengths. Which wavelengths are involved de- trum to study the atom itself.
pends upon the kind of atom emitting the light. The
mixture of wavelengths provide a unique signature of
that atom, better than a fingerprint, for identifying the
presence of an atom in a gas. In fact, the element helium
(named after the Greek word helios for sun) was first
identified in the sun by a study of the spectrum of light
from the sun. Only later was helium found here on
earth.

The subject of modern astronomy is based on the study

of the spectrum of light emitted by stars. Some stars
consist mostly of hydrogen gas, others a mixture of
hydrogen and helium, while still others contain various
amounts of heavier elements. We learn the composi-
tion of the star by studying the spectrum of light
Figure 24
emitted, and from the composition we can deduce Apparatus to measure hydrogen spectrum.
 33-17

THE HYDROGEN SPECTRUM β and γ are the only lines emitted by pure hydrogen
The apparatus required for studying the hydrogen gas in the visible part of the electromagnetic spectrum.
spectrum can be as simple as the hydrogen source, Their wavelengths are
meter stick and diffraction grating shown in the photo- λ α = 6.56 × 10 – 5 cm
graph of Figure (24). The hydrogen source consists of
a narrow glass tube filled with hydrogen gas, with λ β = 4.86 × 10 – 5 cm
metal electrodes at the ends of the tube. When a high
voltage is applied to the electrodes, an electric current λ γ = 4.34 × 10 – 5 cm (5)
flows through the gas, heating it and causing it to emit When actually performing the experiment shown in
light. The diffraction grating is placed in front of the Figure (24), there are some steps one should take to
hydrogen tube, and the meter stick is used to measure improve the accuracy of the results. As shown in
the location of the maxima. Figure (27), a small arrowhead is placed on the grating
The setup of the apparatus is illustrated in Figure (25) itself. You then place your eye behind the meter stick
and the resulting spectrum in Figure (26). In this and move your head and the slider on the meter stick
spectrum we are looking at the first maxima on the left until the point on the slider lines up with the arrowhead
side of the meter stick as shown in Figure (25). The on the grating and with the spectral line you are trying
leftmost line, the one bent the farthest is a deep red line to measure.
which is called the hydrogen α line, and labeled by α α β γ
in the photograph. The next line is a spurious line
caused by impurities in the hydrogen tube. More to the
right is a bright, swimming-pool blue line called hydro-
gen β . Much harder to see is the third line called
hydrogen γ , a deep violet line near the short wave-
length end of the visible spectrum. The three lines α ,

Figure 26
Photograph of the α , β and γ lines
hydrogen tube in the hydrogen spectrum.
(top view)

diffraction grating

gliding pointer
D

meter stick
Ymax
eye
central maximum
Figure 25
To determine the wavelength of light using a diffraction
grating, you need to measure the distance Ymax to the
first maximum, and the distance D shown. To measure
Ymax , slide the pointer along the meter stick until it lines Figure 27
up with the first maximum. Looking through the grating, move your eye so that the
spectral line is centered over the pointer as shown.
33-18 Light Waves

Rather than trying to measure the distance Ymax from Exercise 8

the central maximum to the spectral line, it is more
Derive a formula for the wavelength λ of a spectral line
accurate to measure the distance 2Y max from the first
in terms of the distance Y2 max from the central maximum
maximum on the left to the first maximum on the right, to the second maxima of the line. The second maxima
and then divide by 2. The wavelength of the line under of the bright lines of an atomic spectra are quite easily
study is then given by the diffraction grating formula, seen using the apparatus of Figure (24).
Equation 3
λ = Ymax d (3 repeated)
2 2 The Experiment on Hydrogen Spectra
D +Ymax
You should carry out the following steps when doing
where d is the separation of the slits in the grating and the hydrogen spectrum experiment shown in Figure
D the distance from the grating to the meter stick. (25).

(When you first perform this experiment you may be (1) Determine the wavelength of all the spectral lines
confused by where the central maximum is. If you look you can see, and compare your results with those given
straight through the grating at the tube, all you see is the in Equation 5. Measure distances between first maxima,
tube. But that is the central maximum. It looks like the not to the central maxima.
tube because all the colors go straight through the
grating. There is no separation of colors or distortion (2) Measure the distances to the second maxima for the
of the image. To see a spectrum you have to look lines you can see out there and compute the corre-
through the grating but far off to the side from the tube.) sponding wavelengths using our results from Exercise
8. Compare these wavelengths with those you get
using the first maxima.
 33-19

The Balmer Series It was the Swiss school teacher Johann Balmer who in
There are many spectral lines emitted by the hydrogen 1885 discovered a formula for the wavelengths of the
atom. Only three, however, are in the visible part of the spectral lines seen in Figure (28). The wavelength of
spectrum. The complete spectrum consists of a num- the m th line (m=3 for H3, m=4 for H4, etc.) is given by
ber of series of lines, and the three visible lines belong the formula
to the series called the Balmer series. The red line, m2
hydrogen α , is the longest wavelength line in the λ m = 3.6456 × 10 – 5 cm × (6)
m2 – 4
Balmer series, next comes the blue hydrogen β , then
the violet hydrogen γ . Then there are many lines of the Equation 6 is known as the Balmer formula.
Balmer series out in the ultraviolet, which we cannot
For m=3 we get from the Balmer formula
see by eye, but which we can record on photographic
film. λ H 3 = 3.6456 × 10 – 5 cm × 9
9–4
Figure (28) shows part of the spectrum of light from a (6a)
= 6.56 × 10 – 5 cm
hydrogen star. These lines are in the ultraviolet and are
all part of the Balmer series. Slightly different naming which agrees with Equation 5 for hydrogen α . Each
is used here. In the notation of Figure (28), we should higher value of m gives us the wavelength of a new line.
call the red hydrogen α line H3, the blue β line H4, and At large values of m the factor m 2/ m 2 – 4 ap-
the violet γ line H5. In Figure (28), the first 6 Balmer proaches 1, and the lines get closer and closer together
lines are missing. Here we see lines H9 through H40. as seen in Figure (28). The end is at 3.65 × 10 – 5 cm
As the lines increase in number they get closer and where m is very large.
closer together. The whole series ends with very many,
very closely spaced lines near 3.65 × 10 – 5 cm . It is Exercise 9
called a series because the lines converge to a final (a) Use Equation 6 to calculate the wavelengths of the
wavelength in much the same way that many math- β and γ lines of the hydrogen spectrum and compare
ematical series converge to a final value. the results with Equation 5.

(b) Calculate the wavelength of H40 and compare your

results with Figure (28).

–5 –5 –5 –5 –5
3.65 10 3.70 10 3.75 10 wavelength 3.80 10 3.85 10

H40 H30 H20 H15 H14 H13 H12 H11 H10 H9

Figure 28
Spectrum of the star HD193182, showing ultra violet hydrogen lines near the limit of the Balmer
series. This series of lines begins in the visible part of the spectrum with the lines we have called
α , β , and γ , (which would be called H3, H4, and H5 in this diagram), and goes on to the ultra
violet. The lines get closer and closer together, until the end just beyond the point labeled H40. The
Swiss school teacher Johann Balmer discovered a formula for the wavelengths of these lines.
33-20 Light Waves

THE DOPPLER EFFECT As a result the wavelengths in front and back of the
One phenomena of wave motion that is particularly source are
easy to visualize is the Doppler effect. As you can see v
in Figure (29), if the wave source is moving, the λ front = λ 0 – ∆λ = λ 0 1 – vsource (9a)
wave
wavelength of the waves is compressed in front of the
source and stretched out behind. This result, which is v
obvious for water waves, also applies to sound waves λ back = λ 0 + ∆λ = λ 0 1+ vsource (9b)
wave
in air and to light waves moving through space.
If we are in front of the moving source, the wave period
To analyze the effect, we first note that if the source is Tfront we observe is the time it takes the shortened
at rest, then the waves all travel out from the source at wavelength λ front to pass us at a speed vwave , which is
a speed vwave , have a wavelength λ 0 and a period T0
given by λ λ v
Tfront = v front = v 0 1 – vsource
wave
sec = λ 0 cm / cycle = λ 0 sec
wave wave
T0 cycle vwave cm/sec vwave cycle (7)
v
If the source is moving forward at a speed vsource , then Tfront = T 0 1 – vsource (10a)
wave
during one period T0 the source will move forward a
distance x = vsourceT0 . But this is just the amount ∆λ where we now replaced λ 0 vwave by T0 . In the back,
by which the wavelength is shortened in front and the period is extended to
stretched out in back. Thus
λ0 v
∆λ = vsourceT 0 = vsource v (8) Tback = T 0 1 + vsource (10b)
wave wave

where we used Equation 7 to replace T0 by λ 0 vwave .

d = 150µ

d = 250µ

Figure 29 Figure 30
When the source of the wave is moving, the When the source is moving faster than the waves,
wavelengths are compressed in front and
d = 300µ
the waves build up on the front edge to create a
stretched out behind. shock wave. For supersonic flight, this shock wave
produces the sonic boom.
 33-21

If the speed of the source approaches the speed of the Stationary Source
wave, as in the case of a jet airplane approaching the and Moving Observer
speed of sound, the wavelength in front goes to zero. At If the source is at rest but we, the observer, are moving,
speeds greater than the speed of the wave, as in super- there is also a Doppler effect. In the case of water or
sonic flight, there are no waves ahead of the source; sound waves, if we are moving through the medium
instead, the leading edge of the waves pile up as shown toward the source, then the wave crests pass by us at an
in Figure (30) to create what is called a shock wave. increased relative speed vrel = vwave + vus . Even
This shock wave is responsible for the sonic boom we though the wavelength is unchanged, the increased
hear when a jet passes overhead at supersonic speeds. speed of the wave will carry the crests by faster, giving
us an apparently shorter period and higher frequency.
Exercise 10
There is a simple experiment you can perform to ob-
If our velocity through the medium is small compared
serve the Doppler effect. Stand beside a road and have to the wave speed, then we observe essentially the same
a friend drive by at about 40 mi/hr while blowing the car decrease in period and increase in frequency as in the
horn. As the car passes, the pitch of the horn will case when the source was moving. In particular,
suddenly drop because the wavelength of the sound Equation 10 is approximately correct.
waves, which was shortened as the car approached, is
lengthened after it passes. The shorter, higher-pitched On the other hand, when the waves are in water or air
sound waves change to longer, lower-pitched waves. and the relative speed of the source and observer
approaches or exceeds the wave speed, there can be a
For this exercise, assume the car is owned by a musi- considerable difference between a moving source and
cian, and the car horn plays the musical note A at a a moving observer. As illustrated in Figure (31a), if the
frequency of 440 cycles per second.
source is moving faster than the wave speed, there is a
a) What is the wavelength of a 440 cycle/sec note, if the shock wave and the observer detects no waves until the
speed of sound is 1000 ft/sec? source passes. But if the source is at rest as in Figure
(31b), there is no shock front and the observer moves
b) What is the wavelength of the note we hear if the car
through waves before getting to the source, even if the
is approaching at a speed of 40 miles/hr?
observer is moving faster than the wave speed.
c) What is the frequency we hear if the car is approach-
ing at 40 miles/hr?

d) What is the frequency we hear when the car is going

away from us at 40 miles/hr?

a) moving source b) moving observer

Figure 31
For waves in water or air, there can be a significant difference between a moving
source with a stationary observer, and a moving observer with a stationary source,
even though the source and observer have the same relative velocity in the two cases.
For light, the principle of relativity requires that the two cases be identical.
33-22 Light Waves

Doppler Effect for Light Equations 9 and 10, modified this way, are correct as
When a source of light waves is moving toward or long as the source is not moving too fast. However if the
away from us, there is also a Doppler effect. If the source is moving relative to us at a speed approaching
source is moving toward us, the wavelengths we see are the speed of light, there is one more relativistic effect
shortened. This means that the color of the light is that we have to take into account . Remember that a
shifted toward the blue. If the source is moving away, moving clock runs slow by a factor of 1 – v2 /c2 . If the
the wavelengths are stretched out, become longer, and source is radiating a light wave of period T0 , then that
the color shifts toward the red. When the speed of the period can be used in the construction of a clock. If we
source is considerably less than the speed of light, observe the source go by at a speed vsource , the period
Equations 9 and 10 correctly give the observed wave- T0 must appear to us to increase to T0′ given by
length λ and period T in terms of the source’s wave- T0
length λ 0 and period and T0 . T0′ = (see Eq 1-11)
2
1 – vsource /c 2
Principle of Relativity From our point of view, the source is radiating light of
There is one fundamental difference, however, be- period T0′ . This is the light whose wavelength is
tween the Doppler effect for water and sound waves, stretched or compressed, depending on whether the
and the Doppler effect for light waves. For water and source is moving away from or towards us. Thus we
sound waves we could distinguish between a source at should use T0′ instead of T0 in Equation 10.
rest with a moving observer and an observer at rest with
a moving source. If the source were at rest, it was at rest Replacing T0 by T0′ in Equation 10 gives
relative to the medium through which the wave moves.
We got different results depending on whether it was T0 vsource
the source or the observer that was at rest. Tfront = 1– c (11a)
2
1 – vsource /c 2
In the case of light, the medium through which light
moves is space. According to the principle of relativ-
ity, one cannot detect uniform motion relative to space. T0 vsource
Tback = 1+ c (11b)
Since it is not possible to determine which one is at rest 2
1 – vsource /c 2
and which one is moving, we must have exactly the
same Doppler effect formula for the case of a stationary where vsource is the speed of the source relative to us,
source and a moving observer, or vice versa. The and we have set vwave = c. Equations (11) are the
Doppler effect formula can depend only on the relative relativistic Doppler effect equations for light. They are
velocity of the source and observer. applicable for any source speed, even if the source is
One way to use the principle of relativity is to always moving relative to us at speeds approaching the speed
assume that you yourself are at rest relative to space. of light. The corresponding wavelengths are
(No one can prove you are wrong.) This suggests that cm × T
λ front = c sec sec cm
front cycle = cTfront cycle
we should start from Equations 9 and 10, which were
derived for a stationary observer, and replace vwave by
the speed of light c, and interpret vsource as the relative λ front = cTfront
(12)
velocity between the source and the observer. λ back = cTback
 33-23

Exercise 11 Until the 1960s astronomers did not have much need
Using Equations (12), express λ front and λ back in
for the relativistic Doppler shift equations. The non
terms of λ 0 , vsource and c. relativistic Equations 10 were generally adequate be-
cause we did not observe stars or galaxies moving
Exercise 12 relative to us at speeds greater than 10 to 20% the speed
In Figure 31b, where we picture a stationary source and
of light. But that changed dramatically with the discov-
a moving observer, the waves pass by the observer at ery of quasars in 1963. Quasars are now thought to be
a speed vwave + vobserver . Why can’t this picture be brilliant galaxies in the early stages of formation. They
applied to light, simply replacing vwave by c and letting can be seen from great distances and are observed to
vobserver be the relative velocity between the source move away from us at speeds as great as 95% the speed
and observer. of light. To analyze such motion, the relativistic
formulas, Equations 11 and 12 are clearly needed.

Doppler Effect in Astronomy

Exercise 13
The Doppler effect has become one of the most pow- The most rapidly receding galaxy observed by the
erful tools astronomers use in the study of the universe. spring of 1995 is the galaxy named 8C 1435 + 63 shown
Assuming that distant stars and galaxies are made up of in the photograph of Figure (32) taken by the Keck
the same matter as nearby stars, we can compare the telescope in Hawaii. Much of the light from this galaxy
spectral lines emitted by distant galaxies with the is radiated by hydrogen gas. This galaxy is moving
corresponding spectral lines radiated by elements here away from us at a speed vsource = .95c.
on earth. A general shift in the wavelengths to the blue
(a) Assuming that the hydrogen in this galaxy
or the red, indicates that the source of the waves is
radiates the same spectrum of light as the hydrogen
moving either toward or away from us. Using Equa-
gas in our discharge tube of Figure (24), what are the
tions 12 we can then quite accurately determine how wavelengths of the first three Balmer series lines λ α , λ β ,
fast this motion toward or away from us is. and λ γ , by the time these waves reach us. (They will be
greatly stretched out by the motion of the galaxy.)

(b) Astronomers use the letter z to denote the relative

shift of the wavelength of light due to the Doppler effect.
I.e.,

λ – λ0 astronomers
z = ∆λ = notation for (13)
λ0 λ0 the red shift

where λ 0 is the wavelength of the unshifted spectral

line, and λ is the Doppler shifted wavelength we see.
What is z for galaxy 8C 1435 + 63?

Figure 32
The most distant galaxy observed as of January
1995. This galaxy, given the romantic name 8C
1435+63, was photographed by the Keck telescope
in Hawaii. The two halves of the distant galaxy are
indicated by the white bracket. The galaxy is
moving away from us at 95% the speed of light.
33-24 Light Waves

The Red Shift and the That the universe is not static was discovered by
Expanding Universe Doppler shift measurements. In the 1920s, the astrono-
In 1917 Albert Einstein published his relativistic theory mer Edwin Hubble observed that spectral lines from
of gravity, known as General Relativity. In applying distant galaxies were all shifted toward the red, and that
his theory of gravity to the behavior of the stars and the farther away the galaxy was, the greater the red
galaxies in the universe, he encountered what he thought shift. Interpreting the red shift as being due to the
was a serious problem with the theory. Any model of Doppler effect meant that the distant galaxies were
the universe he constructed was unstable. The galaxies moving away from us, and the farther away a galaxy
tended either to collapse in upon themselves or fly was, the faster it was moving.
apart. He could not find a solution to his equations that
Hubble was the first astronomer to develop a way to
represented the stable unchanging universe everyone
measure the distance out to other galaxies. Thus he
knew was out there.
could compare the red shift or recessional velocity to
Einstein then discovered that he could add a new term the distance the galaxy is away from us. He found a
to his gravitational equations. By properly adjusting simple rule known as Hubble’s law. If you look at
the value of this term, he could construct a model of the galaxies twice as far away, they will be receding from
universe that neither collapsed or blew up. This term, us twice as fast. Roughly speaking, he found that a
that allowed Einstein to create a static model of the galaxy .1 billion light years away would be receding at
universe, became known as the cosmological con- 1% the speed of light; a galaxy .2 billion light years
stant. away at 2% the speed of light, etc. In the 1930s,
construction of the 200 inch Mt. Palomar telescope was
In later life, Einstein said that his introduction of the started. It was hoped that this telescope (completed in
cosmological constant was the greatest mistake he ever 1946) would be able to observe galaxies as far away as
made. The reason is that the universe is not static. 2 billion light years. Such galaxies should be receding
Instead it is expanding. The galaxies are all flying apart at the enormous speeds of approximately 20% the
like the debris from some gigantic explosion. The speed of light.
expansion, or at least instability of the universe, could
have been considered one of the predictions of Einstein’s With the discovery of quasars, we have been able to
theory of gravity, had Einstein not found his cosmo- observe much more distant galaxies, with far greater
logical constant. (Later analysis showed that the static recessional velocities. As we have just seen, the galaxy
model, obtained using the cosmological constant, was 8C 1435 + 63, photographed by the 10 meter (400 inch)
not stable. The slightest perturbation would cause it to telescope in Hawaii, is receding from us at a speed of
either expand or contract.) 95% the speed of light. To analyze the Doppler effect
for such a galaxy, the fully relativistic Doppler effect
formula, Equation 12 is needed; non relativistic ap-
proximations will not do.

Hubble’s law raises several interesting questions. First,

it sounds as if we must be at the center of everything,
since the galaxies in the universe appear to all be
moving away from us. But this is simply a conse-
quence of a uniform expansion. Someone in a distant
galaxy will also observe the same Hubble law.
 33-25

To see how a uniform expansion works, mark a number One of the things to remember when we look at distant
of equally spaced dots on a partially blown up balloon. galaxies is that we are not only looking far away, but we
Select any one of the dots to represent our galaxy, and are also looking back in time. When we look at a galaxy
then start blowing up the balloon to represent the 10 billion light years away, we are looking at light
expansion of the universe. You will notice that dots emitted 10 billion years ago, when the universe was 10
twice as far away move away twice as fast, no matter billion years younger. Recent studies have clearly
which dot you selected. Hubble’s law is obeyed from shown that galaxies 10 billion years away look differ-
the point of view of any of the dots on the balloon. (You ent than nearby galaxies. Over the past 10 billion years
can see this expansion in Figure (33), where we started the universe has evolved; galaxies have aged, becom-
with an array of light colored dots, and uniformly ing more symmetric and less violent.
expanded the array to get the black dots.)
To predict what we will find as we look back in time,
Another interesting question is related to nature’s look at ever more distant galaxies, imagine that we take
speed limit c. We cannot keep looking out twice as far a moving picture of the universe and run the moving
to see galaxies receding twice as fast, because we picture backwards.
cannot have galaxies receding faster than the speed of
light. Something special has to happen when the If we reverse the moving picture of expanding galaxies,
recessional speeds approach the speed of light, as they we see contracting galaxies. They are all contracting
have in the case of 8C 1435 + 63. This appears to place back to one point in space and time. Go back to that
a limit on the size of the universe we can observe. point and run the movie forward, and we see all of the
universe rushing out of that point, apparently the
consequence of a gigantic explosion. This explosion
has become known as the Big Bang. (The name Big
Bang was a derisive expression coined by the astrono-
mer Fred Hoyle who had a competing theory of the
origin of the universe.)

The idea that the universe started in a big bang, pro-

vides a simple picture of the Hubble law. From our
point of view, galaxies emerged at various speeds in all
directions from the Big Bang. Those that were moving
away from us the fastest just after the explosion are now
the farthest away from us. Galaxies moving away
twice as fast are now twice as far away.

In the next chapter we will have more to say about the

origin of the universe and evidence for the Big Bang.
We will also introduce another way to interpret the
Doppler effect and its relationship to the expansion of
the universe.
d 2d 3d

Figure 33
During a uniform expansion, neighboring galaxies
move a certain distance away (d), galaxies twice as
far away move twice as far (2d), etc. This is
Hubble’s law for the expanding universe.
33-26 Light Waves

A CLOSER LOOK AT In Figure (3) reproduced here, we have a deceptively

INTERFERENCE PATTERNS simple picture of the single slit diffraction pattern. In
the photograph a wave is impinging upon a slit whose
Our focus so far in this chapter has been on the
width is less than one wavelength, with the result that
application of the wave properties of light to the study
we get a simple circular wave emerging on the other
of physical phenomena such as atomic spectra and the
side. In Figures (6), (7) and (8) we look at what happens
expansion of the universe. We now wish to turn our
when the slit becomes wider than a wavelength. These
attention to a more detailed study of the wave phenom-
are all views of the wave pattern close to the slit. We
ena itself. We will first take a closer look at the single
slit diffraction pattern that serves as the envelope of the see that as the slit becomes wider, more of the wave
passes through undisturbed, creating a more or less
multiple slit patterns we saw back in Figure (18). We
distinct shadow effect. Nevertheless we always see
will then discuss an experimental technique for accu-
rately recording various interference patterns produced circular waves at the edge of the shadow. If you
carefully look at Figure (6), reproduced below, you can
by laser beams. We will then end the chapter with a
see lines of nodes coming out of the slit that is about 2
demonstration of how Fourier analysis can be used to
predict the structure of the interference patterns we wavelengths wide.
observe. In Figure (34), we have the diffraction pattern pro-
duced by a laser beam passing through a 50 micron
The reason for these studies is to strengthen intuition
wide single slit and striking a screen 10 meters away.
about the behavior of waves. The remainder of the text
deals with the inherent wave nature of matter, and here This is a reproduction of the single slit pattern that acts
as an envelope for the multiple slit patterns seen in
we wish to develop the conceptual and experimental
Figure (18). A 50 micron slit is 50 × 10 – 6 meters or
tools to study this wave nature.
500 × 10 – 5 cm wide. This is nearly a hundred times
Single Slit Diffraction Pattern
greater than the 6.4 × 10 – 5 cm wavelength of the laser
light passing through the slit. Thus we are dealing with
In the 1600s, Francesco Maria Grinaldi discovered that
slits that are about 100 wavelengths wide.
light going through a fine slit cannot be prevented from
spreading on the other side. He named this phenom-
enon diffraction. Independently Robert Hook, of
Hook’s law fame, made the same observation and
provided a wave like explanation. The clearest expla-
nation comes from the Huygens construction illus-
trated in Figures (3) through (8).

Figure 6 (repeated)
The pattern becomes more complex when the slit is
Figure 3 (repeated) wider than a wavelength. Here you can begin to
The simple diffraction pattern you get when the see lines of nodes emerging from the slit.
slit is narrow compared to a wavelength.
 33-27

The fact that the diffraction pattern was photographed Analysis of the Single Slit Pattern
10 meters from the slits means that we are looking at the In our discussion of diffraction gratings, we estimated
pattern nearly 20 million wavelengths away from the the width of the maxima by determining how far from
slits. Thus the ripple tank photographs of Figures (5) the center of the maxima the intensity first went to zero,
through (8), showing diffraction patterns within a few where we first got complete cancellation. This oc-
wavelengths of the slits, are not a particularly relevant curred where light from pairs of slits cancelled. In our
guide as to what we could expect to see 20 million example of Figure (21), light from slit 1 cancelled that
wavelengths away. from slit 501, from slit 2 with slit 502, etc., all the way
down to slits 500 and 1000.
The general features of the diffraction pattern in Figure
(34) is that we have a relatively broad central maxi- We can use a similar analysis for the single slit pattern,
mum, with nodes on either side. Then there are dimmer except the one big slit is broken up, conceptually, into
and narrower maxima on either side. There is a series many narrow slits, as illustrated in Figure (35). Sup-
of these side maxima that extend out beyond the pose, for example, we think of the one wide slit of width
photograph of Figure (34). w as being broken up into 1000 neighboring individual
slits. The individual slits are so narrow that each piece
If the slit were narrow compared to a wavelength, if the of wave front in them should act as a source of a pure
wave spread out as in Figures (3), then we would get circular wave as shown back in Figure (3).
just one broad central maxima. Only when the slit is
wider than a wavelength do we get the minima we see Now consider the light heading out in such a direction
in Figure (34). These minima result from the interfer- that the wave from the first conceptual slit is half a
ence and cancellation of waves from different parts of wavelength λ/2 , in front of the wave from the middle
the slit. What we wish to do now is to show how this slit, number 501. When the waves from these two
cancellation occurs and predicts where the minima will "slits" strike the screen they will cancel. Similarly
be located. waves from slits 2 and 502 will cancel, as will those
from 3 and 503, etc., down to 500 and 1000. Thus is the
direction where the path length difference from the
edge to the center of the opening is half a wavelength
λ/2 . Between the two edges of the opening, the path
w = 50 microns
length difference to this minimum is λ ,as shown.
D = 10 meters
Ymin = 13 cm
Ymin
1
2
w m
3
4 inimu
irst m
to f
5
6

D 499
500
λ/2
501
w 502
Figure 34 503
For this single slit laser beam diffraction
Figure 35
pattern, the slit was about 100 wavelengths
Conceptually break
wide (w), and the screen was about 20
the single slit up into
million wavelengths away (D).
many individual slits. 998
We get a minimum 999
when light from the 1000 λ
conceptual slits
cancels in pairs.
33-28 Light Waves

From Figure (36), we can calculate the height of the RECORDING DIFFRACTION
first minimum using the familiar similar triangles we GRATING PATTERNS
have seen in previous analysis. The small right triangle Another way to record the diffraction pattern is to use
near the slit has a short side of length λ and a hypot-
a device called a photoresistor. A photoresistor is an
enuse equal to the slit width w. The big triangle has a
inexpensive resistor whose resistance R p varies de-
short side equal to Ymin and a hypotenuse given by the pending upon the intensity of the light striking the
Pythagorean theorem as D2 +Ymin 2
. Usually Ymin
resistor. If you place the photoresistor in the circuit
will be much smaller than the distance D, so that we can
shown in Figure (37), along with a fixed resistance R 2 ,
replace D2 +Ymin 2
by D, to get a battery of voltage Vb , and an oscilloscope, you can
measure with the oscilloscope the intensity of light
λ Ymin
≈
Ymin
w = D
striking the photoresistor.
2
D 2 + Ymin
or The analysis of the circuit in Figure (37) is as follows.
The resistors R p and R 1 are in series and thus have an
distance to the effective resistance R = R p + R 1 . The current i in the
≈ λD
first minima of
Y min w a single slit (14) circuit is thus i = Vb /R = Vb / R p + R 1 . Thus as the
diffraction pattern photoresistor's resistance R p changes with changes in
light intensity, the current i will also change. Finally
Exercise 14 the voltage V1 that the oscilloscope sees across the
fixed resistor R 1 is given by Ohm's law as V1 = iR 1 .
To obtain the single slit diffraction pattern seen in Figure
(34), we used a slit 50 microns wide located 10 meters
Thus as i changes, V1 changes and we see the change
from the screen. The distance Ymin to the first minimum on the oscilloscope.
was about 13 cm. Use this result to determine the i
wavelength of the laser light used. Compare your
answer with your results from Exercises 5 and 6, where R2
the same wavelength light was used.

Vb
2 photo-
2
+ Ymin resistor VRp= iRp
D Ymin Rp V
scope
D
w
λ Figure 37
The photoresistor circuit. By making R2 considerably
bigger than the photoresistor resistance Rp , the current
Figure 36
i stays relatively constant. As a result, the voltage
Similar triangles for calculating the distance to the
Vp = i Rp is nearly proportional to Rp . (We used
first minimum of a single slit diffraction pattern.
Rp = 6.8K Ω and the EG&G opto VT30N4
photoresistor.)
 33-29

For a number of years we tried various ways of moving A photoresistor is sensitive to the intensity or energy
the photoresistor through the diffraction pattern in density of the light striking it. And the intensity is
order to record the intensity of the light in the diffrac- proportional to the square of the amplitude of the waves
tion pattern. We tried mounting the photoresistor on xy in the beam. Thus in Figure (39) we are looking at a
recorders and various home-built devices, but there graph of the square of the wave amplitude in a single slit
was always some jitter and the results were only fair. diffraction pattern. It is reasonable that the intensity
The solution, as it turns out, is not to move the photo- should be proportional to the square of the amplitude,
resistor, but move the diffraction pattern across a fixed because amplitudes can be positive or negative, but
photoresistor instead. This is easily done using a rotat- intensities are always positive. You cannot have a
ing mirror, a mirror attached to a clock motor as shown negative intensity, and you do not get one if you square
in Figure (38). We have found that if you use a motor the amplitude since squares of real numbers are always
with a speed of 1/2 revolutions per minute, you have positive.
plenty of time to make a stable noise-free recording of
the bottom. Using the recording oscilloscope
MacScope, we recorded the single slit diffraction pat-
tern seen in Figure (39).

Figure 39 a,b
Single slit diffraction pattern. Data from
the project by Cham, Cole, and Layang.

Figure 38
The rotating mirror. (We were careful to make sure
that the axis of rotation was accurately perpendicular
to the base. If it isn’t, the laser beam wobbles up and
down.)

Figure 39c
Single slit diffraction pattern with the amplitude
of the voltage amplified so that we can see the
side lobes. Data from the project by Cham,
Cole, and Layang.
33-30 Light Waves

Exercise 15
In Figure (40) we study the interference pattern pro-
mirror rotating at
duced by a laser beam passing through three equally 1/2 revolution/hour
spaced slits. Figure (40a) shows the experimental
beam rotating at
setup and (40b) the shape of the slits through which the 1 revolution/hour
laser beam went. Figure (40c) is a photograph of the
interference pattern, and Figure (40d) is a recording in 3 slits
which the voltage is proportional to the intensity of the
light striking a stationary photoresistor. The beam
rotated at a rate of 1 revolution per hour, sweeping the photo resistor

laser
beam past the photo resistor. (The mirror only turned at behind slit
.5 revolutions per hour, but the reflected beam rotates
twice as fast as the mirror. You can see this by the fact
that when the mirror turns 45° the beam rotates 90°. )
a) Experimental setup
To Calculate the wavelength of the laser light from the
experimental data in Figure 40, first note that the beam
is sweeping past the photoresistor at a speed 3 slits
2 π r (cm / revolution)
vbeam = = 2 π r sec
cm line thickness = 53µ
3600 (sec / revolution) 3600
spacing between line
where r is the distance from the axis of the mirror to the
centers = 160µ
photoresistor. If it takes a time T for two maxima to
sweep past the photoresistor, then the distance Ymax
b) The slits. ( 1µ = 10– 6 m = 10– 4 cm )
between the maxima is

Ymax = vbeam ×T = 2 π r ×T
3600
If the slits are close to the mirror, then r is also equal to
the distance D from the slits to the photoresistor (screen).
The wavelength is then given by Equation 3a as
c) 3 slit diffraction pattern
λ = Ymax × d ≈ Ymax × dr
D

= 2 π r ×T × d
3600 r

The factors of r cancel, and we are left with

λ = 2 π d ×T (15)
3600

Thus if the slits are close to the mirror, we do not need

to know the distance to the photoresistor. (You can see
that if the rotating beam is twice as long, the end travels
twice as fast. But the maxima are twice as far apart, thus
it takes the same length of time for the maxima to pass
the photoresistor.)

Use the results of Figure 40d to determine the wave- d) Voltage recording on photoresistor
length of the laser beam. (In this experiment, the slits
were close to the mirror.) Figure 40
Recording the 3 slit interference pattern
(More to come on the use of Fourier analysis to predict
diffraction patterns.)
Chapter 34
Photons

CHAPTER 34 PHOTONS

The effort to determine the true nature of light has been Many of these problems were cleared up by a picture
a fitful process in the history of physics. Newton and developed by Max Planck and Einstein, a picture in
Huygens did not agree on whether light was a wave or which light consisted of beams of particles which
consisted of beams of particles. That issue was appar- became known as photons. The photon picture imme-
ently settled by Thomas Young's two-slit experiment diately explained the ejection of electrons from a metal
performed in 1801, nearly three quarters of a century surface and the spectrum of radiation from a heated
after Newton's death. Young's experiment still did not solid object. In the past few years the observation of
indicate what light was a wave of. That insight had to photons coming in uniformly from all directions in
come from Maxwell's theory of 1860 which showed space has led to a new and surprisingly well confirmed
that light was a wave of electric and magnetic fields. picture of the origin of the universe.
In the late 1800s there were dramatic confirmations of In this chapter we will discuss the properties of photons
Maxwell's theory. In 1888 Heinrich Hertz observed and how discovering the particle nature of light solved
radio waves, the expected low frequency component of some outstanding problems of the late nineteenth cen-
the electromagnetic spectrum. As we have seen from tury. We will finish with a discussion of what photons
our own experiments, the electric and magnetic fields have told us about the early universe.
in a radio wave can be measured directly.
What we will not discuss in this chapter is how to
But as the nineteenth century was ending, not all reconcile the two points of view about light. How could
predictions of Maxwell's theory were as successful. light behave as a wave in Thomas Young's experiment,
Applications of Maxwell's equations to explain the and as a particle in experiments explained by Einstein.
light radiated by matter were not working well. No one How could Maxwell's theory work so well in some
understood why a heated gas emitted sharp spectral cases and fail completely in others? These questions,
lines, and scientists like Boltzman were unable to which puzzled physicists for over a quarter of a cen-
explain important features of light radiated by hot solid tury, will be the topic of discussion in the chapter on
objects. The fact that Boltzman could get some features quantum mechanics.
right, but not others, made the problem more vexing.
Even harder to understand was the way beams of light
could eject electrons from the surface of a piece of
metal, a phenomenon discovered in 1897 by Hertz.
34-2 Photons

BLACKBODY RADIATION There are a few simple rules governing blackbody

When we studied the spectrum of hydrogen, we saw radiation. One is that the wavelength of the most
that heated hydrogen gas emits definite spectral lines, intense radiation, indicated by λ max in Figure (1), is
the red hydrogen α , the blue hydrogen β and violet inversely proportional to the temperature. The explicit
hydrogen γ . Other gases emit definite but different formula, known as Wein's displacement law turns out
spectral lines. But when we look through a diffraction to be
grating at the heated tungsten filament of a light bulb,
we see something quite different. Instead of sharp λ max = 2.898 mmK (1)
T
spectral lines we see a continuous rainbow of all the
colors of the visible spectrum. Another difference is where λ max is in millimeters and the temperature T is
that the color of the light emitted by the filament in kelvins. For T= 5800K, Equation 1 gives
changes as you change the temperature of the filament.
If you turn on the light bulb slowly, you first see a dull λ max 5800K = 2.898mmK = 5 × 10 – 4mm
red, then a brighter red, and finally the filament be- 5800K
comes white hot, emitting the full spectrum seen in = 5.0 × 10 – 5cm
white light. In contrast, if you heat hydrogen gas, you
see either no light, or you see all three spectral lines at which is the expected result.
definite unchanging wave lengths. While λ max changes with temperature, the relative
Some complications have to be dealt with when study- shape of the spectrum of radiated intensities does not.
ing light from solid objects. The heated burner on an Figure (1) is a general sketch of the blackbody radiation
electric stove and a ripe McIntosh apple both look red, spectrum. To determine the blackbody spectrum for
but for obviously different reasons. The skin of the another temperature, first calculate the new value of
McIntosh apple absorbs all frequencies of visible light λ max using Equation 1 then shift the horizontal scale in
except red, which it reflects. A stove burner, when it is Figure (1) so that λ max has this new value.
cool, looks black because it absorbs all wavelengths of
λ max classical theory
light equally. When the black stove burner is heated, radiation
the spectrum of light is not complicated by selective intensity
absorption or emission properties of the surface that
might enhance the radiation at some frequencies. The blackbody spectrum
for an object at a
light emitted by a heated black object has universal temperature of 5800K
characteristic properties that do not depend upon what like our sun.
kind of black substance is doing the radiating. The light
yellow

from such objects is blackbody radiation.

blue

red
ultra violet

One reason for studying blackbody radiation is that you

visible infra red
can determine the temperature of an object from the spectrum
light it emits. For example, Figure (1) shows the
intensity of light radiated at different wavelengths by a 1 2 3 4 5 6 7 8 9 10 12 14 16 10–5 cm
wavelength
tungsten filament at a temperature of 5800 kelvins. The
greatest intensity is at a wavelength of 5 × 10 – 5cm , Figure 1
Blackbody spectrum at 5800 degrees on the kelvin
the middle of the visible spectrum at the color yellow. scale. The solid line is the experimental curve, the
If we plot intensities of the various wavelengths radi- dotted line represents the prediction of Newtonian
ated by the sun, you get essentially the same curve. As mechanics combined with Maxwell’s equations. The
a result we can conclude that the temperature of the classical theory agrees with the experimental curve
surface of the sun is 5800 kelvins. It would be hard to only at long wavelengths.
make this measurement any other way.
 34-3

Knowledge of the blackbody spectrum is particularly touch the burner you feel infrared radiation which is
useful in astronomy. Most stars radiate a blackbody being emitted faster than it is being absorbed. By the
spectrum of radiation. Thus a measurement of the time the burner becomes red hot, the amount of radia-
value of λ max determines the temperature of the sur- tion it emits greatly exceeds the amount being ab-
face of the star. There happens to be quite a variation sorbed.
in the surface temperature and color of stars. This may
seem surprising at first, because most stars look white. In 1879, Joseph Stefan discovered that the total inten-
But this is due to the fact that our eyes are not color sity, the total energy emitted per second in blackbody
sensitive in dim light. The variation in the color of the radiation was proportional to the fourth power of the
stars can show up much better in a color photograph. temperature, to T 4 where T is in kelvins. Five years
later Ludwig Boltzman explained the result theoreti-
As an example of the use of Equation 1, suppose you cally. This result is thus known as the Stefan-Boltzman
observe a red star that is radiating a blackbody spec- law.
trum with λ max = 7.0 × 10 – 5cm . The surface tem-
perature should then be given by As an example of the use of the Stefan-Boltzman law,
suppose that two stars are of the same size, the same
2.898mmK = 4140 kelvin surface area, but one is a red star at a temperature of
T =
7.0 × 10 – 4mm 4,000K while the other is a blue star at a temperature of
10,000K. How much more rapidly is the hot blue star
radiating energy than the cool red star?
Exercise 1
The ratio of the rates of energy radiation is equal to the
(a) What is the surface temperature of a blue star whose
ratio of the fourth power of the temperatures. Thus
most intense wavelength is λmax = 4 × 10– 5cm ?

(b) What is the wavelength λmax of the most intense energy radiated
radiation emitted by an electric stove burner that is at a by blue star 4
Tblue 10,000K
4
= 4 =
temperature of 600° C (873K)? energy radiated Tred 4,000K
by red star
Another feature of blackbody radiation is that the = 2.5 4 ≈ 40
intensity of the radiation increases rapidly with tem-
perature. You see this when you turn up the voltage on We see that the blue star must be burning its nuclear fuel
the filament of a light bulb. Not only does the color 40 times faster than the red star.
change from red to white, the bulb also becomes much
brighter.

The net amount of radiation you get from a hot object

is the difference between the amount of radiation
emitted and the amount absorbed from the surround-
ings. If the object is at the same temperature as its
surroundings, it absorbs just as much radiation as it
emits, with the result that there is no net radiation. This
is why you cannot feel any heat from an electric stove
burner before it is turned on. But after the burner is
turned on and its temperature rises above the room
temperature, you begin to feel heat. Even if you do not
34-4 Photons

Planck Blackbody Radiation Law Planck succeeded in the following way. According to
Boltzman used a combination of Maxwell's equations, Maxwell's theory of light, the amount of radiation
Newtonian mechanics, and the theory of statistics to emitted or absorbed by a charged particle was related
show that the intensity of blackbody radiation in- to the acceleration of the particle, and that could vary
creased as the fourth power of intensity. But neither he continuously. Planck found that he could get his
nor anyone else was able to derive the blackbody empirical formula if he assumed that the electrons in a
radiation spectrum shown in Figure (1). There was solid emitted or absorbed radiation only in discrete
some success in predicting the long wavelength side of packets. The energy in each packet had to be propor-
the curve, but no one could explain why the intensity tional to the frequency of the radiation being emitted
curve dropped off again at short wavelengths. and absorbed. Planck wrote the formula for the energy
of the packets in the form
In 1900 Max Planck tried a different approach. He first
found an empirical formula for a curve that matched the E = hf (2)
blackbody spectrum. Then he searched for a derivation where f is the frequency of the radiation. The propor-
that would lead to his formula. The idea was to see if tionality constant h became known as Planck's con-
the laws of physics, as they were then known, could be stant.
modified in some way to explain his empirical black-
body radiation curve. For over two decades physicists had suspected that
something was wrong either with Newtonian mechan-
ics, Maxwell's equations, or both. Maxwell was unable
to derive a formula that explained the specific heat of
gases (except the monatomic noble gases), and no one
had the slightest idea why heated gases emitted sharp
spectral lines. Planck's derivation of the blackbody
radiation formula was the first successful derivation of
a phenomena that could not be explained by Newtonian
mechanics and Maxwell's equations.
But what did it mean that radiation could be emitted or
absorbed only in discrete packets or quanta as Planck
called them? What peculiar mechanism lead to this
quantization of the emission and absorption process?
Planck did not know.
 34-5

THE PHOTOELECTRIC EFFECT To perform the photoelectric effect experiment, clean

1905 was the year in which Einstein cleared up several the surface of a piece of zinc metal by scrubbing it with
outstanding problems in physics. We have seen how steel wool, and charge the zinc with a negative charge.
his focus on the basic idea of the principle of relativity We can be sure that the charge is negative by going
lead to his theory of special relativity and a new back to Ben Franklin's definition. If you rub a rubber
understanding of the structure of space and time. rod with cat fur, a negative charge will remain on the
Another clear picture allowed Einstein to explain why rubber rod. Then touch the rubber rod to the piece of
light was emitted and absorbed in discrete quanta in zinc, and the zinc will become negatively charged. The
blackbody radiation. The same idea also explained a presence of charge will be detected by the spreading of
process called the photoelectric effect, a phenomenon the gold leaves.
first encountered in 1887 by Heinrich Hertz. Now shine a beam of light at the charged piece of zinc.
In the photoelectric effect, a beam of light ejects For a source of light use a carbon arc that is generated
electrons from the surface of a piece of metal. This when an electric current jumps the narrow gap between
phenomenon can be easily demonstrated in a lecture, two carbon electrodes. The arc is so bright that you do
using the kind of equipment that was available to Hertz. not need to use a lens to focus the light on the zinc. The
You start with a gold leaf electrometer like that shown setup is shown in Figure (3).
in Figure (2), an old but effective device for measuring When the light is shining on the zinc, the gold leaves
the presence of electric charge. (This is the apparatus start to fall toward each other. Shut off or block the light
we used in our initial discussion of capacitors.) If a and the leaves stop falling. You can turn on and off the
charged object is placed upon the platform at the top of light several times and observe that the gold leaves fall
the electrometer, some of the charge will flow down to only when the light is shining on the zinc. Clearly it is
the gold leaves that are protected from air currents by the light from the carbon arc that is discharging the zinc.
a glass sided container. The gold leaves, each receiving
the same sign of charge, repel each other and spread
apart as shown. Very small amounts of charge can be
detected by the spreading of the gold leaves.

Figure 2
The gold leaf electrometer.
This is the same apparatus we
used back in Figure 26-28 in
our study of capacitors.

Figure 3 carbon arc

Photoelectric effect experiment.
light source
34-6 Photons

A simple extension to the experiment is to see what The question is whether the electric force is capable of
happens if the zinc is given a positive charge. Follow- ejecting an electron from the metal surface. A certain
ing Ben Franklin's prescription, we can obtain a posi- amount of energy is required to do this. For example,
tive charge by rubbing a glass rod with a silk cloth. in our electron gun experiment we had to heat the
Then touch the positively charged glass rod to the zinc filament in order to get an electron beam. It was the
and again you see the gold leaves separate indicating thermal energy that allowed electrons to escape from
the presence of charge. Now shine the light from the the filament. We now want to know whether the
carbon arc on the zinc and nothing happens. The leaves oscillating electric force of the light wave can supply
stay spread apart, and the zinc is not discharged by the enough energy to an electron for the electron to escape.
light.
There are two obvious conclusions we should reach.
When we charge the zinc with a negative charge, we are One is that we do not want the frequency of oscillation
placing an excess of electrons on the zinc. From to be too high, because the direction of the electric field
Gauss's law we know that there cannot be any net reverses on each half cycle of the oscillation. The
charge inside a conductor, thus the excess negative electron is pushed one way, and then back again. The
charge, the extra electrons, must be residing in the longer the time it is pushed in one direction, the lower
surface of the metal. The light from the carbon arc, the frequency of the oscillation, the more time the
which discharges the zinc, must therefore be knocking electron has to pick up speed and gain kinetic energy.
these extra electrons out of the metal surface. When we If the frequency is too high, just as the electron starts to
charged the zinc positively, we created a deficiency of move one way, it is pushed back the other way, and it
electrons in the surface, and no electrons were knocked does not have time to gain much kinetic energy.
out.
The second obvious conclusion is that we have a better
In the context of Maxwell's equations, it is not particu- chance of ejecting electrons if we use a more intense
larly surprising that a beam of light should be able to beam of light. With a more intense beam, we have a
knock electrons out of the surface of a piece of metal. stronger electric field which should exert a stronger
According to Maxwell's theory, light consists of a wave force on the electron, producing a greater acceleration
of electric and magnetic fields. An electron, residing and giving the electron more kinetic energy. An
on the surface of the zinc, should experience an oscil- intense enough beam might supply enough kinetic
lating electric force when the light shines on the zinc. energy for the electrons to escape.
The frequency of oscillation should be equal to the
frequency of the light wave, and the strength of the In summary, we expect that light might be able to eject
electric field should be directly related to the intensity electrons from the surface of a piece of metal if we use
of the light. (We saw earlier that the intensity of the a low enough frequency and an intense enough beam
light should be proportional to the square of the mag- of light. An intense beam of red light should give the
nitude of the electric field.) best results.

These predictions, based on Maxwell's equations

and Newtonian mechanics, are completely wrong!
 34-7

Let us return to our photoelectric effect demonstration. photoelectric effect occurred when a particle of light, a
During a lecture, a student suggested that we make the photon, struck an electron in the surface of the metal.
light from the carbon arc more intense by using a All the energy of the photon would be completely
magnifying glass to focus more of the arc light onto the absorbed by the electron. If this were enough energy
zinc. The more intense beam of light should discharge the electron could escape, if not, it could not.
the zinc faster.
The idea that light actually consisted of particles ex-
When you use a magnifying glass, you can make the plains why Planck had to assume that in blackbody
light striking the zinc look brighter. But something radiation, light could only be emitted or absorbed in
surprising happens. The zinc stops discharging. The quantum units. What was happening in blackbody
gold leaves stop falling. Remove the magnifying glass radiation, photons, particles of light, were being emit-
and the leaves start to fall again. The magnifying glass ted or absorbed. As a result, Planck's formula for the
prevents the discharge. energy of the quanta of emitted and absorbed radiation,
must also be the formula for the energy of a photon.
You do not have to use a magnifying glass to stop the Thus Einstein concluded that a photon's energy is given
discharge. A pane of window glass will do just as well. by the equation
Insert the window glass and the discharge stops. Re-
move it, and the gold leaves start to fall again.
Einstein's photoelectric
E photon = hf effect formula (3)
How could the window glass stop the discharge? The
window glass appears to have no effect on the light
where again f is the frequency of the light and h is
striking the zinc. The light appears just as bright. It was
Planck's constant. Equation 3 is known as Einstein's
brighter when we used the magnifying glass, but still no
photoelectric effect formula.
electrons were ejected. The prediction from Maxwell's
theory that we should use a more intense beam of light With Equation 3, we can begin to understand our
does not work for this experiment. photoelectric effect demonstration. It turns out that
visible photons do not have enough energy to knock an
What the window glass does is block ultraviolet radia-
electron out of the surface of zinc. There are other
tion. It is ultraviolet radiation that tans your skin (and
metals that require less energy and for these metals
can lead to skin cancer). It is difficult to get a tan indoors
visible light will produce a photoelectric effect. But for
from sunlight that has gone through a window because
zinc, visible photons do not have enough energy. Even
the glass has blocked the ultraviolet component of the
making the visible light more intense using a magnify-
sun's radiation. Similarly the pane of window glass, or
ing glass does not help. It is only the higher frequency,
the glass in the magnifying lens, used in the photoelec-
more energetic, ultraviolet photons that have enough
tric effect experiment, prevents ultraviolet radiation
energy to kick an electron out of the surface of zinc. We
from the carbon arc from reaching the zinc. It is the high
blocked these energetic photons with the window glass
frequency ultraviolet radiation that is ejecting electrons
and the magnifying glass.
from the zinc, not the lower frequency visible light.
This is in direct contradiction to the prediction of In 1921, Einstein received the Nobel prize, not for the
Maxwell's theory and Newton's laws. special theory of relativity which was still controver-
sial, nor for general relativity, but for his explanation of
Einstein's explanation of the photoelectric effect is
the photoelectric effect.
simple. He assumed that Newton was right after all, in
that light actually consisted of beams of particles. The
34-8 Photons

PLANCK'S CONSTANT h It is not hard to see that Planck's constant also has the
Planck's constant h, the proportionality constant in dimensions of angular momentum. Recall that the
Einstein's photoelectric effect formula, appears no- angular momentum L of an object is equal to the
where in Newtonian mechanics or Maxwell's theory of object's linear momentum p = mv times its lever arm
electricity and magnetism. As physicists were to r⊥ about some point. Thus the formula for angular
discover in the early part of the twentieth century, momentum is
Planck's constant appears just when Newtonian me-
L = pr⊥ = m kg v meter
sec × r⊥ meter
chanics and Maxwell's equations began to fail. Some-
thing was wrong with the nineteenth century physics, m2
= mvr⊥ kg sec
and Planck's constant seemed to be a sign of this failure.
We get the same dimensions if we write Planck's
The value of Planck's constant is
constant in the form
h = 6.63 × 10 – 34 joule sec (4) 2
h joule sec = h kg m 2 × sec
where the dimensions of h have to be an energy times sec
a time, as we can see from the photoelectric formula
m2
= h kg sec (5)
cycles
E joules = h joule sec × f sec (3a)
where we used the fact that the dimensions of energy
The dimensions check because cycles are dimension- are a mass times a velocity squared.
less. A fundamental constant of nature with the dimensions
of angular momentum is not something to be expected
in Newtonian mechanics. It suggests that there is
something special about this amount of angular mo-
mentum, 6.63 × 10 – 34 kg m 2 /sec of it, and
nowhere in Newtonian mechanics is there any reason
for any special amount. It would be Neils Bohr in 1913
who first appreciated the significance of this amount of
angular momentum.
 34-9

PHOTON ENERGIES Converting this to electron volts, we get

Up to a point we have been describing the electromag- 3.03 × 10 – 19 joules
netic spectrum in terms of the frequency or the wave- E H α line =
1.6 × 10 – 19 joules/eV
length of the light. Now with Einstein's photoelectric
formula, we can also describe the radiation in terms of E H α line = 1.89 eV (7)
the energy of the photons in the radiation. This can be
convenient, for we often want to know how much That is a convenient result. It turns out that the visible
energy photons have. For example, do the photons in spectrum ranges from about 1.8 eV for the long wave-
a particular beam of light have enough energy to kick length red light to about 3.1 eV for the shortest wave-
an electron out of the surface of a given piece of metal, length blue photons we can see. It requires 3.1 eV to
or to break a certain chemical bond? remove an electron from the surface of zinc. You can
see immediately that visible photons do not quite have
For visible light and nearby infrared light, the frequen- enough energy. You need ultraviolet photons with an
cies are so high that describing the light in terms of energy greater than 3.1 eV.
frequency is not particularly convenient. We are more
likely to work in terms of the light's wavelength and the Exercise 2
photon's energy, and want to go back and forth between The blackbody spectrum of the sun corresponds to an
the two. Using the formula object whose temperature is 5800 kelvin. The predomi-
cycles nant wavelength λmax for this temperature is
f sec = c meters/sec 5.0 × 10– 5cm as we saw in the calculation following
λ meters/cycle
Equation 1. What is the energy, in electron volts, of the
which we can get from dimensions, we can write the photons of this wavelength?
photoelectric formula in the form
Exercise 3
E = hf = hc (6) The rest energy of an electron is .51MeV = 5.1 × 105 eV .
λ What is the wavelength, in centimeters, of a photon
Using MKS units in Equation 6 for h, c, and λ , we end whose energy is equal to the rest energy of an electron?
up with the photon energy expressed in joules. But a
joule, a huge unit of energy compared to the energy of
a visible photon, is also inconvenient to use. A far more
convenient unit is the electron volt. To see why, let us
calculate the energy of the photons in the red hydrogen
α line, whose wavelength was 6.56 × 10 – 5cm or
6.56 × 10 – 7m . First calculating the energy in joules,
we have

E H α line = hc
λα
6.63 × 10 – 34 joule sec × 3 × 10 8m / sec
=
6.56 – 7 × 10 – 7m
= 3.03 × 10 – 19 joules
34-10 Photons

We will often want to convert directly from a photon’s Exercise 7

wavelength λ in centimeters to its energy E in electron A 100 watt bulb uses 100 joules of energy per second.
volts. This is most easily done by starting with the For this problem, assume that all this energy went into
formula E = hc/λ and using conversion factors until E emitting yellow photons at a wavelength of
is in electron volts when λ is in centimeters. We get λ = 5.88 × 10– 5cm .

E = hc (a) What is the energy, in eV and joules, of one of these

λ photons?
cm
6.63 × 10 – 34 joule sec × 3 × 10 10 sec (b) How many of these photons would the bulb radiate
=
λ cm in one second?

= 1.989 × 10 – 23 joule cm × 1 (c) From the results of part (b), explain why it is difficult
λ cm joule
1.6 × 10 – 19 to detect individual photons in a beam of light.
eV
The desired formula is thus Exercise 8
Radio station WBZ in Boston broadcasts at a frequency
E photon in eV = 12.4 × 10 eV ⋅ cm
–5 of 1050 kilocycles at a power of 50,000 watts.
(8)
λ in cm (a) How many photons per second does this radio
station emit?
As an example in the use of Equation 8, let us recalcu- (b) Should these photons be hard to detect individually?
late the energy of the H α photons whose wavelength
is 6.56 × 10 – 5cm . We get immediately Exercise 9
In what part of the electromagnetic spectrum will pho-
E Hα = 12.4 × 10 –eV
–5
cm = 1.89eV tons of the following energies be found?
6.56 × 10 cm5
(a) 1 eV (e) 5 eV
which is our previous result.
(b) 2.1 eV (f) 1000 eV
Exercise 4 (c) 2.5 eV (g) .51 × 106eV .51 MeV
The range of wavelengths of light in the visible spectrum (d) 3 eV (h) 4.34 × 10– 9eV
is from 7 × 10– 5cm in the red down to 4 × 10– 5cm in the
blue. What is the corresponding range of photon (The rest energy of the electron is .51 MeV.)
energies?
Exercise 10
Exercise 5 (a) Calculate the energy, in eV, of the photons in the
three visible spectral lines in hydrogen
(a) It requires 2.20 eV to eject an electron from the
surface of potassium. What is the longest wavelength
λ α red = 6.56 × 10– 5cm
light that can eject electrons from potassium?
λ β blue = 4.86 × 10– 5cm
(b) You shine blue light of wavelength 4 × 10– 5cm at
potassium. What is the maximum kinetic energy of the λ γ violet = 4.34 × 10– 5cm
ejected electrons?
It requires 2.28 eV to eject electrons from sodium.
Exercise 6
(b) The red Hα light does not eject electrons from
The human skin radiates blackbody radiation corre-
sodium. Explain why.
sponding to a temperature of 32°C. (Skin temperature
is slightly lower than the 37°C internal temperature.) (c) The Hβ and Hγ lines do eject electrons. What is the
What is the predominant energy, in eV of the photons maximum kinetic energy of the ejected electrons for
radiated by a human? (This is the energy correspond- these two spectral lines?
ing to λmax for this temperature.)
 34-11

PARTICLES AND WAVES problem, you first had to calculate the energy of a 1.05
We gain two different perspectives when we think of megacycle photon using Einstein's formula
the electromagnetic spectrum in terms of wavelengths E photon = hf . This turns out to be about
and in terms of photon energies. The wavelength 7 × 10 – 28 joules. The radio station is radiating 50,000
picture brings to mind Young's two slit experiment and joules of energy every second, and thus emitting
Maxwell's theory of electromagnetic radiation. In the 7 × 10 31 photons per second. It is hard to imagine an
photon picture we think of electrons being knocked out experiment in which we can detect individual photons
of metals and chemical bonds being broken. These when so many are being radiated at once. Any experi-
pictures are so different that it seems nearly impossible ments should detect some kind of average effect, and
to reconcile them. Reconciling these two pictures will, that average effect is given by Maxwell's equations.
in fact, be the main focus of the remainder of the text. When we get up to visible photons, whose energies are
For now we seek to answer a more modest question. in the 2—3 eV range and wavelengths of the order of
How can the two pictures coexist? How could some 5 × 10 – 5cm , it is reasonably easy to find experiments
experiments, like our demonstration of the photoelec- that can detect either the particle or the wave nature of
tric effect exhibit only the particle nature and com- light. With a diffraction grating we have no problem
pletely violate the predictions of Maxwell's equations, measuring wavelengths in the range of 10 – 5cm . With
while other experiments, like our measurements of the the photoelectric effect, we can easily detect individual
magnetic field of a radio wave, support Maxwell's photons in the 2-3 eV range.
equations and give no hint of a particle nature? As we go to shorter wavelengths, individual photons
In Figure (4) we show the electromagnetic spectrum have more energy and the particle nature begins to
both in terms of wavelengths and photon energies. It is dominate. To detect the wave nature of X rays, we need
in the low energy, long wavelength region, from radio something like a diffraction grating with line spacing of
waves to light waves, that the wave nature of the the order of the X ray wavelength. It turns out that the
radiation tends to dominate. At shorter wavelengths regular lines and planes of atoms in crystalline materi-
and higher photon energies, from visible light through als act as diffraction gratings allowing us to observe the
γ rays, the particle nature tends to dominate. The wave nature of X ray photons. But when we get up into
reason for this was well illustrated in Exercise 8. the γ ray region, where photons have energies compa-
rable to the rest energies of electrons and protons, all we
In Exercise 8 you were asked to calculate how many observe experimentally are particle reactions. At these
photons were radiated per second by radio station high energies, the wave nature of the photon is basically
WBZ in Boston. The station radiates 50,000 watts of a theoretical concept used to understand the particle
power at a frequency of 1.05 megacycles. To solve the reactions.

visible ultraviolet
radio, television, radar, microwaves light rays gamma rays

wavelength, cm
10 6 5 4 3 2 10 1 10 -1 -2 -3 -4 -5 -6 -7 -8 -9 -10 -11 -12

10 -10 -9 -8 -7 -6 -5 -4 -3 -2 10 -1 1 10 2 3 4 5 6 7 8
energy, eV

Figure 4 infrared rays X-rays

The electromagnetic spectrum.
34-12 Photons

While it is a rule of thumb that at wavelengths longer PHOTON MASS

than visible light, the wave nature of electromagnetic The basic idea behind Einstein's famous formula
radiation dominates, there are important exceptions. E = mc 2 is that energy is mass. The factor c 2 is a
The individual photons in the WBZ radio wave can be conversion factor to go between energy measured in
detected! You might ask, what kind of experiment can grams and energy measured in ergs. If we had used a
detect an object whose energy is only 7 × 10 – 28 different set of units, for example, measuring distances
joules. This, however, happens to be the amount of in feet, and time in nanoseconds, then the numerical
energy required to flip the spin of an electron or a value of c would be 1, and Einstein's equation would be
nucleus in a reasonably sized magnetic field. This spin E = m, the more revealing statement.
flip process for electrons is called electron spin reso-
nance and for nuclei, nuclear spin resonance. In Photons have energy, thus they have mass. If we
Chapter 38 we will discuss an electron spin resonance combine the photoelectric formula E = hf with
experiment that is easily performed in the lab. Nuclear E = mc 2 , we can solve for the mass m of a photon of
spin resonance, as you may be aware, is the basis of frequency f. The result is
magnetic resonance imaging, an increasingly impor-
tant medical diagnostic tool. E = hf =m photonc 2

The truly amazing feature of the magnetic resonance

experiments is that Maxwell's equations and Einstein's m photon = hf2 (10)
c
photoelectric effect formula make the same predic-
tions! Einstein's photoelectric effect formula is easier We can also express the photon mass in terms of the
to use and will be the way we analyze the electron spin wavelength λ , using f/c = 1 / λ
resonance experiment. Maxwell's equations and the
classical pictures of angular momentum and gyro- m photon = hc cf = h (11)
scopes facilitate the more detailed analysis needed for cλ
the imaging apparatus. Texts describing the imaging The idea that photons have mass presents a certain
apparatus use the classical approach. For this discus- problem. In our earliest discussions of mass in Chapter
sion the important point is that the two points of view 6, we saw that the mass increased with velocity,
come together in this low energy, long wavelength increasing without bounds as the speed of the object
limit. approached the speed of light. The formula that
described this increase in mass was

m0
m = (6-14)
1 – v 2 /c 2
where m 0 is the mass of the particle at rest and m its
mass when traveling at a speed v.

The obvious problems with photons is that they are

light—and therefore travel at the speed of light. Ap-
plying Equation 6-14 to photons gives

m0 m0 m
m photon = = = 0 (12)
1 – c 2 /c 2 1–1 0

a rather embarrassing result. The divisor in Equation

12 is exactly zero, not approximately zero. Usually
division by 0 is a mathematical disaster.
 34-13

There is only one way Equation 12 can be salvaged. For particles with rest mass, we used the formulas
The numerator m 0 must also be identically zero. Then E = mc 2 , m = m 0 / 1 – v 2 /c 2 to get the formulas for
Equation 12 gives m = 0/0, an undefined, but not the rest energy and the kinetic energy of the particle. In
disastrous result. The numerical value of 0/0 can be particular we got the approximate formula 1/2 m 0v 2
anything -1, 5, 10 – 17 , anything you want. In other for the kinetic energy of a slowly moving particle. For
words if the rest mass m 0 of a photon is zero, Equation photons, the formula m0 / 1 – v 2/c 2 does not apply,
12 says nothing about what the actual mass m photon is. there is no such thing as a slowly moving photon, and
Equation 12 only tells us that the rest mass of a photon the kinetic energy formula 1/2 mv 2 is completely
must be zero. wrong! For photons, all the energy is kinetic energy,
and the formula for the photon's kinetic energy is given
Stop a photon and what do you have left? Heat! In the by Einstein's photoelectric effect formula
daytime many billions of photons strike your skin E = hf = hc/λ . The energy of a photon is determined
every second. But after they hit nothing is left except by its frequency, not its speed.
the warmth of the sunlight. When a photon is stopped
it no longer exists—only its energy is left behind. That
is what is remarkable about photons. Only if they are Photon Momentum
moving at the speed of light do they exist, carry energy While photons have no rest mass, and do not obey
and have mass. This distinguishes them from all the Newton's second law, they do obey what turns out to be
particles that have rest mass and cannot get up to the a quite simple set of rules of mechanics. Like their
speed of light. massive counterparts, photons carry energy, linear
momentum, and angular momentum all of which are
An interesting particle is the neutrino. We are not sure conserved in interactions between particles. The for-
whether a neutrino (there are actually 3 different kinds mulas for these quantities can all be obtained straight-
of neutrinos) has a rest mass or not. If neutrinos have forwardly from Einstein's photoelectric formula E = hf
no rest mass, then they must travel at the speed of light, and energy formula E = mc 2 .
and obey the same mechanics as a photon. The
evidence is highly suggestive of this interpretation. We We have already combined these two equations to
saw, for example, that neutrinos from the 1987 super- obtain Equation 11 for the mass of a photon
nova explosion raced photons for some 100,000 years,
and took within an hour of the same amount of time to m photon = h (11a)
λc
get here. That is very close to the speed of light. If the
neutrinos took a tiny bit longer to reach us, if they To find the momentum of the photon, we multiply its
moved at slightly less than the speed of light, then they mass by its velocity. Since all photons move at the
would have to have some rest mass. The rest mass of same speed c, the photon momentum p photon is given
an individual neutrino would have to be extremely by
small, but there are so many neutrinos in the universe
that their total mass could make up a significant frac-
p photon = m photonc = h
tion of the mass in the universe. This might help λ (13)
explain some of the missing mass in the universe that
astronomers are worrying about. At the present time, In the next few chapters, we will find that Equation 13
however, all experiments are consistent with the idea applies to more than just photons. It turns out to be one
that a neutrino's rest mass is exactly zero. of the most important equations in physics.
34-14 Photons

In our discussion of systems of particles in Chapter 11, black side more than the reflecting side. Air molecules
we had an exercise where a boy washing a car, was striking the black side are heated, gain thermal energy,
squirting the hose at the door of the car. The water and bounce off or recoil from the vane with more speed
striking the door carried a certain amount of momen- than molecules bouncing off the cooler reflecting side.
tum per second, and as a result exerted a force F = dp/dt It is the extra speed of the recoil of the air molecules
on the door. The exercise was to calculate this force. from the black side that turns the vane. This thermal
effect is stronger than the force exerted by the light
When you shine a beam of light at an object, if the beam itself.
photons in the beam actually carry momentum p = h/λ
then the beam should exert a force equal to the rate at We can see from the example of the radiometer that the
which momentum is being absorbed by the object. If measurement of the force exerted by a beam of light,
the object absorbs the photon, like a black surface measuring the so-called pressure of light, must be done
would, the momentum delivered is just the momentum in a good vacuum during a carefully controlled experi-
of the photons. If it is a reflecting surface, then we have ment. That measurement was first made by Nichols
to include the photon recoil, and the momentum trans- and Hull at Dartmouth College in 1901. While
ferred is twice as great. Maxwell's theory of light also predicts that a beam of
light should exert a force, we can now interpret the
There is a common toy called a radiometer that has 4 Nichols and Hull experiment as the first experimental
vane structures balanced on the tip of a needle as shown measurement of the momentum carried by photons.
in Figure (5). One side of each vane is painted black,
while the other side is reflecting. If you shine a beam The first experiments to demonstrate that individual
of light at the vanes, they start to rotate. If, however, photons carried momentum were carried out by Arthur
you look at the apparatus for a while, you will notice Compton in 1923. In what is now known as the
that the vanes rotate the wrong way. They move as if Compton scattering or the Compton effect, X ray
the black side were being pushed harder by the beam of photons are aimed at a thin foil of metal. In many cases
light than the reflecting side. the X ray photons collide with and scatter an electron
rather than being absorbed as in the photoelectric
In the toy radiometers, it is not the force exerted by the effect. Both the struck electron and the scattered
light, but the fact that there are some air molecules photon emerge from the back side of the foil as illus-
remaining inside the radiometer, that causes the vanes trated in Figure (6).
to rotate. When the light strikes the vanes, it heats the
The collision of the photon with the electron in the
metal foil is in many ways similar to the collision of the
two steel balls studied in Chapter 7, Figures (1) and (2).
The energy of the X ray photons used by Compton were
of the order of 10,000 eV while the energy of the
electron in the metal is of the order of 1 or 2 eV. Thus
the X ray photon is essentially striking an electron at
rest, much as the moving steel ball struck a steel ball at
rest in Figure (7-2).

In both the collision of the steel balls and in the

Compton scattering, both energy and linear momen-
tum are conserved. In particular the momentum carried
in by the incoming X ray photon is shared between the
scattered X ray and the excited electron. This means
Figure 5
The radiometer.
 34-15

that the X ray photon loses momentum in the scattering The core will become so bright, emit so much light, that
process. Since the photon's momentum is related to its the pressure of the escaping light will lift the surface of
wavelength by p = h/λ , a loss in momentum means an the sun out into space. As a result the sun will expand
increase in wavelength. Thus, if the photon mechanics until it engulfs the orbit of the earth. At this point the
we have developed applies to X ray photons, then the sun will have become what astronomers call a red giant
scattered X rays should have a slightly longer wave- star. Because of its huge surface area it will become
length than the incident X rays, a result which Compton thousands of times brighter than it is now.
observed.
The red giant phase does not last long, only a few
According to Maxwell's theory, if a light wave im- million years. If the sun were bigger than it is, the
pinges on a metal, it should start the electrons oscillat- released gravitational potential energy would be enough
ing at the frequency of the incident wave. The oscillat- to ignite helium and nuclear fusion would continue.
ing electrons should then radiate light at the same But the red giant phase for the sun will be near the end
frequency. This radiated light would appear as the of the road. The sun will gradually cool and shrink,
scattered light in Compton's experiments. Thus becoming a white dwarf star about the size of the earth,
Maxwell's theory predicts that the scattered X rays and finally a black ember of about the same size.
should have the same wavelength as the incident wave,
a result which is not in agreement with experiment. The pressure of light played an even more important
role in the evolution of the early universe. The light
While the experiments we have just discussed involved from the big bang explosion that created the universe
delicate measurements in order to detect the photon was so intense that for the first 1/3 of a million years, it
momentum, in astronomy the momentum of photons knocked the particles of matter around and prevented
and the pressure of light can have dramatic effects. In the formation of stars, and galaxies. But a dramatic
about 5 billion years our sun will finish burning the event occurred when the universe reached an age of
hydrogen in its core. The core will then cool and start 1/3 of a million years. That was the point where the
to collapse. In one of the contradictory features of universe had cooled enough to become transparent. At
stellar evolution, the contracting core releases gravita- that point the light from the big bang decoupled from
tional potential energy at a greater rate than energy was matter and stars and galaxies began to form. We will
released by burning hydrogen. As a result the core discuss this event in more detail shortly.
becomes hotter and much brighter than it was before.
scattered electron
electrons after collision
slab of matter

electron
beam of X rays initially at rest

incoming
photon
scattered
photon

X rays which have collided with electrons in

the slab are scattered out of the main beam.
These X rays lose momentum, with the
result that their wavelength is longer than
those that were not scattered.

(a) observation of Compton scattering (b) collision of photon and electron

resulting in Compton scattering
Figure 6 .
Compton scattering.
34-16 Photons

ANTIMATTER The problem Dirac faced was that he could not ignore
The fact that photons have no rest mass and travel only the negative mass solution. If he started with a collec-
at the speed of light makes them seem quite different tion of positive mass particles and let them interact, the
from particles like an electron or proton that have rest equation predicted that negative mass particles would
mass and make up the atoms and molecules. The appear, would be created. He could not avoid them.
distinction fades somewhat when we consider a pro- Through a rather incredible trick, Dirac was able to
cess in which a photon is transformed into two particles reinterpret the negative mass solutions as positive mass
with rest mass. The two particles can be any particle- solutions of another kind of matter—antimatter. In this
antiparticle pair. Figure (7) is a bubble chamber interpretation, every elementary particle has a corre-
photograph of the creation of an electron-positron pair sponding antiparticle. The antiparticle had the same
by a photon. rest mass but opposite charge from its corresponding
In 1926 Erwin Schrödinger developed a wave equation particle. Thus a particle-antiparticle pair could be
to describe the behavior of electrons in atoms. The first created or annihilated without violating the law of
equation he tried had a serious problem; it was a conservation of electric charge.
relativistic wave equation that appeared to have two In 1927 when Dirac proposed his theory, no one had
solutions. One solution represented the ordinary elec- seen any form of antimatter, and no one was sure of
trons he was trying to describe, but the other solution exactly what to look for. The proton had the opposite
appeared to represent a particle with a negative rest charge from the electron, but its mass was much
mass. Schrödinger found that if he went to the non greater, and therefore it could not be the electron's anti-
relativistic limit, and developed an equation that ap- particle. If the electron antiparticle existed, it would
plied only to particles moving at speeds much less than have to have the same positive charge as the proton, but
the speed of light, then the negative rest mass solutions the same mass as an electron. In 1932 Carl Anderson
did not appear. The non relativistic equation was at Caltech found just such a particle among the cosmic
adequate to describe most chemical phenomena, and is rays that rain down through the earth's atmosphere.
the famous Schrödinger equation. That particle is the positron which is shown being
A year later, Paul Dirac developed another relativistic created in the bubble chamber photograph of
wave equation for electrons. The equation was specifi- Figure (7).
cally designed to avoid the negative mass solutions, but (In the muon lifetime moving picture, discussed in
the techniques used did not work. Dirac's equation Chapter 1, positively charged muons were stopped in
correctly predicted some important relativistic phe- the block of plastic, emitting the first pulse of light.
nomena, but as Dirac soon found out, the negative mass When a positive muon decays, it decays into a positron
solutions were still present. and a neutrino. It was the positron that made the second
Usually one ignores undesirable solutions to math- flash of light that was used to measure the muon's
ematical equations. For example, if you want to solve lifetime.)
for the hypotenuse of a triangle, the Pythagorean In the early 1950s, the synchrotron at Berkeley, the one
theorem tells you that c 2 = a 2 + b 2 . This equation has shown in Figure (28-27b) was built just large enough
two solutions, c = a 2 + b 2 and c = – a 2 + b 2 . to create antiprotons, and succeeded in doing so. Since
Clearly you want the positive solution, the negative then we have created antineutrons, and have observed
solution in this case is irrelevant. antiparticles corresponding to all the known elemen-
tary particles. Nature really has two solutions—matter
and antimatter.
 34-17

The main question we have now concerning antimatter Exercise 11

is why there is so little of it around at the present time. Since an electron and a positron have opposite charge,
In the very early universe, temperatures were so high they attract each other via the Coulomb electric force.
that there was a continual creation and annihilation of They can go into orbit forming a small atom like object
particles and antiparticles, with roughly equal but not called positronium. It is like a hydrogen atom except
exactly equal, numbers of particles and antiparticles. that the two particles have equal mass and thus move
There probably was an excess of particles over antipar- about each other rather than having one particle sit at
ticles in the order of about one part in 10 billion. In a the center. The positronium atom lasts for about a
short while the universe cooled to the point where microsecond, whereupon the positron and electron
annihilation became more likely than creation, and the annihilate each other, giving off their rest mass energy
particle-antiparticle pairs annihilated. What was left in the form of photons. The rest mass energy of the
behind was the slight excess of matter particles, the electron and positron is so much greater than their
orbital kinetic energy, that one can assume that the
particles that now form the stars and galaxies of the
positron and electron were essentially at rest when they
current universe.
annihilated. In the annihilation both momentum and
In 1964, James Cronin and Val Fitch, while working on energy are conserved.
particle accelerator experiments, discovered interac- (a) Explain why the positron and electron cannot anni-
tions that lead to an excess of particles over antipar- hilate, forming only one photon. (What conservation law
ticles. It could be that these interactions were active in would be violated by a one photon annihilation?)
the very early universe, creating the slight excess of
matter over antimatter. But on the other hand, there (b) Suppose the positronium annihilated forming two
photons. What must be the energy of each photon in
may not have been time for known processes to create
eV? What must be the relative direction of motion of the
the observed imbalance. We do not yet have a clear two photons?
picture of how the excess of matter over antimatter
came about. The answer to part (b) is that each photon must have an
energy of .51 MeV and the photons must come out in
exactly opposite directions. By detecting the emerging
photons you can tell precisely where the positronium
e– e+ annihilated. This phenomenon is used in the medical
imaging process called positron emission tomography
electron
positron or PET scans.

photon

Figure 7
Creation of positron-electron pair. A photon enters from the bottom
of the chamber and collides with a hydrogen nucleus. The nucleus
absorbs some of the photon’s momentum, allowing the photon’s
energy to be converted into a positron-electron pair. Since a photon
is uncharged, it leaves no track in the bubble chamber; the photon’s
path is shown by a dotted line. (Photograph copyright The Ealing
Corporation, Cambridge, Mass.)
34-18 Photons

INTERACTION OF PHOTONS energy shows up in the form of increased velocity, and

AND GRAVITY is given by the formula 1/2 mv 2 . For a photon, all of
whose energy is kinetic energy anyway, the kinetic
Because photons have mass, we should expect that
energy gained from the fall shows up as an increased
photons should interact with gravity. But we should be
frequency of the photon.
careful about applying the laws of Newtonian gravity
to photons, because Newtonian gravity is a non relativ- Using Einstein's formula E = hf for the kinetic energy
istic theory, while photons are completely relativistic of a photon, we predict that the photon energy at the
particles. bottom is given by
If we apply the ideas of Newtonian gravity to photons, E bottom = hfbottom = hftop + m photon gy (14)
which we will do shortly, we will find that we get
agreement with experiment if the photons are moving where we are assuming that the same formula mgy for
parallel to the gravitational force, for example, falling gravitational potential energy applies to both rocks and
toward the earth. But if we do a Newtonian type of photons.
calculation of the deflection of a photon as it passes a
star, we get half the deflection predicted by Einstein's Since m photon = hf/c 2 , the mass of the photon changes
general theory of relativity. It was in Eddington's slightly as the photon falls. But for a 22 meter deep
famous eclipse expedition of 1917 where the full shaft, the change in frequency is very small and we can
deflection predicted by Einstein's theory was observed. quite accurately use hftop/c 2 for the mass of the photon
This observation, along with measurements of the in Equation 14. This gives
precession of Mercury's orbit, were the first experi- hftop
mental evidence that Newton's theory of gravity was hfbottom = hftop + gy
c2
not exactly right.
Cancelling the h's, we get
In 1960, R. V. Pound and G.A. Rebka performed an
experiment at Harvard that consisted essentially of gy
fbottom = ftop 1 + (15)
dropping photons down a well. What they did was to c2
aim a beam of light of precisely known frequency down as the formula for the increase in the frequency of the
a vertical shaft about 22 meters long, and observed that photon. This is in agreement with the results found by
the photons at the bottom of the shaft had a slightly Pound and Rebka.
higher frequency, i.e., had slightly more energy than
when they were emitted at the top of the shaft.
Exercise 12
The way you can use Newtonian gravity to explain (a) Show that the quantity gy/c2 is dimensionless.
their results is the following. If you drop a rock of mass
(b) What is the percentage increase in the frequency of
m down a shaft of height h, the rock's gravitational
the photons in the Pound-Rebka experiment? (Answer:
potential energy mgh at the top of the shaft is converted
2.4 × 10– 13% . This indicates how extremely precise the
to kinetic energy at the bottom. For a rock, the kinetic experiment had to be.)
 34-19

To calculate the sideways deflection of a photon pass- It was hard to believe that something not much bigger
ing a star, we could use Newton's second law in the than a star could be as bright as a galaxy. There were
form F = dp/dt to calculate the rate at which a side- suggestions that the red shift detected by Maarten
ways gravitational force added a sideways component Schmidt was due to something other than the expan-
to the momentum of the photon. The gravitational sion of the universe. Perhaps quasars were close by
force would be Fg = m photong , with m photon = hf/c 2. objects that just happened to be moving away from us
The result, as we have mentioned, is half the deflection at incredible speeds. Perhaps they were very massive
predicted by Einstein's theory of gravity and half that objects so massive that the photons escaping from the
observed during Eddington's eclipse expedition. object lost a lot of their energy and emerged with lower
frequencies and longer wavelengths. (This would be
The gravitational deflection of photons, while difficult the opposite effect than that seen in the Pound-Rebka
to detect in 1917, has recently become a useful tool in experiment where photons falling toward the earth
astronomy. In 1961, Allen Sandage at Mt. Palomar gained kinetic energy and increased in frequency.)
Observatory discovered a peculiar kind of object that
seemed to be about the size of a star but which emitted Over the years, no explanation other than the expansion
radio waves like a radio galaxy. In 1963 Maarten of the universe satisfactorily explained the huge red
Schmidt photographed the spectral lines of a second shifts seen in quasars, but there was this nagging doubt
radio star and discovered that the spectral lines were all about whether the quasars were really that far away.
shifted far to the red. If this red shift were caused by the Everything seemed to fit with the model that red shifts
Doppler effect, then the radio star would be moving were caused by an expanding universe, but it would be
away from the earth at a speed of 16% the speed of light. nice to have direct proof.
If the motion were due to the expansion of the universe, The direct proof was supplied by gravitational lensing,
then the radio star would have to be between one and a consequence of the sideways deflection of photons as
two billion light years away. An object that far away, they pass a massive object. In 1979, a photograph
and still visible from the earth, would have to be as revealed two quasars that were unusually close to each
bright as an entire galaxy. other. Further investigation showed that the two qua-
sars had identical red shifts and emitted identical
The problem was the size of the object. The intensity spectral lines. This was too much of a coincidence. The
of the radiation emitted by these radio stars was ob- two quasars had to be two images of the same quasar.
served to vary significantly over times as short as
weeks to months. This virtually guarantees that the
object is no bigger than light weeks or light months
across, because the information required to coordinate
a major change in intensity cannot travel faster than the
speed of light. Thus Schmidt had found an object, not
much bigger than a star, radiating as much energy as
the billions of stars in a galaxy. These rather dramatic
objects, many more of which were soon found, became
known as quasars, which is an abbreviation for quasi
stellar objects.
34-20 Photons

How could two images of a single quasar appear side galaxy that is deflecting the quasar's light. This proved
by side on a photographic plate? The answer is that the quasars are distant objects and that the red shift
illustrated in Figure (8). Suppose the quasar were is definitely due to the expansion of the universe.
directly behind a massive galaxy, so that the light from
the quasar to the earth is deflected sideways as shown. Evidence over the years has indicted that quasars are
Here on earth we could see light coming from the the cores of newly formed galaxies. Quasars tend to be
quasar from 2 or more different directions. The tele- distant because most galaxies were formed when the
scope forms images as if the light came in a straight line. universe was relatively young. If all quasars we see are
Thus in Figure (8), light that came around the top side very far away, the light from them has taken a long time
of the galaxy would look like it came from a quasar to reach us, thus they must have formed a long time ago.
located above the actual quasar, while light that came The fact that we see very few nearby quasars means that
around the bottom side would look as if it came from most galaxy formation has already ceased.
another quasar located below the actual quasar.
Although we have photographed galaxies for over a
This gravitational lensing turned out to be a more hundred years, we know surprisingly little about them,
common phenomena than one might have expected. especially what is at the core of galaxies. Recent
More than a dozen examples of gravitational lensing evidence indicates that at the core of the galaxy M87
have been discovered in the past decade. Figure (9), an there is a black hole whose mass is of the order of
image produced by the repaired Hubble telescope, millions of suns. The formation of such a black hole
shows a quasar surrounded by four images of itself. would produce brilliant radiation from a very small
The four images were formed by the gravitational region of space, the kind of intense localized radiation
lensing of an intermediate galaxy. seen in quasars. At this point we only have proof for
one black hole at the center of one galaxy, but the pieces
The importance of gravitational lensing is that it pro- are beginning to fit together. Something quite spec-
vides definite proof that the imaged objects are more tacular may be at the center of most galaxies, and
distant than the objects doing the imaging. The quasar quasars are probably giving us a view of the formation
in Figure (9) must be farther away from us than the of these centers.

quasar
massive
galaxy
image

Figure 8
A galaxy, acting as a lens, can
produce multiple images of a
distant quasar. observer

Figure 9
Hubble telescope photograph of a distant quasar
surrounded by 4 images of the quasar. This is
known as the Einstein cross.
 34-21

EVOLUTION OF THE UNIVERSE There is another way to view the red shift that gives the
The two basic physical ideas involved in understand- same results but provides a more comprehensive pic-
ing the early universe are its expansion, and the idea ture of the evolution of the universe. Consider a galaxy
that the universe was in thermal equilibrium. Before that is, for example, receding from us at 10% the speed
we see how these concepts are applied, we wish to of light. According to the Doppler effect, the wave-
develop a slightly different perspective of these two length of the light from that galaxy will be lengthened
concepts. First we will see how the red shift of light can by a factor of 10%.
be interpreted in terms of the expansion of the universe. Where is that galaxy now? If the galaxy were moving
Then we will see that blackbody radiation can be away from us at 10% the speed of light, it has traveled
viewed as a gas of photons in thermal equilibrium. away from us 1/10th as far as the light has traveled in
With these two points of view, we can more easily reaching us. In other words the galaxy is 10% farther
follow the evolution of the universe. away now than when it emitted the light. If the reces-
sional motion of the galaxy is due to the expansion of
Red Shift and the the universe, then the universe is now 10% bigger than
Expansion of the Universe it was when the galaxy emitted the light.
The original clue that we live in an expanding universe
In this example, the universe is now 10% bigger and the
was from the red shift of light from distant galaxies. We
wavelength of the emitted light is 10% longer. We can
have explained this red shift as being caused by the
take the point of view that the wavelength of the light
Doppler effect. The distant galaxies are moving away
was stretched 10% by the expansion of the universe.
from us, and it is the recessional motion that stretches
the wavelengths of the radiated waves, as seen in the In other words it makes no difference whether we say
ripple tank photograph back in Figure (33-29) repro- that the red shift was caused by the 10% recessional
duced here. velocity of the galaxy, or the 10% expansion of the
universe. Both arguments give the same answer.
When we are studying the evolution of the universe, it
is easier to use the idea that the universe's expansion
stretches the photon wavelengths. This is especially
true for discussions of the early universe where reces-
sional velocities are close to the speed of light and
relativistic Doppler calculations would be required.

Figure 33-29
The doppler effect
34-22 Photons

Another View of Blackbody Radiation

E photon = 3 kT
The surface of the sun provides an example of a hot gas 2
more or less in thermal equilibrium. Not only are the E photon = hc
ordinary particles, the electrons, the protons, and other λ max
nuclei in thermal equilibrium, so are the photons, and
Combining these equations gives
this is why the sun emits a blackbody spectrum of
radiation. Blackbody radiation at a temperature T can hc = 3 kT
be viewed as a gas of photons in thermal equilibrium at λ max 2
that temperature.
λ max = 2hc
3kT
In our derivation of the ideal gas law, we were surpris-
ingly successful using the idea that the average gas Putting in numbers gives
molecule had a thermal kinetic energy 3/2 kT. In a
2 × 6.63 × 10 – 34 joule sec × 3 × 10 8 m
s
similar and equally naive derivation, we can explain λ max =
joule
one of the main features of blackbody radiation from 3 × 1.38 × 10 – 23 ×T
the assumption that the average or typical photon in K
blackbody radiation also has a kinetic energy 3/2 kT. = .0096 meter ⋅ K
T
The main feature of blackbody radiation, that could not Converting from meters to millimeters gives
be explained using Maxwell's theory of light, was the
fact that there was a peak in the blackbody spectrum. λ max = 9.60 mm ⋅ K our estimate (16)
There is a predominant wavelength which we have T for λmax
called λ max that is inversely proportional to the tem-
perature T. The precise relationship given by Wein's While this is not the exact result, it gives us the picture
displacement law is that there should be a peak in the blackbody spectrum
around λ max . The formula gives the correct tempera-
λ max = 2.898 mmK (1) repeated ture dependence, and the constant is only off by a factor
T of 3.3. Not too bad a result considering that we did not
a result we stated earlier. The blackbody radiation deal with relativistic effects and the distribution of
peaks around λ max as seen in Figure (1) reproduced energies in thermal equilibrium. None of these results
here. can be understood without the photon picture of light.

If blackbody radiation consists of a gas of photons in λ max classical theory

radiation
thermal equilibrium at a temperature T, we can assume intensity
that the average photon should have a kinetic energy
like 3/2 kT. (The factor 3/2 is not quite right for blackbody spectrum
relativistic particles, but close enough for this discus- for an object at a
temperature of 5800K
sion.) Some photons should have more energy, some like our sun.
less, but there should be a peak in the distribution of
yellow

photons around this energy. Using Einstein's photo-

blue

red
ultra violet

electric effect formula we can relate the most likely

photon energy to a most likely wavelength λ max . We
visible infra red
have spectrum

1 2 3 4 5 6 7 8 9 10 12 14 16 10–5 cm
wavelength

Figure 34-1 (reproduced)

Blackbody radiation spectrum showing the peak at λ max .
(The classical curve goes up to infinity at λ = 0.)
 34-23

MODELS OF THE UNIVERSE very strong, but short range, attractive nuclear force.
As we saw in Chapter 33, Hubble was able to combine Such a strong collision is required to overcome the
his new distance scale for stars and galaxies with Coulomb barrier, that fusion is a rare event in the
Doppler shift measurements to discover that the uni- lifetime of any particular solar proton. On the average,
verse is expanding, that the farther a galaxy is away a solar proton can bounce around about 30 million
from us, the faster it is moving away from us. Another years before fusing. There are, of course, many protons
property of the interaction of light with matter, the in the sun, so that many such fusions are occurring at
blackbody spectrum discussed at the beginning of this any one time.
chapter, provided a critical clue to the role of this Just after two protons fuse, electric potential energy is
expansion in the history of the universe. To see why, released when one of the protons decays via the weak
it is instructive to look at the evolution of our picture of interaction into a neutron, electron, and a neutrino. The
the universe, to see what led us to support or reject electron and neutrino are ejected, leaving behind a
different models of its large scale structure. deuterium nucleus consisting of a proton and a neutron.
This reaction is the source of the neutrinos radiated by
Powering the Sun the sun.
In the 1860s, Lord Kelvin, for whom the absolute
Within a few seconds of its creation, the deuterium
temperature scale is named, did a calculation of the age
nucleus absorbs another proton to become a helium 3
of the sun. Following a suggestion by Helmholtz,
nucleus. Since helium 3 nuclei in the sun are quite rare,
Kelvin assumed that the most powerful source of
it is on the average several million years before the
energy available to the sun was its gravitational poten-
helium 3 nucleus collides with another helium 3 nucleus.
tial energy. Noting the rate at which the sun was
The result of this collision is the very stable helium 4
radiating energy, Kelvin estimated that the sun was no
nucleus and the ejection of 2 protons. The net result of
older than half a billion years. This was a serious
all these steps is the conversion of 4 protons into a
problem for Darwin, whose theory of evolution re-
helium 4 nucleus with the release of .6% of the proton’s
quired considerably longer times for the processes of
rest mass energy in the form of neutrinos and photons.
evolution to have taken place. During their lifetimes
neither Darwin or Kelvin could explain the apparent Not only did Beta’s theory provide an explanation for
discrepancy of having fossils older than the sun. the source of the sun’s energy, it also demonstrated
how elements can be created inside of a star. It raised
This problem was overcome by the discovery that the
the question of whether all the elements could be
main source of energy of the sun was not gravitational
created inside stars. Could you start with stars initially
potential energy, but instead the nuclear energy re-
containing only hydrogen gas and end up with all the
leased by the fusion of hydrogen nuclei to form helium
elements we see around us?
nuclei. In 1938 Hans Beta worked out the details of
how this process worked. The reaction begins when
two protons collide with sufficient energy to overcome
the Coulomb repulsion and get close enough to feel the
34-24 Photons

Abundance of the Elements answer was supplied by E. E. Salpeter in 1952 who

From studies of minerals in the earth and in meteorites, showed that two helium 4 nuclei could produce an
and as a result of astronomical observations, we know unstable beryllium 8 nucleus. In a dense helium rich
considerable detail about the abundances of the ele- stellar core, the beryllium 8 nucleus could, before
ments around us. As seen in the chart of Figure (9), decaying, collide with another helium 4 nucleus form-
hydrogen and helium are the most abundant elements, ing a stable carbon 12 nucleus. One result is that
followed by peaks at carbon, oxygen, iron and lead. elements between helium and carbon are skipped over
There is a noticeable lack of lithium, beryllium and in the element formation process, explaining the ex-
boron, and a general trailing off of the heavier ele- ceptionally low cosmic abundances of lithium, beryl-
ments. Is it possible to explain not only how elements lium and boron seen in Figure (10).
could be created in stars, but also explain these ob-
served abundances as being the natural result of the The biggest barrier to explaining element formation in
nuclear reactions inside stars? stars is the fact that the iron 56 nucleus is the most stable
of all nuclei. Energy is released if the small nuclei fuse
The first problem is the fact that there are no stable together to create larger ones, but energy is also re-
nuclei with 5 or 8 nucleons. This means you cannot leased if the very largest nuclei split up (as in the case
form a stable nucleus either by adding one proton to a of the fission of uranium in an atomic bomb or nuclear
helium 4 nucleus or fusing two helium 4 nuclei. How, reactor). To put it another way, energy is released
then, would the next heavier element be formed in a star making nuclei up to iron, but it costs energy to build
that consisted of only hydrogen and helium 4? The nuclei larger than iron. Iron is the ultimate ash of
nuclear reactions. How then could elements heavier
12 than iron be created in the nuclear furnaces of stars?
10
Hydrogen

10
10 Helium

Oxygen
8
10 Carbon
Neon

6 Silicon Iron
10
Sulphur
Cosmic Abuncance

4
10

2
10
Lithium Boron
lead
1 Berylium
Thorium

–2
Uranium
10
2 3 4 5 6 7 8 9 10 20 30 40 50 60 80 100
Atomic Number
Figure 10
Abundance of the elements
 34-25

In 1956 the element technetium 99 was identified in the Knowing that the universe is expanding, it seems to be
spectra of a certain class of stars. Technetium 99, a contradiction to propose that the universe is steady
heavier than iron, is an unstable element with a half life state—i.e., that on the average, it is unchanging. If the
of only two hundred thousand years. On a cosmic time universe is expanding and galaxies are flying apart,
scale, this element had to have been made quite re- then in a few billion years the galaxies will be farther
cently. Thus elements heavier than iron are now being separated from each other than they are today. This is
created in some kind of a process. hardly a steady state picture.

Soon after the observation of technetium, the British The steady state theory got around this problem by
astronomer Geoffrey Burbidge, looking over recently proposing that matter was continually being created to
declassified data from the Bikini Atoll hydrogen bomb replace that being lost due to the expansion. Consider,
tests, noticed that one of the elements created in the for example, a sphere a billion light years in diameter,
explosion, californium 254, had a half life of 55 days. centered on the earth. Over the next million years a
Burbidge realized that this was also the half life of the certain number of stars will leave the sphere due to the
intensity of a recently observed supernova explosion. expansion. To replace this matter flowing out of the
This suggested that the light from the supernova was sphere, the steady state theory assumed that hydrogen
powered by decaying californium 254. That meant that atoms were continually being created inside the sphere.
it was the supernova explosion itself that created the All that was needed was about one hydrogen atom to be
very heavy californium 254, and probably all the other created in each cubic kilometer of space every year.
elements heavier than iron. The advantage of constructing a model like the steady
In 1957 Geoffrey and Margaret Burbidge, along with state theory is that the model makes certain definite
the nuclear physicist William Fowler at Caltech and the predictions that can be tested. One prediction is that all
British astronomer Fred Hoyle, published a famous the matter around us originated in the form of the
paper showing how the fusion process in stars could hydrogen atoms that are assumed to be continually
explain the abundances of elements up to iron, and how created. This implies that the heavier elements we see
supernova explosions could explain the formation of around us must be created by ongoing processes such
elements heavier than iron. This was one of the as nuclear reactions inside stars. This provided a strong
important steps in the use of our knowledge of the incentive for Hoyle and others to see if nuclear synthe-
behavior of matter on a small scale, namely nuclear sis inside stars, starting from hydrogen, could explain
physics, to explain what we see on a large scale—the the observed abundance of elements.
cosmic abundance of the elements. Another prediction of the steady state model is that
galaxies far away must look much like nearby galaxies.
The Steady State When you look far away, you are also looking back in
Model of the Universe time. If you look at a galaxy one billion light years
A model of the universe, proposed in 1948 by Fred away, you are seeing light that started out a billion years
Hoyle, Herman Bondi and Thomas Gold, fit very well ago. Light reaching us from a galaxy 10 billion light
with the idea that all the elements in the universe years away started out 10 billion light years ago. If the
heavier than hydrogen, were created as a result of universe is really in a steady state, then galaxies 10
nuclear reactions inside stars. This was the so-called billion years ago should look much like galaxies do
steady state model. today.
34-26 Photons

THE BIG BANG MODEL as big as it is now, wavelengths of light would contract
The discovery of the expansion of the universe sug- to 1/10 their original size.
gests another model of the universe, namely that the
In the Einstein photoelectric effect formula,
universe started in one gigantic explosion, and that the
expansion we now see is the result of the pieces from Ephoton = hf = hc/λ , the shorter the photon wavelength,
the more energetic the photons become. This suggests
that explosion flying apart.
that as we compress the universe in the time reversal
To see why you are led to the idea of an explosion, moving picture, photon energies increase. If there is no
imagine that you take a motion picture of the expanding limit to the compression, then there is no limit to how
universe and then run the motion picture backwards. If much the photon energies increase.
the expansion is uniform, then in the reversed motion
picture we see a uniform contraction. The particles in Now introduce the idea of thermal equilibrium. If we go
this picture are the galaxies which are getting closer and back to a very small universe, we have very energetic
photons. If these photons are in thermal equilibrium
closer together. There is a time, call it t = 0, when all
the galaxies come together at a point. Now run the with other forms of matter, as they are inside of stars,
motion picture forward and the galaxies all move out as then all of the matter has enormous thermal energy, and
the temperature is very high. Going back to a zero sized
if there were an explosion at that point.
universe means going back to a universe that started out
The explosion of the universe was first proposed by the at an infinite temperature. Fred Hoyle thought that this
Belgian priest and mathematician Georges Lemaître in picture was so ridiculous that he gave the explosion
the late 1920s. It was, in fact, Lemaître who explained model of the universe the derisive name the “Big Bang”
Hubble’s red shift versus distance data as evidence for model. The name has stuck.
the expansion of the universe. In the late 1920s not
much was known about nuclear physics, even the The Helium Abundance
neutron had not yet been discovered. But in the 1940s In the mid 1950s the cosmological theory taken seri-
after the development of the atomic fission bomb and ously by most physicists was the steady state theory. In
during the design of the hydrogen fusion bomb, physi- the late 1940s George Gamov had suggested that the
cists gained considerable experience with nuclear reac- elements had been created in the big bang when the
tions in hot, dense media, and some, George Gamov in universe was very small, dense and hot. But the work
particular, began to explore the consequences of the of Hoyle and Fowler was showing that the abundance of
idea that the universe started in an initial gigantic the elements could much more satisfactorily be ex-
explosion. plained in terms of nuclear synthesis inside of stars.
A rough picture of the early universe in the explosion This nuclear synthesis also explained the energy source
model can be constructed using the concepts of the in stars and the various stages of stellar evolution. What
Doppler effect and thermal equilibrium. Let us see how need was there to propose some gigantic, cataclysmic
this works. explosion?

We have seen that the red shift of the spectral lines of Hoyle soon found a need. Most of the energy released
light from distant galaxies can be interpreted as being in nuclear synthesis in stars results from the burning of
caused by the stretching of the wavelengths of the light hydrogen to form helium. By observing how much
due to the expansion of the universe. In a reverse energy is released by stars, you can estimate how much
motion picture of the universe, distant galaxies would helium should be produced. By the early 1960s Hoyle
be coming toward us and the wavelengths of the began to realize that nuclear synthesis could not pro-
spectral lines would be blue shifted. We would say that duce enough helium to explain the observed cosmic
the universe was contracting, shrinking the wavelength abundance of 25%. In a 1964 paper with R. J. Taylor,
of the spectral lines. The amount of contraction would Hoyle himself suggested that perhaps much of the
depend upon how far back toward the t = 0 origin we helium was created in an initial explosion of the uni-
went. If we went back to when the universe was 1/10 verse.
 34-27

Cosmic Radiation THE THREE DEGREE RADIATION

In a talk given at Johns Hopkins in early 1965, In 1964, two radio astronomers working for Bell Labs,
Princeton theoretician P. J. E. Peebles suggested that Arno Penzias and Robert Wilson, began a study of the
the early universe must have contained a considerable radio waves emitted from parts of our galaxy that are
amount of radiation if the big bang model were away from the galactic plane. They expected a faint
correct. If there were little radiation, any hydrogen diffuse radiation from this part of the galaxy and
present in the early universe would have quickly fused planned to use a sensitive low noise radio antenna
to form heavier elements, and no hydrogen would be shown in Figure (11), an antenna left over from the
left today. This directly contradicts the observation Echo satellite experiment. (In that early experiment
that about 75% of the matter we see today consists of on satellite communication, a reflecting balloon was
hydrogen. If, however, there were a large amount of placed in orbit. The low noise antenna was built to
radiation present in the early universe, the energetic detect the faint radio signals that bounced off the
photons would bust up the larger nuclei as they balloon.)
formed, leaving behind hydrogen.
Since the kind of signals Penzias and Wilson expected
Peebles proposed that this radiation, the cosmic pho- to detect would look a lot like radio noise, they had to
tons which prevented the fusion of hydrogen in the be careful that the signals they recorded were coming
early universe still exist today but in a very altered from the galaxy rather than from noise generated by
form. There should have been little change in the the antenna or by electronics. To test the system, they
number of photons, but a great change in their energy. looked for signals at a wavelength of 7.35 cm, a
As the universe expanded, the wavelength of the wavelength where the galaxy was not expected to
cosmic photons should be stretched by the expansion, produce much radiation. They found, however, a
greatly reducing their energy. If the photons were in stronger signal than expected. After removing a pair of
thermal equilibrium with very hot matter in the early pigeons that were living in the antenna throat, cleaning
universe, they should still have a thermal black body out the nest and other debris which Wilson referred to
spectrum, but at a much lower temperature. He as “a white dielectric material”, and taking other steps
predicted that the temperature of the cosmic radiation to eliminate noise, the extra signal persisted.
should have dropped to around 10 kelvin. His col-
leagues at Princeton, P. G. Roll and D .T. Wilkinson
were constructing a special antenna to detect such
radiation. All of this work had been suggested by R.
H. Dicke, inventor of the key microwave techniques
needed to detect ten degree photons.

Peebles was not the first to suggest that there should be

radiation left over from the big bang. That was first
suggested in a 1948 paper by George Gamov and
colleagues Ralph Alpher and Robert Herman in a
model where all elements were to be created in the big
bang. A more realistic model of the big bang proposed
by Alpher and Herman in 1953 also led to the same
prediction of cosmic radiation. In both cases, it was
estimated that the thermal radiation should now have
a temperature of 5 kelvin. In the early 1950s, Gamov,
Alpher and Herman were told by radio astronomers
that such radiation could not be detected by equipment
then available, and the effort to detect it was not Figure 11
pursued. Peebles was unaware of these earlier predic- Penzias and Wilson, and the Holmdel radio telescope.
tions.
34-28 Photons

If the 7.35 cm wavelength signal were coming from the various balloon and rocket based experiments, which
galaxy, there should be regions of the galaxy that lifted antennas above the earth’s atmosphere, verified
produced a stronger signal than other regions. And the that the radiation detected by Penzias and Wilson was
neighboring galaxy Andromeda should also be a local- part of a complete blackbody spectrum of radiation at
ized source of this signal. However Penzias and a temperature of 2.74 kelvin.
Wilson found that the 7.35 cm signal was coming in
uniformly from all directions. The radiation had to be In 1989, NASA orbited the COBE (Cosmic Back-
coming in from a much larger region of space than our ground Explorer) satellite to make a detailed study of
galaxy. the cosmic background radiation. The results from this
satellite verified that this radiation has the most perfect
Studies of the signal at still shorter wavelengths showed blackbody spectrum ever seen by mankind. The
that if the signal were produced by a blackbody spec- temperature is 2.735 kelvin with variations of the order
trum of radiation, the effective temperature would be of one part in 100,000. The questions we have to deal
about 3.5 kelvin. Penzias talked with a colleague who with now are not whether there is light left over from the
had talked with another colleague who had attended big bang, but why it is such a nearly perfect blackbody
Peebles’ talk at Johns Hopkins on the possibility of spectrum.
radiation left over from the big bang. Penzias and
Wilson immediately suspected that the signal they
Thermal Equilibrium of the Universe
were detecting might be from this radiation.
That the cosmic background radiation has nearly a
Penzias and Wilson could detect only the long wave- perfect blackbody spectrum tells us that at some point
length tail and of the three degree radiation. Three in its history, the universe was in nearly perfect thermal
degree radiation should have a maximum intensity at a equilibrium, with everything at one uniform tempera-
wavelength given by the Wein formula, Equation 1, ture. That is certainly not the case today. The cosmic
radiation is at a temperature of 2.735 kelvin, Hawaii has
an average temperature of 295 kelvin, and the tempera-
λ max = 2.898 mm K
T ture inside of stars ranges up to billions of degrees.
(17)
There must have been a dramatic change in the nature
= 2.898 mm K ≈ 1mm of the universe sometime in the past. That change
3K
occurred when the universe suddenly became transpar-
Radiation with wavelengths in the 1 mm region cannot ent at an age of about 700,000 years.
get through the earth’s atmosphere. As a result Penzias
and Wilson, and others using ground based antennas, To see why the universe suddenly became transparent,
could not verify that the radiation had a complete and why this was such an important event, it is instruc-
blackbody spectrum. From 1965 to the late 1980s, tive to reconstruct what the universe must have been
like at still earlier times.
 34-29

THE EARLY UNIVERSE There is essentially the same number of nucleons in this
Imagine that we have a videotape recording of the early universe as there are today. Since the thermal
evolution of the universe. We put the tape in our VCR energy of 40 MeV is much greater than the 1.3 MeV
and see that the tape has not been rewound. It is mass difference between neutrons and protons, there is
showing our current universe with stars, galaxies and enough thermal energy to freely convert protons into
the cosmic radiation at a temperature of 2.735 k. You neutrons, and vice versa. As a result there are about
can calculate the density of photons in the cosmic equal numbers of protons and neutrons. There is also
radiation, and compare that with the average density of about the same number of thermal photons in this early
protons and neutrons (nucleons) in the stars and galax- universe as there are today, about 100 billion photons
ies. You find that the photons outnumber the nucleons for each nucleon.
by a factor of about 10 billion to 1. Although there are While there is not much change from today in the
many more photons than nucleons, the rest energy of a number of nucleons or photons in our .01 second
proton or neutron is so much greater than the energy of universe, there is a vastly different number of electrons.
a three degree photon that the total rest energy of the The thermal photons, with an average energy of 40
stars and galaxies is about 100 times greater than the MeV, can freely create positron and electron pairs. The
total energy in the cosmic radiation. rest energy of a positron or an electron is only .5 MeV,
Leaving the VCR on play, we press the rewind button. thus only 1 MeV is required to create a pair. The result
The picture is not too clear, but we can see general is that the universe at this time is a thermal soup of
features of the contracting universe. The galaxies are photons, positrons and electrons—about equal num-
moving together and the wavelength of the cosmic bers. There are also many neutrinos left over from an
radiation is shrinking. Since the energy of the cosmic earlier time. All of those species outnumber the few
nucleons by a factor of about 100 billion to one.
photons is given by Einstein’s formula E = hc λ , the
shrinking of the photon wavelengths increases their
energy. On the other hand the rest mass energy of the Excess of Matter over Antimatter
stars and galaxies is essentially unaffected by the If you look closely and patiently count the number of
contraction of the universe. As a result the energy of the positrons and electrons in some region of space, you
cosmic photons is becoming a greater and greater share will find that for every 100,000,000,000 positrons,
of the total energy of the universe. When the universe there are 100,000,000,001 electrons. The electrons
has contracted to about 1/100th of its present size, when outnumber the positrons by 1 in 100 billion. In fact, the
the universe is about 1/2 million years old, the cosmic excess number of negative electrons is just equal to the
photons have caught up to the matter particles. At number of positive protons, with the result that the
earlier times, the cosmic photons have more energy universe is electrically neutral.
than other forms of matter.
The tiny excess of electrons over positrons represents
an excess of matter over antimatter. In most particle
The Early Universe reactions we study today, if particles are created, they
As the tape rewinds our attention is diverted. When we are created in particle, antiparticle pairs. The question
look again at the screen, we see that the tape is showing is then, why does this early universe have a tiny excess
a very early universe. The time indicated is .01 of matter particles over antimatter particles? What in
seconds! The temperature has risen to 100 billion the still earlier universe created this tiny imbalance?
degrees, and the thermal photons have an average There is a particle reaction, caused by the weak inter-
energy of 40 million electron volts! We obviously action, that does not treat matter and antimatter sym-
missed a lot in the rewind. Stopping the tape, we then metrically. This reaction, discovered by Val Fitch in
run it forward to see what the universe looks like at this 1964, could possibly explain how this tiny imbalance
very early stage. came about. It is not clear whether there was enough
time in the very early universe for Fitch’s reaction to
create the observed imbalance.
34-30 Photons

An excellent guidebook for our video tape is Steven Frame #4 (13.82 seconds)
Weinberg’s The First Three Minutes. Weinberg was At a time of 13.82 seconds, Weinberg’s fourth frame,
one of the physicists who discovered the connection the temperature has dropped to 3 billion kelvin, corre-
between the weak interaction and electromagnetism. sponding to an average thermal energy of 1 MeV per
Weinberg breaks up the first three minutes of the life of particle. With any further drop in temperature, the
the universe into five frames. We happened to have average thermal photon will not have enough energy to
stopped the tape recording at Weinberg’s frame #1. To create positron electron pairs. The result is that vast
see what we missed in our fast rewind, we will now run numbers of positrons and electrons are beginning to
the tape forward, picking up the other four frames in the annihilate each other. Soon there will be equal num-
first three minutes as well as important later events. bers of electrons and protons, and the only particles
remaining in very large numbers will be neutrinos and
Frame #2 (.11 seconds) thermal photons.
As we run the tape forward, the universe is now
expanding, the wavelength of the thermal photons is By this fourth frame, the percentage of neutrons has
getting longer, and their temperature is dropping. When dropped to 17%. The temperature of 3 billion degrees
the time counter gets up to t = .11 seconds, the tempera- is low enough for helium nuclei to survive, but helium
ture has dropped to 30 billion kelvin and the average nuclei do not form because of the deuterium bottle-
energy of the thermal photons has dropped to 10 MeV. neck. When a proton and neutron collide, they can
Back at frame #1, when the thermal energy was 40 easily form a deuterium nucleus. Although deuterium
MeV, there were roughly equal numbers of protons and is stable, it is weakly bound. At a temperature of 3
neutrons. However, the lower thermal energy of 10 billion kelvin, the thermal protons quickly break up any
MeV is not sufficiently greater than the 1.3 MeV deuterium that forms. Without deuterium, it is not
proton-neutron mass difference to maintain the equal- possible to build up still larger nuclei.
ity.
Frame #5 (3 minutes and 2 seconds)
In the many rapid collisions where protons are being Going forward to a time of 3 minutes and 2 seconds, the
converted into neutrons and vice versa (via the weak universe has cooled to a billion kelvin, the positrons
interaction), there is a slightly greater chance that the and most electrons have disappeared, and the only
heavier neutron will decay into a lighter proton rather abundant particles are photons, neutrinos and antineutri-
than the other way around. As a result the percentage nos. The neutron proton balance has dropped to 14%
of neutrons has dropped to 38% by the time t = .11 neutrons. While tritium (one proton and two neutrons)
seconds. and helium 4 are stable at this temperature, deuterium
is not, thus no heavier nuclei can form.
Frame #3 (1.09 seconds)
Aside from the drop in temperature and slight decrease A short time later, the temperature drops to the point
in the percentage of neutrons, not much else happened where deuterium is stable. When this happens, neu-
as we went from frame #1 at .01 seconds to frame #2 at trons can combine with protons to form deuterium and
.11 seconds. Starting up the tape player again, we go tritium, and these then combine to form helium 4.
forward to t = 1.09 seconds, Weinberg’s third frame. Almost immediately the remaining nearly 13% neu-
The temperature has dropped to 10 billion kelvin, trons combine with an equal number of protons to form
which corresponds to a thermal energy of 4 MeV. This most of the 25% abundance of cosmic helium we see
is not too far above the 1 MeV threshold for creating today. This is where the helium came from that Hoyle
positron electron pairs. As a result the positron electron could not explain in terms of nuclear synthesis inside of
pairs are beginning to annihilate faster than they are stars.
being created. Also by this time the percentage of
neutrons has dropped to 24%.
 34-31

Because there are no stable nuclei with 5 or 8 nucleons, A similar transition takes place in the early universe.
there is no simple route to the formation of still heavier By the age of about 700,000 years, the universe cools
elements. At a temperature of a billion degrees, the to a temperature of 3,000K. Before that the universe is
universe is only about 70 times hotter than the core of an opaque plasma like the inside of the sun. The
our sun, cooler than the core of hot stars around today photons in thermal equilibrium with the matter par-
that are fusing the heavier elements. As a result, nuclear ticles have enough energy to bust up any complete
synthesis in the early universe stops at helium 4 with a atoms and any gravitational clumps that are trying to
trace of lithium 7. One of the best tests of the big bang form.
theory is a rather precise prediction of the relative
abundances of hydrogen, deuterium, helium 4 and When the universe drops to a temperature below 3,000
lithium 7, all left over from the early universe. When kelvin, the hydrogen gas forms neutral atoms and
the formation of these elements is complete, the uni- becomes transparent. (The 25% helium had already
verse is 3 minutes and 46 seconds old. become neutral some time earlier). As a result the
universe suddenly becomes transparent, and the ther-
mal photons decouple from matter.
Decoupling (700,000 years)
Continue running the tape forward, and nothing of From this decoupling on, there is essentially no inter-
much interest happens for a long time. The thermal action between the thermal photons and any form of
photons still outnumber the nucleons and electrons by matter. All that happens to the photons is that their
a factor of about 10 billion to one, and the constant wavelength is stretched by the expansion of the uni-
collisions between these particles prevent the forma- verse. This stretching preserves the blackbody spec-
tion of atoms. What we see is a hot, ionized, nearly trum of the photons while lowering the effective black-
uniform plasma consisting of photons, charged nuclei body temperature. This blackbody spectrum is now at
and separate electrons. As time goes on, the plasma is the temperature of 2.735K, as observed by the COBE
expanding and cooling. satellite.

When you look at the sun, you see a round ball with an When the matter particles are decoupled, freed from
apparently sharp edge. But the sun is not a solid object the constant bombardment of the cosmic photons,
with a well defined surface. Instead, it is a bag of mostly gravity can begin the work of clumping up matter to
hydrogen gas held together by gravity. It is hottest at form stars, globular clusters, black holes and galaxies.
the center and cools off as you go out from the center. All these structures start to form after the decoupling,
At what appears to us to be the surface, the temperature after the universe is 700,000 years old. It is this
has dropped to about 3,000 kelvin. formation of stars and galaxies that we see as we run the
tape forward to our present day.
At a temperature above 3,000 kelvin, hydrogen gas
becomes ionized, a state where an appreciable fraction Looking out with ever more powerful telescopes is
of the electrons are torn free from the proton nuclei. essentially equivalent to running our videotape back-
When the gas is ionized, it is opaque because photons wards. The farther out we look, the farther back in time
can interact directly with the free charges present in the we see. Images from the Hubble telescope are giving
gas. Below a temperature of 3,000 kelvin, hydrogen us a view back toward the early universe when galaxies
consists essentially of neutral atoms which are unaf- were very much younger and quite different than they
fected by visible light. As a result the cooler hydrogen are today. The most distant galaxy we have identified
gas is transparent. The apparent surface of the sun so far emitted light when the universe was 5% of its
marks the abrupt transition from an opaque plasma, at current size.
temperatures above 3,000K, to a transparent gas at
temperatures below 3,000K.
34-32 Photons

What happens when we build still more powerful

telescopes and look still farther back? When we look
out so far that the universe is only 700,000 years old, we
are looking at the universe that has just become trans-
parent. We can see no farther! To look farther is like
trying to look down inside the surface of the sun.

In fact we do not need a more powerful telescope to see

this far back. The three degree cosmic background
radiation gives us a fantastically clear, detailed photo-
graph of the universe at the instant it went transparent.

The horn antenna used by Penzias and Wilson was the

first device to look at a small piece of this photograph.
The COBE satellite looked at the whole photograph,
but with rather limited resolution. COBE detected
some very tiny lumpiness, temperature variations of
about one part in 100,000. This lumpiness may have
been what gravity needed to start forming galaxies. A
higher resolution photograph will be needed to tell for
sure.

Guidebooks
We ran the videotape quite rapidly without looking at
many details. Our focus has been on the formation of
the elements and the three degree radiation, two of the
main pieces of evidence for the existence of a big bang.
We have omitted a number of fascinating details such
as how dense was the early universe, when did the
neutrinos decouple from matter, and what happened
before the first frame? There are excellent guidebooks
that accompany this tape where you can find these
details. There is Weinberg’s The First Three Minutes
which we have mentioned. The 1993 edition has an
addendum that introduces some ideas about the very,
very early universe, when the universe was millions of
times younger and hotter than the first frame. Perhaps
the best guidebook to how mankind came to our current
picture of the universe is the book by Timothy Ferris
Coming of Age in the Milky Way. Despite the title, this
is one of the most fascinating and readable accounts
available. In our discussion we have drawn much from
Weinberg and Ferris.
Chapter 35
Bohr Theory of Hydrogen

CHAPTER 35 BOHR THEORY OF

HYDROGEN

The hydrogen atom played a special role in the history continuously, it had to have special values, be quan-
of physics by providing the key that unlocked the new tized in units of Planck's constant divided by 2π , h/2π .
mechanics that replaced Newtonian mechanics. It In Bohr's theory, the different allowed orbits corre-
started with Johann Balmer's discovery in 1884 of a sponded to orbits with different allowed values of
mathematical formula for the wavelengths of some of angular momentum.
the spectral lines emitted by hydrogen. The simplicity
of the formula suggested that some understandable Again we see Planck's constant appearing at just the
mechanisms were producing these lines. point where Newtonian mechanics is breaking down.
There is no way one can explain from Newtonian
The next step was Rutherford's discovery of the atomic mechanics why the electrons in the hydrogen atom
nucleus in 1912. After that, one knew the basic could have only specific quantized values of angular
structure of atoms—a positive nucleus surrounded by momentum. While Bohr's model of hydrogen repre-
negative electrons. Within a year Neils Bohr had a sented only a slight modification of Newtonian me-
model of the hydrogen atom that "explained" the chanics, it represented a major philosophical shift.
spectral lines. Bohr introduced a new concept, the Newtonian mechanics could no longer be considered
energy level. The electron in hydrogen had certain the basic theory governing the behavior of particles
allowed energy levels, and the sharp spectral lines and matter. Something had to replace Newtonian
were emitted when the electron jumped from one mechanics, but from the time of Bohr's theory in 1913
energy level to another. To explain the energy levels, until 1924, no one knew what the new theory would be.
Bohr developed a model in which the electron had
certain allowed orbits and the jump between energy In 1924, a French graduate student, Louis de Broglie,
levels corresponded to the electron moving from one made a crucial suggestion that was the key that led to
allowed orbit to another. the new mechanics. This suggestion was quickly
followed up by Schrödinger and Heisenberg who de-
Bohr's allowed orbits followed from Newtonian me- veloped the new mechanics called quantum mechan-
chanics and the Coulomb force law, with one small but ics. In this chapter our focus will be on the develop-
crucial modification of Newtonian mechanics. The ments leading to de Broglie's idea.
angular momentum of the electron could not vary
35-2 Bohr Theory of Hydrogen

THE CLASSICAL HYDROGEN ATOM For an electron in a circular orbit, predicting the motion
With Rutherford's discovery of the atomic nucleus, it is quite easy. If an electron is in an orbit of radius r,
became clear that atoms consisted of a positively moving at a speed v, then its acceleration a is directed
charged nucleus surrounded by negatively charged toward the center of the circle and has a magnitude
electrons that were held to the nucleus by an electric 2
force. The simplest atom would be hydrogen consist- a = vr (2)
ing of one proton and one electron held together by a
Using Equation 1 for the electric force and Equation 2
Coulomb force of magnitude
for the acceleration, and noting that the force is in the
2 p Fe Fe e
Fe = e2 (1) same direction as the acceleration, as indicated in
r r Figure (2), Newton's second law gives
(For simplicity we will use CGS units in describing the
hydrogen atom. We do not need the engineering units, F = m a
and we avoid the complicating factor of 1/4πε0 in the
electric force formula.) As shown in Equation 1, both the e2 = m v2
r (3)
proton and the electron attract each other, but since the r2
proton is 1836 times more massive than the electron, the
proton should sit nearly at rest while the electron orbits One factor of r cancels and we can immediately solve
around it. for the electron's speed v to get v 2 = e 2/mr, or
velectron = e (4)
Thus the hydrogen atom is such a simple system, with mr
known masses and known forces, that it should be a
The period of the electron's orbit should be the distance
straightforward matter to make detailed predictions about
2πr travelled, divided by the speed v, or 2πr/v sec-
the nature of the atom. We could use the orbit program
onds per cycle, and the frequency should be the inverse
of Chapter 8, replacing the gravitational force GMm/r 2
of that, or v/2πr cycles per second. Using Equation 4
by e 2 /r 2 . We would predict that the electron moved in
an elliptical orbit about the proton, obeying all of Kepler's
for v, we get
laws for orbital motion. frequency of
= v = e (5)
electron in orbit 2πr 2πr mr
There is one important point we would have to take into
account in our analysis of the hydrogen atom that we did According to Maxwell's theory, this should also be the
not have to worry about in our study of satellite motion. frequency of the radiation emitted by the electron.
The electron is a charged particle, and accelerated charged
particles radiate electromagnetic waves. Suppose, for v0
example, that the electron were in a circular orbit moving v = v0sin(ωt)
e e
at an angular velocity ω as shown in Figure (1a). If we
were looking at the orbit from the side, as shown in Figure
(1b), we would see an electron oscillating up and down p p
with a velocity given by v = v0 sin ωt .

In our discussion of radio antennas in Chapter 32, we saw

that radio waves could be produced by moving electrons
up and down in an antenna wire. If electrons oscillated up a) electron in b) side view of
and down at a frequency ω , they produced radio waves circular orbit circular orbit
of the same frequency. Thus it is a prediction of Maxwell's
equations that the electron in the hydrogen atom should Figure 1
emit electromagnetic radiation, and the frequency of the The side view of circular motion
radiation should be the frequency at which the electron is an up and down oscillation.
orbits the proton.
 35-3

Electromagnetic radiation carries energy. Thus, to see constant GMm by e2 . Making this same substitution
what effect this has on the electron’s orbit, let us look in the potential energy formula gives
at the formula for the energy of an orbiting electron. 2 electrical potential energy
PE = – re of the electron in the (7)
From Equation 3 we can immediately solve for the hydrogen atom
electron's kinetic energy. The result is
Again the potential energy is zero when the particles
1 mv 2 = e 2 electron
kinetic (6) are infinitely far apart, and the electron loses potential
2 2r energy energy as it falls toward the proton. (We used this result
The electron also has electric potential energy just as an in the analysis of the binding energy of the hydrogen
earth satellite had gravitational potential energy. The molecule ion, explicitly in Equation 18-15.)
formula for the gravitational potential energy of a
satellite was The formula for the total energy E total of the electron in
hydrogen should be the sum of the kinetic energy,
potential energy
= – GMm
r (10-50a) Equation 6, and the potential energy, Equation 7.
of an earth satellite
potential
where M and m are the masses of the earth and the E total = kinetic
energy + energy
satellite respectively. This is the result we used in
2 2
Chapter 8 to test for conservation of energy (Equations = e – er
8-29 and 8-31) and in Chapter 10 where we calculated 2r
the potential energy (Equations 10-50a and 10-51). 2
Etotal = – e
total energy
The minus sign indicated that the gravitational force is (8)
2r of electron
attractive, that the satellite starts with zero potential
energy when r = ∞ and loses potential energy as it falls
The significance of the minus (–) sign is that the
in toward the earth.
electron is bound. Energy is required to pull the
We can convert the formula for gravitational potential electron out, to ionize the atom. For an electron to
energy to a formula for electrical potential energy by escape, its total energy must be brought up to zero.
comparing formulas for the gravitational and electric
We are now ready to look at the predictions that follow
forces on the two orbiting objects. The forces are
from Equations 5 and 8. As the electron radiates light
2
Fgravity = GMm ; Felectric = e2 it must lose energy and its total energy must become
r2 r more negative. From Equation 8 we see that for the
Since both are 1/r 2 forces, we can go from the gravi- electron's energy to become more negative, the radius
tational to the electric force formula by replacing the r must become smaller. Then Equation 5 tells us that
as the radius becomes smaller, the frequency of the
v radiation increases. We are lead to the picture of the
e electron spiraling in toward the proton, radiating even
a higher frequency light. There is nothing to stop the
Fe
process until the electron crashes into the proton. It is
an unambiguous prediction of Newtonian mechanics
p r and Maxwell's equations that the hydrogen atom is
unstable. It should emit a continuously increasing
frequency of light until it collapses.
Figure 2
For a circular orbit, both the acceleration a and the
force F point toward the center of the circle. Thus we
can equate the magnitudes of F and ma.
35-4 Bohr Theory of Hydrogen

Energy Levels The question is, why does the electron in hydrogen emit
By 1913, when Neils Bohr was trying to understand the only certain energy photons? The answer is Bohr's
behavior of the electron in hydrogen, it was no surprise main contribution to physics. Bohr assumed that the
that Maxwell's equations did not work at an atomic electron had, for some reason, only certain allowed
scale. To explain blackbody radiation and the photo- energies in the hydrogen atom. He called these allowed
electric effect, Planck and Einstein were led to the energy levels. When an electron jumped from one
picture that light consists of photons rather than energy level to another, it emitted a photon whose
Maxwell's waves of electric and magnetic force. energy was equal to the difference in the energy of the
two levels. The red 1.89 eV photon, for example, was
To construct a theory of hydrogen, Bohr knew the radiated when the electron fell from one energy level to
following fact. Hydrogen gas at room temperature another level 1.89 eV lower. There was a bottom,
emits no light. To get radiation, it has to be heated to lowest energy level below which the electron could not
rather high temperatures. Then you get distinct spectral fall. In cold hydrogen, all the electrons were in the
lines rather than the continuous radiation spectrum bottom energy level and therefore emitted no light.
expected classically. The visible spectral lines are the
H α , H β and H γ lines we saw in the hydrogen spec- When the hydrogen atom is viewed in terms of Bohr’s
trum experiment. These and many infra red lines we energy levels, the whole picture becomes extremely
saw in the spectrum of the hydrogen star, Figure (33- simple. The lowest energy level is at -13.6 eV. This is
28) reproduced below, make up the Balmer series of the total energy of the electron in any cold hydrogen
lines. Something must be going on inside the hydrogen atom. It requires 13.6 eV to ionize hydrogen to rip an
atom to produce these sharp spectral lines. electron out.

Viewing the light radiated by hydrogen in terms of 0

Einstein's photon picture, we see that the hydrogen –.544 n=5
–.850 n=4
atom emits photons with certain precise energies. As Figure 3
–1.51 n=3
an exercise in the last chapter you were asked to Energy level
calculate, in eV, the energies of the photons in the H α , diagram for the
H β and H γ spectral lines. The answers are hydrogen atom.
All the energy
–3.40 Hα Hβ Hγ n=2
E Hα = 1.89 eV levels are given by
the simple formula
E Hβ = 2.55 eV En = – 13.6/n 2 eV.
All Balmer series
E Hγ = 2.86 eV (9) lines result from
jumps down to the
n = 2 level. The 3
jumps shown give
rise to the three
visible hydrogen
lines.

–5 –5 –5 –5 –5
3.65 10 3.70 10 3.75 10 wavelength 3.80 10 3.85 10

H40 H30 H20 H15 H14 H13 H12 H11 H10 H9

Figure 33-28
Spectrum of a hydrogen star
–13.6 n=1
 35-5

The first energy level above the bottom is at –3.40 eV All of the lines in the Balmer result from jumps down
which turns out to be (–13.6/4) eV. The next level is at to the second energy level. For historical interest, let us
–1.51 eV which is (–13.6/9) eV. All of the energy see how Balmer's formula for the wavelengths in this
levels needed to explain every spectral line emitted by series follows from Bohr's formula for the energy
hydrogen are given by the formula levels. For Balmer's formula, the lines we have been
calling H α , H β and H γ are H 3 , H 4 , H 5 . An arbitrary
E n = – 13.62 eV (10) line in the series is denoted by H n , where n takes on the
n
values starting from 3 on up. The Balmer formula for
where n takes on the integer values 1, 2, 3, .... These the wavelength of the H n line is from Equation 33-6
energy levels are shown in Figure (3).
λ n = 3.65 × 10 – 5cm × n2 (33-6)
n2 – 4
Exercise 1
Use Equation 10 to calculate the lowest 5 energy levels Referring to Bohr's energy level diagram in Figure (3),
and compare your answer with Figure 3. consider a drop from the nth energy level to the second.
The energy lost by the electron is ( E n – E 2 ) which has
Let us see explicitly how Bohr's energy level diagram the value
explains the spectrum of light emitted by hydrogen. If, energy lost by
for example, an electron fell from the n=3 to the n=2 electron going
E n – E 2 = 13.62eV – 13.62eV from nth to
level, the amount of energy E 3–2 it would lose and n 2
therefore the energy it would radiate would be second level

This must be the energy E H n carried out by the

E 3–2 = E 3 – E 2
photon in the H n spectral line. Thus
= – 1.51 eV – ( – 3.40 eV) 1 1
(11) E H n = 13.6 eV – 2
= 1.89 eV 4 n
energy lost in falling
=
from n = 3 to n = 2 level n2 – 4 (12)
= 13.6 eV
which is the energy of the red photons in the H α line. 4n 2
We now use the formula
Exercise 2
λ = 12.4 × 10 cm ⋅ eV
–5
Show that the Hβ and Hγ lines correspond to jumps to (34-8)
the n = 2 level from the n = 4 and the n = 5 levels
E photon in eV
respectively.
relating the photon's energy to its wavelength. Using
From Exercise 2 we see that the first three lines in the Equation 12 for the photon energy gives
Balmer series result from the electron falling from the
λ n = 12.4 × 10 cm ⋅ eV 4n
–5 2
third, fourth and fifth levels down to the second level, 2
13.6 eV n –4
as indicated by the arrows in Figure (3).
λ n = 3.65 × 10 – 5cm n2
n2 – 4
which is Balmer's formula.
35-6 Bohr Theory of Hydrogen

It does not take great intuition to suspect that there are Another series, all of whose lines lie in the infra red, is
other series of spectral lines beyond the Balmer series. the Paschen series, representing jumps down to the
The photons emitted when the electron falls down to n = 3 energy level at -1.55 eV, as indicated in Figure (5).
the lowest level, down to -13.6 eV as indicated in There are other infra red series, representing jumps
Figure (4), form what is called the Lyman series. In this down to the n = 4 level, n = 5 level, etc. There are many
series the least energy photon, resulting from a fall from series, each containing many spectral lines. And all
-3.40 eV down to -13.6 eV, has an energy of 10.2 eV, these lines are explained by Bohr's conjecture that the
well out in the ultraviolet part of the spectrum. All the hydrogen atom has certain allowed energy levels, all
other photons in the Lyman series have more energy, given by the simple formula En = (– 13.6/n 2) eV .
and therefore are farther out in the ultraviolet. This one simple formula explains a huge amount of
experimental data on the spectrum of hydrogen.
It is interesting to note that when you heat hydrogen and
see a Balmer series photon like H α , H β or H γ ,
Exercise 3
eventually a 10.2 eV Lyman series photon must be
Calculate the energies (in eV) and wavelengths of the 5
emitted before the hydrogen can get back down to its
longest wavelength lines in
ground state. With telescopes on earth we see many
hydrogen stars radiating Balmer series lines. We do not (a) the Lyman series
see the Lyman series lines because these ultraviolet
(b) the Paschen series
photons do not make it down through the earth's
atmosphere. But the Lyman series lines are all visible On a Bohr energy level diagram show the electron
using orbiting telescopes like the Ultraviolet Explorer jumps corresponding to each line.
and the Hubble telescope.
Exercise 4
0 In Figure (33-28), repeated 2 pages back, we showed
–.544 n=5 the spectrum of light emitted by a hydrogen star. The
–.850 n=4 lines get closer and closer together as we get to H40 and
–1.51 n=3 just beyond. Explain why the lines get closer together
and calculate the limiting wavelength.

–3.40 n=2

Figure 4 0
The Lyman series Figure 5 –.278 n=7
consists of all jumps –.378 n=6
The Paschen series –.544 n=5
down to the –13.6eV consists of all jumps n=4
–.850
level. (Since this is as down to the n = 3
far down as the level. These are all in
electron can go, this the infra red. –1.51 n=3
level is called the
“ground state”.)

–13.6 n=1
 35-7

THE BOHR MODEL of hydrogen, we saw that an electron in an orbit of

Where do Bohr's energy levels come from? Certainly radius r had a total energy E(r) given by
not from Newtonian mechanics. There is no excuse in total energy of
2
Newtonian mechanics for a set of allowed energy E(r) = – e an electron in
a circular orbit (8 repeated)
levels. But did Newtonian mechanics have to be 2r
of radius r
rejected altogether? Planck was able to explain the
blackbody radiation formula by patching up classical If the electron can have only certain allowed energies
physics, by assuming that, for some reason, light was E n = –13.6/n 2 eV , then if Equation (8) holds, the
emitted and absorbed in quanta whose energy was electron orbits can have only certain allowed orbits of
proportional to the light's frequency. The reason why radius r n given by
Planck's trick worked was understood later, with 2
En = – e (13)
Einstein's proposal that light actually consisted of 2r n
particles whose energy was proportional to frequency. The r n are the radii of the famous Bohr orbits. This
Blackbody radiation had to be emitted and absorbed in leads to the rather peculiar picture that the electron can
quanta because light itself was made up of these quanta. exist in only certain allowed orbits, and when the
electron jumps from one allowed orbit to another, it
By 1913 it had become respectable, frustrating per- emits a photon whose energy is equal to the difference
haps, but respectable to modify classical physics in in energy between the two orbits. This model is
order to explain atomic phenomena. The hope was that indicated schematically in Figure (6).
a deeper theory would come along and naturally ex-
plain the modifications.
Exercise 5
What kind of a theory do we construct to explain the From Equation 13 and the fact that E1 = – 13.6 eV ,
allowed energy levels in hydrogen? In the classical calculate the radius of the first Bohr orbit r1 . [Hint: first
picture we have a miniature solar system with the convert eV to ergs.] This is known as the Bohr radius
proton at the center and the electron in orbit. This can and is in fact a good measure of the actual radius of a
be simplified by restricting the discussion to circular cold hydrogen atom. [The answer is
°
r1 = .529 × 10 cm = .529A .] Then show that rn = n2 r1 .
–8
orbits. From our earlier work with the classical model

Figure 6
The Bohr orbits are determined by
equating the allowed energy Lyma
ns
Ba

E n = – 13.6 n 2 to the energy E n = – e2 2rn e

lm
rie

er s

for an electron in an orbit of radius rn.

s

Paschen series

The Lyman series represents all jumps r1

eries

down to the smallest orbit, the Balmer

series to the second orbit, the Paschen r2
series to the third orbit, etc. (The radii in
this diagram are not to scale, the radii r n
increase in size as n 2, as you can easily
r3
show by equating the two values for E n.)
35-8 Bohr Theory of Hydrogen

Angular Momentum in the Bohr Model The next step is to express r in terms of the angular
Nothing in Newtonian mechanics gives the slightest momentum L. Squaring Equation 13 gives
hint as to why the electron in hydrogen should have L 2 = e 2 mr
only certain allowed orbits. In the classical picture or
there is nothing special about these particular radii. 2
r = L2 (16)
But ever since the time of Max Planck, there was a e m
special unit of angular momentum, the amount given
Finally we can eliminate the variable r in favor of the
by Planck's constant h. Since Planck's constant keeps
angular momentum L in our formula for the electron's
appearing whenever Newtonian mechanics fails, and
total energy. We get
since Planck's constant has the dimensions of angular
momentum, perhaps there was something special about
2
E = –e
the electron's angular momentum when it was in one of total energy
of the electron 2r
the allowed orbits.
= – e2
We can check this idea by re expressing the electron's
2 L 2 e2 m
total energy not in terms of the orbital radius r, but in
terms of its angular momentum L. 2 2
= –e e m
2 L2
We first need the formula for the electron's angular
4
momentum when in a circular orbit of radius r. Back = – e 2m (17)
in Equation 4, we found that the speed v of the electron 2L
was given by
In the formula – e 4m/2L 2 for the electron's energy,
v = e (4 repeated) only the angular momentum L changes from one orbit
mr
to another. If the energy of the nth orbit is E n , then
Multiplying this through by m gives us the electron's there must be a corresponding value L n for the angular
linear momentum mv momentum of the orbit. Thus we should write
4
mv = me = e m
r (14) En = – e m (18)
mr 2L2n
The electron's angular momentum about the center of v
the circle is its linear momentum mv times the lever
arm r, as indicated in the sketch of Figure (7). The result L = mvr m
is
r
L = mv r = e m r
r
(15)
= e mr
where we used Equation 14 for mv.
Figure 7
Angular momentum of a particle
moving in a circle of radius r.
 35-9

At this point, Bohr had the clue as to how to modify This quantity, Planck's constant divided by 2π , ap-
Newtonian mechanics in order to get his allowed pears so often in physics and chemistry that it is given
energy levels. Suppose that angular momentum is the special name “h bar” and is written h
quantized in units of some quantity we will call L0 . In
h ≡ h "h bar " (23)
the smallest orbit, suppose it has one unit, i.e., 2π
L1 = 1 × L0 . In the second orbit assume it has twice as Using h for L0 in the formula for E n , we get Bohr's
much angular momentum, L2 = 2 L0 . In the nth orbit formula
it would have n units
4
quantization E n = – e m2 12 (24)
Ln = nL0 of angular (19) 2h n
momentum
where e 4m/2h 2, expressed in electron volts, is 13.6 eV.
Substituting Equation 19 into Equation 18 gives This quantity is known as the Rydberg constant.
4 [Remember that we are using CGS units, where e is in
En = – e m 1
2 n2
(20) esu, m in grams, and h is erg-sec.]
2L 0

as the total energy of an electron with n units of angular Exercise 6

momentum. Comparing Equation 20 with Bohr's Use Equation 21 to evaluate L0 .
energy level formula
Exercise 7
E n = –13.6 eV 12 (10 repeated) What is the formula for the first Bohr radius in terms of the
n
electron mass m, charge e, and Planck's constant h.
we see that we can explain the energy levels by Evaluate your result and show that
° . (Answer: r = h2/e2m .)
r1 = .51 × 10– 8cm = .51A
assuming that the electron in the nth energy level has n 1

units of quantized angular momentum L0 . We can also

evaluate the size of L0 by equating the constant factors Exercise 8
in Equations 10 and 20. We get Starting from Newtonian mechanics and the Coulomb
force law F = e2/r2 , write out a clear and concise deriva-
e4m = 13.6 eV (21) tion of the formula
2L20 4
En = – e 2m 12
2h n
Converting 13.6 eV to ergs, and solving for L0 gives
Explain the crucial steps of the derivation.
e4m = 13.6 eV × 1.6 × 10 – 12 ergs
2L20 eV A day or so later, on an empty piece of paper and a clean
desk, see if you can repeat the derivation without
With e = 4.8 × 10 – 10esu and m = .911 × 10 – 27gm looking at notes. When you can, you have a secure
in CGS units, we get knowledge of the Bohr theory.
gm cm 2
L 0 = 1.05 × 10 – 27 sec (22)

which turns out to be Planck's constant divided by 2π .

6.63 × 10 – 27gm cm/sec
L0 = h =
2π 2π
gm cm
= 1.05 × 10 – 27 sec
35-10 Bohr Theory of Hydrogen

Exercise 9 DE BROGLIE'S HYPOTHESIS

An ionized helium atom consists of a single electron Despite its spectacular success describing the spectra
orbiting a nucleus containing two protons as shown in of hydrogen and other one-electron atoms, Bohr's
Figure (8). Thus the Coulomb force on the electron has theory represented more of a problem than a solution.
a magnitude It worked only for one electron atoms, and it pointed to
e 2e 2 an explicit failure of Newtonian mechanics. The idea
Fe = 2
= 2e2 of correcting Newtonian mechanics by requiring the
r r
angular momentum of the electron be quantized in
–e units of h , while successful, represented a bandaid
Figure 8 treatment. It simply covered a deeper wound in the
Ionized helium has a theory. For two centuries Newtonian mechanics had
nucleus with two protons, represented a complete, consistent scheme, applicable
surrounded by one electron. 2e
without exception. Special relativity did not harm the
integrity of Newtonian mechanics—relativistic New-
tonian mechanics is a consistent theory compatible
a) Using Newtonian mechanics, calculate the total with the principle of relativity. Even general relativity,
energy of the electron. (Your answer should be – e2/r . with its concepts of curved space, left Newtonian
Note that the r is not squared.) mechanics intact, and consistent, in a slightly altered
form.
b) Express this energy in terms of the electron's angular
momentum L. (First calculate L in terms of r, solve for The framework of Newtonian mechanics could not be
r, and substitute as we did in going from Equations 16 altered to include the concept of quantized angular
to 17.) momentum. Bohr, Sommerfield, and others tried
c) Find the formula for the energy levels of the electron during the decade following the introduction of Bohr's
in ionized helium, assuming that the electron's angular model, but there was little success.
momentum is quantized in units of h.
In Paris, in 1923, a graduate student Louis de Broglie,
d) Figure out whether ionized helium emits any visible had an idea. He noted that light had a wave nature, seen
spectral lines (lines with photon energies between 1.8 in the 2-slit experiment and Maxwell's theory, and a
eV and 3.1 eV.) How many visible lines are there and particle nature seen in Einstein's explanation of the
what are their wavelengths?) photoelectric effect. Physicists could not explain how
light could behave as a particle in some experiments,
Exercise 10 and a wave in others. This problem seemed so incon-
You can handle all single electron atoms in one calcu- gruous that it was put on the back burner, more or less
lation by assuming that there are z protons in the ignored for nearly 20 years.
nucleus. (z = 1 for hydrogen, z = 2 for ionized helium,
z = 3 for doubly ionized lithium, etc.) Repeat parts a), b), De Broglie's idea was that, if light can have both a
and c) of Exercise 9 for a single electron atom with z particle and a wave nature, perhaps electrons can too!
protons in the nucleus. (There is no simple formula for Perhaps the quantization of the angular momentum of
multi electron atoms because of the repulsive force an electron in the hydrogen atom was due to the wave
between the electrons.)
nature of the electron.

The main question de Broglie had to answer was how

do you determine the wavelength of an electron wave?
 35-11

An analogy with photons might help. There is, how- But if the circumference of the circle were an exact
ever, a significant difference between electrons and integral number of wavelengths as illustrated in Figure
photons. Electrons have a rest mass energy and pho- (10), there would be no cancellation. This would
tons do not, thus there can be no direct analogy between therefore be one of Bohr's allowed orbits shown in
the total energies of the two particles. But both particles Figure (6).
have mass and carry linear momentum, and the amount
of momentum can vary from zero on up for both Suppose (n) wavelengths fit around a particular circle
particles. Thus photons and electrons could have of radius r n . Then we have
similar formulas for linear momentum. nλ = 2πr n (25)

Back in Equation 34-13 we saw that the linear momen- Using the de Broglie formula λ = h/p for the electron
tum p of a photon was related to its wavelength λ by the wavelength, we get
simple equation
n hp = 2πr n (26)
de Broglie
λ = h
p wavelength (34-13) Multiplying both sides by p and dividing through by
2π gives
De Broglie assumed that this same relationship also n h = pr n (27)
2π
applied to electrons. An electron with a linear momen-
tum p would have a wavelength λ = h/p . This is now Now h/2π is just h , and pr n is the angular momentum
called the de Broglie wavelength. This relationship L n (momentum times lever arm) of the electron. Thus
applies not only to photons and electrons, but as far as Equation 27 gives
we know, to all particles! nh = pr n = L n (28)
With a formula for the electron wavelength, de Broglie Equation 28 tells us that for the allowed orbits, the
was able to construct a simple model explaining the orbits in which the electron wave does not cancel, the
quantization of angular momentum in the hydrogen angular momentum comes in integer amounts of the
atom. In de Broglie's model, one pictures an electron angular momentum h . The quantization of angular
wave chasing itself around a circle in the hydrogen momentum is thus due to the wave nature of the
atom. If the circumference of the circle, 2πr did not electron, a concept completely foreign to Newtonian
have an exact integral number of wavelengths, then the mechanics.
wave, after going around many times, would eventu-
ally cancel itself out as illustrated in Figure (9).

Figure 9 Figure 10 Figure 10a--Movie

De Broglie picture of an If the circumference of the orbit is The standing waves on a
electron wave cancelling an integer number of wavelengths, circular metal band nicely
itself out. the electron wave will go around illustrate de Broglie’s waves
without any cancellation.
35-12 Bohr Theory of Hydrogen

When a graduate student does a thesis project, typically metal had crystallized during the heating, and the
the student does a lot of work under the supervision of peculiar scattering had occurred as a result of the
a thesis advisor, and comes up with some new, hope- crystallization. Davisson and Germer then prepared a
fully verifiable, results. What do you do with a student target consisting of a single crystal, and studied the
that comes up with a strange idea, completely unveri- peculiar scattering phenomena extensively. Their ap-
fied, that can be explained in a few pages of algebra? paratus is illustrated schematically in Figure (11), and
Einstein happened to be passing through Paris in the their experimental results are shown in Figure (12). For
summer of 1924 and was asked if de Broglie's thesis their experiment, there was a marked peak in the
should be accepted. Although doubtful himself about scattering when the detector was located at an angle of
a wave nature of the electron, Einstein recommended 50° from the incident beam.
that the thesis be accepted, for de Broglie just might be
right. Davisson presented these results at a meeting in Lon-
don in the summer of 1927. At that time there was a
In 1925, two physicists at Bell Telephone Laboratories, considerable discussion about de Broglie's hypothesis
C. J. Davisson and L. H. Germer were studying the that electrons have a wave nature. Hearing of this idea,
surface of nickel by scattering electrons from the Davisson recognized the reason for the scattering peak.
surface. The point of the research was to learn more The atoms of the crystal were diffracting electron
about metal surfaces in order to improve the quality of waves. The enhanced scattering at 50° was a diffrac-
switches used in telephone communication. After tion peak, a maximum similar to the reflected maxima
heating the metal target to remove an oxide layer that we saw back in Figure (33-19) when light goes through
accumulated following a break in the vacuum line, they a diffraction grating. Davisson had the experimental
discovered that the electrons scattered differently. The evidence that de Broglie's idea about electron waves
was correct after all.
electron gun

detector θ=
Reflected 50
°
maximum

θ
electron
beam

transmitted
maximum
Figure 33-19
Figure 12
nickel crystal Laser beam Plot of intensity vs. angle for electrons scattered by a
impinging on nickel crystal, as measured by Davisson and Germer.
Figure 11 a diffraction The peak in intensity at 50° was a diffraction peak
Scattering electrons from the grating. like the ones produced by diffraction gratings. (The
surface of a nickel crystal.
intensity is proportional to the distance out from the
origin.)
Chapter 36
Scattering of Waves

CHAPTER 36 SCATTERING OF
WAVES
We will briefly interrupt our discussion of the hydrogen
atom and study the scattering of waves by atoms. It was
the scattering of electron waves from the surface of a
nickel crystal that provided the first experimental
evidence of the wave nature of electrons. Earlier
experiments involving the scattering of x rays had
begun to yield detailed information about the atomic
structure of crystals.

Our main focus in this chapter will be an experiment

developed in the early 1960s by Harry Meiners at R .P.
I., that makes it easy for students to study electron
waves and work with de Broglie's formula λ = h/p .
The apparatus involves the scattering of electrons from
a graphite crystal. The analysis of the resulting diffrac-
tion pattern requires nothing more than a combination
of the de Broglie formula with the diffraction grating
formula discussed in Chapter 33. We will use Meiner's
experiment as our main demonstration of the wave
nature of the electron.
36-2 Scattering of Waves

SCATTERING OF A WAVE If the scattering object is smaller in size than the

BY A SMALL OBJECT wavelength of the wave, as in Figure (1), the scattered
waves contain essentially no information about the
The first step in studying the scattering of waves by
shape of the object. For this reason, you cannot study
atoms is to see what happens when a wave strikes a
the structure of something that is much smaller than the
small object, an object smaller in size than the wave-
wavelength of the wave you are using for the study.
length of the wave. The result can be seen in the ripple
Optical microscopes, for example, cannot be used to
tank photographs shown in Figure (1). In (1a), an
study viruses, because most viruses are smaller than the
incident wave is passing over a small object. You can
wavelength of visible light. (Very clever work with
see scattered waves emerging from the object. In (1b),
optical microscopes allows one to see down to about
the incident wave has passed, and you can see that the
1/10th of the wavelength of visible light, to see objects
scattered waves are a series of circular waves, the same
like microtubules.)
pattern you get when you drop a stone into a quiet pool
of water.
incident wave incident wave

a) Incident and scattered wave together. b) After the incident wave has passed.

Figure 1
If the scattering object is smaller than a wavelength,
we get circular scattered waves that contain little or
no information about the shape of the object.
 36-3

REFLECTION OF LIGHT waves do not cancel. Instead they add up to produce a

Using the picture of scattering provided by Figure (1), new, transmitted wave whose crests move slower than
we can begin to understand the reflection of visible the speed of light. This apparent slowing of the speed
light from a smooth metal surface. Suppose we have a of light, due to the interference of transmitted and
long wavelength wave impinging on a metal surface scattered waves, leads to the bending of a beam of light
represented by a regular array of atoms, as illustrated in as it enters or leaves a transparent medium. It is this
Figure (2). As the wave passes over the array of atoms, bending that allows one to construct lenses and
circular scattered waves emerge. As seen in Figure optical instruments.
(2a), the scattered waves add up to produce a reflected
wave coming back out of the surface. The angles Exercise 1
labeled θ i and θ r in Figure (2b) are what are called the Using Figure (2), prove that the angle of incidence
angle of incidence and angle of reflection , respec- equals the angle of reflection.
tively. Since the scattered waves emerge at the same
speed as the incident wave enters, it is clear from the
geometry that the angle of incidence is equal to the angle of θi θr angle of
angle of reflection. That is the main rule governing the incidence reflection
reflection of light.

What happens inside the material depends upon details

of the scattering process. Note that the reflected
wavefront inside the material coincides with the inci-
dent wave. For a metal surface, the phases of the mirror
scattered waves are such that the reflected wave inside Figure 2b
When light reflects from a mirror, the angle
just cancels the incident wave and there is no wave
of incidence equals the angle of reflection.
inside. All the radiation is reflected. For other types of
material that are not opaque, the incident and scattered

reflected wave
angle of θi θr angle of
incidence reflection

incident wave

Figure 2a
A reflected wave is produced when the incident wave is
scattered by many atoms. From this diagram, you can see
why the angle of incidence equals the angle of reflection.
36-4 Scattering of Waves

X RAY DIFFRACTION When a beam of x rays is sent through a crystal

If the wavelength of the light striking a crystal becomes structure, the x rays will reflect from the planes of
comparable to the spacing between atoms, we get a atoms within the crystal. The process, called Bragg
new effect. The scattered waves from adjacent atoms reflection, is illustrated for the example of a cubic
begin to interfere with each other and we get diffraction lattice in Figure (3). The dotted lines connect lines of
patterns. atoms, which are actually planes of atoms if you
consider the depth of the crystal. An incident wave
The spacing between atoms in a crystal is of the order
o
coming into the crystal can be reflected at various
of a few angstroms. (An angstrom, abbreviated A , is angles by various planes, with the angle of incidence
10 – 8cm . An angstrom is essentially the diameter of a equal to the angle of reflection in each case.
hydrogen atom.) Light with this wavelength is in the x
ray region. Using Einstein's formula E = hf = hc/λ , When the wavelength of the incident radiation is
but in the form comparable to the spacing between atoms, we get a
strong reflected beam when the reflected waves from
E (in eV ) = 12.4 × 10 eV⋅cm
–5
one plane of atoms are an integral number of wave-
λ in cm lengths behind the reflected waves from the plane
o above as illustrated in Figure (4). If it is an exact
we see that photons with a wavelength of A 2 have an
integral wavelength, then the reflected light from all the
energy
parallel planes will interfere constructively giving us
°
= 12.4 × 10 – 8eV⋅cm
photon with 2A –5 an intense reflected wave. If, instead, there is a slight
E wavelength 2 × 10 cm mismatch, then light from relatively distant planes will
cancel in pairs and we will not get constructive interfer-
= 6,200 eV (1) ence. The argument is similar to the one used to find the
This is a considerably greater energy than the 2 to 3 eV maxima in a diffraction grating.
of visible photon.
incident X rays

incident X rays

reflected X rays

reflected X rays

Planes of atoms

Figure 3 Figure 4
Planes of atoms act like mirrors reflecting X rays. When the incident X ray wavelength equals the spacing
between one of the sets of planes, the reflected waves
add up to produce a maxima.
 36-5

Thus with Bragg reflection you get an intense reflec- The main use of x ray diffraction has been to determine
tion only from planes of atoms, and only if the wave- the structure of crystals. From the location of the dots
length of the x ray is just right to produce the construc- in the x rays' diffraction photograph, and a knowledge
tive interference described above. As a result, if you of the wavelength of the x rays, you can figure out the
send an x ray beam through a crystal, you get diffrac- orientation of and spacing between the planes of atoms.
tion pattern consisting of a series of dots surrounding By using various wavelength x rays, striking the crystal
the central beam, like those seen in Figure (5). Figure at different angles, it is possible to decipher complex
(5a) is a sketch of the setup and (5b) the resulting crystal structures. Figure (6) is one of many x ray
diffraction pattern for x rays passing through a silver diffraction photographs taken by J. C. Kendrew of a
bromide crystal whose structure is shown in (5c). crystalline form of myoglobin. Kendrew used these x
ray diffraction pictures to determine the structure of the
crystal ys
cted ra myoglobin molecule shown in Figure (17-3). Kendrew
refle was awarded the 1962 Nobel prize in chemistry for this
incident X ray
film work.

a) An incident beam of X rays is diffracted

by the atoms of the crystal.

Figure 6
One of the X ray diffraction photographs used by
Kendrew to determine the structure of the
Myoglobin molecule.

b) X ray diffraction pattern produced by a silver

bromide crystal. (Photograph courtesy of R. W.
Christy.)

c) The silver
bromide crystal is
a cubic array with
alternating silver
and bromine
atoms.

Figure 5
X ray diffraction study of a silver bromide crystal. Figure 17-3
The Myoglobin molecule, whose structure
was determined by X ray diffraction studies.
36-6 Scattering of Waves

Diffraction by Thin Crystals Now suppose a beam of waves is impinging upon the
The diffraction of waves passing through relatively crystal as indicated in Figure (7b). The impinging
thin crystals can also be analyzed using the diffraction waves will scatter from the lines of atoms, producing
grating concepts discussed in Chapter 33. Suppose for an array of circular waves as shown.
example, we had a thin crystal consisting of a rectangu-
lar array of atoms as shown in Figure (7a). The edge Compare this with Figure (8), a sketch of waves
view of the array is shown in (7b). Here each dot emerging from a diffraction grating. The scattered
represents the end view of a line of atoms. waves from the lines of atoms, and the waves emerg-
ing from the narrow slits have a similar structure and
therefore should produce similar diffraction patterns.
lines of
atoms

edge
view

alternate
lines of
atoms
Figure 7a
Front view of a rectangular array
of atoms in a thin crystal.

Figure 9
lines of atoms Various lines of atoms can imitate
slits in a diffraction grating.

incident
wave

Figure 10a
scattered waves
Figure 7b A laser beam sent through a
Edge view with an incident wave. Each dot now single grating. The lines of the
represents one of the line of atoms in Figure (7a). grating were 25 microns wide,
spaced 150 microns apart.

diffraction grating

incident
wave

emerging waves
Figure 8
The waves emerging from a diffraction grating have a
similar structure as waves scattered by a line of atoms.
 36-7

There is one major difference between the array of Exercise 2

atoms in Figure (7) and the diffraction grating of Figure In Figure (10a) the maxima seen in the photograph are
(8). In the crystal structure there are numerous sets of 1.68 cm apart and the distance from the grating to the
lines of atoms, some of which are indicated in Figure screen is 4.00 meters. The wavelength of the laser
(9). Each of these sets of lines of atoms should act as beam is 6.3 × 10 – 5 cm. What is the spacing between the
an independent diffraction grating, producing its own slits of the diffraction grating?
diffraction pattern. The main sets of lines are horizon-
tal and vertical, thus the main diffraction pattern we Exercise 3
should see should look like that produced by two In Figure (11), a laser beam is sent through two crossed
diffraction gratings crossed at right angles. Sending a diffraction gratings of different spacing. Which image,
laser beam through two crossed diffraction gratings (a) or (b) is oriented correctly? (What happens to the
produces the image shown in Figure (10). In Figure spacing of the maxima when you make the grating lines
(10a), the laser beam is sent through a single grating. In closer together?)
(10b) we see the effect of adding another grating
crossed at right angles.
Figure 11
Two diffraction gratings with
different spacing are crossed. As
shown, the vertical lines are
farther apart than the horizontal
ones. Which of the two images of
the resulting diffraction pattern
has the correct orientation?

Figure 10b
A laser beam sent through
crossed diffraction gratings.
Again the lines of the grating
were 25 microns wide, spaced
150 microns apart. a)

b)
36-8 Scattering of Waves

THE ELECTRON The Graphite Crystal

DIFFRACTION EXPERIMENT Graphite makes an ideal substance to study by electron
One of the main differences between the scattering of scattering because graphite crystals come in thin sheets.
x rays and of electrons is that x ray photons interact less A graphite crystal consists of a series of planes of
strongly with atoms, with the result that x rays can carbon atoms. Within one plane the atoms have the
penetrate deeply into matter. This enables doctors to hexagonal structure shown in Figure (12), reminiscent
photograph through flesh to observe broken bones, or of the tiles often seen on bathroom floors. The spacingo

engineers to photograph through metal looking for between neighboring atoms in each hexagon is 1.42 A
hidden flaws. Electrons interact strongly with atoms, as indicated at the bottom of Figure (12).
do not penetrate nearly as deeply, and therefore are well The atoms within a plane are very tightly bound
suited for the study of the structure of surfaces or thin together. The hexagonal array forms a very strong
crystals where you get considerable scattering from a framework. The planes themselves are stacked ono top
few layers of atoms. of each other at the considerable distance of 3.63 A as
indicated in Figure (13). The forces between these
planes are weak, allowing the planes to easily slide over
each other. The result is that graphite is a slippery
substance, making an excellent dry lubricant. In con-
trast, the strength within a plane makes graphite an
excellent strengthening agent for epoxy. The resulting
carbon filament epoxies, used for constructing racing
boat hulls, light airplanes and stayless sailboat masts, is
one of the strongest plastics available.

plane
separation
o
= 3.63 A

Figure 13
d1

effective gratings Edge view of the graphite crystal, showing the

planes of atoms. The planes can easily slide over
o each other, making the substance slippery.
1.42 A
d1 3A
2.1
=

Figure 12
The hexagonal array of atoms in one layer of a
graphite crystal. Lines of atoms in this crystal
act as crossed diffraction gratings.
 36-9

The Electron Diffraction Tube Electron Wavelength

The electron diffraction experiment where we sent a The accelerating voltage required to produce a good
beam of electrons through a graphite crystal, can be diffraction pattern is in the range of 6,000 volts. As our
viewed either as an experiment to demonstrate the first step in the analysis, let us use the de Broglie
wave nature of electrons or as an experiment to study wavelength formula to calculate the wavelength of
the structure of a graphite crystal. Perhaps both. 6,000 eV electrons.

The apparatus, shown in Figure (14), consists of an The rest energy of an electron is .51 MeV, or 510,000
evacuated tube with an electron gun at one end, a eV, far greater than the 6,000 eV we are using in this
graphite target in the middle, and a phosphor screen at experiment. Since the 6,000 eV kinetic energy is much
the other end. A finely collimated electron beam can be less than the rest energy, we can use the nonrelativistic
aimed to strike an individual flake of graphite, produc- formula 1/2 mv 2 for kinetic energy. First converting
ing a single crystal diffraction pattern on the phosphor 6,000 eV to ergs, we can equate that to 1/2 mv 2 to
screen. Usually you hit more than one crystal and get calculate the speed v of the electron. We get
a multiple image on the screen, but with some adjust- ergs
ment you can usually obtain a single crystal image. 6000 eV × 1.6 × 10 – 12 = 1/2 m ev 2 (2)
eV
With the electron mass m e = .911 × 10 – 27gm , we get
phosphor
screen graphite crystal
2 × 6000 × 1.6 × 10 – 12ergs
v2 =
.911 × 10 – 27gm
electron gun 2
= 21.1 × 10 18 cm 2
sec
v = 4.59 × 10 9cm/sec (3)
electron beam which is slightly greater than 10% the speed of light.

18 cm The next step is to calculate the momentum of the

diffracted electron for use in de Broglie's formula. We have
electrons
p = mv
cm
= .911 × 10 – 27gm × 4.59 × 10 9 sec (4)
gm cm
= 4.18 × 10 – 18 sec
Finally using de Broglie's formula we have
6.63 × 10 – 27gm cm 2/sec
λ = hp =
4.18 × 10 – 18gm cm/sec
°
λ electron = 1.59 × 10 – 9cm = .159 A (5)
Thus the wavelength of the electrons we are using in
this experiment is about one tenth the spacing between
Figure 14
atoms in the hexagonal array.
Electron diffraction apparatus. An electron beam,
produced by an electron gun, strikes a graphite Exercise 4
crystal located near the center of the evacuated tube. Calculate the wavelength of a 6000 eV photon. What
The original beam and the scattered electrons strike would cause such a difference in the wavelengths of a
a phosphor screen located at the end of the tube. photon and an electron of the same energy?
36-10 Scattering of Waves

The Diffraction Pattern

What should we see when a beam of waves is diffracted
by the hexagonal array of atoms in a graphite crystal?
Looking back at the drawing of the graphite crystal,
Figure (12), we see that there are prominent sets of lines
of atoms in the hexagonal array. To make an effective
Figure 15a
Single grating diffraction grating, the lines of atoms have to be equally
diffraction spaced. We have marked three sets of equally-spaced
pattern. lines of atoms, each set being at an angle of 60° from each
other. We expect that these lines of atoms should
produce a diffraction pattern similar to three crossed
diffraction gratings.

In Figure (15), we are looking at the diffraction we get

when a laser beam is sent through three crossed diffrac-
tion gratings. In (15a), we have 1 diffraction grating. In
(15b) a second grating at an angle of 60° has been added.
In (15c) we have all three gratings, and see a hexagonal
Figure 15b array of dots surrounding the central beam, the central
Two grating maximum.
diffraction
pattern. Figure (16) is the electron diffraction pattern photo-
graphed from the face of the electron diffraction tube
shown in Figure (14). We clearly see an hexagonal array
of dots expected from our diffraction grating analysis.
On the photograph we have superimposed a centimeter
scale so that measurements may be made from this
photograph.

Figure 15c
Diffraction
pattern from
three crossed
gratings.

o
1.42 A
d1 3A
2.1
=

o
d1

effective gratings 0 1 2 3 4 cm

Figure 12 (section) Figure 16

Three sets of lines of atoms act as three crossed Diffraction pattern produced by a beam of electrons
diffraction gratings with 2.13 angstrom spacing. passing through a single graphite crystal. The energy
of the electrons was 6000 eV.
 36-11

The electron diffraction apparatus allows us to move Using the diffraction formula, Equation 33-3, and
the beam around, so that we can hit different parts of the noting that Ymax << D, we have
target. In Figure (16), we have essentially hit a single
crystal. When the electron beam strikes several graph- λ = Ymax d ≈ Ymax d
ite crystals at the same time, we get the more complex 22
D + Ymax D
pattern seen in Figure (17).
= 1.33 cm × 2.13 × 10 cm
–8
18 cm
Analysis of the Diffraction Pattern
Let us begin our analysis of the diffraction pattern by λ = 1.57 × 10 – 9cm (6)
selecting one set of dots in the pattern that would be
produced by one set of lines of atoms in the crystal. The
which agrees well with Equation 5, the calculation of
dots and the corresponding lines of atoms are shown in
the electron wavelength using the de Broglie wave-
Figure (18). In (18a) we see that the spacing Ymax
length formula.
between the dots on the screen is 1.33 cm. These
horizontal dots correspond to the maxima for a vertical
set of lines of atoms indicated in (18c). In (18b) we are
reminded that the distance from the target to the screen
is 18 cm. Using the diffraction grating formula, we can 0 1 2 3 4 cm
calculate the wavelength of the electron waves that Figure 18a
produce this set of maxima. The diffraction grating maxima from one set of lines in
the graphite crystal. You can see that 3y max = 4cm, so
that y max = 1.33 cm.

diffraction
grating
maxima
18 cm

diffracted
electrons
graphite
crystal

Figure 18b
Top view of the electron diffraction apparatus..

Figure 17
Diffraction pattern produced by a beam of electrons
passing through multiple graphite crystals.

d1 =
o

2.13 A
d1

Figure 18c
The vertical lines of atoms in the graphite crystal that
produce the horizontal row of dots seen in (a).
36-12 Scattering of Waves

Other Sets of Lines

With a careful analysis of the lines of atoms in the

d2
hexagonal ray of atoms, one can explain all the dots of
the diffraction pattern of Figure (16). For example, in
Figure (19) we see that there is another set of lines that
are rotated at an angle of 30° and more closely spaced
than our original set. In Figure (20), we have high-
lighted a set of dots in the diffraction pattern that are
rotated by an angle of 30° and more widely spaced than
d1

effective gratings effective gratings

set 1 set 2 the dots we have been analyzing. Since more closely
o
spaced lines in a grating produce more widely spaced
1.42 A
maxima, we should suspect that the highlighted maxima
d2 = o result from this new set of lines. The point of Exercise
1.22 A 5 is to see if this is true.
d1 3A
2.1
=

o Exercise 5
(a) Explain why more closely spaced atoms should
produce more widely spaced dots in the diffraction
Figure 19
pattern.
It is easy to find a second set of effective
gratings, rotated 30° from the first set, (b) Assuming that the dots highlighted in Figure (20) are
and with a narrower spacing. produced by the lines of atoms shown by dotted lines in
set 2 of the effective gratings, calculate the wavelength
of the waves producing the dots. Compare your results
Figure 20 with our previous analysis.
We have highlighted the
maxima produced by this
set of lines. Note that the Exercise 6
more narrowly spaced Suppose that a beam of neutrons rather than electrons
lines produce more were fired at the graphite crystal. Assuming that neu-
widely spaced maxima. trons also obey the de Broglie relationship λ = h/p , what
should be the kinetic energy, in eV, of the neutrons in
order to produce the same diffraction pattern with the
same spacing between dots?

0 1 2 3 4 cm
 36-13

Student Projects Several students doing project work with these gratings
The crossed diffraction gratings used to obtain the and dot patterns suspected that some patterns were not
various laser diffraction patterns in this chapter, were as good as they should be and took microscope photo-
created using the Adobe Illustrator program, and then graphs of them. They found that lines or dots as small
printed on film using a Linatronic imagesetter at a local as 10 microns wide tended to be filled in and blotchy,
desktop publishing company. The one micron resolu- but lines or dots 25 microns wide came out fairly well
tion of the imagesetter allowed us to construct various as can be seen in Figures (21) and (22). Figure (23) is
grating and dot patterns that produced reasonable the laser diffraction pattern produced by a laser beam
diffraction patterns with a laser. passing through the hexagonal dot pattern of Figure
(22).

Figure 21 Figure 22
Microphotograph of the three crossed diffraction Microphotograph of a hexagonal dot pattern. The dots
gratings. The lines are 25 microns wide and 100 are 25 microns in diameter and 100 microns apart.
microns apart, on centers. (Student project by Brady (Student project by Brady Beale and Amy Coughlin.)
Beale and Amy Coughlin.)

Figure 15c Figure 23

Diffraction pattern produced by a laser beam going Diffraction pattern produced by a laser beam going
through the three crossed gratings of Figure 21. through the hexagonal dot pattern of Figure 22.
36-14 Scattering of Waves

Student project by Gwendylin Chen lines of atoms diffraction grating

In our discussion of the diffraction of waves by the

atoms of a crystal, we pointed out that waves should
emerge from a line of atoms in much the same wah that
incident incident
they do from the slits of a diffraction grating. The two
wave wave
situations were illustrated in Figures (7b and 8) repro-
duced below.

That a slit and a line produce similar diffraction pat-

terns was clearly illustrated in a project by Gwendylin scattered waves emerging waves
Chen. While working with a laser, she observed that Figure 7b Figure 8
when the beam passed over a strand of hair it produced Edge view of a thin The waves emerging from
crystal with an incident a diffraction grating have
a single slit diffraction pattern superimposed on the
wave. Each dot now a structure similar to the
image of the beam itself. Here we have reproduced represents one of the line waves scattered by a line of
Gwendylin’s experiment. Figure (24) is a photograph of atoms in the crystal. atoms.
of a slit made from two scapel blades, and a strand of
Gwendylin’s hair. We tried to make the width of the slit slit formed by two scapels
the same as the width of the hair. The two circles
indicate where we aimed the laser for the two diffrac- laser through slit
tion patterns.

The results are seen in Figure (25).The diffraction

patterns are almost identical. The only difference is laser past hair
that when the beam passes over the hair, it continues on
landing in the center of the diffraction pattern.
strand of hair
Figure 24
Slit and hair used to produce diffractiopn patterns.
The circles indicate where we aimed the laser.

a) single slit diffraction pattern

b) diffraction pattern produced by strand of hair

Figure 25
incident laser beam
Comparason of diffeaction patterns.
Chapter 37
Lasers, a Model Atom
and Zero Point Energy
CHAPTER 37 LASERS, A MODEL ATOM
AND ZERO POINT ENERGY
“Once at the end of a colloquium I heard Debye saying introductory physics and chemistry texts, with the
something like: ‘Schrödinger, you are not working result that students must simply be shown the solutions
right now on very important problems...why don’t you without being told how to get them. We will have to do
tell us some time about that thesis of de Broglie, which the same in the next chapter.
seems to have attracted some attention?’ So in one of
the next colloquia, Schrödinger gave a beautifully clear In this chapter we will study a model atom, one in which
account of how de Broglie associated a wave with a we can see how the particle-wave nature of the electron
particle, and how he could obtain the quantization leads directly to quantized energy levels and atomic
rules ... by demanding that an integer number of waves spectra. The basic idea, which we illustrate with the
should be fitted along a stationary orbit. When he had model atom, is that whenever you have a wave confined
finished, to some region of space, there will be a set of allowed
ChapterDebye6 casually remarked that he thought this
way of talking was rather childish ... To deal properly standing wave patterns for that wave. Whether the
Mass
with waves, one had to have a wave equation.” patterns are complex or simple depends upon the way
the wave is confined. If the wave is also a particle, like
FELIX BLOCK, in an address to the American Physi- an electron or photon, you can then use the particle
cal Society in 1976. wave nature to calculate the energy of the particle in
each of the allowed standing wave patterns. These
Schrödinger took Debye’s advice, and in the following energy values are the quantized energy levels of the
months devised a wave equation for the electron wave, particle.
an equation from which one could calculate the elec-
tron energy levels. The structure of the hydrogen atom An example of a set of simple standing waves that are
was a prediction of the equation without arbitrary easily analyzed is found in the laser. It is essentially the
assumptions like those needed for the Bohr theory. The laser standing wave patterns that we use for our model
wave nature of the electron turned out to be the key to atom. For this reason we begin the chapter with a
the new mechanics that was to replace Newtonian discussion of the laser and how the photon standing
mechanics as the fundamental theory. waves are established. In the model atom the photon
standing waves of the laser are replaced by electron
In the next chapter we will take a look at some of the standing waves.
electron wave patterns determined by Schrödinger’s
equation, and see how these patterns, when combined An analysis of the model atom shows why any particle,
with the Pauli exclusion principle and the concept of when confined to some region of space, must have a
electron spin, begin to explain the chemical properties non zero kinetic energy. The smaller the region of
of atoms and the structure of the periodic table. space, the greater this so called zero point kinetic
energy. When these ideas are applied to the atoms in
The problem one encounters when discussing the liquid helium, we see why helium does not freeze even
application of Schrödinger’s equation to the hydrogen at absolute zero. We also see why the entropy definition
atom, is that relatively complex mathematical steps of temperature must be used at these low temperatures.
are required in order to obtain the solutions. These
steps are usually beyond the mathematical level of most
37-2 Lasers, a Model Atom and Zero Point Energy

THE LASER AND The purity of the light in a laser beam depends upon the
STANDING LIGHT WAVES standing light wave pattern created by the two mirrors,
and upon a quantum mechanical effect discovered by
The laser, the device that is at the heart of your CD
Einstein in 1915.
player and fiber optics communications, provides a
common example of a standing light wave. In most Einstein found that there were two distinct ways an
cases a laser consists of two parallel mirrors with excited atom could radiate light, either by spontaneous
standing light waves trapped between the mirrors as emission or stimulated emission. An example of
illustrated in Figure (1). The light comes from radiation spontaneous emission is when an excited atom is all by
emitted by excited atoms that are located within the itself and eventually drops down to a lower energy
standing wave. level. The emitted photon can come out in any direc-
tion and can be Doppler shifted.
How the light radiated by the excited atoms ends up in
a standing wave is a story in itself. An atom excited to If, however, a photon with the right energy passes by
a high energy level can drop down to a lower level by the excited atom, there is some chance that the atom
emitting a photon whose energy is the difference in will emit a photon exactly like the one passing by. This
energy of the two levels. This photon will have the is called stimulated emission. (The energy of the pass-
wavelength of the spectral line associated with those ing photon has to be close to the energy the atom would
two levels. naturally radiate.)
Spectral lines are not absolutely sharp. For example, It is the process of stimulated emission that can lead to
due to the Doppler effect, thermal motion slightly shifts a laser beam. Suppose we have a gas of excited atoms
the wavelength of the emitted radiation. If the atom is located between parallel mirrors. At first the atoms
moving toward you when it radiates, the wavelength is radiate spontaneously in all directions. (We assume
shifted slightly towards the blue. If moving away, the that there is some mechanism to excite the atoms).
shift is toward the red. In addition the photons are After a while one of the photons hits a mirror straight
radiated in all directions, and waves from different on and starts reflecting back and forth between the
photons have different phases. Even in a sharp spectral parallel mirrors. As the photon moves back and forth,
line the light is a jumble of directions and phases, giving it passes by an excited atom, stimulating that atom to
what is called incoherent light. emit an identical photon.
In contrast the light in a laser beam travels in one Now there are two identical photons bouncing back
direction, the phases of the waves are lined up and there and forth. Each is likely to stimulate another atom to
is almost no spread in photon energies. This is the beam emit an identical photon, and we have four identical
of coherent light which made it so easy for us to study photons, etc. Soon there are so many identical photons
interference effects like those we saw in the two slit and moving through the excited atoms that there is little
multiple slit diffraction patterns. These patterns would chance that an atom can radiate spontaneously. All the
be much more difficult to observe if we had to use radiation is stimulated and all the photons are identical
incoherent light. to the one that started bouncing back and forth between
the mirrors.
optically excited standing partially reflecting
flat mirror gas atoms light wave optically flat mirror
The mirrors on the ends of the laser are not perfect
reflectors, a few percent of the photons striking the
laser beam mirror pass through, forming the beam produced by the
laser. The photons lost to the laser beam are continually
Figure 1
Laser consisting of two parallel mirrors with replaced by new identical photons being emitted by
standing light waves trapped between the mirrors stimulated emission. One of the tricky technical parts
of constructing a laser is to maintain a continuous
supply of excited atoms. There are various ways of
doing this that we need not discuss here.
 37-3

Photon Standing Waves Because of the simple geometry, we do not need to

The photons bouncing back and forth between the solve a wave equation to determine the shape of these
mirrors in a laser are in an allowed standing wave standing light waves. The waves are sinusoidal, and the
pattern. Back in Chapter 15 in our discussion of allowed wavelength are given by the same formula as
standing waves on a guitar string, we saw that only for the allowed waves on a guitar string, namely
certain standing wave patterns were allowed, those wavelength of the
shown in Figure (15-15) reproduced here which had an λ n = 2D
n nth standing wave (1)
integral or half integral number of wavelengths be-
where D is the separation between the mirrors.
tween the ends of the string. For photons trapped
between two mirrors, the allowed standing wave pat-
terns are also those with an integral or half integral mirror mirror
optically flat surface
number of wavelengths between the mirrors, as indi- with reflective coating
λ
cated schematically in Figure (2).

bridge string nut

standing light wave

D

D
first
harmonic or λ 1 = 2D
fundamental 1
1
2 λ1

second λ 2 = 2D
harmonic 2
2
2 λ1 λ = 2D/1

third λ 3 = 2D
harmonic 3
3
2 λ1

fourth λ 4 = 2D
harmonic 4
4
2 λ1 λ = 2D/2

Figure 15-15 (reproduced)

On a guitar string only certain standing wave patterns
which have an integral or half intergral number of
wavelengths between the ends of the string are allowed.

λ = 2D/3
Figure 2
Three longest wavelength standing wave patterns
for a light beam trapped between two mirrors.
37-4 Lasers, a Model Atom and Zero Point Energy

Photon Energy Levels mirrors is D, then from Equation 1 ( λ n = 2D n ), and

The special feature of the standing light wave is that the Equation 2 ( En = hc λ n ), we find that the quantized
light has both a wave and a particle nature. Equation 1, values of E n are
which tells us the allowed wavelengths, is all we need
En = hc = hc
to know about the wave nature of the light. The particle λn 2D
nature is described by Einstein’s photoelectric effect n
formula E = hf = hc λ . Applying this formula to the
photons in the standing wave, we find that a photon En = n hc (3)
2D
with an allowed wavelength λ n has a corresponding
energy E n given by From Equation 3 we can construct an energy level
diagram for the photons trapped between the mirrors.
E n = hc (2) In contrast to the energy level diagram for the hydrogen
λn
atom, the photon energies start at zero because there is
Because only certain wavelengths λ n are allowed, no potential energy. We see that the levels are equally
only certain energy photons, those with an energy E n spaced, a distance hc/2D apart.
are allowed between the mirrors. We can say that the
photon energies are quantized. If the separation of the
Exercise 1
o
If you could have two mirrors 1A apart (the size of a
En = hc/λ n hydrogen atom) what would be the energy, in eV, of the
lowest 5 energy levels for a photon trapped between the
mirrors?
E4 = 4(hc/2D)

A MODEL ATOM
Now imagine that we replace the photons trapped
E3 = 3(hc/2D)
between two mirrors with an electron between parallel
walls located a distance D apart, as shown in Figure (4).
For this model, the allowed standing wave patterns are
again similar to the guitar string standing waves. The
E2 = 2(hc/2D) allowed electron wavelengths are

allowed wavelength
λ n = 2D
n of an electron trapped (1a)
between two walls

;; ;
E1 = 1(hc/2D) The difference between having a photon trapped be-
D tween mirrors and an electron between walls, is the
formula for the energy of the particle. If the energy of

;; ;
the electron is non relativistic, then the formula for its
E=0
D

;; ;
photon energy levels
electron
Figure 3
Energy level diagram for a photon
trapped between two mirrors. Figure 4
Electron trapped between two walls
 37-5

kinetic energy is 1/2 mv 2 , not the Einstein formula If the electron is in one of the higher levels and falls to
E = hc/λ that applies to photons. The difference arises a lower one, it will get rid of its energy by emitting a
because the electron has a rest mass while the photon photon whose energy is equal to the difference in the
does not. energy of the two levels. Thus the trapped electron
should emit a spectrum of radiation with sharp spectral
For the electron trapped between walls, there is no lines, where the lines correspond to energy jumps
electric potential energy like there was in the hydrogen between levels just as in the hydrogen atom. Thus the
atom. Thus we can take 1/2 mv 2 as the formula for the electron trapped between plates is effectively a model
electron’s total energy, ignoring the electron’s rest atom, complete with an energy level diagram and
mass energy as we usually do in non relativistic calcu- spectral lines.
lations.
2
E n = n E1
To relate the kinetic energy to the electron’s allowed
wavelength λ n , we use de Broglie’s formula p = h/λ .
The easy way to do this is to express the energy E 4 = 16E1
1/2 mv 2 in terms of the electron’s momentum p = mv.
We get

E = 1/2 mv 2 = 1 mv
2
2m
p2
E = (4)
2m
Next use the de Broglie formula p = h/λ to give us
2
h/λ n h2
En = = (5)
2m 2mλ 2n E 3 = 9E1
as the formula for the energy of an electron of wave-
length λ n .
Finally use Equation 1, λ n = 2D/n , for allowed elec-
tron wavelengths to get

En = h2
2
2m 2D
n
E 2 = 4E1

E n = n2 h2 (6)
8mD2 h2
E1 = 2
8mD
This is our equation for the energy levels of an electron D
trapped between two plates separated by a distance D. E=0
The corresponding energy level diagram is shown in
electron energy levels
Figure (5). The energy levels go up as n 2 instead of
being equally spaced as they were in the case of a Figure 5
photon trapped between two mirrors. Energy level diagram for an electron
trapped between two walls.
37-6 Lasers, a Model Atom and Zero Point Energy

Our model atom is not just a fantasy. With the Exercise 2

techniques used to fabricate microchips, it has been Assume that an electron is trapped between two walls
possible to construct tiny boxes, the order of a few a distance D apart. The distance D has been adjusted
angstroms across, and trap electrons inside. An elec- so that the lowest energy level is E1 = 0.375eV .
tron microscope photograph of these quantum dots as
they are called, is shown in Figure (6). The allowed (a) What is D?
standing wave patterns are reasonably well represented (b) What are the energies, in eV, of the photons in the
by the sine wave patterns of Figure (5), where D is the six longest wavelength spectral lines radiated by this
smallest dimension of the box. Thus we predict that system? Draw the energy level diagram for this system
electrons trapped in these boxes should have allowed and show the electron jumps corresponding to each
energies En close to those given by Equation (6), and spectral line.
emit discrete line spectra like an atom. This is precisely
(c) What are the corresponding wavelengths, in cm, of
what they do. (Some of the low energy jumps are
these six spectral lines?
shown in Figure 7.)
(d) Where in the electromagnetic spectrum (infra red,
In calculating with the model atom we have not fudged visible, or ultra violet) do each of these spectral lines lie?
in any way by modifying Newtonian mechanics or If any of these lines are visible, what color are they?
even picturing a wave chasing itself around in a circle. (Partial answer: the photon energies are 1.125, 1.875,
We see a spectrum resulting purely from a combination 2.625, 3.00, 3.375, and 4.125 eV)
of the particle nature and the wave nature of electrons
and photons, where the connection between the two Exercise 3
points of view is de Broglie’s formula p = h λ . Explain why an electron, confined in a box, cannot sit at
rest. This is an important result whose consequences
will be discussed next. Try to answer it now.

E n = n2 E1
E4 = 16E1

E3 = 9E1

E2 = 4E1

Figure 6
2
Grid of quantum dots. These cells are made on a E1 = h 2
silicon wafer with the same technology used in 8mD
E=0
making electronic chips. An electron trapped in
one of these cells has energy levels similar to some electron transitions
those of our model atom. (See Scientific
Figure 7
American, Jan. 1993, p118.)
When an electron falls from one energy level to
another, the energy of the photon it emits equals
the energy lost by the photon.
 37-7

ZERO POINT ENERGY Exercise 4

One of the immediate consequences of the particle- o
In liquid helium, the helium atoms are about 3A apart
wave nature of the electron is that a confined electron and the atoms have a mass essentially equal to 4 times
can never be at rest. The smaller the confinement, the the mass of a proton.
greater the kinetic energy the electron must have. This
follows from the fact that at least half a wavelength of (a) what is the zero point energy, in ergs, of helium
the electron’s wave must fit within the confining atoms in liquid helium?
region. If D is the length of the smallest dimension of
(b) at what temperature T is the helium atom's thermal
the confining region, then the electron’s wavelength
kinetic energy 3/2 kT equal to the zero point energy
cannot be greater than 2D. But the smaller D is, the calculated in part (a)? [Answer: (a) 9.1 × 10– 16ergs ,
shorter the electron’s wavelength is, the greater its (b) 4.42 kelvin.]
kinetic energy.
Helium is an especially interesting substance to study
The de Broglie wavelength formula λ = h/p applies
at low temperatures because it is the only substance that
not only to photons and electrons, but to any particle,
remains a liquid all the way down to absolute zero. The
even an entire atom. As a result, an atom confined to
only way you can freeze helium is to take it down to
a region of size D should have a wavelength no greater
very low temperatures, and then squeeze it at relatively
than λ 1 = 2D , and thus a minimum kinetic energy
high pressure.
E min = h2 (7)
8m atomD 2 In all other substances, at low enough temperatures the
atoms settle down to a solid array. To melt the solid,
where we simply replaced the electron's mass by the you have to add enough thermal energy to disrupt the
atom' mass in Equation 7. Equation 7 is somewhat molecular bonds that hold the atoms in a more or less
approximate if the atom is confined on all sides in a fixed array.
three dimensional box, but it is reasonably accurate if
D is the smallest dimension of the box. Why can't helium atoms be cooled to the point where
molecular forces dominate and the atoms form a solid
An atom in a solid or a liquid is an example of a particle array? Part of the answer is that the molecular forces
confined in a box. The atom is confined by its neigh- between helium atoms are very weak, the weakest
boring atoms as illustrated in Figure (8). We may think there is between any atoms. Consequently you have to
of its neighbors as forming a box of size D where D is go to very low temperatures before helium gas even
the average spacing between atoms. Thus atoms in becomes a liquid. At atmospheric pressure, helium
solids or liquids have a minimum kinetic energy given becomes a liquid at 4.5 kelvins. To turn liquid helium
by Equation 7, and the atoms must be in continual into a solid you should have to go to still lower
motion no matter how low the temperature! Cooling temperatures.
the solid cannot get rid of this so-called zero point
energy. From Exercise 4, you saw that, in one sense, you cannot
get helium to a lower temperature, at least as far as the
Figure 8 kinetic energy of the atoms is concerned. The zero
A helium atom in liquid point energy of the atoms is as big as the thermal energy
helium is confined by its our that the atoms would have at a few kelvin-- 4.4 kelvin
neighbors. As a result it atom
by our rough estimate in Exercise 4. As a result, cooling
has a zero point energy
the helium further cannot remove enough kinetic en-
like an electron
confined between walls. ergy to allow the helium liquid to freeze. Helium thus
remains a liquid all the way down to absolute zero.
neighboring atoms
37-8 Lasers, a Model Atom and Zero Point Energy

Definition of Temperature TWO DIMENSIONAL

This discussion raises interesting questions about the STANDING WAVES
very concept of temperature. Our initial experimental In our discussion of percussion instruments in Chapter
definition of temperature was the ideal gas thermom- 16, we saw that a drumhead has a set of allowed
eter, which, as we saw from the derivation of the ideal standing wave patterns or normal modes, in some ways
gas law, is based on the thermal kinetic energy of the like the standing waves or normal modes on a guitar
particles. The simple idea of absolute zero was the string. On a guitar string we have one dimensional
point where all the thermal kinetic energy was gone and waves, while the drumhead has the two dimensional
the atoms were at rest. Now we see that no matter how wave patterns. The six lowest frequency patterns are
much thermal kinetic energy we try to remove, zero shown in Figure (16-41) repeated here. We could
point or "quantum kinetic energy" remains. This is not excite and observe individual standing waves using the
a problem at ordinary temperatures, but it can signifi- apparatus shown in Figure (16-40) also shown again
cantly affect the behavior of matter at temperatures here.
close to absolute zero.
That we get the same kind of standing wave patterns on
At low temperatures, the ideal gas thermometer is not an atomic scale is seen in Figure (9), which is a recent
adequate, and a new definition of temperature is needed. tunneling microscope image of an electron standing
That new definition is provided by the efficiency of wave on the surface of a copper crystal. The standing
Carnot's heat engine. As we suggested in Chapter 17, wave, which is formed inside a corral of 48 iron atoms,
this gives us a definition of temperature based, not on has the same shape as one of the allowed standing
the kinetic energy of the molecules, but upon the degree waves on a drumhead. (This particular standing wave
of randomness or disorder. A system at absolute zero pattern is excited because the average wavelength in
is as perfectly ordered as it can be. If zero point energy the standing wave is closest to the wavelength of the
is required by the particle wave nature of the atoms, if conduction electrons at the surface of the copper.)
it cannot be removed, then the most organized, least
disordered state of the system must include this zero A colleague Geoff Nunes, who works with scanning
point energy. Helium can go to its most ordered state microscopes, describes the image: “The incredible
at absolute zero, retain its zero point energy, and remain power of today’s personal computers has been made
a liquid. possible by our ability to make smaller and smaller
transistors. The smallest transistor one could imagine
building would be made up of single atoms. In a
dramatic series of experiments at IBM, Don Eigler and
his co-workers have shown how to use a tunneling
microscope to move and arrange single atoms.”

Figure 16-41 (reproduced)

Standing waves on a drumhead.
 37-9

“ This picture (Figure 9) shows a ridge of 48 iron atoms a beautifully symmetric pattern. This pattern occurs
arranged in a circle on the surface of a copper crystal. often in the physical world. For example it is the shape
Electrons in the copper are reflected from these iron that the head of a drum forms when struck. You can
atoms much as the waves on the surface of a pond are easily observe a similar pattern by gently skidding the
reflected from anything at the surface: rocks, weeds, base of a Styrofoam cup full of coffee across the surface
the shoreline. Inside the ring, the electron waves form of a table.”

plywood frame
rubber
sheet

speaker

strobe
light

Figure 16-40 (reproduced)

Exciting and observing the standing waves on a drumhead.

Figure 9
Conduction electrons on the surface of a copper crystal, forming a standing wave
inside a corral of 48 iron atoms. The shape is the same as one of the symmetric
standing waves on a drumhead. (Photo credits: Crommie and Eigler/IBM.)
Chapter 38
Atoms

CHAPTER 38 ATOMS

The focus of this chapter will be on the allowed electron To determine electron waves in hydrogen, our only
standing wave patterns in hydrogen, as calculated by option is to rely on Schrödinger’s equation. The
Schrödinger’s wave equation. We will not work with resulting standing waves are three dimensional in
Schrödinger’s equation itself, which involves deriva- shape, and do not have sharp edges like the drumhead
tives in both space and time, and requires fairly ad- waves. The electron in hydrogen is confined by the
vanced mathematical techniques to handle. But this is electric force of the nucleus, in what physicist Jay
not a terrible loss, because the resulting wave patterns Orear called a “fuzzy walled box”. Even though the
are well known, and are all we need in order to walls are not rigid, the standing waves have precise
understand much of the structure and behavior of shapes.
atoms.
Although the standing wave patterns we will discuss
As we saw at the end of the last chapter, when we go to were calculated for the hydrogen atom, the general
two dimensions, the standing wave patterns become features of these patterns apply to the electrons in
more complex. For example, to find the drumhead larger atoms. We will find that when we include Pauli’s
standing wave patterns, we either had to do an experi- exclusion principle and the concept of electron spin, we
ment to observe the patterns, or solve a wave equation can begin to see how the electron wave patterns
to calculate them. determine the chemical properties of atoms and the
structure of the periodic table.
38-2 Atoms

SOLUTIONS OF SCHRÖDINGER’S Each of the allowed standing wave patterns in hydrogen

EQUATION FOR HYDROGEN has a distinct set of values of the quantum numbers n, ,
and m. Figure (1) shows six of the allowed patterns.
In the Bohr theory, the energy levels for hydrogen were
What we have drawn is the intensity of the wave pattern
determined by the assumption that the electron’s angu-
as it would be seen if we looked through the wave. When
lar momentum L was quantized in units of h. The the side and top view are different we show both to help
electron’s angular momentum was L1 = h in the low- visualize the three dimensional structure of the wave.
est energy level, L2 = 2h in the second level, etc.
Assuming circular orbits, and applying classical me- The pattern on the bottom row labeled by the quantum
chanics, this led to the set of energy levels En given by numbers (n = 1, = 0, m = 0) is a spherical ball with a
4 fuzzy edge. The radius of the ball is about equal to the
E1 = – e 2m = – 13.6 eV Bohr radius of .529 angstroms. Schrödinger’s equation
2h
allows us to calculate the energy of the electron in this
E1 pattern and the result is -13.6 eV, the same as the lowest
En = (1)
n2 energy state of the electron in the Bohr theory. This is the
which gave us the values E1 = –13.6 eV , standing wave pattern for an electron in the ground state
E2 = – 3.40 eV , etc., that explained the hydrogen – cool, transparent hydrogen.
spectra.
On the second row in Figure (1) there are four distinct
De Broglie’s contribution was to show that one could patterns, all with n = 2 but with different values of and
understand the reason for quantization of angular mo- m. Schrödinger’s equation predicts that the energy of
mentum by assuming that the electron had a wave an electron in hydrogen is given, in general, by the
nature, with the electron’s wavelength λ related to its formula
momentum p by the formula E 4
En = 21 ; E1 = – e 2m (3)
n 2h
λ = h
p de Broglie formula (2)
where n is the “n” quantum number we have been
The quantization of angular momentum came from the discussing. Since these are the same values we got from
picture that an integral number of wavelengths fit around the Bohr theory, Schrödinger’s equation predicts all the
one of the allowed circular orbits. energy levels needed to explain the entire spectrum of
light radiated by hydrogen. Because of Equation 3, it is
Following Debye’s suggestion (see the introduction to reasonable to call n the energy quantum number for
Chapter 37), Schrödinger developed a wave equation hydrogen.
with which he was able to solve for the allowed standing
wave patterns of the electron in a hydrogen atom. Doing The first big surprise from Schrödinger’s equation is that
this required no arbitrary assumptions like circular orbits we can have several standing wave patterns all represent-
or the quantization of angular momentum. The wave ing an electron with the same energy. In the n = 2 energy
patterns are simply solutions of the wave equation. level, there are four distinct patterns representing an
electron with the energy E 2 = –3.40eV. One of these
Schrödinger’s equation has a surprisingly large number patterns has quantum numbers =0,
of solutions for the allowed standing waves of an electron m = 0. The other three have quantum numbers = 1 , m
in hydrogen. These waves are characterized by three = 1, m = 0, and m = –1.
numbers commonly given the names “n”, “ ”, and “m”.
It turns out that for the wave to be an acceptable solution, When we get up to the third energy level, n = 3, there are
a solution that does not have an infinite value at some nine patterns all with an energy of -1.51 eV. There is one
point, the numbers n, , and m have to have integer pattern with = 0 , m = 0; three patterns with = 1 , m =
values. These integer numbers have become known as (1, 0, –1); and five patterns with =2,
quantum numbers. m = (2, 1, 0, –1, –2). As we go up in energy, we get an
ever increasing number of patterns. The general rule is
that can range from zero up to n –1, and the m values
can range from + to – .
 38-3

There are 8 more n = 3 patterns

E = –1.51eV

in addition to the one shown.

The and m quantum numbers
are
= 1; m = 1, 0, –1
= 2; m = 2, 1, 0, –1, –2.

(i)
n = 3, = 0, m = 0
E = –3.40eV

top view (c) top view (e) top view (g)

z z z

(b)
n = 2, = 0, m = 0

side view (d) side view (f ) side view (h)

n = 2, = 1, m = 1 n = 2, = 1, m = 0 n = 2, = 1, m = –1
E = –13.6eV

(a)
n = 1, = 0, m = 0

Figure 1
The lowest energy standing wave patterns in hydrogen. The intensity
is what you would see looking through the wave.
38-4 Atoms

Exercise 1 The = 0 Patterns

a) For the n = 4 energy level, where E4 = –.85eV, there For each energy level n, there is one wave pattern with
are 16 allowed standing wave patterns. What are the = 0 . We have shown the first three = 0 patterns in
values of the quantum numbers and m for these Figure (1). All = 0 wave patterns are spherically
patterns? symmetric. The n = 1, = 0 pattern, for the ground
b) How many allowed patterns are there, what are the
state electron, is a fuzzy spherical ball with a diameter
values of and m, and what is the energy, for the n = 5 of about one angstrom. The n = 2, = 0 pattern is a
standing wave patterns? spherical ball surrounded by a spherical shell. Between
the ball and the shell, at a radius r = 1.06 angstroms, the
wave has value of zero. We can call this a spherical
node.

In the n = 3, = 0 pattern we have a spherical ball

surrounded by two spherical shells. There are now two
nodes
third
spherical nodes, the inner one located at r = 1.00
harmonic angstroms and the outer one at r = 3.75 angstroms. As
we go up higher in energy, we get one more spherical
node for each step up in energy level. In Figure (2), we
n=3 =0
compare the three lowest energy = 0 wave patterns
with the three lowest frequency standing wave patterns
on a guitar string. While the patterns look quite
node different, both have the feature that as we go up one step
second
harmonic
in energy or frequency, we get one more node in the
wave pattern. The first harmonic (n = 1) has no nodes
between the ends of the string. The second harmonic
n=2 =0 (n = 2) has one node, while the third harmonic (n = 3)
has two nodes, etc.

One of the key features of angular momentum is that it

first represents a rotation about an axis. In Newtonian
harmonic or
fundamental mechanics, we defined the direction of the angular
momentum vector L as being the direction of the axis
n=1 =0
of rotation. Because all the hydrogen wave patterns
with = 0 are spherically symmetric, they have no
Figure 2 preferred axis about which the electron could have
Comparison of the L = 0 electron standing
angular momentum.
wave patterns with the guitar standing
waves. Each step up in energy level or
harmonic introduces one more node.
 38-5

The ≠ 0 Patterns Later in this chapter we will see that the fact that = 0
patterns have a maximum at the nucleus while the ≠ 0
In all the ≠ 0 patterns, like the three n = 1, = 1 patterns
shown in the middle row of Figure (1), there is a special
patterns have a node there, plays an important role in
axis about which the electron can have angular momen- the electron structure and chemical properties of atoms.
tum. This suggests that the quantum number is related Quantized Projections
to the electron’s angular momentum, and that when
of Angular Momentum
= 0 , the electron has no angular momentum. This is
different from the Bohr picture where the electron’s A clue to understanding the ≠ 0 wave patterns can be
angular momentum started with one unit L = h in the obtained from a more detailed look at the two doughnut
lowest energy level, two units L = 2h in the second level, shaped patterns in Figure (1), the patterns labeled by the
etc. The Bohr theory did not allow for zero orbital angular quantum numbers n = 2, = 1 , m = +1 and m = –1.
momentum orbits while Schrödinger’s equation tells us While the m = +1 and m = –1 patterns look the same,
that there is a zero angular momentum wave pattern in a more detailed calculation with the Schrödinger equa-
each energy level. tion shows in the m = +1 pattern the electron is traveling
around the doughnut in a counterclockwise direction,
Intensity at the Origin while in the m = –1 pattern the electron is traveling
Another general feature of the hydrogen wave patterns is clockwise.
that all = 0 patterns have a maximum intensity at the
origin, at the nucleus, while all the ≠ 0 patterns have a These two patterns have an axis of symmetry which we
node there. The node at the origin for ≠ 0 patterns has have labeled the z axis, that passes up through the center
a simple classical explanation. The classical formula for of the doughnut. (These axes are shown as white dotted
angular momentum is the linear momentum p times the lines in the side views of these patterns, Figures 1d and
lever arm r⊥ . In order for the electron to have non zero 1h.) Further calculation with Schrödinger’s equation
angular momentum about the nucleus, it must have a non shows that the electron in the = 1 , m = 1 pattern
zero lever arm r⊥ and therefore cannot be at the nucleus. (counterclockwise motion), the electron has a z compo-
(One has to be careful applying Newtonian arguments to nent of angular momentum precisely equal to one unit
atomic phenomena. In the next chapter we will see a h.
similar argument fail when we discuss electron spin). for the = 1
Lz = h (4a)
m = 1 pattern

For the clockwise motion, the = 1 , m = –1 pattern, the

z component of angular momentum is minus one unit
h
for the = 1
Lz = –h (4b)
m = –1 pattern

The pattern in between, the one that looks like two

top view (c) top view (g) tennis balls, one on top of the other, described by the
quantum numbers = 1 , m = 0 turns out to have no
z z
angular momentum in the z direction.
for the = 1
Lz = 0 (4c)
m = 0 pattern
We see that the “m” quantum number tells us how
many units of angular momentum the electron has in
the z direction.
side view (d) side view (h)
n = 2, = 1, m = 1 n = 2, = 1, m = –1
Figures 1c,d,g, and h repeated
38-6 Atoms

There is somewhat of an analogy between the three over, you can change the z component Lz from +L
n = 2, = 1 patterns in Figure (1) and the bicycle wheel when it is pointing up to any value down to –L when the
demonstration we discussed in Chapter 7 (Figures 7-15 wheel is pointing down. Any value between +L and –
and 7-16). In Figure (3) we compare the m = 1 pattern L is allowed.
with a bicycle wheel whose angular momentum L
points in the + z direction, the m = –1 pattern with a For the hydrogen atom, an electron in the second
bicycle wheel whose angular momentum points in the energy level has only three = 1 wave patterns, only
– z direction, and the m = 0 pattern with a bicycle wheel three distinct z projections of angular momentum, each
that has no component of angular momentum in the z differing by one unit of angular momentum h . There
direction. is no wave pattern for Lz equal to some fractional value
of h—the projections of angular momentum are
The analogy shown in Figure (3) actually demonstrates quantized! There is absolutely nothing in Newtonian
how different angular momentum is on an atomic scale mechanics that prepares us for understanding how
from what we are familiar with on a human scale. The projections of angular momentum can be quantized. It
most striking difference is that you can point a bicycle is strictly a consequence of the wave nature of the
wheel in any direction you want. By turning the wheel electron, and the fact that a confined wave has only
certain allowed standing wave patterns.

Figure 3 n = 2, = 1, m = 1 n = 2, = 1, m = 0 n = 2, = 1, m = –1
There are three top top top
standing wave
patterns for a second
energy level, unit
angular momentum
electron.
Schrödinger’s
equation tells us that
the z axis projection
of angular z L z
momentum in the
three patterns are 1
unit, 0 units, and –1 z
units . There are no L
intermediate values,
because there are no
other wave patterns.
In comparison, a L
bicycle wheel has not
only the three
projections of side z side z side z
angular momentum
shown, but also
many intermediate
values.
 38-7

The Angular Momentum There was no guarantee that angular momentum had to
Quantum Number behave on an atomic scale, just the way we expected it
In Figure (3) we showed the different orientations of a to from our experience with large scale phenomena.
bicycle wheel with a total angular momentum of All we need to do is understand the transition from large
magnitude L. The z component of the bicycle wheel’s to small scale phenomena. In the case of angular
angular momentum ranges from +L when the axis is momentum, we can picture the bicycle wheel as having
pointing up to –L when the axis is pointing down. For a huge angular momentum quantum number . As a
the hydrogen electron in Figure (3), = 1 for all the result there are a huge number of allowed projections,
patterns, and the z projection of the electron’s angular with m ranging from + to – , which allows us to
momentum ranges from +1 unit for the m = 1 pattern rotate the bicycle wheel axis in an apparently continu-
down to –1 unit for the m = –1 pattern. This suggests ous fashion. And there is essentially no difference
that the quantum number represents the total angular between and + 1 , thus the maximum z projec-
momentum of the electron while m represents the tion of angular momentum essentially equals the total
allowed z projections. angular momentum L.

This interpretation is almost right. The quantum

number is related to the electron’s total angular mo- Other notation
mentum, but the value of the total angular momentum Further notation that some readers may have encoun-
is not quite equal to units of angular momentum as tered, are the names (s waves) for the = 0 patterns,
one might expect. Solving Schrödinger’s equation for (p waves) for = 1 patterns and (d waves) for = 2
the magnitude L of the electron’s orbital angular mo- patterns. These names, which are fairly common, have
mentum about the proton gives the result a rather obscure historical origin. Using this notation,
total angular one can, for example, refer to an electron in an n = 3,
L = +1 h momentum of = 2 pattern as a 3d wave. The ground state of
the electron (5) hydrogen is a 1s wave.
For large values of , the difference between and
+ 1 is slight and the z projections of angular Exercise 2
momentum can range essentially from + h to – h as Using the s,p,d notation, what would we call the waves
one would expect from the bicycle wheel analogy. But shown in Figure (3) ?
for small (non zero) values of , the total angular
momentum is significantly larger than the maximum z
projection. For = 1 , the maximum z projection is h ,
while the total angular momentum is
L= 1 1+1 h= 2 h.
38-8 Atoms

An Expanded Energy Level Diagram Another advantage of the energy level diagram of
In our discussion of the Bohr theory, we drew an energy Figure (4) is related to the fact that when an electron in
level diagram so that we could study transitions from an atom radiates a photon, the electron’s value almost
one level to another in order to predict the energy of the always changes by one unit. This is because a photon
photons the atom could emit. The diagram, like the one carries out angular momentum, and to conserve angu-
in Figure (35-3), is quite simple with one line for each lar momentum, the electron’s angular momentum has
energy level. to change. The common transitions represent not only
steps up and down, but one step sideways.
With the Schrödinger equation we discover that there
are numerous standing wave patterns for each energy (It is not impossible for an electron to emit a photon and
level. The simple energy level diagram of Figure (35- not change its angular momentum , it just a much less
3) does not give a hint of the multiple wave patterns. likely event. We only see such ∆ = 0 transitions,
Only the energy quantum number n is shown, there called “forbidden transitions”, when the electron has
being no indication of the and m quantum numbers no where to jump and change its value by one unit.
(which were unknown when Bohr developed his theory). For example, if the electron, for some reason, ends up
in the n = 2, = 0 state, the only lower energy state is
It is traditional (and convenient) to expand the energy the n = 1, = 0 state. The electron cannot fall there and
level diagram as we have done in Figure (4), to distin- change by one unit. As a result the electron hangs up
guish not only the energy quantum numbers n, but also in the n = 2, = 0 state for a much longer time than it
the angular momentum quantum numbers . We might would if a ∆ = 1 transition were available.)
be tempted to expand the diagram further and
display the separate projections m, but this
would make the diagram too complex. (In =0 =1 =2 =3
Figure (4) we indicated the z projections by
including some sketches of the lower energy n = 5
n=4
wave patterns. Such sketches are not usually
n=3 E3 = –1.51eV
included in energy level diagrams.) 3 patterns 5 patterns

One advantage of the expanded energy level

diagram is that it illustrates graphically that the n = 2 E2 = –3.40eV
maximum value of goes up only to n–1. It
shows that there is one = 0 pattern for n = 1, an
= 0 and an = 1 pattern for n = 2, etc. When m =1
you look at this diagram, you have to remember m =0
m = –1
that for each line, the z projections m can range
from m = + down to m = – in unit steps.

Figure 4
An expanded hydrogen energy
level diagram, including some
sketches of the lower energy
standing wave patterns..

n=1 E1 = –13.6eV
 38-9

MULTI ELECTRON ATOMS This is not a good prediction. Experimentally, the amounts
Straightforward techniques can be used to solve of energy needed to remove electrons from lithium one at
Schrödinger’s equation for one electron atoms like hydro- a time are 5.39 eV, 75.26 eV and 121.8 eV. While two
gen. To deal with a two electron atom like helium, we of lithium’s electrons are tightly bound, one is very
have to take into account not only the attraction between loosely bound, requiring less than half the energy to
the electrons and the nucleus, but also the repulsion remove than the hydrogen electron. A possible explana-
between electrons. This makes Schrödinger’s equation tion for the loose binding of lithium’s third electron is that,
more difficult to solve. One has to either use approxima- for some reason, that electron did not fall down to the
tion techniques or a computer. lowest energy E 1 type of standing wave pattern. It
appears to be hung up in the much higher energy, less
However, for all atoms, there are certain properties that tightly bound E 2 type of standing wave, one of the four
can be understood in terms of the general structure of the E 2 patterns seen in Figure (1).
hydrogen standing wave patterns, rather than from de-
tailed calculations. We can learn enough from these
Pauli Exclusion Principle
general properties to begin to see why atoms behave as
But why couldn’t the third lithium electron fall down to
they do in chemical reactions.
the low energy E 1 pattern? In 1925, two separate ideas
To study multi electron atoms, imagine that we start with provided the explanation. Wolfgang Pauli proposed that
hydrogen and add electrons one at a time (also increasing no two electrons were allowed to be in exactly the same
the number of protons and neutrons in the nucleus to keep state. This is known as the Pauli exclusion principle. But
the atom electrically neutral and the nucleus stable). We the exclusion principle seems to go too far, because in
will assume that as we add each electron, it falls down to helium, both electrons are in the same E 1 , = 0 standing
the lowest energy wave pattern available. wave pattern. If you cannot have two electrons in exactly
the same state in an atom, then something must be
If we start with a nucleus with one proton, and drop in one different about the two electrons in helium.
electron, the electron eventually falls down to the E 1 ,
= 0 standing wave pattern shown in Figure (1a).
Electron Spin
Add a proton to form a helium nucleus, drop in another To explain what the difference between the two electrons
electron, and we can expect the electron to also fall down might be, two graduate students, Samuel Goudsmit and
to the lowest energy E 1 standing wave pattern. The extra George Uhlenbeck, proposed that the electron was like a
Coulomb attractive force of the two protons in the nucleus spinning top with its own internal angular momentum.
strengthens the binding of the electrons, but the repulsive This became known as spin angular momentum. The
force between the two electrons weakens it. Experimen- special feature of the electron’s spin is that it has two
tally, it takes 24.6 eV to remove an electron from helium, allowed projections, which we call spin up and spin
while only 13.6 eV are needed for hydrogen. Thus the down. In helium you could have two electrons in the same
electrons are more tightly bound in helium, and we see E 1 wave pattern if they had different spin projections, for
that the extra Coulomb attraction to the nucleus is more then they would not be in identical states.
important than the repulsion between electrons.
Because the electron spin has only two allowed projec-
Using helium as a guide, we should expect that when we tions, we cannot add a third electron to the E 1 wave
go to lithium with 3 protons in the nucleus, the increased pattern. Lithium’s third electron must stop at one of the
Coulomb attraction to the nucleus should cause lithium’s higher energy E 2 standing wave patterns. Its energy is
three electrons to be even more tightly bound than helium’s much less negative and therefore this electron is much less
two. This would lead us to predict that it takes even more tightly bound than the first two electrons that went down
than 24.6 eV to pull one of the electrons out of lithium. to the E 1 wave pattern.
38-10 Atoms

THE PERIODIC TABLE A closer look also uncovers some exceptions to the rule
As we go to larger atoms, adding electrons one at a time, that the lower energy levels fill first. The most notable
the E 2 standing wave patterns begin to fill up. Since exception is at potassium, where the E 4 patterns with
there are four E 2 patterns, each with two allowed spin = 0 begin to fill before the E 3 patterns with = 2 .
states, up to 8 electrons can fit there. When the E 2 To understand why the binding energy gradually in-
patterns are full, when we get to the element neon with creases as an energy level fill up, and why the E 3 , = 2
two E 1 electrons and eight E 2 electrons, we have an patterns fill up late, we have to take a closer look at the
inert noble gas that is chemically similar to helium. structure of the electron wave patterns and see how this
Adding one more electron by going to sodium, the structure affects the binding energy. To do this it is
eleventh electron has to go up into the E 3 energy level useful to introduce the concepts of electron screening
since both the E 1 and E 2 patterns are full. This and effective nuclear charge.
eleventh electron in sodium is loosely bound like the
third electron in lithium, with the result that both
lithium and sodium have similar chemical properties. Electron Screening
They are both strongly reactive metals. In our discussion of the binding energy of the two
electrons in helium, we pointed out that there was a
Table 1 shows the electron structure and the binding competition between the increased Coulomb attractive
energy of the last electron for the first 36 elements in the force to the nucleus and the repulsion between the
periodic table. The general features of this table are that electrons. We could see that the increased attraction
the lowest energy levels fill up first, and there is a large was more important because helium’s two electrons
drop in binding energy when we start filling a new are each more tightly bound to the nucleus than
energy level. We see these drops when we go from the hydrogen’s one. It requires 24.5 eV to remove an
inert gases helium, neon, and argon to the reactive electron from helium and only 13.6 eV from hydrogen.
metals lithium, sodium and potassium. We can see that
this sudden change in binding energy leads to a signifi- The following argument provides an explanation of
cant change in the chemical properties of the atom. this increased binding of helium’s electron. Since the
two electrons are in the same E 1 wave pattern, half the
A closer look at Table 1 shows that there is a relatively time a given electron is closer to the nucleus than its
steady uniform increase in the electron binding as the partner and feels the full force of the nuclear charge
energy level fills up. The binding energy goes from +2e. But half the time it is farther away, and the net
5.39 eV for lithium in fairly equal steps up to 21.56 eV charge attracting it toward the nucleus is +2e reduced
for neon as the E 2 energy level fills. The pattern is by the other electron’s charge –1e for a total +1e. Thus,
more or less repeated as we go from 5.14 eV for sodium on the average the electron sees an effective charge of
up to 15.76 eV for argon while filling the E 3 energy approximately 1.5 e. This is greater than the charge +1e
level. It repeats again in going from 4.34 eV for seen by the single electron in hydrogen, and thus results
potassium up to the 14.00 eV for krypton. in a stronger binding energy. What we have done is to
account for the repulsion of the other electron by saying
that the other electron screens the nucleus, reducing the
nuclear charge from +2e to an effective value of
approximately 1.5e.
 38-11

Binding
energy
of last E1 E2 E3 E4 Energy level En
electron Angular momentum
Z Element in eV 0 0 1 0 1 2 0 1 quantum number
1 H Hydrogen 13.60 1
2 He Helium 24.58 2

3 Li Lithium 5.39 1
4 Be Beryllium 9.32 2

Helium core
5 B Boron 8.30 2 1
6 C Carbon 11.26 2 2
7 N Nitrogen 14.54 2 3
8 O Oxygen 13.61 2 4
9 F Fluorine 17.42 2 5
10 Ne Neon 21.56 2 6

11 Na Sodium 5.14 1
12 Mg Magnesium 7.64 2
13 Al Aluminum 5.98 2 1
14 Si Silicon 8.15 2 2
15 P Phosphorus 10.55 Neon core 2 3
16 S Sulfur 10.36 2 4
17 Cl Chlorine 13.01 2 5
18 A Argon 15.76 2 6

19 K Potassium 4.34 1
20 Ca Calcium 6.11 2
21 Sc Scandium 6.56 1 2
22 Ti Titanium 6.83 2 2
23 V Vanadium 6.74 3 2
24 Cr Chromium 6.76 5 1
25 Mn Manganese 7.43 5 2
26 Fe Iron 7.90 6 2
27 Co Cobalt 7.86 7 2
28 Ni Nickel 7.63 Argon core 8 2
29 Cu Copper 7.72 10 1
30 Zn Zinc 9.39 10 2
31 Ga Gallium 6.00 10 2 1
32 Ge Germanium 7.88 10 2 2
33 As Arsenic 9.81 10 2 3
34 Se Selenium 9.75 10 2 4
35 Br Bromide 11.84 10 2 5
36 Kr Krypton 14.00 10 2 6

Table 1
Electron binding energies. Adapted from Charlotte E Moore,
Atomic Energy Levels, Vol II, National Bureau of Standards
Circular 467, Washington, D.C.,1952.
38-12 Atoms

Effective Nuclear Charge This estimate of 30.6 eV is about 25% too high since the
To see what effect changing the nuclear charge has on experimental value is only 24.6 eV. We can take this
the binding energy, we can go back to our Bohr theory to imply that our estimate of z eff = 1.5 e for helium
calculations for different one electron atoms. In Exer- was a bit too crude. Our simple arguments about
cises 9 and 10 of Chapter 35 we found that the ground screening are not a substitute for an accurate calcula-
state energy of a single electron in an atom where the tion using Schrödinger’s equation.
nucleus had z protons was
What we can do, however, is to turn our approach
ground state
around and use the experimental values of the binding
E 1 = z 2 × –13.6 eV energy in a single (6)
electron atom energy to calculate an effective nuclear charge z eff .
Doing this for helium gives
While Equation 6 was derived from the Bohr theory, it
gave results in excellent agreement with experiment as
E1 neutral
helium
= z2eff × –13.6 eV
one can easily see from working Exercise 3.
– 24.6 eV = z2eff × –13.6 eV
Exercise 3
Table 2 lists the binding energy for the last electron for
zeff = 24.6 = 1.34 (8)
the elements hydrogen through boron. This is the 13.6
binding energy when all the other electrons have al-
The value of 1.34 is not too far off our original guess of
ready been removed. For each element, check the
prediction that the binding energy is given by
1.5. The result tells us that the electron screening is a
Equation 6. bit more effective than we had predicted.
Binding energy
z Element of last electron Lithium
1 Hydrogen 13.6 eV We will now see that various features of the periodic
2 Helium 54.14 eV table begin to make sense when viewed in terms of
3 Lithium 121.8 eV electron screening and the structure of the electron
4 Beryllium 216.6 eV wave patterns. Let us start off with lithium where the
5 Boron 338.5 eV last electron is in the E 2 , = 0 pattern and has a
Table 2
binding energy of 5.39 eV. Since this electron is in an
E 2 energy level, our formula for z eff should be
Equation 6 suggests that if an electron in a multi electron
atom sees an effective nuclear charge z eff e , the E 1 lithium = z 2eff × –3.40 eV
electron binding energy should be approximately z2eff
times the energy the electron would have in the same – 5.39 eV = z 2eff × –3.40 eV
energy level in hydrogen. Trying out this idea on
helium, where we estimated z eff to be about 1.5e, we get z eff = 5.39 = 1.26 (9)
3.40
E1 neutral
helium
= z2eff × –13.6 eV where we used - 3.40 eV rather than - 13.6 eV because
we are discussing an E 2 electron.
2
= 1.5 × –13.6 eV

= – 30.6 eV
estimated
= binding energy (7)
of helium
 38-13

In a nucleus with 3 protons, why does the E 2 electron At boron, both the E 2 , = 0 wave patterns are already
only see an effective charge of 1.26e? The answer lies full and the electron has to go into one of the E 2 , = 1
in the shape of the E 1 and E 2 wave patterns. The first patterns. All the electron standing wave patterns with a
two electrons in lithium are in the E 1 pattern of Figure non zero amount of angular momentum have a node at
(1a) reproduced below. It consists of a small spherical the origin. The more the angular momentum, the more
ball centered on the nucleus. The third electron, the one spread out the node and the farther the electron is kept
whose binding energy we are discussing, is in the E 2 , away from the nucleus. An electron in an ≠ 0 wave
= 0 pattern of Figure (1b). This pattern consists of a pattern will thus be effectively screened by electrons in
larger spherical ball surrounded by a spherical shell. = 0 wave patterns where the electron spends a lot of
The electron in this pattern spends a considerable time right down at the nucleus. Thus we expect that
amount of the time outside the smaller spherical ball of electrons in ≠ 0 patterns to be less tightly bound than
the two E 1 electrons. Thus much of the time the third those in the = 0 pattern of the same energy level. This
electron sees only an effective nuclear charge of about shows up with the drop in binding energy in going from
(1.0 e). Some of the time, however, the third electron beryllium to boron.
is also down at the nucleus feeling the full nuclear
charge of (3e). That the average nuclear charge seen by Up to Neon
the third electron is (1.26e) is not too difficult to believe. For all atoms beyond helium, there is a core consisting
of the nucleus and the two tightly bound E 1 electrons.
As the charge on the nucleus increases, the size of the
E 1 patterns shrink, and are penetrated less and less by
the outer electrons. We can think of this helium core as
acting as the effective nucleus for the larger atoms.

As we go from boron to neon, the charge on the helium

core increases from 3e to 8e as the E 2 , = 1 patterns fill
Figure 1a Figure 1b
up. This increase in the charge of the core causes a more
or less gradual increase in the binding energy, from 8.30
E1 , =0 E2 , =0
eV up to 21.56 eV. The one exception is the slight drop
in binding energy as we go from nitrogen to oxygen.
Beryllium The arguments we have made so far are not detailed
When we went from one E 1 electron in hydrogen to enough to explain this drop.
two E 1 electrons in helium, the binding energy about
doubled, from 13.6 eV to 24.6 eV. In going from one
Sodium to Argon
E 2 electron in lithium to two E 2 electrons in beryl-
lium, the binding energy increases from 5.39 eV to 9.32 We get the expected large drop in binding energy as we
eV. Again the electron binding energy almost doubled go from neon to sodium and start filling the E 3 patterns.
as we went from one to two electrons in the same The E 3 , = 0 patterns are full at magnesium and we get
energy level. a small drop in binding energy as the non zero angular
momentum patterns E 3 , = 1 start to fill at aluminum.
Again the angular momentum keeps the electrons away
Boron from the nucleus and increases the screening. As the
When we go from beryllium to boron, we add a third E 3 , = 1 patterns fill up, they are building a structure
electron to the E 2 energy level. From our experience on the ever shrinking neon core. The charge on the neon
with beryllium, we expect another significant increase core increases from (3e) at aluminum to (8e) at argon,
in binding energy, up to perhaps 13 eV or 14 eV. But again causing a gradual buildup of the electron binding
instead the binding energy drops from 9.32 eV down to energy from 5.98 eV to 15.76 eV. There is even the
8.30 eV. Something broke the pattern and caused this slight glitch going from phosphorous to sulfur that
drop. mirrors the glitch from nitrogen to oxygen.
38-14 Atoms

Potassium to Krypton Summary

The first major break in the pattern of filling the lower At this point it should be clear that the structure of the
energy levels first occurs at potassium. At potassium periodic table of the elements arises from the allowed
the E 3 , = 2 levels remain unfilled while the last electron standing wave patterns. Because of the exclu-
electron goes into the higher energy level E 4 , = 0 sion principle, no two electrons can be in the same state.
pattern. At this point the screening due to angular But because electron spin has two allowed states, up to
momentum has become more important than the en- two electrons can fit into each standing wave pattern.
ergy level. The = 2 patterns have such a big fat node
at the nucleus that an = 2 electron cannot get near the In general, as we go to atoms with more electrons, the
nucleus to feel the now large nuclear charge. Even lowest energy patterns fill up first, and there is a
though an E 4 , = 0 electron is in a higher energy level, significant change in chemical properties when a new
its wave pattern has a non zero value right down at the energy level begins to fill. But the angular momentum
nucleus. Some of the time this electron feels the full of the wave pattern also plays a significant role. The
charge of (19e) for potassium, and this increases the = 0 patterns can penetrate down to the nucleus,
binding beyond that of the E 3 , = 2 patterns. where the electron feels the full strength of the nuclear
charge. The ≠ 0 patterns have a node at the nucleus,
At calcium, the E 4 , = 0 pattern is full, and now the and the full nuclear charge is screened by = 0 elec-
five E 3 , = 2 , m = +2, +1 , 0, -1. -2 patterns begin to trons.
fill up. There is room for 10 electrons in these 5
patterns, and that takes us down to zinc. As the E 3 , The effect of angular momentum shows up most
= 2 patterns fill up underneath the E 4 , = 0 pattern, noticeably at potassium and calcium, where the two
there is little change in the outer electron structure and E 4 , = 0 patterns fill before the E 3 , = 2 patterns.
the binding energy increases slowly. The result is that Because of the extra angular momentum of the = 2
the 10 elements from scandium to zinc have similar electrons, the = 2 patterns have an extra large node at
chemical properties—all are metals. In some periodic the nucleus, keeping these electrons farther away and
tables, these elements are shown as the first set of more effectively screened.
transition elements. As we get to the heavier elements in the periodic table,
As we go from gallium to krypton we have the familiar those beyond krypton, the energy levels get closer
pattern of the E 4 , = 1 states filling up. There is a together and the binding energy depends more on the
gradual increase in binding energy from 6.00 eV at detailed structure of the wave patterns. As a result it
gallium to 14.00 eV at the noble gas krypton. There is becomes more difficult to predict how the wave pat-
even the slight glitch in binding energy going from terns will be filled and to estimate what the binding
arsenic to selenium that mirrors the glitches from energies should be. But despite this, we have been able
phosphorus to sulfur, and from nitrogen to oxygen. to go a long way in explaining the structure of the
periodic table from a few simple arguments about the
shape of the electron standing waves in hydrogen, and
the idea of electron screening.
 38-15

IONIC BONDING By looking at the number of loosely bound electrons in

In 1871 the Russian chemist Dimitri Mendeleyev a metal, or the number of empty slots in a non metal (the
worked out the periodic table of the elements from an number of electrons required to get to the next noble
analysis of the atomic weights and chemical reactions gas), you can predict the kind of compounds an element
of the elements. Here we will reverse Mendeleyev’s can form. For example, sodium, magnesium, and
approach and use Table 1, our shortened version of the aluminum have one, two and three loosely bound
periodic table, to explain some of the typical chemical electrons respectively, while oxygen has two empty
reactions and chemical compounds. slots. (Oxygen has six E 2 electrons, and needs two
more to fill up the E 2 standing wave patterns). When
As an example, suppose we placed a sodium atom next you completely burn the three metals, the oxides you
to a chlorine atom, what would happen? The sodium end up with are Na 2O , MgO and Al 2O 3 . In Na 2O ,
atom has one loosely bound electron in the E 2 , = 0 two sodium atoms each contribute one electron to fill
wave pattern. The binding energy of this electron is oxygen’s two slots. In MgO magnesium’s two loosely
5.14 eV. The chlorine atom has seven E 2 electrons all bound electrons are taken up by one oxygen atom. In
tightly bound because of the increase in the effective Al 2O 3 , two aluminum atoms each supply three elec-
nuclear charge seen by these electrons. It requires trons, these six electrons are then taken up by three
13.01 eV to remove an electron from chlorine. oxygen atoms. There is no simpler way for all the
aluminum’s loosely bound electrons to completely fill
If the sodium and the chlorine atom are brought close all of oxygen’s empty slots.
enough together, the loosely bound outer sodium elec-
tron can lose energy by moving into the remaining E 2 Hydrogen has one moderately bound electron which it
wave pattern in the chlorine atom. We end up with a can give up in some chemical reaction and act like a
negative chlorine ion Cl – , where all the E 2 patterns metal. An example is hydrochloric acid, HCl, where
are full, and a positively charged sodium ion Na + the chlorine ion has grabbed the hydrogen electron.
which has lost its outer electron. These charged ions
then attract each other electrically to form a sodium Hydrogen can also behave as a non metal. When it
chloride molecule NaCl which is common table salt. combines with active metals like lithium and sodium,
hydrogen grabs the metal’s loosely bound electron to
Sodium chloride is a typical example of ionic bonding. complete its E 1 standing wave pattern. The results are
The class of elements like lithium, sodium, magne- the compounds lithium and sodium hydride, LiH and
sium, aluminum, etc. that have one, two, or even three NaH.
loosely bound electrons, tend to give up these electrons
in a chemical reaction. These are called metals. Those More important to life are the bonds like those between
elements like oxygen, fluorine and chlorine, which hydrogen and carbon atoms which are not ionic in
have nearly full wave patterns and tightly bound elec- nature. Neither atom has a strong preference to give up
trons, tend to take up electrons in a chemical reaction or grab electrons. Instead the bonding results from the
and are called non metals. When metals and non sharing of electrons. This is the covalent bonding that
metals combine, held together by ionic bonding, you we described in our discussion of the hydrogen mol-
get a compound called a salt. ecule in Chapter 18.
Chapter 39
Spin

CHAPTER 39 S P I N

In the last chapter we saw that the basic structure of the

periodic table follows from the idea that up to two
electrons can fit into any given standing electron wave
pattern. If you consider the Pauli exclusion principle
which says that two electrons cannot be in exactly the
same state, then you have to find some difference
between the two electrons that can occupy the same
wave pattern. Gaudsmit and Uhlenbeck introduced
the concept of electron spin to explain this difference.
They proposed that the electron had an inherent angu-
lar momentum or spin that had two allowed projec-
tions, and that the difference between the two electrons
in one wave pattern was their spin projections. We
commonly call these two allowed projections spin up
and spin down.

In this chapter we take a more detailed look at electron

and nuclear spin and the interaction of spin with a
magnetic field. An electron with its spin projection
parallel to the magnetic field gains magnetic energy,
while the opposite projection loses it. The amount of
energy gained or lost is proportional to the strength of
the magnetic field.

The most accurate way to measure the spin magnetic

energy, is to start with an electron in the low energy
state and strike it with a photon. If the energy of the
photon is precisely equal to the energy required to raise
the electron from the low energy spin projection to the
high energy spin projection, the photon can be ab-
sorbed. We say that the photon flips the spin of the
39-2 Spin

electron. Since the energy of a photon is proportional oscillating at the frequency of precession, essentially
to its frequency according to Einstein's photoelectric pushing on the gyroscope once each time it comes
effect formula E=hf, measuring the frequency of the around, the gyroscope can gain energy from the oscil-
electromagnetic radiation that causes a spin flip tells lating field. This is a resonance phenomena; it is like
you how big the magnetic energy is. pushing a kid on a swing. You have to push the kid in
time with the swing in order to increase the amplitude
The energy required to flip the spin of an electron is of the motion.
usually not very large. If the electron is in a magnetic
field of around 10 gauss, typical fields produced by the It turns out that the frequency with which we have to
Helmholtz coils used in several of our laboratory oscillate the magnetic field is the same as the frequency
experiments, then photons in radio waves whose fre- of the photon that can cause the electron spin to flip.
quency is of the order of 30 megacycles have enough The quantum picture of a photon flipping a spin, and
energy to flip the electron spin. Since it is not hard to the classical picture of a precessing gyroscope in
generate electromagnetic waves of this frequency, we resonance with an oscillating magnetic field, gives the
can observe electron spin flip using much of our same results. Thus it is a matter of convenience
standard laboratory equipment. whether you use the classical or quantum picture.
When we talk of electromagnetic waves with frequen- Because the classical picture involves a resonance,
cies of the order of 30 megacycles, we are talking about this spin flip process is called electron spin resonance.
radio waves between the AM and FM broadcast bands. In this chapter we discuss an electron spin resonance
It is such a low frequency that individual photons experiment.
should be very hard to detect. Yet the spin flip experi-
ment does just that. For a given magnetic field, much less energy is re-
quired to flip the spin of a nucleus than of an electron.
At radio wave frequencies, Maxwell's theory and the Measurements of nuclear spin flip energies, in the so-
ideas of classical electric and magnetic fields should called nuclear magnetic resonance experiments, can
work as well as the photon picture. If we treat the be done so accurately that one can study not only the
spinning electron as a classical gyroscope with a spin of the nucleus but also the magnetic environment
magnetic moment, we find that a magnetic field can in which the nucleus sits. Nuclear magnetic resonance
exert a torque on the gyroscope, causing the gyroscope forms the basis of magnetic resonance imaging which
to precess. If we add an oscillating magnetic field, has become such an important diagnostic tool in
medicine.
 39-3

THE CONCEPT OF SPIN Perhaps the best way to view electron spin is to realize
A spinning top has an inherent angular momentum, but that we are dealing with a wave equation, and wave
if you try to picture an electron as a spinning top, you equations have specific allowed standing waves as
run into conceptual problems. First of all, if you have solutions. While electron spin does not come from
a spinning top, you can orient the top in any direction Schrödinger's wave equation, it does from Dirac's
you please. The top's angular momentum vector can more accurate relativistic wave equation. From Dirac's
point up, down, sideways to the left, sideways to the equation, we find that the electron has an inherent
right. But when we describe the electron's spin, we angular momentum of h/2 , with two possible projec-
have only two orientations, up and down. We ran into tions along the z axis, +h/2 and – h/2 . These are the
the puzzling idea of quantized projections of angular two allowed states of the electron. They are not
momentum in our interpretation of the allowed stand- different standing wave shapes like the hydrogen stand-
ing wave patterns of the hydrogen atom. But the idea ing waves of Figure (1) of the last chapter, but they are
of the electron's spin or rotational axis only pointing up different solutions to Dirac's wave equation.
or down seems even more counter intuitive. One of the surprises is that the spin angular momentum
Another blow at our classical intuition for angular of the electron is half a unit h/2 . The quantity h is not
momentum is our current theoretical picture of the the smallest amount of quantized angular momentum,
electron as a point particle. No experiment has demon- h/2 is. The standard terminology is to say that the
strated any finite size to the electron, and the theory that electron has half a unit of angular momentum, that it is
treats the electron as a point particle, quantum electro- a "spin 1/2" particle. The orbital angular momentum,
dynamics, is the most accurately tested theory in all of representing the motion of the electron around a nucleus,
physics. (String theory allows for some size for an is quantized in units of h . Only spin angular momen-
electron, radii of the order of 10 – 72 cm, but there are tum can come in half integer units.
no experimental tests of string theory.) While the electron's spin has a half integer value, its
If an electron has no radius, how can it have an inherent projection along the z axis changes by an integer value.
angular momentum? Angular momentum is linear In our discussion of the angular momentum of the
momentum times a lever arm. How can there be hydrogen standing waves, we saw that an electron in a
angular momentum if the particle has no radius, no certain energy level E n , with a total angular momen-
lever arm? The classical picture of electron spin tum , could have z projections ranging from , – 1 ,
resembling that of a spinning tops leaves a lot to be – 2 down to – . The allowed projections changed in
desired. However, despite the problems one encoun- units. The same is true for the electron spin, the allowed
ters, this picture does lead to some useful insights projections are +1/2 and – 1/2 , a change of one unit.
which we will mention shortly.
39-4 Spin

INTERACTION OF THE Magnetic Moments

MAGNETIC FIELD WITH SPIN and the Bohr Magneton
One of the predictions of the Dirac equation for elec- The formula for the Bohr magneton has its origin in a
trons is that the electron spin interacts with a magnetic combination of classical physics with the Bohr theory.
field. The state with the spin parallel to the magnetic Back in Chapter 31 we observed that if you place a loop
field gains magnetic energy while the other state loses of wire in a magnetic field B , and then run an electric
it. The amount of energy gained or lost is proportional current i through the wire, the magnetic field can exert
to the strength B of the magnetic field, and the propor- a torque on the loop. We found that if you curled the
tionality constant turns out to be a quantity called the fingers of your right hand in the direction the current is
Bohr magneton, designated by the symbol µ B . going around the loop, then the magnetic torque tended
to orient the loop so that your thumb pointed parallel to
magnetic the magnetic field. We called this the low energy
energy of +µB B spin parallelto B
Emag = (1) orientation of the loop. To turn the loop over to the high
electron – µB B spin opposite to B energy orientation required an amount of work that was
spin
proportional to the current i, the strength of the mag-
µ B = eh/2m Bohr magneton netic field B, and to the area A of the loop. The explicit
formula for the amount of work required was
= 5.79 × 10 – 5 eV/tesla (2)
The amount of energy required to flip an electron from energy required
= 2 (iA) B
its low energy state to the high energy state is thus to turn loop over
energy required to We defined the product of the current i times the area
∆E mag = 2µ BB flip the electron spin (3) A as the magnetic moment µ of the current loop.
in a magnetic field
µ ≡ iA (see 31-34)
Since a Bohr magneton is 5.79 × 10 – 5eV/tesla we
can express Equation 3 numerically as which gave us the formula

energy required
∆E mag = 11.6 × 10 – 5B eV (4) = 2 µB (31-36a)
to turn loop over
where B has to be expressed in tesla.
Later in the chapter we considered a special kind of
If you wish to measure magnetic fields in gauss, then current loop consisting of a charge q moving at a speed
convert µ B to eV/gauss: v in a circular orbit of radius r. We found that the
magnetic moment µ = iA of this special current loop
µ B = 5.79 × 10 – 5 eV × 1 could be written in the form
tesla 10 4 gauss q
tesla µ = mvr (31-38a)
2m
eV
µ B = 5.79 × 10 – 9 gauss (5) However mvr is the angular momentum L (we called
it J back there because we were using L for induc-
tance). Thus the formula for the magnetic moment µ
of an orbiting charge can be written
q
µ = L (31-39)
2m
 39-5

The above result is strictly classical. If we jump ahead Electron Spin Resonance Experiment
to the Bohr picture where we are dealing with an The basic idea of the electron spin resonance experi-
electron whose charge is q = –e, and angular momen- ment is to flip the spin of an electron by striking the
tum L is quantized in units of h , then we find that the electron with a photon. The electron’s spin will flip
magnetic moment is quantized in units of only if the photon’s energy hf is equal to the magnetic
spin flip energy 2µ BB. Thus we wish to test the rela-
µB = e h (6) tionship
2m
where this unit of magnetic moment µB is called a Bohr hf = 2µ BB spin flip requirement (7)
magneton. This is where the name and formula for the
Bohr magneton originated. The same constant ap- where f is the frequency of the photon.
peared in Dirac’s equation for the energy required to
flip the spin of the electron. Exercise 2
a) An electron is placed in a 10 gauss magnetic field.
The minus sign of the electron charge means that the How much energy, in eV, is required to flip the electron
high energy orientation of the electron is when the spin from its low energy to its high energy state? (Answer:
is parallel to the magnetic field. 1.15 × 10– 7 eV ).

b) You wish to flip the spin of the electron in part (a) by

Exercise 1
striking it with a photon. Assume that the photon is
The formulas absorbed by the electron, and that all the photon's
Emagnetic = 2µ B B
energy goes into flipping the electron's spin. What
wavelength photon should you use? (Answer:
µB = e h λ = 1071 cm )
2m
c) What is the frequency of the photon in part (b)?
came mostly from Chapter 31 where we were working
(Answer: 28 megacycles).
in MKS units. As a result, we need to use MKS units to
evaluate µ B . (The constant µ B has a different formula
in CGS units.) Exercise 3
The student FM radio station at Dartmouth College
We can get the dimensions of µB from the equation broadcasts on a frequency of 99.4 megacycles. If you
Emagnetic = 2µ B B , or wished to use this frequency radiation to flip the spin of
an electron in a magnetic field B, what should be the
Emagnetic joules strength of B? Give the answer in gauss.
µB =
2B tesla
In our discussion of the particle nature of light, we
Thus when you use MKS units to evaluate µ B = eh 2m ,
pointed out that because a radio wave consists of so
your answer comes out in joules/tesla rather than
eV/tesla. To get the final answer in eV/tesla, you then
many photons of such low energy, it would be difficult
use the conversion factor 1.6 × 10– 19 joules/eV. to detect individual photons, and thus the wave nature
of radio waves should predominate. However the spin
With this background, show that of the electron is just the right detector for these low
energy photons. An electron spin flip experiment can
µ B = 5.79 × 10– 5 eV be viewed as an experimental detection of the indi-
tesla
vidual photons in a radio wave.
(You will get a value of µ B = 5.82 × 10– 5 eV/tesla, which
differs slightly due to the way we have rounded off the
constants.)
39-6 Spin

Nuclear Magnetic Moments Sign Conventions

Both the proton and neutron are spin 1/2 particles, To handle the fact that the electrons and protons have
which means that they each have a spin angular mo- opposite charges and therefore opposite magnetic
mentum with two allowed spin states, spin up and spin moments, the following sign conventions are generally
down. If you place either of these particles in a used
magnetic field, one of the projections will gain mag-
netic energy while the other loses it. magnetic
energy E magnetic = – µ ⋅ B
of spin
For an electron, the high magnetic energy state was
when the spin pointed parallel to the magnetic field.
– µB spin parallel to B
Because the proton has the opposite charge from the = (10)
electron, the opposite orientation is the high magnetic + µB spin opposite to B
energy state.
where the electron, proton and neutron have the follow-
If the Dirac equation is applied accurately to a proton, ing magnetic moments
then the formula for the proton’s magnetic moment
would be one nuclear magneton µN defined by the µ e = – 1.00114 µ B electron
magnetic moment (11)
equation
proton
µ p = 2.79 µ N magnetic moment (9)
eh definition of the
µN = nuclear magneton (8)
2m proton
µ n = – 1.19 µ N neutron
magnetic moment (12)
which is the Bohr magneton formula with the electron
mass replaced by the proton mass. Since the proton is where the Bohr magneton µ B is
1836 times heavier than an electron, a nuclear magne-
ton is 1/1836 times smaller than a Bohr magneton. µB = eh = 5.79 × 10 – 5 eV (13a)
2m electron tesla
The Dirac equation, however, does not give the correct
and the nuclear magneton µ N is
value for the proton’s magnetic moment µ p . The
experimental value is eh
µN = = 3.15 × 10 – 8 eV (13b)
2m proton tesla
µ p = 2.79 µ N (9)
Note that by putting a – (minus) sign in the formula for
The fact that the Dirac equation is off by a factor of 2.79 E magnetic , and making the electron’s magnetic mo-
is one indication that the proton is a more complex ment negative, we still have the result that the electron’s
object than the electron. spin magnetic energy is positive when the electron’s
spin is parallel to B .
(The Dirac equation is not exact even for the electron.
The experimental value for the electron’s magnetic
moment is 1.00114 Bohr magnetons. The correction of
.00114 Bohr magnetons is accurately explained by the
theory of quantum electrodynamics.)
 39-7

Where the Dirac equation completely fails is in the case Exercise 4

of the neutron. If the neutron were a simple uncharged (a) Express the magnetic moment of the proton in
particle, it would have no magnetic moment. The fact eV/tesla .
that it does have a magnetic moment suggests that,
while it has no net charge, it must be some kind of (b) A proton is in a 1 tesla magnetic field. How much
composite object with charged particles inside. We energy, in eV, is required to flip the spin of the proton?
now know that this suggestion is correct. The neutron (c) What is the wavelength and frequency of a photon
is made up of three quarks, one up quark with a charge that can flip the spin of the proton in part (b)? (Answer:
+2/3 e and two down quarks with a charge -1/3 e. (a) 8.79 × 10– 8 eV/tesla , (b) 17.6 × 10– 8 eV , (c) 706 cm
While there is no net charge, the quarks contribute to and 42.5 megacycles.)
magnetic energy. (The proton, which consists of two
up quarks and one down quark, has a total charge of - Exercise 5
2/3e + 2/3e - 1/3e = +e.) What strength magnetic field should you use so that
photons from the student FM radio station (99.4 mega-
As we mentioned, the difference between the Bohr cycles) can flip the spin of the proton? (Answer: 2.34
magneton µ B and the nuclear magneton µ N is due to tesla.)
the mass difference between the electron and the
proton. Since a proton is 1836 times as massive as an
electron, the nuclear magneton is 1836 times smaller
than the Bohr magneton. The result is that the magnetic
moments of protons, neutrons, and nuclei in general are
typically an order of a thousand times smaller than the
electron magnetic moment. To get the same magnetic
spin energies as you do for electrons, you thus need
magnetic fields of the order of a thousand times stron-
ger when working with nuclei.
39-8 Spin

Classical Picture of In Equation 14, Planck's constant h cancels and we get

Magnetic Resonance
frequency of radio wave
In the appendix to this chapter we work out the f = e B photon that can flip (15)
classical picture of the interaction of the electron's spin 2πm an electron spin
with a magnetic field. One pictures the spinning
electron as acting as a tiny current loop with a mag- as the relationship between the frequency f of the radio
netic moment µ as described at the end of Chapter 31. wave and the strength of the magnetic field B.
The current loop also has an angular momentum L
which makes it act like a gyroscope. If you place the The fact that Planck's constant cancelled in Equation 15
current loop in a magnetic field, as shown in Figure suggests that a classical analysis might give similar
(1), the magnetic field exerts a torque and one predicts results. In the appendix, we analyze the behavior of a
that the current loop should precess about the mag- current loop consisting of a particle of charge q and
netic field lines. The precession is analogous to the mass m, travelling in a circular orbit. The magnetic
precession of the bicycle wheel gyroscope studied in moment µ of the loop points along the axis of the orbit.
Chapter 12. If the loop is placed in a magnetic field B oriented
perpendicular to µ , then the loop will precess around
In this classical picture, if you subject the precessing the magnetic field line at a precessional frequency
current loop to the electromagnetic field of a radio given by the formula
wave, whose frequency f is equal to the precessional
frequency fp of the loop, the loop can gain energy precessional frequency
f = e B of a current loop in (16)
from the radio wave. This is a resonance phenomena, 2πm a magnetic field
where the push of the fields of the radio wave have to
match the timing of the precession of the loop. It is which is the same formula as Equation 15 for the
analogous to pushing a child on a swing, where you frequency of a radio wave photon that can flip the spin
have to time your pushes with the motion of the child of an electron. In the classical picture, if we superim-
in order to add energy to the motion. pose a radio wave at this frequency, we get a resonance
The classical picture of a precessing current loop between the frequency of the radio wave and the
gradually gaining energy from a radio wave, and the precessional frequency of the current loop, enabling
quantum picture of an electron spin being flipped by the radio wave to add energy to the current loop.
a photon, happen to lead to nearly the same predic- The classical calculations have certain errors that have
tions. If we start with the condition hf = 2µ BB for the to be corrected on an ad hoc basis. The current loop
photon energy to match the spin flip energy, then model leads to a relationship between the loop's angu-
replace µ B by eh/2m = e h/2π /2m = eh /4 π m , we lar momentum L and its magnetic moment µ . If we
get evaluate µ experimentally from the relationship
hf = 2µ BB = 2 eh B (14) E mag = – µ ⋅ B , and set L = h/2 for a spin one half
4πm
particle, the classical relationship is off by a factor of 2
for electrons and 2 x 2.79 for protons. These errors are
B accounted for by introducing a fudge factor called the
Laudé g factor to correct the value of µ . Since the
ωp classical picture has fundamental problems, such as no
Figure 1 hint of quantization of angular momentum, and no
Current loop in a explanation of how a particle of zero radius can have
magnetic field. angular momentum, it is surprising that the semi clas-
µ sical picture works as well as it does.
i
 39-9

ELECTRON SPIN When two electrons are in the same standing wave, we
RESONANCE EXPERIMENT say that the electrons are paired. In order to see the spin
The point of the electron spin resonance experiment is flip energy, we need a substance with an unpaired
electron, a substance where the electron spin can be
to detect the electron spin flip energy ∆E = 2µ BB
flipped without otherwise disturbing the structure of
predicted by the Dirac equation using photons of
energy E = hf. You might try to do this by placing a the substance. Such unpaired electrons can be quite
chemically active and are known as free radicals. An
container of hydrogen in a magnetic field B and radi-
example of a substance with such an unpaired electron
ating the hydrogen with a radio wave. If the frequency
f of the radio wave were such that the photon energies is the crystalline organic chemical diphenyl-picryl-
hydrazyl or DPPH for short. Since free radicals cause
hf equalled the spin flip energy 2µ BB , then perhaps we
cancer, when we use this substance in our electron spin
could detect radio wave energy being absorbed as
hydrogen atom electrons in the low energy spin state resonance experiment, we seal it in a small glass vial to
keep from coming in contact with it.
were flipped over to the high energy spin state.
To perform the magnetic resonance experiment, we
Such an experiment will not work because of an
need to both create the radio waves and detect the
interesting quantum mechanical effect. Hydrogen
atoms form hydrogen molecules consisting of 2 pro- energy lost to the electrons being flipped. Both of these
steps can be accomplished by placing the glass vial
tons surrounded by 2 electrons. In the ground state of
containing the DPPH inside the coil of a resonant LC
the hydrogen molecule, both electrons are in the lowest
energy standing wave pattern allowed for the mol- circuit. The circuit, oscillating at its resonant frequency
(f) is the source of the photons of energy hf. Detecting
ecule. This is the electron cloud we sketched in Figure
the drain of energy from the coil when hf equals 2µ BB
(19-8) in our discussion of the molecular forces be-
is the way we detect the spin flips.
tween hydrogen atoms.
The easiest way to perform the experiment is to get the
The Pauli exclusion principle requires that no two
LC circuit oscillating at some frequency (f), and then
electrons be in exactly the same state. If the two
change the magnetic field strength B until 2µ BB = hf.
electrons in the hydrogen molecule are in the same
To detect the loss of energy at this point, we use a
standing wave pattern, then they must have opposite
specially designed LC circuit that is barely oscillating.
spins in order to satisfy the exclusion principle.
The circuit is designed to stop oscillating if any energy
If we try to flip one of the electron spins with a radio is being drained from the circuit. To detect whether or
wave, the spin flipped electron cannot stay in the low not the circuit is oscillating, another circuit detects the
energy standing wave pattern, for then we would have amplitude of the oscillation and puts out a DC voltage
two electrons with the same spin in the same wave proportional to that amplitude. The DC voltage can
pattern. In order to flip the spin of one of the electrons, then be displayed on an oscilloscope.
we must supply not only the spin flip energy 2µ BB , but
also enough energy to raise the electron into a higher
energy standing wave pattern. Going to a higher
energy standing wave requires much more energy than
flipping a spin, thus photons with an energy hf equal to
2µ BB will have no effect on hydrogen molecules.
39-10 Spin

With this arrangement we can sweep the magnitude of was met six times producing the six resonant peaks. To
B through the value 2µ BB = hf and watch on an see how the condition was met, consider the detailed
oscilloscope the amplitude of the oscillation of the diagram of the center peaks shown in Figure (3). The
circuit. The result is shown in the solid curve of Figure first peak, at the time of 18.7 milliseconds, occurred
(2). We see six peaks because the conditions 2µ BB = hf when the magnetic field had a magnitude of 8.4 gauss.
was met six times during the 40 milliseconds shown in [We calculated this from 105 × 10 – 3volts 0.1Ω
the diagram. = 1.05 amps, and then B(gauss) = 8i(t) = 8 × 1.05amps
= 8.4 gauss.] At time t = 20.1 milliseconds, the
To produce the magnetic field B, the probe containing magnitude of B goes down through zero, and then
the vial of DPPH was placed at the center of our familiar reaches a magnitude of –8.4 gauss at a time t = 21.5
Helmholtz coils as shown in Figure (4). As in our milliseconds. We get a peak at both +8.4 gauss and –8.4
magnetic field mapping experiment [Figure (24) on gauss because the resonance does not depend upon
page 30-24], we power the helmholtz coils with a 60Hz which of the two ways the magnetic field was pointing.
current i(t) to produce a sinusoidally varying magnetic Looking back at Figure (2), we see that B had a
field. This current passes through a 0.1Ω resistor so that magnitude of + or – 8.4 gauss six times, which is why
we can measure the strength of the magnetic field by we got the six peaks.
plotting the voltage V(t) across the resistor. The result
is the dashed curve seen in Figure (2). We can estimate For the experiment shown in Figures (2) and (3), the
the strength of B by using i(t) = V(t)/0.1Ω and then photons in the resonant LC circuit, the photons flipping
remembering that for these coils the magnetic field B the spin of the DPPH electrons, had a frequency f = 28
in gauss is about equal to 8i(t). megacycles. We can use the fact that these photons
flipped the spins when the magnetic field was about 8.4
As the magnetic field went through somewhat more gauss to calculate our value of the electron magnetic
than one cycle in Figure (2), the condition 2µ BB = hf moment µ B. We have

Figure 2
The solid curve shows the resonant peaks while the dashed
curve is proportional to the strength of the magnetic field.
 39-11

2µ BB = hf

µ B = hf
2B
Using the values
6.63 × 10 – 34 joule sec
h = = 4.14 ×10 – 15 eV sec
–19 joules
1.6 × 10
eV
1
f = 28 × 10 6 sec

tesla = 8.4 × 10 –4tesla

B = 8.4 gauss × 10 –4 gauss

we get
1
4.14 × 10 – 15 eV sec × 28 × 10 6 sec Figure 4
µB =
2 × 8.4 × 10 –4tesla Electron spin resonance apparatus. The coil containing
the vial of DPPH is at the tip of the probe, which is at
the center of the Helmholtz coils. The capacitor of the
µ B. = 6.9 × 10 – 5 eV our
result (17) LC circuit, and the controls, are at the other end of the
tesla probe. The Helmholtz coils are being driven by a 60Hz
This result is nearly 20% above the known value alternating current. As a result, the magnitude of the
magnetic field sweeps back and forth through the
accepted
µ B. = 5.79 × 10 – 5 eV value (5) resonant value.
tesla

105 millivolts
= 8.4 gauss

–105 millivolts
= 8.4 gauss
pointing the
other way

Figure 3
We get a resonant peak for both orientations of the magnetic field.
39-12 Spin

The source of this error is in our measurement of the Exercise 6

strength of the magnetic field. We have relied on the In our electron spin resonance experiment, we saw that
accuracy of the value of the 0.1Ω resistor through an electron in a magnetic field had two energy states,
which the helmholtz coil current i(t) passes, and then and that the difference in the energy between the states
used the approximate formula B(gauss) = 8i(t). One was proportional to the strength B of the magnetic field.
can obtain much more accurate results using the preci- We measured this energy by placing the electrons in an
sion search coil shown in Figure (5). This is a 100 turn oscillating electromagnetic field of a given frequency
coil wound on a 1 inch (2.54cm) plastic rod. Using the (around 30 megacycles) and observing at what values
techniques discussed in the magnetic mapping experi- of the magnetic field we got a transition between the two
ment, one can accurately relate the voltage VR mea- states.
sured across the 0.1Ω resistor to the actual value of B. In Figure (6) we have a somewhat similar situation
We have encouraged students who wish to do a project except the object being studied is a HD (Hydrogen-
involving electron spin resonance to see how accurate Deuterium) molecule. In this molecule the Hydrogen
a value of µ B they can obtain using this precision nucleus (a proton) weakly interacts with the Deuterium
search coil. nucleus (a proton and a neutron). These two nuclei form
a system with several energy states or levels. If you
apply a magnetic field B to the HD molecule, the
difference in the energy between the states is related to
the strength of B. The energy difference between the
states can be measured by applying an oscillating
electromagnetic field of a given frequency and observ-
ing at what values of the magnetic field we get a
transition between states.

Figure 5
100 turn search coil for accurately
determining the magnetic field.

0 H(mG)
–20 –10 10 20

H. Benoit and P. Piejus,Compt. Rend. 265B, 101 (1967).

"Spectre de R.M.N. de HD á une Fréquence de 54 Hz."
Transition from the 3/2,1/2 to the 3/2,3/2 states of the
nuclei in an HD molecules at 20 kelvins. The nuclei
were pre alligned (polarized) in an 8 killogauss field.

Figure 6
NMR data on liquid Hydrogen-Deuterium.
 39-13

Figure (6) is a nuclear magnetic resonance scan of the

HD molecules. As in our electron spin resonance
experiment, the molecules are placed in an oscillating
electromagnetic field of a given frequency, and the
strength of a uniform magnetic field B is varied. Two
resonance peaks are observed, but they represent the
same transition between the energy levels, since the
transition does not depend upon the sign of the mag-
netic field.

One of the main differences between the electron

system we studied in the lab and the HD molecule, is
that the energy level splitting is really really small in the
HD molecule experiment compared to the splitting we
observed in the electron experiment. Instead of fields
of tens of gauss and frequencies of around 30 Mega-
cycles, in the HD experiment of Figure (6) , the fre-
quency was 54 cycles per second and the field B was
just under 20 milligauss (.020 gauss or .000002 tesla)!
This example demonstrates the enormous range of
applicability of the magnetic resonance experiments.

For this exercise, we want you to calculate the splitting

between the two energy levels involved in the reso-
nance transitions seen in Figure (6). Give the answer in
ergs or joules, and in electron volts. Comment on the
reasonableness of your answer—do you think your
result is too big, too small, or perhaps OK.
39-14 Spin

APPENDIX But if the angular momentum of the particle is quan-

tized, if the magnitude of E mag = –µ ⋅ B cannot change,
CLASSICAL PICTURE OF then our current loop analysis and magnetic energy
MAGNETIC INTERACTIONS formula E mag = –µ ⋅ B has a better chance of working.
At the end of Chapter 31, we discussed the magnetic There is no reason to expect any classical formulas to
moment of a current loop, deriving the formulas apply to atomic or subatomic systems. What we are
looking for are those that do.
A = area of current loop
To apply classical formulas to particle spins, we have
µ ≡ iA magnetic moment (31-34) to fudge the relationship between the particle's mag-
netic moment µ and its spin angular momentum. As a
τ=µ×B magnetic torque (31-35) general relationship between magnetic moment µ and
angular momentum L , we will rewrite Equation 31-39
E mag = – µ ⋅ B magnetic energy (31-37) in the form
q
These equations applied to a current loop whose cur- µ = g L (A-1)
2m
rent i and area A were unaffected by the magnetic
field. We only allowed the magnetic field to change the where g is our fudge factor. It is the factor we have to
orientation of the loop via the torque τ . introduce to make classical calculations give the cor-
rect results.
We then treated a charged particle in a circular orbit as
a current loop. Using the definition µ = iA , we found The value of g depends upon the kind of system we are
that the orbiting particle had a magnetic moment µ talking about. If we are talking about the angular
related to its angular momentum L by momentum of an electron in orbit about a nucleus, then
g=1 and the classical equations work. If we are talking
q
µ = L (31-39) about the spin angular momentum of an electron, then
2m
g=2. For electron spin, the classical formulas are off by
where q is the charge and m the mass of the particle. a factor of 2. For the proton, g has to have the value
2 × 2.79 in order to get the proton magnetic moment
To use the magnetic energy formula E mag = –µ ⋅ B, given in Equation 9. Since protons and neutrons are
we have to make the same assumptions about the composite particles made from quarks it should not be
orbiting particle as we did about the current loop. surprising that g should have a peculiar value. This
Namely we have to assume that the magnetic field factor g is called either the gyromagnetic ratio or
alters only the orientation of the orbit and not the Landé g factor. Fudge factors sound better if you give
particle's speed v or orbital radius r. Since L = mvr, them impressive names.
we are thus assuming that the magnetic field does not
affect the magnitude of the particle's angular momen- Combining our modified Equation A-1 for µ with 31-
tum. 37 for E, we get
q
For a classical particle, such an argument is not reason- E mag = – µ ⋅ B = g L⋅B (A-2)
2m
able. If we turn on a magnetic field, we change the
magnetic flux through the orbit and thus by Faraday's as the semi classical formula for the magnetic energy of
law induce a voltage around the orbit. This induced a particle in a magnetic field. We say semi classical
voltage should affect both the particle's speed and because of the correction factor g.
orbital radius.
 39-15

If the particle has an angular momentum L and the To go to the quantum picture, multiply Equation A-4
magnetic field exerts a torque τ , the particle should through by Planck's constant h, to get
precess like a gyroscope. Thus we can compare a q
bicycle wheel gyroscope subject to a gravitational hf precession = g h B
2π 2m
torque to a current loop of magnetic moment µ subject
to a magnetic torque. To simplify the analysis, we are qh
= g B (A-5)
assuming that the magnetic moment µ lies in the plane 2m
perpendicular to B so the magnetic torque τ = µ × B Applying this to an electron spin, setting q and m to the
has a magnitude µB . charge and mass of an electron, we get
classical theory
Following the standard analysis of a gyroscope, we hf precession = gµ B B (A-6)
with factor g
predict that the magnetic moment vector µ should
precess about the magnetic field vector B at a rate where µ B = eh/2m is the Bohr magneton.
ωprecession given by
µB In the quantum picture, the electron gains magnetic
ωprecession = τ = (12-58) energy if the photons in the radio wave have the right
L L
a result we derived back in Chapter 12. Using our semi amount of energy to flip the spin of the electron. The
Dirac equation gave the spin flip energy as
classical formula A1 for µ , we get
qL B energy required
ωprecession = g ∆E = 2µ BB to flip the (3 repeated)
2m L electron spin
The L’s cancel, and we are left with
In Equation 7, we equated this energy to the photon
q
ωprecession = g B (A-3) energy to get
2m
set spin flip
The quantity ωprecession is the precessional frequency hf = 2µ BB energy equal to (7 repeated)
in radians per second. To convert this to cycles per photon energy
second, we divide by 2π radians/cycle to get
as the formula giving the frequency of the radio wave
ωprecession g q that can add magnetic energy to the electron. Compar-
fprecession = = B (A-4)
2π 2π 2m ing Equation (7) with the semi classical result (A-6), we
as the precessional frequency of a charged orbiting or see that the Landé g factor, the gyromagnetic ratio g,
spinning particle in a magnetic field. has to be set equal to 2 for the semi classical theory to
agree with the Dirac equation
Applying Equation A-4 to the spin of a particle, we
predict classically that if the particle is subject to an g = 2
gyromagneticratio
(A-7)
alternating electric and magnetic fields of frequency f, for the electron spin
there will be a resonance and the particle can gain
magnetic energy if the frequency f equals the preces-
sional frequency fprecession .
Chapter 40
Quantum
Mechanics
CHAPTER 40 QUANTUM MECHANICS

That light had both a particle and a wave nature

became apparent with Einstein’s explanation of the
photoelectric effect in 1905. One might expect that
such a discovery would lead to a flood of publications
speculating on how light could behave both as a
particle and a wave. But no such response occurred.
The particle wave nature was not looked at seriously
for another 18 years, when de Broglie proposed that
the particle wave nature of the electron was respon-
sible for the quantized energy levels in hydrogen. Even
then there was great reluctance to accept de Broglie’s
proposal as a satisfactory thesis topic.

Why the reluctance? Why did it take so long to deal

with the particle-wave nature, first of photons then of
electrons? What conceptual problems do we encoun-
ter when something behaves both as a particle and as
a wave? How are these problems handled? That is the
subject of this chapter.
40-2 Quantum Mechanics

TWO SLIT EXPERIMENT much less than D, which it is for most 2 slit experi-
Of all the experiments in physics, it is perhaps the 2 slit ments, then the hypotenuse of the big triangle is ap-
experiment that most clearly, most starkly, brings out proximately D and equating corresponding sides of the
the problems encountered with the particle-wave na- similar triangles gives us the familiar relationship
ture of matter. For this reason we will use the 2 slit
experiment as the basis for much of the discussion of λ/2 = y min
this chapter. d D

Let us begin with a review of the 2 slit experiment for 2y mind

water and light waves. Figure (1) shows the wave λ = (1)
D
pattern that results when water waves emerge from 2
slits. The lines of nodes are the lines along which the Figure (3a) is the pattern we get on a screen if we shine
waves from one slit just cancel the waves coming from a laser beam through 2 slits. To prove that the dark
the other. Figure (2) shows our analysis of the 2 slit bands are where the light from one slit cancels the light
pattern. The path length difference to the first mini- from the other, we have in Figure (3b) moved a razor
mum must be half a wavelength λ/2 . This gives us the blade in front of one of the slits. We see that the dark
two similar triangles shown in Figure (2). If y min is bands disappear, and we are left with a one slit pattern.
The dark bands disappear because there is no longer
any cancellation of the waves from the 2 slits.

Figure 3a
Two slit interference pattern for light. The closely
spaced dark bands are where the light from one slit
Figure 1 cancels the light from the other.
Water waves emerging from two slits.

θ ymin
θ
d

λ/2 (path length difference)

razor blade
D
Figure 3b
Figure 2 Move a razor blade in front of one of the
Analysis of the two slit pattern. We get a minimum slits, and the closely spaced dark bands
when the path length difference is half a wavelength. disappear. There is no more cancellation.
 40-3

In 1961, Claus Jönsson did the 2 slit experiment using Imagine yourself shrunk down in size so that you could
electrons instead of light, with the results shown in stand in front of a small section of the photographic
Figure (4). Assuming that the electron wavelength is screen in Figure (3a). Small enough that you want to
given by the de Broglie formula p = h/λ , the dark avoid being hit by one of the photons on the laser beam.
bands are located where one would expect waves from As you stand at the screen and look back at the slits, you
the 2 slits to cancel. The 2 slit experiment gives the see photons being sprayed out of both slits as if two
same result for light and electron waves. machine guns were firing bullets at you, but you
discover that there is a safe place to stand. There are
these dark bands where the particles fired from one slit
The Two Slit Experiment
cancel the particles coming from the other.
from a Particle Point of View
In Figure (3a), the laser interference patterns were Then one of the slits is closed, there is no more
recorded on a photographic film. The pattern is re- cancellation, the dark bands disappear as seen in Figure
corded when individual photons of the laser light strike (3b). There is no safe place to stand when particles are
individual silver halide crystals in the film, producing being fired at you from only one slit. It is hard to
a dark spot where the photon landed. Where the image imagine in our large scale world how it would be safe
shows up white in the positive print, many photons to have two machine guns firing bullets at you, but be
have landed close together exposing many crystal lethal if only one is firing. It is hard to visualize how
grains. machine gun bullets could cancel each other. But the
particle wave nature of light seems to require us to do
In a more modern version of the experiment one could so. No wonder the particle nature of light remained an
use an array of photo detectors to count the number of enigma for nearly 20 years.
photons landing in each small element of the array. The
number of counts per second in each detector could
then be sent to a computer and the image reconstructed Two Slit Experiment—One Particle
on the computer screen. The result would look essen- at a Time
tially the same as the photograph in Figure (3a). You might object to our discussion of the problems
involved in interpreting the two slit experiment. After
The point is that the image of the two slit wave pattern all, Figure (1) shows water waves going through two
for light is obtained by counting particles, not by slits and producing an interference pattern. The waves
measuring some kind of a wave height. When we look from one slit cancel the waves from the other at the lines
at the two slit experiment from the point of view of of nodes. Yet water consists of particles—water
counting particles, the experiment takes on a new molecules. If we can get a two slit pattern for water
perspective. molecules, what is the big deal about getting a two slit
pattern for photons ? Couldn’t the photons somehow
interact with each other the way water molecules do,
and produce an interference pattern?

Photons do not interact with each other the way water

molecules do. Two laser beams can cross each other
with no detectable interaction, while two streams of
water will splash off of each other. But one still might
Figure 4 suspect that the cancellation in the two slit experiment
Two slit experiment using electrons. (By C. Jönsson) for light is caused by some kind of interaction between
the photons. This is even more likely in the case of
electrons, which are strongly interacting charged par-
ticles.
40-4 Quantum Mechanics

In an earlier text, we discussed the possibility of an

experiment in which electrons would be sent through a
two slit array, one electron at a time. The idea was to
eliminate any possibility that the electrons could pro-
duce the two slit pattern by bouncing into each other or
interacting in any way. Since the experiment had not
a) 10 dots yet been done, we drew a sketch of what the results
should look like. That sketch now appears in a number
of introductory physics texts.

When he saw the sketch, Lawrence Campbell of the

Los Alamos Scientific Laboratories did a computer
simulation of the experiment. We will first discuss
Campbell’s simulation, and then compare the simula-
b) 100 dots tion with the results of the actual experiment which was
performed in 1991.

It is not too hard to guess some of the results of sending

electrons through two slits, one at a time. After the first
electron goes through you end up with one dot on the
screen showing where the electron hit. The single dot
is not a wave pattern. After two electrons, two dots; you
c) 1000 dots cannot make much of a wave pattern out of two dots.
If, after many thousands of electrons have hit the
screen, you end up with a two slit pattern like that
shown in Figure (4), that means that none of the
electrons land where there will eventually be a dark
band. You know where the first dot, and the second dot
cannot be located. Although two dots do not suggest a
d) 10000 dots wave pattern, some aspects of the wave have already
imposed themselves by preventing the dots from being
located in a dark band.

To get a better idea of what is happening, let us look at

Campbell’s simulation in Figure (5). In (5a), and (5b)
we see 10 dots and 100 dots respectively. In neither is
Predicted pattern there an apparent wave pattern, both look like a fairly
random scatter of dots. But by the time there are 1000
dots seen in (5c), a fairly distinctive interference pattern
is emerging. With 10,000 dots of (5d), we see a fairly
close resemblance between Campbell’s simulation
and Jönsson’s experimental results. Figure (5e) shows
the wave pattern used for the computer simulation.
Experimental results by C. Jönsson
Figure 5 Although the early images in Figure (5) show nearly
Computer simulation of the 2 slit random patterns, there must be some order. Not only
electron diffraction experiment, as if the do the electrons not land where there will be a dark
electrons had landed one at a time. band, but they must also accumulate in greater numbers
 40-5

where the brightest bands will eventually be. If this

were a roulette type of game in Las Vegas, you should
put your money on the center of the brightest band as
being the location most likely to be hit by the next
electron.

Campbell’s simulation was done as follows. Each

point on the screen was assigned a probability. The
probability was set to zero at the dark bands and to the
greatest value in the brightest band. Where each
electron landed was randomly chosen, but a random-
ness governed by the assigned probability.

How to assign a probability to a random event is

illustrated by a roulette wheel. On the wheel, there are
100 slots, of which 49 are red, 49 black and 2 green.
Thus where the ball lands, although random, has a 49%
chance of being on red, 49% on black, 2% on green, and
0% on blue, there being no blue slots.

In the two slit simulation, the probability of the electron

landing at some point was proportional to the intensity
of the two slit wave pattern at that point. Where the
wave was most intense, the electron is most likely to
land. Initially the pattern looks random because the
electrons can land with roughly equal probability in
any of the bright bands. But after many thousands of
electrons have landed, you see the details of the two slit
wave pattern. The dim bands are dimmer than the
bright ones because there was a lower probability that
the electron could land there.

Figure (6) shows the two slit experiment performed in

1991 by Akira Tonomura and colleagues. The experi-
ment involved a novel use of a superconductor for the
two slits, and the incident beam contained so few
electrons per second that no more than one electron was
between the slits and the screen at any one time. The
screen consisted of an array of electron detectors which
recorded the time of arrival of each electron in each
detector. From this data the researchers could recon-
struct the electron patterns after 10 electrons (6a), 100
electrons (6b), 3000 electrons (6c), 20,000 electrons
(6d) and finally after 70,000 electrons in Figure (6e).
Just as in Campbell’s simulation, the initially random Figure 6
looking patterns emerge into the full two slit pattern Experiment in which the 2 slit electron interference
when enough electrons have hit the detectors. pattern is built up one electron at a time. ( A.
Tonomura, J. Endo, T. Matsuda, T. Kawasaki,
American Journal of Physics, Feb. 1989. See also
Physics Today, April 1990, Page 22.)
40-6 Quantum Mechanics

Born’s Interpretation Photon Waves

of the Particle Wave Both electrons and photons have a particle-wave nature
In 1926, while calculating the scattering of electron related by the de Broglie formula p = h/λ , and both
waves, Max Born discovered an interpretation of the produce a two slit interference pattern. Thus one would
electron wave that we still use today. In Born’s picture, expect that the same probability interpretation should
the electron is actually a particle, but it is the electron apply to electron waves and light waves.
wave that governs the behavior of the particle. The
electron wave is a probability wave governing the We have seen, however, that a light wave, according to
probability of where you will find the electron. Maxwell’s equations, consists of a wave of electric and
magnetic fields E and B . These are vector fields that
To apply Born’s interpretation to the two slit electron at each point in space have both a magnitude and a
experiment, we do what Campbell did in the simulation direction. Since probabilities do not point anywhere,
of Figure (5). We first calculate what the wave pattern we cannot directly equate E and B to some kind of
at the screen would be for a wave passing through the probability.
two slits. It is the two slit interference pattern we have
seen for water waves, light waves and electron waves. To see how to interpret the wave nature of a photon, let
We then interpret the intensity of the pattern at some us first consider something like a radio wave or a laser
point on the screen as being proportional to the beam that contains many billions of photons. In our
probability that the electron will land at that point. We discussion of capacitors in Chapter 27, we saw that the
cannot predict where any given electron will actually energy density in a classical electric field was given by
land, any more than we can predict where the ball will
Energy ε 0E 2 energy density in
end up on the roulette wheel. But we can predict what = an electric field (27-36)
density 2
the pattern will look like after many electrons have
landed. If we repeat the experiment, the electrons will where E 2 = E ⋅ E . In an electromagnetic wave there
not land in the same places, but eventually the same two are equal amounts of energy in the electric and the
slit pattern will result. magnetic fields. Thus the energy density in a classical
electromagnetic field is twice as large as that given y
Exercise 1 Equation 27-36, and we have
Figure (36-16) reproduced here, shows the diffraction joules energy density in an
pattern produced when a beam of electrons is scat- E = ε 0E 2 electricomagneticwave
meter 3
tered by the atoms of a graphite crystal. Explain what
you would expect to see if the electrons went through If we now picture the electromagnetic wave as consist-
the graphite crystal one at a time and you could watch ing of photons whose energy is given by Einstein’s
the pattern build up on the screen. Could you market photoelectric formula
this apparatus in Las Vegas, and if so, how would you joules
use it? E photon = hf
photon
then the density of photons in the wave is given by
ε 0E 2 joules meter 3
Figure 36-16 n =
Diffraction hf joules photon
pattern
produced by ε0 E2 photons densityof photons
electrons n = in an electromagnetic (1)
hf meter3 wave of frequency f
passing
through a where f is the frequency of the wave.
graphite
crystal.
 40-7

In Exercise 2, we have you estimate the density of

photons one kilometer from the antenna of the student number ofphotons
AM radio station at Dartmouth College. The answer per cubic centimeter = number 12
at 1km
at 1 million kilometers 10
turns out to be around .25 billion photons/cc, so many
= .25 ×1210
9
photons that it would be hard to detect them individu-
ally. 10
photons
= .00025 (2)
Exercise 2 cm 3
To estimate the density of photons in a radio wave, we
can, instead of calculating E for the wave, simply use In the classical picture of Maxwell’s equations, the
the fact that we know the power radiated by the station. radio wave has a continuous electric and magnetic field
As an example, suppose that we are one kilometer away even out at 1 million kilometers. You could calculate
from a 1000 watt radio station whose frequency is the value of E and B out at this distance, and the result
1.4 × 106Hz . A 1000 watt station radiates 1000 joules would be sinusoidally oscillating fields whose struc-
of energy per second or 10– 6 joules in a nanosecond. ture is that shown back in Figure (32-23). But if you
In one nanosecond the radiated wave moves out one went out there and tried to observe something, all you
foot or about 1/3 of a meter. If we ignore spatial would find is a few photons, on the order of .25 per liter
distortions of the wave, like reflections from the ground,
(about one per gallon of space). If you look in 1 cubic
etc., then we can picture this 10– 6 joules of energy as
centimeter of space, chances are you would not find a
being located in a spherical shell 1/3 of a meter thick,
expanding out from the antenna.
photon.

(a) What is the total volume of a spherical shell 1/3 of a So how do you use Maxwell’s equations to predict the
meter thick and 1 kilometer in radius? results of an experiment to detect photons a million
kilometers from the antenna? First you use Maxwell’s
(b) What is the average density of energy, in joules/ m3 equation to calculate E at the point of interest, then
of the radio wave 1 kilometer from the antenna evaluate the quantity (ε 0 E⋅E/hf ) , and finally interpret
(c) What is the energy, in joules, of one photon of the result as the probability of finding a photon in the
frequency 1.4 × 106 Hz ? region of interest. If, for example, we were looking in
a volume of one cubic centimeter, the probability of
(d) What is the average density of photons in the radio finding a photon there would be about .00025 or
wave 1 kilometer from the station? Give the answer first .025%.
in photons/ m3 and then photons per cubic centimeter.
(The answer should be about .25 billion photons/ cm3. ) This is an explicit prescription for turning Maxwell’s
theory of electromagnetic radiation into a probability
Now imagine that instead of being one kilometer from wave for photons. If the wave is intense, as it was close
the radio station, you were a million kilometers away. to the antenna, then (ε 0 E⋅E/hf ) represents the density
Since the volume of a spherical shell 1/3 of a meter of photons. If the wave is very faint, then (ε 0 E⋅E/hf )
thick increases as r 2 , (the volume being 1/3 × 4πr 2 ) becomes the probability of finding a photon in a certain
the density of photons would decrease as 1/r 2 . Thus volume of space.
if you were 10 6 times as far away, the density of
photons would be 10 – 12 times smaller. At one million
kilometers, the average density of photons in the radio
wave would be
40-8 Quantum Mechanics

Reflection and Fluorescence The wavelength of the light from a fluorescent material
An interesting example of the probability interpreta- is not the same wavelength as the incident light. What
tion of light waves is provided by the phenomena of happens is that a photon in the incident beam strikes and
reflection and of fluorescence. excites an individual atom in the material. The excited
atom then drops back down to the ground state radiat-
When a light beam is reflected from a metal surface, the ing two or more photons to get rid of the excitation
angle of reflection, labeled θ r in Figure (7a) is equal to energy. (Ultraviolet light is often used in the incident
the angle of incidence θ i . The reason for this is seen beam, and we see the lower energy visible photons
in Figure (7b). The incident light wave is scattered by radiated from the fluorescing material.)
many atoms in the metal surface. The scattered waves
add up to produce the reflected wave as shown in The reason that fluorescent light emerges in many
Figure (7b). Any individual photon in the incident directions rather than in a reflected beam is that an
wave must have an equal probability of being scattered individual photon in the incident beam is absorbed by
by all of these atoms in order that the scattered probabil- and excites one atom in the fluorescent material. There
ity waves add up to the reflected wave shown in (7b). is no probability that it has struck any of the other
atoms. The fluorescent light is then radiated as a
When you have a fluorescent material, you see a rather circular wave from that atom, and the emerging photon
uniform eerie glow rather than a reflected wave. The has a more or less equal probability of coming out in all
light comes out in all directions as in Figure (8a). directions above the material.
angle of θi θr angle of
incidence reflection incident
beam
flourescent
glow

mirror mirror
Figure 8a
Figure 7a When a beam of light strikes a fluorescent material, we
When a light wave strikes a mirror, the angle see an eerie glow rather than a normal reflected light.
of incidence equals the angle of reflection.

incident
reflected wave photon reemitted wave

excited atom
Figure 8b
Fluorescence occurs when an individual atom is
incident wave excited and radiates its extra energy as two
distinct photons. Since there is no chance that the
radiation came from other atoms, the radiated
wave emerges only from the excited atom.
Figure 7b
The reflected wave results from the scattering of the
incident wave by many atoms. If the incident wave
contains a single photon, that photon must have an
equal probability of being scattered by many atoms
in order to emerge in the reflected wave.
 40-9

A Closer Look at the We will now repeat the experiment with a new twist.
Two Slit Experiment We ask for a volunteer to go inside the box, look at the
While the probability interpretation of electron and slits, and see which one each electron went through.
photon waves provides a reasonable explanation of John volunteers, and we give him a sheet of paper to
some phenomena, the interpretation is not without write down the results. To make the job easier, we tell
problems. To illustrate what these problems are, con- him to just look at the bottom slit on the hour to see if the
sider the following thought experiment. electron went through that slit. If for example he sees
an electron come out of the bottom slit at 9:00 AM, then
Imagine that we have a large box with two slits at one the 9:00 AM electron went through the bottom slit. If
end and a photographic film at the other, as shown in he saw no electron at 10:00 AM, then the 10:00 AM
Figure (9). Far from the slits is an electron gun that electron must have gone through the upper slit.
produces a weak beam of electrons, so weak that on the
average only one electron per hour passes through the If John does his job carefully, what kind of a pattern
slits and strikes the film. For simplicity we will assume should build up on the film after many electrons have
that the electrons go through the slits on the hour, there gone through? If the 9:00 AM electron was seen to pass
being the 9:00 AM electron, the 10:00 AM electron, etc. through the bottom slit, then there is no probability that
it went through the top slit. As a result, a probability
The electron gun is one of the simple electron guns we wave can emerge only from the bottom slit, and there
discussed back in Chapter 28. The beam is so spread out can be no cancellation of probability waves at the
that there is no way it can be aimed at one slit or the photographic film. Since the 10:00 AM electron did not
other. Our beam covers both slits, meaning that each of go through the bottom slit, the probability wave must
the electrons has an equal chance of going through the have emerged only from the top slit and there again can
top or bottom slit. be no cancellation of waves at the photographic film.
We will take the probability interpretation of the elec- If John correctly determines which slit each electron
tron wave seriously. If the electron has an equal went through, there can be no cancellation of waves
probability of passing through either slit, then an equally from the two slits, and we have to end up with a one slit
intense probability wave must emerge from both slits. pattern on the film. Just the knowledge of which slit
When the probability waves get to the photographic each electron went through has to change the two slit
film, there will be bands along which waves from one pattern into a one slit pattern. With Born’s probability
slit cancel waves from the other, and we should even- interpretation of electron waves, just the knowledge of
tually build up a two slit interference pattern on the film. which slit the electrons go through changes the result
of the experiment. Does this really happen, or have we
Suppose that on our first run of the thought experiment, entered the realm of metaphysics?
we do build up a two slit pattern after many hours and
many electrons have hit the film.

film
one slit pattern

electron slits
two slit pattern

gun

John

Figure 9
In this thought experiment, we consider the possibility that someone is
looking at the two slits to see which slit each electron comes through.
40-10 Quantum Mechanics

Let us return to our thought experiment. John has been “But,” Bill responds, “a low frequency photon is a long
in the box for a long time now, so that a number of wavelength photon. Remember that demonstration
electrons have hit the film. We take the film out, where waves were scattered from a tiny object? The
develop it, and clearly see a two slit interference pattern scattered waves were circular, and contained no infor-
emerging. There are the dark bands along which waves mation about the shape of the object (Figure 36-1). You
from one slit cancel the waves from the other slit. can’t use waves to study details that are much smaller
than the wavelength of the wave. That is why optical
Then we go over to the door on the side of the box, open microscopes can’t be used to study viruses that are
it and let John out, asking to see his results. We look at smaller than a wavelength of visible light.”
his sheet of paper and nothing is written on it. “What
were you doing all of that time,” we ask. “What do you “If John’s flashlight,” Bill continues, “produced pho-
mean, what was I doing? How could I do anything? tons whose wavelength was longer than the distance
You were so careful sealing up the box from outside between the two slits, then even if he hit the electron
disturbances that it was dark inside. I couldn’t see a with one of the photons in the wave, John could not tell
thing and just had to wait until you opened the door. which slit the electron came through.”
Not much of a fun experiment.”
“Let us do some calculations,” the professor says. “The
“Next time,” John said, “give me a flashlight so I can most delicate way we can mess up the experiment is to
see the electrons coming through the slits. Then I can hit an electron sideways, changing the electron’s direc-
fill out your sheet of paper.” tion of motion so that if it were heading toward a
maxima, it will instead land in a minima, filling up the
“Better be careful,” Jill interrupts, “about what kind of dark bands and making the pattern look like a one slit
a flashlight you give John. A flashlight produces a pattern. Here is a diagram for the situation (Figure 10).”
beam of photons, and John can only see a passing
electron if one of the flashlight’s photons bounces off “In the top sketch (10a), John is shining his flashlight
the electron.” at an electron that has just gone through the slit and is
heading toward the central maximum. In the middle
“Remember that the energy of a photon is proportional sketch (10b) the photon has knocked the electron
to its frequency. If the photons from John’s flashlight sideways, so that it is now headed toward the first
have too high a frequency, the photon hitting the minimum in the diffraction pattern. Let us assume that
passing electron will change the motion of the electron all the photon’s momentum p photon has been trans-
and mess up the two slit pattern. Give John a flashlight ferred to the electron, so that the electron’s new mo-
that produces low frequency, low energy photons, so mentum is now
he won’t mess up the experiment.”

Figure 36-1 (reproduced)

If an object is smaller than a
wavelength, the scattered
waves are circular and do
not contain information
about the shape of the object.

Incident and scattered wave After incident wave has passed

 40-11

p electron – new = p electron – old + p photon (3)

The angle θ by which the electron is deflected is

electron wave

approximately given by
pelectron
pphoton h/λ photon
photon
θ ≈ p =
electron h/λ electron
John
λ electron
θ ≈ (4)
λ photon
Figure 10a
In order to see the electron, John
where we used the de Broglie formula for the photon
uses a flashlight, and strikes the
electron with a photon. and electron momenta.”

“In the bottom sketch we have the usual analysis of a

two slit pattern. If the angle θ to the first minimum is
small, which it usually is for a two slit experiment, then
by similar triangles we have

λ
electron wave

y min
pelectron-ne
θ
w
pphoton
ymin θ ≈ = electron 2 (5)
D d
pelectron-old
Equating the values of θ from equations (5) and (4), we
get
p
θ ≈ pphoton
λ electron λ
electron-old
θ = = electron (6)
2d λ photon
Figure 10b
Assume the photon’s momentum has been “Look!” Bill says, “ λ electron cancels and we are left
absorbed by the electron. This could deflect
the electron’s path by an angle θ.
with ”

λ photon = 2d (7)

“I told you,” Jill interrupts, “that you had to be careful

D
about what wavelength photons John could use. Here
we see that if John’s photons have a wavelength of 2d
electron wave

θ ymin or less, his photons will carry enough of a punch,

θ enough momentum to destroy the two slit pattern. Be
d
sure John’s photons have a wavelength longer than 2d
λ/2 (path length difference) so that they will be incapable of knocking an electron
from a maxima to a minima.”
θ ≈ ymin ≈ λ electron /2
D d
“No way,” responds Bill. “A wavelength of 2d is
Figure 10c already too big. John cannot use photons with a
Analysis of the two slit pattern. The angle to the wavelength any greater than the slit width d if he wants
first minimum is determined by using similar to see which slit the electron went through. And you
triangles. If the angle θ is small, then sinθ ≈ θ. want him to use photons with a wavelength greater than
2d!”
40-12 Quantum Mechanics

“That’s the dilemma,” the professor replies. “If John “It’s hard to visualize, but you can use a photon picture
uses photons whose wavelength is short enough to see to explain every detail of the interaction between the
which slit the electron went through, he is likely to mess electron and proton in the hydrogen atom. That calcu-
up the experiment and destroy the two slit pattern.” lation was actually done back in 1947. The modern
view is that all electric and magnetic phenomena are
“It looks like the very act of getting information is caused by photons.”
messing up the experiment,” Jill muses.
“If all electric and magnetic phenomena are caused by
“It messes it up if we use photons,” Bill responds.” Let photons,” Jill observes, “then Bill’s capacitor plate and
us work out a better experiment where we do a more voltmeter, which uses electromagnetic phenomena, is
delicate measurement to see which slit the electron based on photons. Since photons obey the de Broglie
went through. Do the experiment so delicately that we relationship, the photons in Bill’s experiment should
do not affect the motion of the electron, but accurately have the same effect as the photons from John’s
enough to see which slit the electron went through.” flashlight. If John’s photons mess up the experiment,
“How would you do that?” Jill asks. Bill’s should too!”

“Maybe I would put a capacitor plate on one of the “I have an idea,” Bill says. “Aren’t there such a thing
slits,” Bill responds, “and record the capacitor voltage. as gravitational waves?”
If the electron went through that slit, the electric field of “Yes,” replies the professor. “They are very hard to
the electron should affect the voltage on the capacitor make, and very hard to detect. We have not been able
and leave a blip on my oscilloscope screen. If I don’t to make or detect them yet in the laboratory. But back
see a blip, the electron went through the other slit.” in the 1970’s Joe Taylor at the University of Massachu-
“Would this measurement affect the motion of the setts discovered a pair of binary neutron stars orbiting
electron?” Jill asks. about each other. Since the stars eclipse each other,
Taylor could accurately measure the orbital period.”
“I don’t see why,” Bill responds.
“According to Einstein’s theory of gravity, the orbiting
“Think about this,” the professor interrupts. “We are neutron stars should radiate gravitational waves and
now interpreting the electric and magnetic fields of a lose energy. Joe Taylor has conclusively shown that
light wave as a probability wave for photons. In this the pair of stars are losing energy just as predicted by
view, all electric and magnetic phenomena are ulti- Einstein’s theory. Taylor got the Nobel prize for this
mately caused by photons. The electric and magnetic work in 1993.”
fields we worked with earlier in the course are now to
“Is Einstein’s theory a quantum theory?” Bill asks.
be thought of as a way of describing the behavior of the
underlying photons.” “What do you mean by that?” Jill asks.
“That’s crazy,” Bill argues. “You mean, for example, “I mean,” Bill responds, “in Einstein’s theory, do
that the good old 1/r 2 Coulomb force law that holds gravitational waves have a particle wave nature like
the hydrogen atom together, is caused by photons? I electromagnetic waves? Are there particles in a gravi-
don’t see how.” tational wave like there are photons in a light wave?”

“Not in Einstein’s theory,” the professor replies.

“Einstein’s theory is strictly a classical theory. No
particles in the wave.”
 40-13

“Then if Einstein’s theory is correct,” Bill continues, “I “Could Einstein be wrong?” Jill asks. “Couldn’t
should be able to make a gravitational wave with a very gravitational waves also have a particle nature? Couldn’t
short wavelength and very little energy.” the gravitational particles also obey the de Broglie
relation?”
“Couldn’t I then use this short wavelength, low en-
ergy, gravitational wave to see which slit the electron “Perhaps,” the professor replies. “For years, physicists
went through? I would make the wavelength much have speculated that gravity should have a particle-
shorter than the slit spacing d so that there would be wave nature. They have even named the particle -- they
no doubt about which slit the electron went through. call it a graviton. One problem is that gravitons should
But I would use a very low energy, delicate wave so that be very, very hard to detect. The only way we know that
I would not affect the motion of the electron.” gravitational waves actually exist is from Joe Taylor’s
binary neutron stars. There are various experiments
“You could do that if Einstein’s theory is right,” the designed to directly observe gravitational waves, but
professor replies. no waves have yet been seen in these experiments.”
“But,” Bill responds, “that allows me to tell which slit “In the case of electromagnetism, we saw electromag-
the electron went through without destroying the two netic radiation -- i.e., light -- long before photons were
slit pattern. What happens to the probability interpre- detected in Hertz’s photoelectric effect experiment.
tation of the electron wave? If I know which slit the After gravitational waves are detected, then we will
electron went through, the probability wave must have have to do the equivalent of a photoelectric effect
come from that slit, and we must get a one slit pattern. experiment for gravity in order to see the individual
If John used gravitational waves instead of light waves gravitons. The main problem here is that the gravita-
in his flashlight, he could observe which slit the elec- tional radiation we expect to see, like that from massive
tron went through without destroying the two slit objects such as neutron stars, is very low frequency
pattern.” radiation. Thus we would be dealing with very low
“You have just stumbled upon one of the major out- energy gravitons which would be hard to detect indi-
standing problems in physics,” the professor replies. vidually.”
“As far as we know there are four basic forces in nature. “And there is another problem,” the professor contin-
They are gravity, the electromagnetic force, the weak ues, “no one has yet succeeded in constructing a
interaction, and the so-called gluon force that holds consistent quantum theory of gravity. There are math-
quarks together. I listed these in the order in which they ematical problems that have yet to be overcome. At the
were discovered.” present time, the only consistent theory of gravity we
“Now three of these forces, all but gravity, are known have is Einstein’s classical theory.”
to have a particle-wave nature like light. All the “It looks like two possibilities,” Jill says. “If the
particles obey the de Broglie relation p = h/λ .” probability interpretation of electron waves is right,
“As a result, if we perform our two slit electron then there has to be a quantum theory of gravity,
experiment, trying to see which slit the electron went gravitons have to exist. If Einstein’s classical theory is
through, and we use apparatus based on non gravita- right, then there is some flaw in the probability interpre-
tional forces, we run into the same problem we had with tation.”
John’s flashlight. The only chance we have for detect- “That is the way it stands now,” the professor replies.
ing which slit the electron went through without mess-
ing up the two slit pattern, is to use gravity.”
40-14 Quantum Mechanics

THE UNCERTAINTY PRINCIPLE moving lengths contracted, and the mass of a moving
We have just seen that, for the probability interpreta- object increased with velocity. This led to the relation-
tion of particle-waves to be a viable theory, there can be ship E = mc 2 between mass and energy, and to the
no way we can detect which slit the electron went connections between electric and magnetic fields. The
through without destroying the two slit pattern. Also simple idea that you cannot measure uniform motion
we have seen that if every particle and every force have has an enormous impact on our understanding of the
a particle wave nature obeying the de Broglie relation- way matter behaves.
ship λ = h/p , then there is no way we can tell which slit Now, with the particle-wave nature of matter, we are
the electron went through without destroying the two encountering an equally universal restriction on what
slit pattern. Both the particle-wave nature of matter, we can measure, and that restriction has an equally
and the probability interpretation of particle waves, important impact on our understanding of the behavior
lead to a basic limitation on our ability to make experi- of matter. Our discussion of the uncertainty principle
mental measurements. This basic limitation was dis- comes at the end of the text rather than at the beginning
covered by Werner Heisenberg shortly before only because it has taken a while to develop the
Schrödinger developed his wave equation for elec- concepts we need to explain this restriction. With the
trons. Heisenberg called this limitation the uncertainty principle of relativity we could rely on the student’s
principle. experience with uniform motion, clocks and meter
When you cannot do something, when there is really no sticks. For the uncertainty principle, we need some
way to do something, physicists give the failure a name understanding of the behavior of particles and waves,
and call it a basic law of physics. We began the text with and as we shall see, Fourier analysis plays an important
the observation that you cannot detect uniform motion. role.
Michaelson and Morley thought they could, repeatedly There are two forms of the uncertainty principle, one
tried to do so, and failed. This failure is known as the related to measurements of position and momentum,
principle of relativity which Einstein used as the foun- and the other related to measurements of time and
dation of his theories of relativity. Throughout the text energy. They are not separate laws, one can be derived
we have seen the impact of this simple idea. When from the other. The choice of which to use is a matter
combined with Maxwell’s theory of light, it implied of convenience. Our discussion of the two slit experi-
that light traveled at the same speed relative to all ment and the de Broglie relationship naturally leads to
observers. That implied moving clocks ran slow, the position-momentum form of the law, while Fourier
analysis naturally introduces the time-energy form.
 40-15

POSITION-MOMENTUM FORM If all forces have a particle nature, and all particles obey
OF THE UNCERTAINTY PRINCIPLE the de Broglie relationship, then the fact that we derived
Equation 10 using photons makes no difference. We
In our two slit thought experiment, in the attempt to see
have to get the same result using any particle, in any
which slit the electron went through, we used a beam
possible kind of experiment. Thus Equation 10 repre-
of photons whose momenta was related to their wave-
length by p = h/λ . The wave nature of the photon is sents a fundamental limitation on the measurement
process itself!
important because we cannot see details smaller than a
wavelength λ when we scatter waves from an object. Equation 10 is not like any formula we have previously
When we use waves of wavelength λ , the uncertainty dealt with in the text. It gives you an estimate, not an
in our measurement is at least as large as λ . Let us call exact value. Often you will see the formula written
the uncertainty in the position measurement ∆x . ∆p∆x ≥ h with h = h/2π , rather than h, appearing on
the right side. Whether you use h or h depends upon
However when we use photons to locate the electron,
how you wish to define the uncertainties ∆p and ∆x .
we are slugging the electron with particles, photons of
But it is not necessary to be too precise. The important
momentum p photon = h/λ . Since we do not know
point is that the product ∆p∆x must be at least of the
where the photons are within a distance λ , we do not
order of magnitude h. It cannot be h/100 or something
know exactly how the electron was hit and how much
smaller.
momentum it absorbed from the photon. The electron
could have absorbed the full photon momentum p photon The gist of the uncertainty principle is that the more
or none of it. If we observe the electron, we make the accurately you measure the position of the particle, the
electron’s momentum uncertain by an amount at least more you mess up the particle’s momentum. Or, the
as large as p photon . Calling the uncertainty in the more accurately you measure the momentum of a
electron’s momentum ∆p electron we have particle, the less you know about the particle’s position.
∆p electron = p photon = h = h (8) Equation (10) is not quite right, because it turns out that
λ ∆x
an accurate measurement of the x position of a particle
multiplying through by ∆x gives does not necessarily mess up the particle’s y compo-
∆p∆x = h (9) nent of momentum, only its x component. A more
accurate statement of the uncertainty principle is
In Equation 9, ∆p and ∆x represent the smallest
possible uncertainties we can have when measuring the ∆p x∆x ≥ h (11a)
position of the electron using photons. To allow for the
fact that we could get much greater uncertainties using ∆p y∆y ≥ h (11b)
poor equipment or sloppy techniques, we will write the
where ∆p x is the uncertainty in the particle’s x compo-
equation in the form
nent of momentum due to a measurement of its x
position–momentum position, and ∆p y is the uncertainty in the y component
∆p∆x ≥ h form of the (10) of momentum resulting from a y position measure-
uncertainty principle ment. The quantities ∆x and ∆y are the uncertainty in
indicating that the product of the uncertainties is at least the x and y measurements respectively.
as large as Planck’s constant h.
40-16 Quantum Mechanics

Single Slit Experiment Equation 13 tells us that the smaller ∆y , i.e., the
In our two slit thought experiment, we measured the narrower the slit, the bigger the uncertainty ∆p y we
position of the electron by hitting it with a photon. create in the particle’s y momentum. This is what
Another way to measure the position of a particle is to happens if the particle’s motion is governed by its wave
send it through a slit. For example, suppose a beam of nature.
particles impinges on a slit of width (w) as illustrated in
Figure (11). We know that any particle that makes it to In Figure (12a) we have a ripple tank photograph of a
the far side of the slit had, at one time, been within the wave passing through a moderately narrow slit. The
slit. At that time we knew its y position to within an wave on the far side of the slit is seen to spread out a bit.
uncertainty ∆y equal to the width (w) of the slit. We can calculate the amount of spread by noting that
the beam is mostly contained within the central maxi-
∆y = w (12) mum of the single slit diffraction pattern.
This is an example of a position measurement with a y
precisely known uncertainty ∆y .

According to the uncertainty principle, the particle’s y

component of momentum is uncertain by an amount
∆p y given by Equation 11b as

∆p y ≥ h = w
h (13) w
∆y

Figure 11
When a particle goes through the slit, its y position is
known to within an uncertainty ∆y = w .
 40-17

Now suppose that this wave represented a beam of In Figure (13a), we see a wave passing through a
photons or electrons. On the right side of the slit, all the narrower slit than the one in Figure (12a). With the
particles have a definite x component of momentum narrower slit, we have made a more precise measure-
p x = h/λ and no y momentum. The uncertainty in the ment of the particle’s y position. We have reduced the
y momentum is zero. uncertainty ∆y = w . According to the uncertainty
principle ∆p y ≥ h/∆y , a decrease in ∆y should in-
Once the particle’s have gone through the slit, the beam crease the uncertainty ∆p y in the particle’s y momen-
spreads out giving the particles a y momentum. tum. But an increase in ∆p y means that the beam
should spread out more, which is what it does in Figure
Since you do not know whether any given particle in
(13). In going from Figure (12) to Figure (13), we have
the beam will go straight ahead, up or down, the spread
cut the slit width about in half and about doubled the
of the beam introduces an uncertainty ∆p y in the
spread. I.e., cutting ∆y in half doubles ∆p y as
particle’s y momentum. This spread is illustrated in
expected.
Figure (12b).

ymin

Figure 12a Figure 13a

Wave picture of the single slit experiment, wide slit. Narrow the slit and the wave spreads out.

θ min
θ min ymin ∆py
pi ∆py pi

pi
pi

Figure 12b Figure 13b

Particle picture of the single slit experiment. With a narrower slit, ∆ p y increases.
40-18 Quantum Mechanics

The particles entered the slit as plane waves with only

Example 1
an x component of momentum given by de Broglie’s
We can use our analysis of the single slit pattern in
formula
Chapter 33 to show that ∆py and ∆y are related by the
uncertainty principle. We saw that when a wave goes px = h (17)
λ
through a single slit of width w, the distance ymin to the
first minimum is given by Using this formula for px in Equation 16 gives
ymin = λD
w (33-14) λ ∆py λ∆py
w = h/ λ = h (18)
where λ is the wavelength and D the distance to the
screen as shown in Figure (14a). The angle to the first The λ's cancel, and we are left with
minimum is given by
∆py w = h (19)
y λD/w λ
tan θmin = min = = w (14) But the slit width w is ∆y , the uncertainty in the y
D D
measurement, thus
∆py∆y = h (20)
There is an equal sign in Equation 20 because this
Ymin particular measurement of the y position of the particle
θ min causes the least possible uncertainty in the particle’s y
w component of momentum. (Note that the x component
of the particle’s momentum is more or less unaffected
by the slit. The wave has the same wavelength λ before
and after going through the slit. It is the y component of
D
momentum that changed from zero on the left side to
± ∆py on the right.)
Figure 14a
After the beam emerges from the slit, the momenta of Exercise 3
the particles spreads out through the same angle θmin
A microwave beam, consisting of 1.24 × 10– 4eV pho-
as indicated in Figure (14b). From that figure we have
tons impinges on a slit of width (w) as shown in Figure
∆p y 14a.
tan θmin = (15)
px
(a) What is the momentum px of the photons in the laser
beam before they get to the slit?
θ min
∆py (b) When the photon passes through the slit, their y
px position is known to an uncertainty ∆y = w, the slit width.
Before the photons get to the slit, their y momentum has
the definite value py = 0. Passing through the slit makes
Figure 14b
the photon’s y momentum uncertain by an amount ∆py .
where ∆p y represents the possible spread in the Using the uncertainty principle, calculate what the slit
particle’s y momenta. Equating values of tan θmin from width (w) must be so that ∆py is equal to the photon’s
Equation 14 and Equation 15 gives original momentum px . How does w compare with the
λ ∆py wavelength λ of the laser beam?
w = p (16)
x
(c) If ∆py becomes as large as the original momentum
px , what can you say about the wave pattern on the right
side of Figure (11)? Is this consistent with what you
know about waves of wavelength λ passing through a
slit of this width? Explain.
 40-19

TIME-ENERGY FORM OF For our example of the time-energy form of the uncer-
THE UNCERTAINTY PRINCIPLE tainty principle, we wish to consider the nature of the
The second form of the uncertainty principle, which photons in a 2 femtosecond long laser pulse. If we want
to measure the energy of the photons in such a pulse, we
perhaps has an even greater impact on our understand-
only have 2 femtoseconds to make the measurement
ing of the behavior of matter, involves the measure-
ment of the energy of a particle, and the time available because that is how long the pulse takes to go by us. In
the notation of the uncertainty principle
to make the measurement. The shorter the time
available, the less accurate the energy measurement time available to
is. If ∆E is the uncertainty in the results of our energy ∆t = 2 × 10 –15sec measure the energy
of the photons (22)
measurement, and ∆t the time we had to make the = 2 femtoseconds
in our laser pulse
measurement, then ∆E and ∆t are related by
Let us suppose that the laser produces red photons
∆E∆t ≥ h (21) whose wavelength is 6.2 × 10 – 5cm , about the wave-
length of the lasers we have been using. According to
One can derive this form of the uncertainty principle our usual formula for calculating the energy of the
from ∆p∆x ≥ h , but we can gain a better insight into photons in such a laser beam we have
the relationship by starting with an explicit example.
E photon = 12.4 × 10 – eV
–5
cm = 2 eV (23)
A device that has become increasingly important in 6.2 × 10 cm5

research, particularly in the study of fast reactions in Now let us use the uncertainty principle in the form
molecules and atoms, is the pulsed laser. The lasers we ∆E ≥ h (24)
have used in various experiments are all continuous ∆t
beam lasers. The beam is at least as long as the distance to calculate the uncertainty in any measurement we
from the laser to the wall. If we had a laser that we could would make the energy of the photons in the 2
turn on and off in one nanosecond, the pulse would be femtosecond laser beam. We have
1 foot or 30 cm long and contain
6.63 × 10 – 27erg sec
30cm ∆E ≥ h ≥
cm = 5 × 10 5wavelengths ∆t 2 × 10 – 15sec
6 × 10 – 5 (25)
wavelength ≥ 3.31 × 10 – 12ergs
Even a picosecond laser pulse which is 1000 times Converting ∆E from ergs to electron volts, we get
shorter, contains 500 wavelengths. Some of the recent
pulsed lasers can produce a pulse 500 times shorter than 3.31 × 10 – 12ergs
∆E ≥ ≈ 2 eV (26)
that, only 2 femtoseconds ( 2 × 10 – 15 seconds) long. 1.6 × 10 – 12ergs/eV
These lasers emit a pulse that is only one wavelength
The uncertainty in any energy measurement we make
long.
of these photons is as great as the energy itself! If we
try to measure the energy of these photons, we expect
the answers to range from E – ∆E = 0 eV up to
E + ∆E = 4 eV . Why does this happen? Why is the
energy of the photons in this beam so uncertain?
Fourier analysis provides the answer.
40-20 Quantum Mechanics

We can see why the energy of the photons in the 2 In Figure (16) we are looking at a waveform consisting
femtosecond pulse is so uncertain by comparing the of a single pulse. This would accurately represent the
Fourier transform of a long laser pulse with that of a output of a red laser that continuously emitted single
pulse consisting of only one wavelength. wavelength pulses spaced 16 wavelengths apart. (Re-
member that our program assumes that the wave shape
Figure (15) shows the Fourier transform of an infinitely is repeated.) From the Fourier analysis box we see that
long sine wave. You will recall that, in the design of the there is a dramatic difference between the composition
MacScope program, it is assumed that we are analyz- of a pure sine wave and of a single pulse. To construct
ing a repeated waveform. If you continuously repeat a single pulse out of sine waves, we have to add up a
the waveform seen in the upper half of the diagram, you slew of harmonics. The single pulse is more like a drum
get an infinitely long cyclic wave which is a pure sine beat while the continuous wave is more like a flute. (In
wave. (Sine waves are by definition infinitely long the appendix we show how the sine wave harmonics
waves.) In effect we have in Figure (15) selected 16 add up to produce a pulse.)
cycles of the pure sine wave, and the Fourier analysis
box shows that we have a pure 16th harmonic. This In Figure (16) we see that the dominant harmonic is still
sine wave has a definite frequency f, and if this repre- around the sixteenth, as it was for the continuous wave,
sented a laser beam, the photons in the beam would but there is a spread of harmonics from near zero up to
have a precise energy given by the formula E = hf. almost the 32nd. For a laser pulse to have this shape,
There is no uncertainty in the energy of this infinitely it must consist of frequencies ranging from near zero up
long sine wave. (It would take an infinite time ∆t to to twice the natural frequency. Each of these frequen-
make sure that the wave was infinitely long, with the cies contains photons whose energy is given by
result ∆E = h/∆t = h/∞ = 0 .) Einstein’s formula E = hf where f is the frequency of the
harmonic.

Figure 15 Figure 16
A pure sine wave has a single frequency. One cycle of a wave is made
up of a spread of harmonics
 40-21

In Figure (17) we have reproduced the Fourier analysis In Figure (18) we analyze a pulse two wavelengths
box of Figure (16), but relabeled the horizontal axis in long. Now we see that the spread of frequencies
electron volts. We have assumed that the 16th har- required to reconstruct this waveform is only half as
monic represented 2 eV photons which would be the wide, ranging from the 8th to 24th harmonic, or from
case if the wave were an infinitely long red laser pulse. 1 eV to 3 eV. We have twice as long to study a 2
Now the diagram represents the density of photons of wavelength pulse, and the uncertainty in energy ∆E is
different energies in the laser beam. While 2 eV is the only about 1 eV, or half as big.
most likely energy, there is a spread of energies ranging
from nearly 0 eV up to nearly 4 eV. If we measure the Going to a 4 wavelength pulse in Figure (19) we see that
energy of a photon in the beam, our answer is 2 eV with by doubling the time available we again cut in half the
an uncertainty of 2 eV, just as predicted by the uncer- uncertainty ∆E in energy. Now the energy varies from
tainty principle ∆E∆t ≥ h . about 1.5 eV to 2.5 eV for a ∆E = .5 eV . This is just
what you expect from ∆E∆t ≥ h . You should now
begin to see that the uncertainty principle is a simple
rule evolving from the wave nature of particles. (By the
way, it would be more accurate to write ∆E∆t = h for
this discussion, because we are describing the very
least uncertainty in energy.)
0eV 1eV 2eV 3eV 4eV
Figure 17
Photon energies in single wavelength
pulse of a red laser beam.

1eV 2eV 3eV 1eV 2eV 3eV

Figure 18 Figure 19
A two cycle wave has half the spread of harmonics. A four cycle wave has a fourth
the spread of harmonics.
40-22 Quantum Mechanics

Probability Interpretation Measuring Short Times

We have interpreted Figure (17) as representing the We have said that the new pulsed lasers produce pulses
spread in energies of the photons in a 2 femtosecond red as short as 2 femtoseconds. How do we know that?
laser pulse. What if the pulse consisted of only a single Suppose we gave you the job of measuring the length
photon? Then how do we interpret this spread in of the laser pulse, and the best oscilloscope you had
energies? The answer is that we use a probability could measure times no shorter than a nanosecond.
interpretation. The photon in the pulse has different This is a million times too slow to see a femtosecond
probabilities of having different energies. pulse. What do you do?
In our discussion of light waves, we saw that the energy If you cannot measure the time directly, you can be
density in a light wave was proportional to the square sneaky and use the uncertainty principle. Send the laser
of the amplitude of the wave. This is reasonable pulse through a diffraction grating, and record the
because while the amplitude of a wave can be positive spread in wavelengths, i.e., the spread in energies of the
or negative, the square of the amplitude, which we call photons in the pulse. If the line is very sharp, if they are
the intensity is always positive. Probabilities, like all red photons of a single wavelength and energy, then
energy densities, also have to be positive, thus we you know that there is no measurable uncertainty ∆E
should associate the probability of a photon as having in the photon energies, and the pulse must last a time
a given frequency with the intensity or square of the ∆t that is considerably longer than 2 femtoseconds. If,
amplitude of the wave of that frequency. on the other hand, the line is spread out from the near
infra red to violet, if the spread in energies is from 1 eV
In Figure (20) we show the intensities (square of the to 3 eV, and the spread is not caused by some other
amplitudes) of the harmonics that make up the single phenomena (like the Doppler effect), then from the
wavelength pulse. (This is plotted automatically by uncertainty principle you know that the pulse is only
MacScope when we click on the button labeled Φ .) about a femtosecond long. (You know, for example, it
We see that squaring the amplitudes narrows the spread. cannot be as long as 10 femtoseconds, or as short as a
tenth of a femtosecond.)
Figure (20) has the following interpretation when
applied to pulses containing a single photon. If we Thus, with the uncertainty principle, you can use a
measure the energy of the photon, we are most likely to diffraction grating rather than a clock or oscilloscope to
get an answer close to 2 eV but there is a reasonable measure very short times. Instead of being an annoying
probability of getting an answer lower than 1 eV or restriction on our ability to make experimental mea-
even higher than 3 eV. The heights of the bars tell us surements, the uncertainty principle can be turned into
the relative probability of measuring that energy for the an important scientific tool for measuring short times
photon. and, as we shall see, short distances.

amplitudes of Exercise 4
the harmonics
An electron is in an excited state of the hydrogen atom,
either the second energy level at -3.40 eV or the third
energy level at -1.51 eV. You wish to do an experiment
to decide which of these two states the electron is in.
What is the least amount of time you must take to make
this measurement?
intensities of
the harmonics
(amplitude
squared)

1eV 2eV 3eV

Figure 20
Intensities of the harmonics are proportional
to the square of the amplitudes.
 40-23

Short Lived Elementary Particles rest mass energy of the Λ (1520). Thus if we collide K –
We usually think of the rest energy of a particle as particles with protons, we expect to create a Λ (1520)
having a definite value. For example the rest energy of particle only if the incoming particles have the right
a proton is 938.2723 × 10 6eV . The proton itself is a total energy.
composite particle made of 3 quarks, and the number
938.2723 MeV represents the total energy of the Figure (22) shows the results of some collision experi-
quarks in the allowed wave pattern that represents a ments, where a K – meson and a proton collided to
proton. This rest energy has a very definite value produce a Λ and two π mesons. The probability of such
because the proton is a stable particle with plenty of a result peaked when the energy of the incoming
time to settle into a precise wave pattern. particles was 1,520 MeV. This peak occurred because
the incoming K – meson and proton created a Λ (1520)
A rather different particle is the so called “Λ (1520)”, particle, which then decayed into the Λ and two π
which is another combination of 3 quarks, but very mesons, as shown in Figure (21). The Λ (1520) was not
short lived. The name comes partly from the fact that observed directly, because its lifetime is too short.
the particle’s rest mass energy is about 1520 million
electron volts (MeV). As indicated in Figure (21), a Figure (22) shows that the energy of the incoming
Λ(1520) can be created as a result of the collision particles does not have to be exactly 1520 MeV in order
between a K – meson and a proton. to create a Λ (1520). The peak is in the range from about
1510 to 1530 Mev, which implies that the rest mass
We are viewing the collision in a special coordinate energy of the Λ (1520) is 1520 MeV plus or minus
system, where the total momentum of the incoming about 10 MeV. From one experiment to another, the
particles is zero. In this coordinate system, the resulting rest mass energy can vary by about 20 MeV. (The
Λ (1520) will be at rest. By conservation of energy, the experimentalists quoted a variation of 16 MeV.)
total energy of the incoming particles should equal the

p p
K
-
p+

a) A K – meson and a proton are about to collide. We are

looking at the collision in a coordinate system where the
total momentum is zero (the so called “center of mass”
system).
1460 1480 1500 1520 1540 1560
Λ center of mass energy, MeV
1520
−
K + p+ Λ(1520) Λ+ π ++π −
b) In the collision a Λ (1520) particle is created.
It is at rest in this center of mass system
Figure 22
π– The probability that a K – meson and a proton collide
to produce a Λ particle, and two π mesons peak at an
Λ energy of 1520 MeV. The peak results from the fact
π+ that a Λ (1520) particle was created and quickly
decayed into the Λ and two π mesons. The
c) The Λ (1520) then quickly probability peaks at 1520 MeV, but can be seen to
decays into a lower energy Λ particle and two π mesons. spread out over a range of about 16MeV. The small
circles are experimental values, the vertical lines
Figure 21 represent the possible error in the value. (Data from
A Λ (1520) particle can be created if the total energy M.B. Watson et al., Phys. Rev. 131(1963).)
(in the center of mass system) of the incoming particles
equals the rest mass energy of the Λ (1520).
40-24 Quantum Mechanics

Why isn’t the peak sharp? Why does the rest mass THE UNCERTAINTY PRINCIPLE
energy of the Λ (1520) particle vary by as much as 16 AND ENERGY CONSERVATION
to 20 MeV from one experiment to another? The
The fact that for short times the energy of a particle is
answer lies in the fact that the lifetime of the Λ (1520)
uncertain, raises an interesting question about basic
is so short, that the particle does not have enough time
physical laws like the law of conservation of energy. If
to establish a definite rest mass energy. The 16 MeV a particle’s energy is uncertain, how do we know that
variation is the uncertainty ∆E in the particles rest mass
energy is conserved in some process involving that
energy that results from the fact that the particle’s
particle? The answer is -- we don’t.
lifetime is limited.
One way to explain the situation is to say that nature will
The uncertainty principle relates the uncertainty in cheat if it can get away with it. Energy does not have
energy ∆E to the time ∆t available to establish that
to be conserved if we cannot do an experiment to
energy. To establish the rest mass energy, time ∆t
demonstrate a lack of conservation of energy.
available is the particle’s lifetime. Thus we can use the
uncertainty principle to estimate the lifetime of the Consider the process shown in Figure (23). It shows a
Λ (1520) particle. With ∆E × ∆t ≈ h we get red, 2 eV photon traveling along in space. Suddenly the
6.63 × 10 – 27erg sec photon creates an positron-electron pair. The rest mass
∆t ≈ h = energy of both the positron and the electron are .51
∆E erg
16 MeV × 1.6×10 – 6 MeV. Thus we have a 2 eV photon creating a pair of
MeV
particles whose total energy is 1.02 × 10 6eV , a huge
∆t ≈ 2.6 × 10 – 22 seconds (27) violation of the law of conservation of energy. A short
time later the electron and positron come back together,
The lifetime of the Λ (1520) particle is of the order of annihilate, leaving behind a 2 eV photon. This is an
10 – 22 seconds ! This is only about 10 times longer equally huge violation of the conservation of energy.
than it takes light to cross a proton! Only by using the
uncertainty principle could we possibly measure such But have we really violated the conservation of energy?
short times. During its lifetime, the positron-electron pair is a com-
posite object whose total energy is uncertain. If the pair
lived a long time, its total energy would be close to the
expected energy of 1.02 × 10 6eV . But suppose the
pair were in existence only for a very short time ∆t,a
time so short that the uncertainty in the energy could be
as large as 1.02 × 10 6eV . Then there is some probabil-
ity that the energy of the pair might be only 2 eV and the
process shown in Figure (2) could happen.
tron oton
elec 2 eV ph
oton
2 eV ph
p o s it r o n
Figure 23
Consider a process where a 2 eV photon
suddenly creates a positron-electron pair. A short
time later the pair annihilates, leaving a 2 eV
photon. In the long range, energy is conserved.
 40-25

The length of time ∆t that the pair could exist and have QUANTUM FLUCTUATIONS
an energy uncertain by 1.02 MeV is AND EMPTY SPACE
6.63 × 10 – 27 erg sec We began the text with a discussion of the principle of
∆t = h = erg
∆E 1.02 MeV × 1.6× 10 – 6 relativity—that you could not detect your own motion
MeV relative to empty space. The concept of empty space
∆t = 4 × 10 – 21sec (32) seemed rather obvious—space with nothing in it. But
Another way to view the situation is as follows. Sup- the idea of empty space is not so obvious after all.
pose the pair in Figure (20) lasted only 4 × 10 – 21 With the discovery of the cosmic background radiation,
seconds or less. Even if the pair had an energy of we find that all the space in this universe is filled with
1.02 × 10 6eV , the lifetime is so short that any mea- a sea of photons left over from the big bang. We can
surement of the energy of the pair would be uncertain accurately measure our motion relative to this sea of
by at least 1.02 × 10 6eV , and the experiment could photons. The earth is moving relative to this sea at a
not detect the violation of the law of conservation of velocity of 600 kilometers per second toward the Vergo
energy. In this point of view, if we cannot perform an cluster of galaxies. While this measurement does not
experiment to detect a violation of the conservation violate the principle of relativity, it is in some sense a
law, then the process should have some probability of measurement of our motion relative to the universe as
occurring. a whole.
Does a process like that shown in Figure (23) actually Empty space itself may not be empty. Consider a
occur? If so, is there any way that we can know that it process like that shown in Figure (24) where a photon,
does? The answer is yes, to both questions. It is an electron, and a positron are all created at some point
possible to make extremely accurate studies of the in space. A short while later the three particles come
energy levels of the electron in hydrogen, and to make back together with the positron and electron annihilat-
equally accurate predictions of the energy using the ing and the photon being absorbed.
theory of quantum electrodynamics. We can view the
binding of the electron in hydrogen as resulting from One’s first reaction might be that such a process is
the continual exchange of photons between the elec- ridiculous. How could these three particles just appear
tron and proton. During this continual exchange, there and then disappear? To do this we would have to violate
is some probability that the photon creates a positron both the laws of conservation of energy and momen-
electron pair that quickly annihilates as shown in tum.
Figure (23). In order to predict the correct values of the
hydrogen energy levels, the process shown in Figure But, of course, the uncertainty principle allows us to do
(23) has to be included. Thus we have direct experi- that. We can, in fact, use the uncertainty principle to
mental evidence that for a short time the particle estimate how long such an object could last. The
antiparticle pair existed. arguments would be similar to the ones we used in the
analysis of the process shown in Figure (23).
electron

photon

p o s itr o n
Figure 24
Quantum fluctuation. The uncertainty principle
allows such an object to suddenly appear, and
then disappear.
40-26 Quantum Mechanics

In the theory of quantum electrodynamics, a com- The answer lies in the potential energy of the gluon
pletely isolated process like that shown in Figure (24) force that holds the quarks together. As we have seen
does not affect the energy levels of the hydrogen atom many times, the potential energy of an attractive force
and should be undetectable in electrical measurements. is negative. In this case the potential energy of the
But such a process might affect gravity. A gravitational gluon force is almost as big in magnitude as the rest
wave or a graviton might interact with the energy of mass of the quarks, reducing the total energy from
such an object. Some calculations have suggested that 1100 MeV to 140 MeV.
such interactions could show up in Einstein’s classical
theory of gravity as a correction to the famous cosmo- Suppose we had an object whose negative potential
logical constant we discussed in Chapter 21. energy was as large as the positive rest mass energy.
Imagine, for example, that the object consisted of a
An object like that shown in Figure (23) is an example collection of point sized elementary particles so close
of what one calls a quantum fluctuation. Here we have together that their negative gravitational potential en-
something that appears and disappears in so-called ergy was the same magnitude as the positive rest mass
empty space. If such objects can keep appearing and and kinetic energy. Suppose such a collection of
disappearing, then we have to revise our understanding particles were created in a quantum fluctuation. How
of what we mean by empty. long could the fluctuation last?
The uncertainty principle allows us to tell the difference Since such an object has no total energy, the violation
between a quantum fluctuation and a real particle. A ∆E of energy conservation is zero, and therefore the
quantum fluctuation like that in Figure (24) violates lifetime ∆t = h/∆E could be forever.
conservation of energy, and therefore cannot last very
long. A real particle can last a long time because energy Suppose the laws of physics required that such a
conservation is not violated. fluctuation rapidly expand, greatly increasing both the
positive rest mass and kinetic energy, while maintain-
However, there is not necessarily that much difference ing the corresponding amount of negative gravitational
between a real object and a quantum fluctuation. To see potential energy. As long as ∆E remained zero, the
why, let us take a closer look at the π meson. The π + expanding fluctuation could keep on going. Perhaps
is a particle with a rest mass energy of 140 MeV, that such a fluctuation occurred 14 billion years ago and we
consists of a quark-antiquark pair. The quark in that live in it now.
pair is the so-called up quark that has a rest mass of
roughly 400 MeV. The other is the antidown quark that
has a rest mass of about 700 MeV. (Since we can’t get
at isolated quarks, the quark rest masses are estimates,
but should not be too far off). Thus the two quarks
making up the π meson have a total rest mass of about
1100 MeV. How could they combine to produce a
particle whose rest mass is only 140 MeV?
 40-27

APPENDIX
HOW A PULSE IS FORMED
FROM SINE WAVES

Figure A1
By selecting more and more harmonics, you can
see how the sine waves add up to produce a pulse.
Chapter on ^ CHAPTER ON GEOMETRICAL OPTICS

Geometrical Optics
For over 100 years, from the time of Newton and To see how wave phenomena can be explained by ray
Huygens in the late 1600s, until 1801 when Thomas tracing, consider the reflection of a light wave by a
Young demonstrated the wave nature of light with his metal surface. When a wave strikes a very small object,
two slit experiment, it was not clear whether light an object much smaller than a wavelength, a circular
consisted of beams of particles as proposed by Newton, scattered wave emerges as shown in the ripple tank
or was a wave phenomenon as put forward by Huygens. photograph of Figure (36-1) reproduced here. But
The reason for the confusion is that almost all common when a light wave impinges on a metal surface consist-
optical phenomena can be explained by tracing light ing of many small atoms, represented by the line of dots
rays. The wavelength of light is so short compared to in Figure (36-2), the circular scattered waves all add
the size of most objects we are familiar with, that light up to produce a reflected wave that emerges at an angle
rays produce sharp shadows and interference and of reflection θr equal to the angle of incidence θi .
diffraction effects are negligible. Rather than sketching the individual crests and troughs
incident wave of the incident wave, and adding up all the scattered
waves, it is much easier to treat the light as a ray that
reflected from the surface. This ray is governed by the
Figure 36-1 law of reflection, namely θr = θi .
An incident
wave passing angle of θi θr angle of
over a small incidence reflection
Light ray
object produces
reflected
a circular
from a
scattered wave.
mirror.

mirror

reflected wave
angle of θi θr angle of
incidence reflection

incident wave
Figure 36-2
Reflection of light. In the photograph, we see an incoming plane wave scattered by a small object. If the
object is smaller than a wavelength, the scattered waves are circular. When an incoming light wave strikes
an array of atoms in the surface of a metal, the scattered waves add up to produce a reflected wave that
comes out at an angle of reflection θ r equal to the angle of incidence θ i .
Optics-2

The subject of geometrical optics is the study of the In most introductory texts, geometrical optics appears
behavior of light when the phenomena can be ex- after Maxwell’s equations and theory of light. There is
plained by ray tracing, where shadows are sharp and a certain logic to this, first introducing a basic theory
interference and diffraction effects can be neglected. for light and then treating geometrical optics as a
The basic laws for ray tracing are extremely simple. At practical application of the theory. But this is clearly
a reflecting surface θr = θi , as we have just seen. When not an historical approach since geometrical optics
a light ray passes between two media of different was developed centuries before Maxwell’s theory. Nor
indexes of refraction, as in going from air into glass or is it the only logical approach, because studying lens
air into water, the rule is n1 sin θ 1 = n2 sin θ 2 , where systems teaches you nothing more about Maxwell’s
n1 and n2 are constants called indices of refraction, equations than you can learn by deriving Snell’s law.
and θ1 and θ2 are the angles that the rays made with Geometrical optics is an interesting subject full of
the line perpendicular to the interface. This is known wonderful applications, a subject that can appear
as Snell’s law. anywhere in an introductory physics course.

This entire chapter is based on the two rules θr = θi We have a preference not to introduce geometrical
and n1 sin θ 1 = n2 sin θ 2 . These rules are all that are optics after Maxwell’s equations. With Maxwell’s
needed to understand the function of telescopes, micro- theory, the student is introduced to the wave nature of
scopes, cameras, fiber optics, and the optical compo- one component of matter, namely light. If the focus is
nents of the human eye. You can understand the kept on the basic nature of matter, the next step is to look
operation of these instruments without knowing any- at the photoelectric effect and the particle nature of
thing about Newton’s laws, kinetic and potential en- light. You then see that light has both a particle and a
ergy, electric or magnetic fields, or the particle and wave nature, which opens the door to the particle-wave
wave nature of matter. In other words, there is no nature of all matter and the subject of quantum me-
prerequisite background needed for studying geo- chanics. We have a strong preference not to interrupt
metrical optics as long as you accept the two rules this focus on the basic nature of matter with a long and
which are easily verified by experiment. possibly distracting chapter on geometrical optics.
 Optics-3

REFLECTION FROM CURVED SURFACES dicular or normal to the part of the ellipse we are
The Mormon Tabernacle, shown in Figure (1), is drawing. The best way to see that the angles θ i and θ r
constructed in the shape of an ellipse. If one stands at are always equal is to construct your own ellipse and
one of the focuses and drops a pin, the pin drop can be measure these angles at various points along the curve.
heard 120 feet away at the other focus. The reason why If a sound wave were emitted from focus 1 in Figure (2),
can be seen from Figure (2), which is similar to Figure the part of the wave that traveled over to point A on the
(8-28) where we showed you how to draw an ellipse ellipse would be reflected at an angle θ r equal to the
with a pencil, a piece of string, and two thumbtacks. angle of incidence θ i , and travel over to focus 2. The
The thumbtacks are at the focuses, and the ellipse is part of the sound wave that struck point B on the ellipse,
drawn by holding the string taut as shown. As you would be reflected at an angle θ r equal to it’s angle of
move the pencil point along, the two sections of string incidence θ i , and also travel over to focus 2. If you
always make equal angles θ i and θ r to a line perpen- think of the sound wave as traveling out in rays, then all
the rays radiated from focus 1 end up at focus 2, and that
is why you hear the whisper there. We say that the rays
are focused at focus 2, and that is why these points are
called focuses of the ellipse. (Note also that the path
lengths are the same, so that all the waves arriving at
focus 2 are in phase.)
pencil point

θi str
Mormon Tabernacle under construction, 1866. θr nor ing
ma

(2) (1)
l

θi
θr
B

Figure 2
Drawing an ellipse using a string and two thumbtacks.
Mormon Tabernacle finished, 1871.

Mormon Tabernacle today. Figure 2a

Figure 1 A superposition of the top half of Figure 2 on Figure 1.
Optics-4

The Parabolic Reflection If parallel rays enter a deep dish parabolic mirror from
You make a parabola out of an ellipse by moving one an angle off axis as shown in Figure (4b), the rays do not
of the focuses very far away. The progression from a focus to a point, with the result that an off axis star
parabola to an ellipse is shown in Figure (3). For a true would appear as a blurry blob. (This figure corresponds
parabola, the second focus has to be infinitely far away. to looking at a star 2.5° off axis, about 5 moon diameters
from the center of the field of view.)
Suppose a light wave were emitted from a star and
traveled to a parabolic reflecting surface. We can think
of the star as being out at the second, infinitely distant,
focus of the parabola. Thus all the light rays coming in parallel rays coming
in from infinity
from the star would reflect from the parabolic surface
and come to a point at the near focus. The rays from the
star approach the reflector as a parallel beam of rays,
thus a parabolic reflector has the property of focusing
parallel rays to a point, as shown in Figure (4a).
focus

circle

ellipse parabolic
reflector

Figure 4a
Parallel rays, coming down the axis
focuses of the parabola, focus to a point.
ellipse

off axis
parallel rays

focus is
not good
other focus focus
at infinity
parabola

deep
parabolic
reflector
Figure 3
Evolution of an ellipse into a parabola. For a Figure 4b
parabola, one of the focuses is out at infinity. For such a deep dish parabola, rays coming
in at an angle of 2.5° do not focus well.
 Optics-5

One way to get sharp images for parallel rays coming As we can see in Figure (4d), light coming from two
in at an angle is to use a shallower parabola as illustrated different stars focus at two different points in what is
in Figure (4c). In that figure, the focal length (distance called the focal plane of the mirror. If you placed a
from the center of the mirror to the focus) is 2 times the photographic film at the focal plane, light from each
mirror diameter, giving what is called an f 2 mirror. In different star, entering as parallel beams from different
Figure (4d), you can see that rays coming in at an angle angles, would focus at different points on the film, and
of 2.5° (blue lines) almost focus to a point. Typical you would end up with a photographic image of the
amateur telescopes are still shallower, around f 8, stars. This is how distant objects like stars are photo-
which gives a sharp focus for rays off angle by as much graphed with what is called a reflecting telescope.
as 2° to 3°.

light from star on axis

f 2 mirror
Figure 4c
A shallow dish is made by using only the shallow bottom of the parabola. Here the focal length is twice the
diameter of the dish, giving us an f2 mirror. Typical amateur telescopes are still shallower, having a focal
length around 8 times the mirror diameter (f8 mirrors). [The mirror in Figure 4b, that gave a bad focus, was
f.125, having a focal length 1/8 the diameter of the mirror.]

light from star #2, 2.5° off

axis

focus #1
f 2 mirror

focus #2

light from star #1, on axis

Figure 4d
We can think of this drawing as representing light coming in from a red star at the center of the field of
view, and a blue star 2.5° (5 full moon diameters) away. Separate images are formed, which could be
recorded on a photographic film. With this shallow dish, the off axis image is sharp (but not quite a point).
Optics-6

MIRROR IMAGES the top and to the bottom of the eye. You can see that
The image you see in a mirror, although very familiar, if you started at the eye and drew the rays back as
is still quite remarkable in its reality. Why does it look straight lines, ignoring the mirror, the rays would
so real? You do not need to know how your eye works intersect at the old source point A as shown by the
to begin to see why. dotted lines in Figure (5b).

Consider Figure (5a) where light from a point source To the eye (or a camera) at point B, there is no
reaches your eye. We have drawn two rays, one from detectable difference between Figures (5a) and (5b). In
the source to the top of the eye, and one to the bottom. both cases, the same rays of light, coming from the
In Figure (5b), we have placed a horizontal mirror as same directions enter the eye. Since the eye has no way
shown and moved the light source a distance h above of telling that the rays have been bent, we perceive that
the mirror equal to the distance it was below the mirror the light source is at the image point A rather than at
before the mirror was inserted. Using the rule that the the source point A′.
angle of incidence equals the angle of reflection, we When we look at an extended object, its image in the
again drew two rays that went from the light source to mirror does not look identical to the object itself. In
eye Figure (6), my granddaughter Julia is holding her right
hand in front of a mirror and her left hand off to the side.
The image of the right hand looks like the left hand. In
B particular, the fingers of the mirror image of the right
hand curl in the opposite direction from those of the
right hand itself. If she were using the right hand rule to
find the direction of the angular momentum of a
rotating object, the mirror image would look as if she
were using a left hand rule.
point source It is fairly common knowledge that left and right are
A reversed in a mirror image. But if left and right are
Figure 5a reversed, why aren’t top and bottom reversed also?
Light from a point source reaching your eye. Think about that for a minute before you go on to the
next paragraph.
eye
A' point source
B
h

mirror
h

mirror image
A
Figure 5b
There is no difference when the source is at point A, Figure 6
or at point A’ and the light is reflected in a mirror. The image of the right hand looks like a left hand.
 Optics-7

To see what the image of an extended object should be, The Corner Reflector
imagine that we place an arrow in front of a mirror as When two vertical mirrors are placed at right angles as
shown in Figure (7). We have constructed rays from shown in Figure (8a), a horizontal ray approaching the
the tip and the base of the arrow that reflect and enter mirrors is reflected back in the direction from which it
the eye as shown. Extending these rays back to the came. It is a little exercise in trigonometry to see that
image, we see that the image arrow has been reversed this is so. Since the angle of incidence equals the angle
front to back. That is what a mirror does. The mirror of reflection at each mirror surface, we see that the
image is reversed front to back, not left to right or top angles labeled θ 1 must be equal to each other and the
to bottom. It turns out that the right hand, when same for the angles θ 2 . From the right triangle ABC,
reversed front to back as in its image in Figure (6), has we see that θ 1 + θ 2 = 90°. We also see that the angles
the symmetry properties of a left hand. If used to define θ 2 + θ 3 also add up to 90°, thus θ 3 = θ 1 , which
angular momentum, you would get a left hand rule. implies the exiting ray is parallel to the entering one.
eye
If you mount three mirrors perpendicular to each other
to form the corner of a cube, then light entering this so
called corner reflector from any angle goes back in the
direction from which it came. The Apollo II astronauts
placed the array of corner reflectors shown in Figure
(8b) on the surface of the moon, so that a laser beam
mirror
from the earth would be reflected back from a precisely
known point on the surface of the moon. By measuring
the time it took a laser pulse to be reflected back from
the array, the distance to the moon could be measured
Figure 7
A mirror image changes front to back, not left to right. to an accuracy of centimeters. With the distance to the
moon known with such precision, other distances in the
solar system could then be determined accurately.
B A
θ1 θ1
θ2
C
θ2 θ3

Figure 8a Figure 8b
With a corner reflector, the light is reflected back Array of corner reflectors left on the moon by the
it the same direction from which it arrived. Apollo astronauts. A laser pulse from the earth,
aimed at the reflectors, returns straight back to the
laser. By measuring the time the pulse takes to go
to the reflectors and back, the distance to that point
on the moon and back can be accurately measured.
Optics-8

MOTION OF LIGHT THROUGH A MEDIUM light, the scattered waves are circular like those in the
We are all familiar with the fact that light can travel ripple tank photograph. The final wave is the sum of the
through clear water or clear glass. With some of the incident and the scattered waves as shown in Figure
new glasses developed for fiber optics communication, (36-1a).
light signals can travel for miles without serious distor- When light passes through a medium like glass or
tion. If you made a mile thick pane from this glass you water, the wave is being scattered by a huge number of
could see objects through it. atoms. The final wave pattern is the sum of the incident
From an atomic point of view, it is perhaps surprising wave and all of the many billions of scattered waves.
that light can travel any distance at all through water or You might suspect that this sum would be very com-
glass. A reasonable picture of what happens when a plex, but that is not the case. At the surface some of the
light wave passes over an atom is provided by the ripple incident wave is reflected. Inside the medium, the
tank photograph shown in Figure (36-1) reproduced incident and scattered waves add up to a new wave of
here. The wave scatters from the atom, and since atoms the same frequency as the incident wave but which
are considerably smaller than a wavelength of visible travels at a reduced speed. The speed of a light wave
in water for example is 25% less than the speed of light
in a vacuum.
incident wave incident wave

a) Incident and scattered wave together. b) After incident wave has passed.
Figure 36-1
If the scattering object is smaller than a wavelength, we get circular scattered waves.
 Optics-9

The optical properties of lenses are a consequence of Index of Refraction

this effective reduction in the speed of light in the lens. The amount by which the effective speed of light is
Figure (9) is a rather remarkable photograph of indi- reduced as the light passes through a medium depends
vidual short pulses of laser light as they pass through both upon the medium and the wavelength of the light.
and around a glass lens. You can see that the part of the There is very little slowing of the speed of light in air,
wave front that passed through the lens is delayed by its about a 25% reduction in speed in water, and nearly a
motion through the glass. The thicker the glass, the 59% reduction in speed in diamond. In general, blue
greater the delay. You can also see that the delay light travels somewhat slower than red light in nearly all
changed the shape and direction of motion of the wave media.
front, so that the light passing through the lens focuses
It is traditional to describe the slowing of the speed of
to a point behind the lens. This is how a lens really
light in terms of what is called the index of refraction of
works.
the medium. The index of refraction n is defined by the
equation

speed of light c
vlight = (1)
in a medium n

The index n has to equal 1 in a vacuum because light

always travels at the speed 3 × 10 8 meters in a vacuum.
The index n can never be less than 1, because nothing
can travel faster than the speed c. For yellow sodium
light of wavelength λ = 5.89 × 10 – 5 cm (589 nanom-
eters), the index of refraction of water at 20° C is
n = 1.333, which implies a 25% reduction in speed. For
diamond, n = 2.417 for this yellow light. Table 1 gives
the indices of refraction for various transparent sub-
stances for the sodium light.

Figure 9 Vacuum 1.00000 exactly

Motion of a wave front through a glass lens. The delay Air (STP) 1.00029
in the motion of the wave front as it passes through the Ice 1.309
glass changes the shape and direction of motion of the Water (20° C) 1.333
wave front, resulting in the focusing of light. (This Ethyl alcohol 1.36
photograph should not be confused with ripple tank Fuzed quartz 1.46
photographs where wavelengths are comparable to the
Sugar solution (80%) 1.49
size of the objects. Here the wavelength of the light is
about one hundred thousand times smaller than the Typical crown glass 1.52
diameter of the lens, with the result we get sharp Sodium Chloride 1.54
shadows and do not see diffraction effects.) Polystyrene 1.55
Heavy flint glass 1.65
Sapphire 1.77
Zircon 1.923
Diamond 2.417
In the 18/February/1999 issue of Nature it was Rutile 2.907
announced that a laser pulse travelled through a gas Gallium phosphide 3.50
of supercooled sodium atoms at a speed of 17 meters Very cold sodium atoms 18000000 for laser pulse
per second! (You can ride a bicycle faster than that.)
This means that the sodium atoms had an index of Table 1
refraction of about 18 million, 7.3 million times Some indices of refraction for yellow sodium light at a
greater than that of diamond! wavelength of 589 nanometers.
Optics-10

Exercise 1a In the muon lifetime picture, one observed how long

What is the speed of light in air, water, crown glass, and
muons lived when stopped in a block of plastic. The
diamond. Express your answer in feet/nanosecond. experiment was made possible by Cerenkov radiation.
(Take c to be exactly 1 ft/nanosecond.) The muons that stopped in the plastic, entered moving
faster than the speed of light in plastic, and as a result
Exercise 1b emitted a flash of light in the form of Cerenkov
In one of the experiments announced in Nature, a laser radiation. When the muon decayed, a charged positron
pulse took 7.05 microseconds to travel .229 millimeters and a neutral neutrino were emitted. In most cases the
through the gas of supercooled sodium atoms. What charged positron emerged faster than the speed of light
was the index of refraction of the gas for this particular in the plastic, and also emitted Cerenkov radiation. The
experiment? (The index quoted on the previous page two flashes of light were detected by the phototube
was for the slowest observed pulse. The pulse we are which converted the light flashes to voltage pulses. The
now considering went a bit faster.) voltage pulses were then displayed on an oscilloscope
screen where the time interval between the pulses could
be measured. This interval represented the time that the
CERENKOV RADIATION muon lived, mostly at rest, in the plastic.
In our discussion, in Chapter 1, of the motion of light
through empty space, we saw that nothing, not even
information, could travel faster than the speed of light.
If it did, we could, for example, get answers to ques-
tions that had not yet been thought of.

When moving through a medium, the speed of a light

wave is slowed by repeated scattering and it is no longer
true that nothing can move faster than the speed of light
in that medium. We saw for example that the speed of
light in water is only 3/4 the speed c in vacuum. Many
elementary particles, like the muons in the muon
lifetime experiment, travel at speeds much closer to c.
When a charged particle moves faster than the speed of
light in a medium, we get an effect not unlike the sonic Figure 33-30
boom produced by a supersonic jet. We get a shock When the source of the waves moves faster than the
wave of light that is similar to a sound shock wave speed of the waves, the wave fronts pile up to produce
(sonic boom), or to the water shock wave shown in a shock wave as shown. This shock wave is the sonic
boom you hear when a jet plane flies overhead faster
Figure (33-30) reproduced here. The light shock wave
than the speed of sound.
is called Cerenkov radiation after the Russian physi-
cist Pavel Cerenkov who received the 1958 Nobel prize
for discovering the effect.
 Optics-11

SNELL’S LAW water has two depths, deeper on the upper part where
When a wave enters a medium of higher index of the waves travel faster, and shallower in the lower part
refraction and travels more slowly, the wavelength of where the waves travel more slowly. You can see that
the wave changes. The wavelength is the distance the the wavelengths are shorter in the lower part, but there
wave travels in one period, and if the speed of the wave are the same number of waves. (We do not gain or loose
is reduced, the distance the wave travels in one period waves at the boundary.) The frequency, the number of
is reduced. (In most cases, the frequency or period of waves that pass you per second, is the same on the top
the wave is not changed. The exceptions are in and bottom.
fluorescence and nonlinear optics where the frequency The only way that the wavelength can be shorter and
or color of light can change.) still have the same number of waves is for the wave to
We can calculate how the wavelength changes with bend at the boundary as shown. We have drawn arrows
wave speed from the relationship showing the direction of the wave in the deep water (the
cm incident wave) and in the shallow water (what we will
vwave
cm sec call the transmitted or refracted wave), and we see that
λ = the change in wavelength causes a sudden change in
cycle sec
T direction of motion of the wave. If you look carefully
cycle
you will also see reflected waves which emerge at an
Setting vwave = c/n for the speed of light in the me- angle of reflection equal to the angle of incidence.
dium, gives for the corresponding wavelength λ n
vwave λ0 Figure (11) shows a beam of yellow light entering a
c/n 1c
λn = = = = (2) piece of glass. The index of refraction of the glass is
T T nT n 1.55, thus the wavelength of the light in the glass is only
where λ 0 = c/T is the wavelength in a vacuum. Thus, .65 times as long as that in air ( n ≈ 1 for air). You can
for example, the wavelength of light entering a dia- see both the bending of the ray as it enters the glass and
mond from air will be shortened by a factor of 1/2.42. also the reflected ray. (You also see internal reflection
and the ray emerging from the bottom surface.) You
What happens when a set of periodic plane waves goes
cannot see the individual wave crests, but otherwise
from one medium to another is illustrated in the ripple
Figures (10) and (11) show similar phenomena.
tank photograph of Figure (10). In this photograph, the
REFLECTED

T ED
EC
FL
RE
INCIDENT TRA INCIDENT
NSM TR
ITTE AN
D SM
(REF (R ITT
EF ED
RAC RA
TED CT
) ED
)

Figure 10 Figure 11
Refraction at surface of water. When the waves enter Refraction at surface of glass. When the light waves
shallower water, they travel more slowly and have a enter the glass, they travel more slowly and have a
shorter wavelength. The waves must travel in a shorter wavelength. Like the water waves, the light
different direction in order for the crests to match up. waves must travel in a different direction in order for
the crests to match up.
Optics-12

Derivation of Snell’s Law Since the triangles ACB and ADB are right triangles in
To calculate the angle by which a light ray is bent when Figure (13a), we have
it enters another medium, consider the diagram in
λ 1 = AB sin θ 1 = λ 0 /n 1 (3)
Figure (12). The drawing represents a light wave,
traveling in a medium of index n 1, incident on a λ 2 = AB sin θ 2 = λ 0 /n 2 (4)
boundary at an angle θ 1. We have sketched successive
where AB is the hypotenuse of both triangles and λ 0 is
incident wave crests separated by the wavelength λ 1.
the wavelength when n 0 = 1. When we divide Equa-
Assuming that the index n 2 in the lower medium is
tion 4 by Equation 5, the distances AB and λ 0 cancel,
greater than n 1, the wavelength λ 2 will be shorter than
and we are left with
λ 1 and the beam will emerge at the smaller angle θ 2.
sin θ 1 n2
To calculate the angle θ 2 at which the transmitted or =
refracted wave emerges, consider the detailed section sin θ 2 n1
of Figure (12) redrawn in Figure (13a). Notice that we or
have labeled two apparently different angles by the
same label θ 1. Why these angles are equal is seen in the n 1 sin θ 1 = n 2 sin θ 2 Snell's law (5)
construction of Figure (13b) where we see that the
angles α and θ 1 are equal. Equation 5, known as Snell’s law, allows us to calculate
the change in direction when a beam of light goes from
Exercise 2 one medium to another.
Show that the two angles labeled θ2 in Figure (13a)

in
must also be equal.

cid e
wa

en
v

t
C
d i cide

λ1 θ1
Figure 13a
re
in

c t nt w

The angles θ1
io

A
n av

involved in
λ 2 θ2
of e

B
the analysis.
θ2 D
C θ1 ref e
λ1
wa
rac
A
v
ted

λ2 B
D
tra

θ2
nsm
itte
dw
ave

Figure 13b
Detail. θ1
α β

Figure 12 α + β = 90°
Analysis of refraction. The crests must match at the θ1 + β = 90°
boundary between the different wavelength waves. α = θ1
 Optics-13

INTERNAL REFLECTION It is easy to calculate the critical angle θ c at which this

Because of the way rays bend at the interface of two complete internal reflection begins. Set the angle of
media, there is a rather interesting effect when light refraction, θ 2 in Figure (14), equal to 90° and we get
goes from a material of higher to a material of lower from Snell’s law
index of refraction, as in the case of light going from n 1 sinθ c = n 2 sinθ 2 = n 2 sin90° = n 2
water into air. The effect is seen clearly in Figure (14).
Here we have a multiple exposure showing a laser n2 n2
beam entering a tank of water, being reflected by a sinθ c = ; θ c = sin –1 (6)
mirror, and coming out at different angles. The outgo- n1 n1
ing ray is bent farther away from the normal as it
emerges from the water. We reach the point where the For light emerging from water, we have n 2 ≈ 1 for air
outgoing ray bends and runs parallel to the surface of and n 1 = 1.33 for water giving
the water. This is a critical angle, for if the mirror is 1
sin –1 θ 2 = = .75
turned farther, the ray can no longer get out and is 1.33
completely reflected inside the surface.
θ c = 48.6° (7)
Anyone who swims underwater, scuba divers espe-
cially, are quite familiar with the phenomenon of
internal reflection. When you look up at the surface of
the water, you can see the entire outside world through
a circular region directly overhead, as shown in Figure
(14a). Beyond this circle the surface looks like a silver
mirror.

Exercise 3
A glass prism can be used as shown in Figure (15) to
Figure 14 reflect light at right angles. The index of refraction ng of
Internal reflection. We took three exposures of a the glass must be high enough so that there is total
laser beam reflecting off an underwater mirror set at internal reflection at the back surface. What is the least
different angles. In the first case the laser beam value ng one can have to make such a prism work?
makes it back out of the water and strikes a white (Assume the prism is in the air where n ≈ 1 .)
cardboard behind the water tank. In the other two
cases, there is total internal reflection at the under
side of the water surface. In the final exposure we
used a flash to make the mirror visible.

Figure 15
Right angled prism. The
.6°
45°
48
index of refraction of the
glass has to be high enough
to cause total internal
reflection.
diver looking up

Figure 14a
When you are swimming under water and look up,
you see the outside world through a round hole.
Outside that hole, the surface is a silver mirror.
Optics-14

Fiber Optics much slower over commercial telephone lines, not

Internal reflection plays a critical role in modern com- much in excess of 30 to 50 thousand bits of information
munications and modern medicine through fiber op- per second (corresponding to 30 to 50 kilobaud).
tics. When light is sent down through a glass rod or These rates are fast enough to carry telephone conver-
fiber so that it strikes the surface at an angle greater than sations or transmit text to a printer, but painfully slow
the critical angle, as shown in Figure (16a), the light for sending pictures and much too slow for digital
will be completely reflected and continue to bounce television signals. High definition digital television
down the rod with no loss out through the surface. By will require that information be sent at a rate of about
using modern very clear glass, a fiber can carry a light 3 million bits or pulses every 1/30 of a second for a baud
signal for miles without serious attenuation. rate of 90 million baud. (Compare that with the baud
rate on your computer modem.) In contrast, fiber optics
The reason it is more effective to use light in glass fibers cables are capable of carrying pulses or bits at a rate of
than electrons in copper wire for transmitting signals, about a billion ( 10 9 ) per second, and are thus well
is that the glass fiber can carry information at a much suited for transmitting pictures or many phone conver-
higher rate than a copper wire, as indicated in Figure sations at once.
(16b). This is because laser pulses traveling through
glass, can be turned on and off much more rapidly than By bundling many fine fibers together, as indicated in
electrical pulses in a wire. The practical limit for Figure (17), one can transmit a complete image along
copper wire is on the order of a million pulses or bits of the bundle. One end of the bundle is placed up against
information per second (corresponding to a baud rate the object to be observed, and if the fibers are not mixed
of one megabit). Typically the information rate is up, the image appears at the other end.

To transmit a high resolution image, one needs a bundle

Figure 16a of about a million fibers. The tiny fibers needed for this
Because of internal are constructed by making a rather large bundle of
reflections, light can small glass strands, heating the bundle to soften the
travel down a glass glass, and then stretching the bundle until the indi-
fiber, even when the vidual strands are very fine. (If you have heated a glass
fiber is bent.
rod over a Bunsen burner and pulled out the ends, you
have seen how fine a glass fiber can be made this way.)

Figure 17
A bundle of glass fibers
can be used to carry an
image from one point to
another. The order of a
million fibers are needed
to carry the medical
Figure 16b images seen on the next
A single glass fiber can carry the same amount page.
of information as a fat cable of copper wires.
 Optics-15

Medical Imaging PRISMS

The use of fiber optics has revolutionized many aspects So far in our discussion of refraction, we have consid-
of medicine. It is an amazing experience to go down ered only beams of light of one color, one wavelength.
and look inside your own stomach and beyond, as the Because the index of refraction generally changes with
author did a few years ago. This is done with a flexible wavelength, rays of different wavelength will be bent
fiber optics instrument called a retroflexion, producing at different angles when passing the interface of two
the results shown in Figure (18). An operation, such as media. Usually the index of refraction of visible light
the removal of a gallbladder, which used to require increases as the wavelength becomes shorter. Thus
opening the abdomen and a long recovery period, can when white light, which is a mixture of all the visible
now be performed through a small hole near the navel, colors, is sent through a prism as shown in Figure (20),
using fiber optics to view the procedure. You can see the short wavelength blue light will be deflected by a
the viewing instrument and such an operation in progress greater angle than the red light, and the beam of light is
in Figure (19). separated into a rainbow of colors.

flexible optical
fiber viewing stomach
scope θ (initia
l)
white

you are
here

duodenum n(red) = 1.516

n(yellow) = 1.522
n(green) = 1.525

re
d
Figure 18 n(blue) = 1.529

blu
Close-up view of the θ(initial) = 30.2°

e
author taken by
photographer Dr. Figure 20
Richard Rothstein. When light is sent through a prism, it is separated into
a rainbow of colors. In this scale drawing, we find that
almost all the separation of colors occurs at the second
surface where the light emerges from the glass.

Figure 19
Gallbladder operation in progress, being viewed by the
rigid laparoscope shown on the right. Such views are
now recorded by high resolution television.
Optics-16

Rainbows What is more important than the fact that the maximum
Rainbows in the sky are formed by the reflection and angle of deviation is 42° is the fact that the rays close to
refraction of sunlight by raindrops. It is not, however, #11 emerge as more or less parallel to each other. The
particularly easy to see why a rainbow is formed. René other rays, like those near #3 for example come out at
Descartes figured this out by tracing rays that enter and diverging angles. That light is spread out. But the light
leave a spherical raindrop. emerging at 42° comes out as a parallel beam. When
you have sunlight striking many raindrops, more yel-
In Figure (21a) we have used Snell’s law to trace the low light is reflected back at this angle of 42° than any
path of a ray of yellow light that enters a spherical drop other angle.
of water (of index n = 1.33), is reflected on the back
side, and emerges again on the front side. (Only a 13
12
11
fraction of the light is reflected at the back, thus the 10
9 42 degrees
reflected beam is rather weak.) In this drawing, the 8
7

angle θ 2 is determined by sin θ 1 = 1.33 sin θ 2 .

6
5
4
3
At the back, the angles of incidence and reflection are 2
1
equal, and at the front we have 1.33 sin θ 2 = sin θ 1 2
(taking the index of refraction of air = 1). Nothing is 3
hard about this construction, it is fairly easy to do with 4
a good drafting program like Adobe Illustrator and a 5
hand calculator.
8

In Figure (21b) we see what happens when a number of 7

13
parallel rays enter a spherical drop of water. (This is Figure 21b
6
similar to the construction that was done by Descartes Light from ray 11 comes out at
in 1633.) When you look at the outgoing rays, it is not 9 the maximum angle of 42°.
Nearby rays come out at nearly
immediately obvious that there is any special direction 10 the same angle, producing a
for the reflected rays. But if you look closely you will 11 parallel beam at an angle of 42°.
12
see that the ray we have labeled #11 is the one that
comes back at the widest angle from the incident ray.

Ray #1, through the center, comes straight back out.

Ray #2 comes out at a small angle. The angles increase
up to Ray #11, and then start to decrease again for Rays
#12 and #13. In our construction the maximum angle,
that of Ray #11, was 41.6°, close to the theoretical value red
of 42° for yellow light.
42° yellow
θ1 blue
θ2
θi
Figure 21a θr
Light ray n = 1.33 42°
reflecting from ang
le of s
a raindrop. un
θ'2
Figure 21c
You will see the yellow part of the rainbow at
θ'1 an angle of 42° as shown above. Red will be
seen at a greater angle, blue at a lesser one.
 Optics-17

Repeat the construction for red light where the index of The Green Flash
refraction is slightly less than 1.33, and you find the The so called green flash at sunset is a phenomenon that
maximum angle of deviation and the direction of the is supposed to be very rare, but which is easy to see if
parallel beam is slightly greater than 42°. For blue light, you can look at a distant sunset through binoculars.
with a higher index, the deviation is less. (Don’t look until the very last couple of seconds so that
you will not hurt your eyes.)
If you look at falling raindrops with the sun at your back
as shown in Figure (21c), you will see the yellow part The earth’s atmosphere acts as a prism, refracting the
of the rainbow along the arc that has an angle of 42° light as shown in Figure (22). The main effect is that
from the rays of sun passing you. The red light, having when you look at a sunset, the sun has already set; only
a greater angle of deviation will be above the yellow, its image is above the horizon. But, as seen in Figure
and the blue will be below, as you can see in Figure (20), the atmospheric prism also refracts the different
(21d). colors in the white sunlight at different angles. Due to
the fact that the blue light is refracted at a greater angle
Sometimes you will see two or more rainbows if the than the red light, the blue image of the sun is slightly
rain is particularly heavy (we have seen up to 7). These higher above the horizon than the green image, and the
are caused by multiple internal reflections. In the green image is higher than the red image. We have over
second rainbow there are two internal reflections and emphasized the displacement of the image in Figure
the parallel beam of yellow light comes out at an angle (22). The blue image is only a few percent of the sun’s
of 51°. Because of the extra reflection the red is on the diameter above the red image. Before the sun sets, the
inside of the arc and the blue on the outside. various colored images are more or less on top of each
other and the sun looks more or less white.
Exercise 4
Next time you see a rainbow, try to measure the angle If it is a very clear day, and you watch the sunset with
the yellow part of the arc makes with the rays of sun binoculars, just as the sun disappears, for about 1/2
passing your head. second, the sun turns a deep blue. The reason is that all
the other images have set, and for this short time only
this blue image is visible. We should call this the “blue
flash”.
blue image greatly exaggerated
separation of
green image sun's images
red image

sun earth

Figure 22
The green flash. You can think
of the white sun as consisting of various
colored disks that add up to white. The earth’s
Figure 21d atmosphere acts as a prism, diffracting the light from the
Rainbow over Cook’s Bay, Moorea. setting sun, separating the colored disks. The blue disk is
the last to set. Haze in the atmosphere can block the blue
light, leaving the green disk as the last one seen.
Optics-18

If the atmosphere is not so clear, if there is a bit of haze LENSES

or moisture as one often gets in the summer, the blue The main impact geometrical optics has had on man-
light is absorbed by the haze, and the last image we see kind is through the use of lenses in microscopes,
setting is the green image. This is the origin of the green telescopes, eyeglasses, and of course, the human eye.
flash. With still more haze you get a red sunset, all the The basic idea behind the construction of a lens is
other colors having been absorbed by the haze. Snell’s law, but as our analysis of light reflected from
Usually it requires binoculars to see the green or blue a spherical raindrop indicated, we can get complex
colors at the instant of sunset. But sometimes the results from even simple geometries like a sphere.
atmospheric conditions are right so that this final light Modern optical systems like the zoom lens shown in
of the sun is reflected on clouds and can be seen without Figure (25) are designed by computer. Lens design is
binoculars. If the clouds are there, there is probably an ideal problem for the computer, for tracing light rays
enough moisture to absorb the blue image, and the through a lens system requires many repeated applica-
resulting flash on the clouds is green. tions of Snell’s law. When we analyzed the spherical
raindrop, we followed the paths of 12 rays for an index
Halos and Sun Dogs of refraction for only yellow light. A much better
Another phenomenon often seen is the reflection of analysis would have resulted from tracing at least 100
light from hexagonal ice crystals in the atmosphere. rays for the yellow index of refraction, and then repeat-
The reflection is seen at an angle of 22° from the sun. ing the whole process for different indices of refraction,
If the ice crystals are randomly oriented then we get a corresponding to different wavelengths or colors of
complete halo as seen in Figure (23a). If the crystals are light. This kind of analysis, while extremely tedious to
falling with their flat planes predominately horizontal, do by hand, can be done in seconds on a modern
we only see the two pieces of the halo at each side of the desktop computer.
sun, seen in Figure (24). These little pieces of rainbow
In this chapter we will restrict our discussion to the
are known as “sun dogs”.
simplest of lens systems in order to see how basic
Figure 23
instruments, like the microscope, telescope and eye,
Halo caused function. You will not learn here how to design a color
by reflection corrected zoom lens like the Nikon lens shown below.
by randomly
oriented
hexagonal ice
crystals.

Figure 24
Sun dogs
caused by
ice crystals
falling flat.

Figure 25
Nikon zoom lens.
 Optics-19

Spherical Lens Surface In Figure (26b) we have restricted the area where the
A very accurate spherical surface on a piece of glass is rays are allowed to enter to a small region around the
surprisingly easy to make. Take two pieces of glass, put center of the surface. To a very good approximation all
a mixture of grinding powder and water between them, these parallel rays come together, focus, at one point.
rub them together in a somewhat regular, somewhat This is the characteristic we want in a simple lens, to
irregular, pattern that one can learn in less than 5 bring parallel incoming rays together at one point as the
minutes. The result is a spherical surface on the two parabolic reflector did.
pieces of glass, one being concave and the other being
convex. The reason you get a spherical surface from Figure (26b) shows us that the way to make a good lens
this somewhat random rubbing is that only spherical using spherical surfaces is to use only the central part
surfaces fit together perfectly for all angles and rota- of the surface. Rays entering near the axis as in Figure
tions. Once the spheres have the desired radius of (26b) are deflected only by small angles, angles where
curvature, you use finer and finer grits to smooth out the we can approximate sin θ by θ itself. When the
scratches, and then jeweler’s rouge to polish the sur- angles of deflection are small enough to use small angle
faces. With any skill at all, one ends up with a polished approximations, a spherical surface provides sharp
surface that is perfectly spherical to within a fraction of focusing. As a result, in analyzing the small angle
a wavelength of light. spherical lenses, we can replace the exact form of
Snell’s law
To see the optical properties of a spherical surface, we n 1 sin θ 1 = n 2 sin θ 2 (5 repeated)
can start with the ray diagram we used for the spherical
raindrop, and remove the reflections by extending the by the approximate equation
refracting medium back as shown in Figure (26a). The Snell's law
result is not encouraging. The parallel rays entering n1 θ1 = n2 θ2 for small (8)
near the center of the surface come together—focus— angles
quite a bit farther back than rays entering near the outer
edge. This range of focal distances is not useful in
optical instruments.

Figure 26a Figure 26b

Focusing properties of a spherical surface. (Not good!) We get a much better focus if we use only a small part
of the spherical surface.
Optics-20

Focal Length of a Spherical Surface Now consider the two triangles reproduced in Figures
Let us now use the simplified form of Snell’s law to (27b) and (27c). Using the small angle approximation
calculate the focal length f of a spherical surface, i.e., tan θ ≈ sin θ ≈ θ , we have for Figure (27b)
the distance behind the surface where entering parallel h h
θ1 ≈ ; α ≈ (11)
rays come to a point. Unless you plan to start making r f
your own lenses, you do not really need this result, but Substituting these values for θ 1 and α into Equation
the exercise provides an introduction to how focal 10 gives
lengths are related to the curvature of lenses. h n1 h h
= + (12)
Consider two parallel rays entering a spherical surface r n2 r f
as shown in Figure (27). One enters along the axis of The height h cancels, and we are left with
the surface, the other a distance h above it. The angle
labeled θ 1 is the angle of incidence for the upper ray, 1 1 n1
= 1–
while θ 2 is the refracted angle. These angles are related f r n2 (13)
by Snell’s law
The fact that the height h cancels means that parallel
n1 θ1 = n2 θ2 rays entering at any height h (as long as the small angle
or approximation holds) will focus at the same point a
n1 distance f behind the surface. This is what we saw in
θ2 = θ
n2 1 (9) Figure (26b).
If you recall your high school trigonometry you will Figure (26b) was drawn for n 1 = 1 (air) and n 2 = 1.33
remember that the outside angle of a triangle, θ 1 in (water) so that n 1 /n 2 = 1/1.33 = .75 . Thus for that
Figure (27a), is equal to the sum of the opposite angles, drawing we should have had
θ 2 and α in this case. Thus
1 1 1 1 1
θ1 = θ2 + α = 1 – .75 = .25 =
f r r r 4
or using Equation 9 for θ 2 or
n1 f = 4r (14)
θ1 = θ +α (10)
n2 1 as the predicted focal length of that surface.
n1
θ1 n2

Figure 27 θ2
parallel rays h
Calculating the θ1 α
focal length f of a r
spherical surface. f
r (radius of sphere)

Figure 27a θ2
θ1 = θ2 + α
h θ1 α
θ1
r
Figure 27b Figure 27c
θ1 ≈ h /r α ≈ h /f h
α
f
 Optics-21

Exercise 5 We get rainbows from raindrops and prisms because

Compare the prediction of Equation 14 with the results
the index of refraction for most transparent substances
we got in Figure (26b). That is, what do you measure for changes with wavelength. As we saw in Exercise 6,
the relationship between f and r in that figure? this causes red light to focus at a different point than
yellow or blue light, (resulting in colored bands around
Exercise 6 the edges of images). This problem is called chromatic
The index of refraction for red light in water is slightly less aberration. The cure for chromatic aberration is to
than the index of refraction for blue light. Will the focal construct complex lenses out of materials of different
length of the surface in Figure (26b) be longer or shorter indices of refraction. With careful design, you can
than the focal length for red light? bring the focal points of the various colors back to-
gether. Some of the complexity in the design of the
Exercise 7 zoom lens in Figure (25) is to correct for chromatic
The simplest model for a fixed focus eye is a sphere of aberration.
index of refraction n2 . The index n2 is chosen so that
parallel light entering the front surface of the sphere Astigmatism is a common problem for the lens of the
focuses on the back surface as shown in Figure (27d). human eye. You get astigmatism when the lens is not
What value of n2 is required for this model to work when perfectly spherical, but is a bit cylindrical. If, for
n1 =1 ? Looking at the table of indexes of refraction, example, the cylindrical axis is horizontal, then light
Table 1, explain why such a model would be hard to from a horizontal line will focus farther back than light
achieve. from a vertical line. Either the vertical lines in the
image are in focus, or the horizontal lines, but not both
at the same time. (In the eye, the cylindrical axis does
Aberrations
not have to be horizontal or vertical, but can be at any
When parallel rays entering a lens do not come to focus angle.)
at a point, we say that the lens has an aberration. We
saw in Figure (26a) that if light enters too large a region There can be many other aberrations depending upon
of a spherical surface, the focal points are spread out in what distortions are present in the lens surface. We
back. This is called spherical aberration. One cure for once built a small telescope using a shaving mirror
spherical aberration is to make sure that the diameter of instead of a carefully ground parabolic mirror. The
any spherical lens you use is small in comparison to the image of a single star stretched out in a line that covered
radius of curvature of the lens surface. an angle of about 30 degrees. This was an extreme
example of an aberration called coma. That telescope
provided a good example of why optical lenses and
mirrors need to be ground very accurately.
n1 = 1 What, surprisingly, does not usually cause a serious
n2 = ?
problem is a small scratch on a lens. You do not get an
image of the scratch because the scratch is completely
out of focus. Instead the main effect of a scratch is to
scatter light and fog the image a bit.

Figure 27d
A simple, but hard to achieve, model for an eye.
Optics-22

Perhaps the most famous aberration in history is the

spherical aberration in the primary mirror of the orbit-
ing Hubble telescope. The aberration was caused by an
undetected error in the complex apparatus used to test
the surface of the mirror while the mirror was being
ground and polished. The ironic part of the story is that
the aberration could have easily been detected using the
same simple apparatus all amateur telescope makers
use to test their mirrors (the so called Foucault test), but
such a simple minded test was not deemed necessary.

What saved the Hubble telescope is that the engineers

found the problem with the testing apparatus, and could
therefore precisely determine the error in the shape of
the lens. A small mirror, only a few centimeters in
diameter, was designed to correct for the aberration in
the Hubble image. When this correcting mirror was
inserted near the focus of the main mirror, the aberra-
tion was eliminated and we started getting the many
fantastic pictures from that telescope.

Another case of historical importance is the fact that

Issac Newton invented the reflecting telescope to avoid
the chromatic aberration present in all lenses at that
time. With a parabolic reflecting mirror, all parallel
rays entering the mirror focus at a point. The location
of the focal point does not depend on the wavelength of
the light (as long as the mirror surface is reflecting at
that wavelength). You also do not get spherical aber-
ration either because a parabolic surface is the correct
shape for focusing, no matter how big the diameter of
the mirror is compared to the radius of curvature of the
surface.

Figure 28
Correction of the Hubble telescope mirror. Top: before
the correction. Bottom: same galaxy after correction.
Left: astronauts installing correction mirror.
 Optics-23

THIN LENSES In Figure (29c), we have moved the two spherical

In Figure (29), we look at what happens when parallel surfaces close together to form what is called a thin
rays pass through the two spherical surfaces of a lens. lens. We have essentially eliminated the distance the
The top diagram (a) is a reproduction of Figure (26b) light travels between surfaces. If the index of refraction
where a narrow bundle of parallel rays enters a new outside the lens is 1 and has a value n inside, and
medium through a single spherical surface. By making surfaces have radii of curvature r 1 and r 2 , then the
the diameter of the bundle of rays much less than the focal length f of the lens given by the equation
radius of curvature of the surface, the parallel rays all
focus to a single point. We were able to calculate where 1 1 1 lens maker 's
= n–1 + (15)
this point was located using small angle approxima- f r1 r2 equation
tions.
Equation 15, which is known as the lens maker’s
In Figure (29b), we added a second spherical surface.
equation, can be derived in a somewhat lengthy exer-
The diagram is drawn to scale for indices of refraction
cise involving similar triangles.
n = 1 outside the gray region and n = 1.33 inside, and
using Snell’s law at each interface of each ray. (The Unless you are planning to grind your own lenses, the
drawing program Adobe Illustrator allows you to do lens maker’s equation is not something you will need
this quite accurately.) The important point to note is to use. When you buy a lens, you specify what focal
that the parallel rays still focus to a point. The differ- length you want, what diameter the lens should be, and
ence is that the focal point has moved inward. whether or not it needs to be corrected for color
aberration. You are generally not concerned with how
the particular focal length was achieved—what combi-
nation of radii of curvatures and index of refraction
were used.

(a) Exercise 8
(a) See how well the lens maker’s equation applies to
our scale drawing of Figure (29c). Our drawing was
done to a scale where the spherical surfaces each had
a radius of r1 = r2 = 37mm , and the distance f from the
(b) center of the lens to the focal point was 55 mm.

(b) What would be the focal length f of the lens if it had

been made from diamond with an index of refraction
n = 2.42?
(c)

Figure 29
A two surface lens. Adding a second surface still leaves
the light focused to a point, as long as the diameter of
the light bundle is small compared to the radii of the
lens surfaces.
Optics-24

The Lens Equation ray, which goes through the center of the lens, is
What is important in the design of a simple lens system undeflected if the lens is thin. The bottom ray goes
is where images are formed for objects that are different through the focal point in front of the lens, and therefore
distances from the lens. Light from a very distant object must come out parallel to the axis behind the lens.
enters a lens as parallel rays and focuses at a distance (Lenses are symmetric in that parallel light from either
equal to the focal length f behind the lens. To locate side focuses at the same distance f from the lens.) The
the image when the object is not so far away, you can image is formed where the three rays from the tip
either use a simple graphical method which involves a merge. To locate the image, you only need to draw two
tracing of two or three rays, or use what is called the lens of these three special rays.
equation which we will derive shortly from the graphi-
cal approach. Exercise 9
(a) Graphically locate the image of the object in Figure
For our graphical work, we will use an arrow for the (31).
object, and trace out rays coming from the tip of the
arrow. Where the rays come back together is where the (b) A ray starts out from the tip of the object in the
image is formed. We will use the notation that the direction of the dotted line shown. Trace out this ray
object is at a distance (o) from the lens, and that the through the lens and show where it goes on the back
image is at a distance (i) as shown in Figure (30). side of the lens.

In Figure (30) we have located the image by tracing In Exercise 9, you found that, once you have located the
three rays from the tip of the object. The top ray is image, you can trace out any other ray from the tip of
parallel to the axis of the lens, and therefore must cross the object that passes through the lens, because these
the axis at the focal point behind the lens. The middle rays must all pass through the tip of the image.

focal
length
f f

image
object
object f
o i
object distance image distance

Figure 30
Locating the image using ray tracing. Three rays are
easy to draw. One ray goes straight through the center
of the lens. The top ray, parallel to the axis, intersects Figure 31
the axis where parallel rays would focus. A ray going Locate the image of the arrow, and then trace the ray
through the left focus, comes out parallel to the axis. starting out in the direction of the dotted line.
The image of the arrow tip is located where these rays
intersect.
 Optics-25

There is a very, very simple relationship between the In Figure (32b) we have selected one of the triangles
object distance o, the image distance i and the lens focal that appears in Figure (32a). The triangle starts at the
length f. It is tip of the object, goes parallel to the axis over to the
image, and then down to the tip of the image. The
1 1 1 the lens length of the triangle is (o + i) and the height of the base
+ = equation
(16)
o i f is (A+B). The lens cuts this triangle to form a smaller
similar triangle whose length is o and base is (A). The
Equation 16 is worth memorizing if you are going to do ratio of the base to length of these similar triangles must
any work with lenses. It is the equation you will use all be equal, giving
the time, it is easy to remember, and as you will see
now, the derivation requires some trigonometry you A = A+B ⇒ A+B = o + i (17)
are not likely to remember. We will take you through o o+i A o
the derivation anyway, because of the importance of
the result. In Figure (32c) we have selected another triangle which
starts where the top ray hits the lens, goes parallel to the
In Figure (32a), we have an object of height A that axis over to the image, and down to the tip of the image.
forms an inverted image of height B. We located the This triangle has a length i and a base of height (A+B)
image by tracing the top ray parallel to the axis that as shown. This triangle is cut by a vertical line at the
passes through the focal point behind the lens, and by focal plane, giving a smaller similar triangle of length
tracing the ray that goes through the center of the lens. f and base (A) as shown. The ratio of the length to base
of these similar triangles must be equal, giving
i A A+B A+B i
o = ⇒ = (18)
f i A f
A A Combining Equations 17 and 18 gives
i o+i i
B = = 1+ (19)
(a) f f o o
Finally, divide both sides by i and we get
1 1 1 lens
= + equation
(16)
f i o
o i
which is the lens equation, as advertised.
A Note that the lens equation is an exact consequence of
A+B
the geometrical construction shown back in Figure
(b) (30). There is no restriction about small angles. How-
ever if you are using spherical lenses, you have to stick
to small angles or the light will not focus to a point.
i

A
A+B
f
(c)

Figure 32
Derivation of the lens equation.
Optics-26

Negative Image Distance Negative Focal Length

The lens equation is more general than you might and Diverging Lenses
expect, for it works equally well for positive and In Figure (33) we got a virtual image by moving the
negative distances and focal lengths. Let us start by object inside the focal length. Another way to get a
seeing what we mean by a negative image distance. virtual image is to use a diverging lens as shown in
Writing Equation 15 in the form Figure (34). Here we have drawn the three special rays,
but the role of the focal point is reversed. The ray
1 1 1 through the center of the lens goes through the center as
= – (16a)
i f o before. The top ray parallel to the axis of the lens
let us see what happens if 1/o is bigger than 1/f so diverges outward as if it came from the focal point on
that i turns out to be negative. If 1/o is bigger than 1/f, the left side of the lens. The ray from the tip of the object
that means that o is less than f and we have placed the headed for the right focal point, comes out parallel to
object within the focal length as shown in Figure (33). the axis. Extending the diverging rays on the right,
back to the left side, we find a virtual image on the left
When we trace out two rays from the tip of the image, side.
we find that the rays diverge after they pass through the
lens. They diverge as if they were coming from a point You get diverging lenses by using concave surfaces as
behind the object, a point shown by the dotted lines. In shown in Figure (34). In the lens maker’s equation,
this case we have what is called a virtual image, which 1 1 1 lensmaker 's
is located at a negative image distance (i). This = n–1 + equation
(15)
f r1 r2
negative image distance is correctly given by the lens
equation (16a). you replace 1/r by – 1/r for any concave surface. If
1/f turns out negative, then you have a diverging lens.
(We will not drag you through another geometrical Using this negative value of f in the lens equation (with
proof of the lens equation for negative image distances. f = – f ) we get
It should be fairly convincing that just when the image
distance becomes negative in the lens equation, the 1 1 1
= – + (16b)
geometry shows that we switch from a real image on i f o
the right side of the lens to a virtual image on the left.)
This always gives a negative image distance i, which
means that diverging lenses only give virtual images.
f virtual
virtual
image image

o
i object

i
Figure 33 o
When the object is located within the focal length, we f
get a virtual image behind the object.
Figure 34
A diverging lens always gives a virtual image.
 Optics-27

Exercise 10 As an example, suppose we have rays converging to a

point, and we insert a diverging lens whose negative
You have a lens making machine that can grind sur-
focal length f = – f is equal to the negative object
faces, either convex or concave, with radius of curva-
tures of either 20 cm or 40 cm, or a flat surface. How
distance o = – o as shown in Figure (37). The lens
many different kinds of lenses can you make? What is equation gives
the focal length and the name of the lens type for each 1 1 1 1 1 1 1
lens? Figure (35) shows the names given to the various = – = – = – (20)
i f o – f – o o f
lens types.
If f = o , then 1/i = 0 and the image is infinitely far
away. This means that the light emerges as a parallel
Negative Object Distance beam as we showed in Figure (37).
With the lens equation, we can have negative image
distances and negative focal lengths, and also negative positive object
distance
object distances as well. o
In all our drawings so far, we have drawn rays coming rays emerging
from real
out of the tip of an object located at a positive object object
distance. A negative object distance means we have a object
lens
virtual object where rays are converging toward the tip
of the virtual object but don’t get there. A comparison
of the rays emerging from a real object and converging
toward a virtual object is shown in Figure (36). The
converging rays (which were usually created by some o
other lens) can be handled with the lens equation by rays converging negative
on a virtual object
assuming that the distance from the lens to the virtual object distance
object is negative. Figure 36
Positive and negative object distances.

negative focal length

f
bi-convex bi-concave planar-convex

o
planar-concave meniscus meniscus negative object distance
convex concave
Figure 35
Various lens types. Note that eyeglasses are Figure 37
usually meniscus convex or meniscus concave. Negative focal length.
Optics-28

Multiple Lens Systems In Figure (38b), we moved the second lens up to within
Using the lens equation, and knowing how to handle 8 cm of the first lens, so that the first image now falls
both positive and negative distances and focal lengths, behind the second lens. We now have a negative object
you can design almost any simple lens system you distance
want. The idea is to work your way through the system, o2 = D – i1 = 8 cm – 23.33 cm = – 15.33 cm
one lens at a time, where the image from one lens
becomes the object for the next. We will illustrate this Using this negative object distance in the lens equation
process with a few examples. gives
1 1 1 1 1
As our first example, consider Figure (38a) where we = – = –
i2 f2 o2 12 –15.33
have two lenses of focal lengths f1 = 10 cm and
f2 = 12 cm separated by a distance D = 40 cm. An 1 1 1
object placed at a distance o1 = 17.5 cm from the first = + =
12 15.33 6.73
lens creates an image a distance i1 behind the first lens.
Using the lens equation, we get i2 = 6.73 cm (24)
1 1 1 1 1
i1 = – = – = (21) In the geometrical construction we find that the still
f1 o1 10 17.5 23.33
inverted image is in fact located 6.73 cm behind the
i1 = 23.33 cm second image.

the same distance we got graphically in Figure (38a). While it is much faster to use the lens equation than
trace rays, it is instructive to apply both approaches for
This image, which acts as the object for the second lens a few examples to see that they both give the same
has an object distance result. In drawing Figure (38b) an important ray was
the one that went from the tip of the original object,
o2 = D – i1 = 40 cm – 23.33 cm = 16.67 cm
down through the first focal point. This ray emerges
This gives us a final upright image at a distance i2 from the first lens traveling parallel to the optical axis.
The ray then enters the second lens, and since it was
given by
parallel to the axis, it goes up through the focal point of
1 1 1 1 1 1 the second lens as shown. The second image is located
= – = – = (22)
i2 f2 o2 12 16.67 42.86 by drawing the ray that passes straight through the
second lens, heading for the tip of the first image.
i2 = 42.86 cm (23) Where these two rays cross is where the tip of the final
image is located.
which also accurately agrees with the geometrical i1
i2
construction. o1 i2
D = 40cm o2
o1 object
i1 o2

object image f1 f2

–17.5 –10 0 8 20
14.73 23.33 cm
f1 f2
Figure 38b
We moved the second lens in so that
–17.5 –10 0 23.33 40 52 82.85 cm
the second object distance is negative.
Figure 38a We now get an inverted image 6.73
Locating the image in a two lens system. cm from the second lens.
 Optics-29

In Figure (38c) we sketched a number of rays passing Using Equation 26 for 1/i1 gives
through the first lens, heading for the first image. These 1 1 1 1
rays are converging on the second lens, which we point = + –
i2 f2 f1 o1
out in Figure (36b) was the condition for a negative
object distance. 1 1 1 1
+ = + (28)
o1 i2 f1 f2
Now o1 is the object distance and i2 is the image
distance for the pair of lenses. Treating the pair of
object lenses as a single lens, we should have
1 1 1
+ = (29)
o1 i2 f
Figure 38c
where f is the focal length of the combined lens.
Two Lenses Together
If you put two thin lenses together, as shown in Figure Comparing Equations 28 and 29 we get
(39), you effectively create a new thin lens with a
different focal length. To find out what the focal length 1 1 1 focal length of two
= + thin lenses together
(30)
of the combination is, you use the lens equation twice, f f1 f2
setting the second object distance o2 equal to minus the
first image distance –i1 . as the simple formula for the combined focal length.

for two lenses

o2 = –i1 together
(25) Exercise 11
(a) Find the image distances i 2 for the geometry of
From the lens equations we have Figures (38), but with the two lenses reversed, i.e., with
1 1 1 f1= 12 cm , f2= 10 cm . Do this for both length D = 40 cm
= – (26) and D = 8 cm.
i1 f1 o1
1 1 1 (b) If the two lenses are put together (D = 0) what is the
= – (27) focal length of the combination?
i2 f2 o2
Setting o2 = – i1 in Equation 27 gives
1 1 1 1 1
= – = +
i2 f2 –i1 f2 i1

i1
o1
–o2

object

object for
Figure 39 1 2 second lens
Two lenses together. Since the object for the second
lens is on the wrong side of the lens, the object distance
o2 is negative in this diagram. If the lenses are close
together, i1 and – o2 are essentially the same.
Optics-30

Magnification Exercise 12
It is natural to define the magnification created by a lens Figures (38) and (40) are scale drawings, so that the
as the ratio of the height of the image to the height of the ratio of image to object sizes measured from these
object. In Figure (40) we have reproduced Figure (38a) drawings should equal the calculated magnifications.
emphasizing the heights of the objects and images.
(a) Calculate the magnifications m1 , m2 and m12 for
We see that the shaded triangles are similar, thus the Figure (38a) or (40) and compare your results with
ratio of the height B of the first image to the height A magnifications measured from the figure.
of the object is (b) Do the same for Figure (38b). In Figure (38b), the
B i1 final image is inverted. Did your final magnification m12
= (31) come out negative?
A o1
We could define the magnification in the first lens as Exercise 13
the ratio of B/A, but instead we will be a bit tricky and Figure (41a) shows a magnifying glass held 10 cm
include a - (minus) sign to represent the fact that the above the printed page. Since the object is inside the
image is inverted. With this convention we get focal length we get a virtual image as seen in the
geometrical construction of Figure (41b). Show that our
–B –i1 definition of formulas predict a positive magnification, and estimate
m1 = = magnificationm
(32)
A o1 the focal length of the lens. (Answer: about 17 cm.)

Treating B as the object for the second lens gives

–C –i2
m2 = = (33)
B o2
The total magnification m12 in going from the object A
to the final image C is
C
m12 = (34)
A
which has a + sign because the final image C is upright.
But
C = –C –B (35)
A B A
Thus we find that the final magnification is the product Figure 41a
of the magnifications of each lens. Using a magnifying glass.

m12 = m1 m2 (36)
virtual
image object
C o
o1 image i
i1 o2
A f
B i2 Figure 41b
When the magnifying glass is less than a focal length
away from the object, we see an upright virtual image.
Figure 40
Magnification of two lenses.
 Optics-31

THE HUMAN EYE Figure (42a) is a sketch of the human eye and Figure
A very good reason for studying geometrical optics is 42b a remarkable photograph of the eye. As seen in
to understand how your own eye works, and how the (42a), light enters the cornea at the front of the eye. The
situation is corrected when something goes wrong. amount of light allowed to enter is controlled by the
opening of the iris. Together the cornea and crystalline
Back in Exercise 6 (p21), lens focuses light on the retina which is a film of nerve
during our early discussion n1 = 1 fibers on the back surface of the eye. Information from
n2 = ?
of spherical lens surfaces, the new fibers is carried to the brain through the optic
we considered as a model nerve at the back. In the retina there are two kinds of
of an eye a sphere of index nerve fibers, called rods and cones. Some of the
Figure 27d
of refraction n 2 , where n 2 roughly 120 million rods and 7 million cones are seen
was chosen so that parallel magnified about 5000 times in Figure (43). The slender
rays which entered the front surface focused on the ones, the rods, are more sensitive to dim light, while the
back surface as shown in Figure (27d). The value of n 2 shorter, fatter, cones, provide our color sensitivity.
turned out to be n 2 = 2.0 . Since the only common
substance with an index of refraction greater than In our discussion of the human ear, we saw how there
zircon at n = 1.923 is diamond at n = 2.417, it would be was a mechanical system involving the basilar mem-
difficult to construct such a model eye. Instead some brane that distinguished between the various frequen-
extra focusing capability is required, both to bring the cies of incoming sound waves. Information from
focus to the back surface of the eye, and to focus on nerves attached to the basilar membrane was then
objects located at various distances. enhanced through processing in the local nerve fibers
before being sent to the brain via the auditory nerve. In
the eye, the nerve fibers behind the retina, some of
ciliary
muscle which can be seen on the right side of Figure (43), also
do a considerable amount of information processing
iris lens optic before the signal travels to the brain via the optic nerve.
central
fovea nerve The way that information from the rods and cones is
pupil processed by the nerve fibers is a field of research.

cornea retina Returning to the front of the eye we have the surface of
the cornea and the crystalline lens focusing light on the
a) retina. Most of the focusing is done by the cornea. The
shape, and therefore the focal length of the crystalline
Figure 42 lens can be altered slightly by the ciliary muscle in
The human eye. The
cornea and the lens
order to bring into focus objects located at different
together provide the distances.
extra focusing power
required to focus light cone
on the retina. rod
(Photograph of the Figure 43
human eye by Rods and
Lennart Nilsson.) cones in the
retina. The
thin ones
are the rods,
the fat ones
b) the cones.
Optics-32

In a normal eye, when the ciliary muscle is in its resting Nearsightedness and Farsightedness
position, light from infinity is focused on the retina as Not all of us have the so called normal eyes described
shown in Figure (44a). To see a closer object, the by Figure (44). There is increasing evidence that those
ciliary muscles contract to shorten the focal length of who do a lot of close work as children end up with a
the cornea-lens system in order to continue to focus condition called nearsightedness or myopia where the
light on the cornea (44b). If the object is too close as in eye is elongated and light from infinity focuses inside
Figure (44c), the light is no longer focused and the the eye as shown in Figure (45a). This can be corrected
object looks blurry. The shortest distance at which the by placing a diverging lens in front of the eye to move
light remains in focus is called the near point. For the focus back to the retina as shown in Figure (45b).
children the near point is as short as 7 cm, but as one
ages and the crystalline lens becomes less flexible, the The opposite problem, farsightedness, where light
near point recedes to something like 200 cm. This is focuses behind the retina as shown in Figure (46a) is
why older people hold written material far away unless corrected by a converging lens as shown in Figure
they have reading glasses. (46b).

a)
Figure 44a
Parallel light rays from a distant object are focuses on
the retina when the ciliary muscles are in the resting
position.

meniscus
concave
Figure 45
b)
Nearsightedness can be corrected by a convex lens..

Figure 44b
The ciliary muscle contracts to shorten the focal length
of the cornea-lens system in order to focus light from a
more nearby object.

a)

Figure 44c
When an object is to close, the light cannot be focused.
The closest distance we. .... meniscus
convex
Figure 46
b)
Farsightedness can be corrected by a convex lens
 Optics-33

THE CAMERA
There are a number of similarities between the human
eye and a simple camera. Both have an iris to control
the amount of light entering, and both record an image
at the focal plane of the lens. In a camera, the focus is
adjusted, not by changing the shape of the lens as in the
eye, but by moving the lens back and forth. The eye is
somewhat like a TV camera in that both record images
at a rate of about 30 per second, and the information is
transmitted electronically to either the brain or a TV
screen. Figure 47a
The Physics department’s Minolta
On many cameras you will find a series of numbers single lens reflex camera.
labeled by the letter f, called the f number or f stop. pentaprism
Just as for the parabolic reflectors in figure 4 (p5), the
f number is the ratio of the lens focal length to the lens
diameter. As you close down the iris of the camera to
reduce the amount of light entering, you reduce the
effective diameter of the lens and therefore increase the
f number.

Exercise 14
film
The iris on the human eye can change the diameter of
retractable
the opening to the lens from about 2 to 8 millimeters. The mirror
total distance from the cornea to the retina is typically
about 2.3 cm. What is the range of f values for the
Figure 47b
human eye? How does this range compare with the The lens system for a Nikon single lens reflex camera.
range of f value on your camera? (If you have one of When you take the picture, the hinged mirror flips out
the automatic point and shoot cameras, the f number of the way and the light reaches the film. Before that,
and the exposure time are controlled electronically and the light is reflected through the prism to the eyepiece.
you do not get to see or control these yourself.)
Optics-34

Depth of Field If the film were ideal, you could tell that the image at
There are three ways to control the exposure of the film positions 1 or 2 was out of focus. But no film or
in a camera. One is by the speed of the film, the second recording medium is ideal. If you look closely enough
is the exposure time, and the third is the opening of the there is always a graininess caused by the size of the
iris or f stop. In talking a picture you should first make basic medium like the silver halide crystals in black and
sure the exposure is short enough so that motion of the white film, the width of the scan lines in an analog TV
camera and the subject do not cause blurring. If your camera, or the size of the pixels in a digital camera. If
film is fast enough, you can still choose between a the image of the arrow tip at position 1 is smaller than
shorter exposure time or a smaller f stop. This choice the grain or pixel size then you cannot tell that the
is determined by the depth of field that you want. picture is out of focus. You can place the recording
medium anywhere between position 1 and 2 and the
The concept of depth of field is illustrated in Figures image will be as sharp as you can get.
(48a and b). In (48a), we have drawn the rays of light
from an object to an image through an f 2 lens, a lens In Figure (48b), we have drawn the rays from the same
with a focal length equal to twice its diameter. (The object passing through a smaller diameter f 8 lens.
effective diameter can be controlled by a flexible Again we show by dotted lines positions 1 and 2 where
diaphragm or iris like the one shown.) If you placed a the image of the arrow point would fill the same size
film at the image distance, the point at the tip of the circle as it did at positions 1 and 2 for the f 2 lens above.
object arrow would focus to a point on the film. If you Because the rays from the f 8 lens fill a much narrower
moved the film forward to position 1, or back to cone than those from the f 2 lens, there is a much greater
position 2, the image of the arrow tip would fill a circle distance between positions 1 and 2 for the f 8 lens.
about equal to the thickness of the three rays we drew
in the diagram.

1 2

debth of field
f 2.8 opening on film side
Figure 48a
A large diameter lens has a narrow depth of field.
Photograph taken at f 5.6.

1 2

f8 opening debth of field

on film side
Figure 48b
Reducing the effective diameter of the
lens increases the depth of field.
Photograph taken at f 22.
 Optics-35

If Figures (48) represented a camera, you would not be everywhere behind the lens. In this limit you do not
concerned with moving the film back and forth. In- even need a lens, a pinhole in a piece of cardboard will
stead you would be concerned with how far the image do. If enough light is available and the subject doesn’t
could be moved back and forth and still appear to be in not move, you can get as good a picture with a pinhole
focus. If the film were at the image position and you camera as one with an expensive lens system. Our
then moved the object in and out, you could not move pinhole camera image in Figure (49) is a bit fuzzy
it very far before it’s image was noticeably out of focus because we used too big a pinhole.
with the f 2 lens. You could move it much farther for
the f 8 lens. (If you are nearsighted you can see how a pinhole
camera works by making a tiny hole with your fingers
This effect is illustrated by the photographs on the right and looking at a distant light at night without your
side of Figures 48, showing a close-up tree and the glasses. Just looking at the light, it will look blurry. But
distant tower on Baker Library at Dartmouth College. look at the light through the hole made by your fingers
The upper picture taken at f 5.6 has a narrow depth of and the light will be sharp. You can also see the eye
field, and the tower is well out of focus. In the bottom chart better at the optometrists if you look through a
picture, taken at f22, has a much broader depth of field small hole, but they don’t let you do that.)
and the tower is more nearly in focus. (In both cases we
focused on the nearby tree bark.)

Camera manufacturers decide how much blurring of

the image is noticeable or tolerable, and then figure out
the range of distances the object can be moved and still
be acceptably in focus. This range of distance is called
the depth of field. It can be very short when the object
is up close and you use a wide opening like f 2. It can
be quite long for a high f number like f 22. The
inexpensive fixed focus cameras use a small enough
lens so that all objects are “in focus” from about 3 feet
or 1 meter to infinity. Figure 49a
We made a pinhole camera by replacing
In the extreme limit when the lens is very small, the the camera lens with a plastic film case
depth of field is so great that everything is in focus that had a small hole poked into the end.

range
for f 11
Figure 48c
Camera lens. This lens is set to f11, and adjusted Figure 49b
to a focus of 3 meters or 10 ft. At this setting, the Photograph of Baker library tower, taken with the
depth of field ranges from 2 to 5 meters. pinhole camera above. If we had used a smaller hole
we would have gotten a sharper focus.
Optics-36

Eye Glasses and a Home Lab they placed the magnifying glass and the eyeglass lens
Experiment together, measured the focal length of the combination,
When you get a prescription for eyeglasses, the optom- and used the formula
etrist writes down number like -1.5, -1.8 to represent 1 1 1
the power of the lenses you need. These cryptic number = + (30 repeated)
f f1 f2
are the power of the lenses measured in diopters. What
a diopter is, is simply the reciprocal of the focal length to calculate the focal length of the lens.
1/f, where f is measured in meters. A lens with a power
(Note that if you measure distances in meters, then
of 1 diopter is a converging lens with a focal length of
1/ f 1 is the power of lens 1 in diopters and 1/ f 2 that of
1 meter. Those of us who have lenses closer to –4 in
lens 2. Equation 30 tells you that the power of the
power have lenses with a focal length of –25 cm, the
combination 1/f is the sum of the powers of the two
minus sign indicating a diverging lens to correct for
lenses.
nearsightedness as shown back in Figure (45).

If you are nearsighted and want to measure the power Exercise 15

of your own eyeglass lenses, you have the problem that Assume that you find a magnifying lens that focuses the
it is harder to measure the focal length of a diverging sun at a distance of 10 cm from the lens. You then
lens than a converging lens. You can quickly measure combine that with one of your (or a friends) eyeglass
the focal length of a converging lens like a simple lenses, and discover that the combination focus at a
magnifying glass by focusing sunlight on a piece of distance of 15 cm. What is the power, in diopters, of
paper and measuring the distance from the lens to (a) The magnifying glass.
where the paper is starting to smoke. But you do not get
a real image for a diverging lens, and cannot use this (b) The combination.
simple technique for measuring the focal length and (c) The eyeglass lens.
power of diverging lenses used by the nearsighted.
Exercise 16 – Home Lab
As part of a project, some students used the following
method to measure the focal length and then determine Use the above technique to measure the power of your
the power in diopters, of their and their friend’s eye- or your friend’s glasses. If you have your prescription
glasses. They started by measuring the focal length f0 compare your results with what is written on the pre-
scription. (The prescription will also contain information
of a simple magnifying glass by focusing the sun. Then
about axis and amount of astigmatism. That you cannot
check as easily.
 Optics-37

THE EYEPIECE Exercise 17 - The Magnifying Glass

When the author was a young student, he wondered There are three distinct ways of viewing an object
why you do not put your eye at the focal point of a through a magnifying glass, which you should try for
telescope mirror. That is where the image of a distance yourself. Get a magnifying glass and use the letters on
object is, and that is where you put the film in order to this page as the object to be viewed.
record the image. You do not put your eye at the image (a) First measure the focal length of the lens by focusing
because it would be like viewing an object by putting the image of a distant object onto a piece of paper. A
your eyeball next to it. The object would be hopelessly light bulb across the room or scene out the window will
out of focus. Instead you look through an eyepiece. do.

The eyepiece is a magnifying glass that allows your eye (b) Draw some object on the paper, and place the paper
to comfortably view an image or small object up close. at least several focal lengths from your eye. Then hold
For a normal eye, the least eyestrain occurs when the lens about 1/2 a focal length above the object as
looking at a distant object where the light from the shown in Figure (51a). You should now see an enlarged
object enters the eye as parallel rays. It is then that the image of the object as indicated in Figure (51a). You are
ciliary muscles in the eye are in a resting position. If the now looking at the virtual image of the object. Check
that the magnification is roughly a factor of 2× .
image or small object is placed at the focal plane of a
lens, as shown in Figure (50), light emerges from the (c) Keeping your eye in the same position, several focal
lens as parallel rays. You can put your eye right up to lengths and at least 20 cm from the paper, pull the lens
that lens, and view the object or image as comfortably back toward your eye. The image goes out of focus
as you would view a distant scene. when the lens is one focal length above the paper, and
then comes back into focus upside down when the lens
parallel is farther out. You are now looking at the real image as
rays indicated in Figure (51b). Keep your head far enough
back that your eye can focus on this real image.
image
or Hold the lens two focal lengths above the page and
small check that the inverted real image of the object looks
object
eye about the same size as the object itself. (As you can see
from Figure (51b), the inverted image should be the
f same size as the object, but 4 focal lengths closer.)

Figure 50
The eyepiece or magnifier. To look at small object, or virtual
image
to study the image produced by another lens or mirror,
place the image or object at the focal plane of a lens, so object
that the light emerges as parallel rays that your eye can f
comfortably focus upon. Figure 51a
Looking at the virtual image.

2f 15 – 25 cm
f

object image

Figure 51b
Looking at the inverted real image.
Optics-38

(d) Now hold the lens one focal length above the page The Magnifier
and put your eye right up to the lens. You are now using When jewelers work on small objects like the innards
the lens as an eyepiece as shown in Figure (50). The of a watch, they use what they call a magnifier which
letters will be large because your eye is close to them,
can be a lens mounted at one end of a tube as shown in
and they will be comfortably in focus because the rays
Figure (53). The length of the tube is equal to the focal
are entering your eye as parallel rays like the rays from
a distant object. When you use the lens as an eyepiece
length of the lens, so that if you put the other end of the
you are not looking at an image as you did in parts (b) tube up against an object, the lens acts as an eyepiece
and (c) of this exercise, instead your eye is creating an and light from the object emerges from the lens as
image on your retina from the parallel rays. parallel rays. By placing your eye close to the lens, you
get a close up, comfortably seen view of the object.
(e) As a final exercise, hold the lens one focal length You may have seen jewelers wear magnifiers like that
above a page of text, start with your eye next to the lens, shown in Figure (54).
and then move your head back. Since the light from the
page is emerging from the lens as parallel rays, the size
lens
of the letters should not change as you move your head
back. Instead what you should see is fewer and fewer
letters in the magnifying glass as the magnifying glass
itself looks smaller when farther away. This effect is

f
seen in Figure (52).

watch

Figure 53
A magnifier.

Figure 54
Jeweler Paul Gross with magnifier
Figure 52
lenses mounted in visor.
When the lens is one focal length from the page, the
emerging rays are parallel. Thus the image letters do
not change size as we move away. Instead the lens
looks smaller, and we see fewer letters in the lens.
 Optics-39

Angular Magnification To calculate the angular magnification we use the small

Basically all the magnifier does is to allow you to move angle approximation sinθ ≈ θ to get
the object close to your eye while keeping the object y
comfortably in focus. It is traditional to define the θm = from Figure 55a
f
magnification of the magnifier as the ratio of the size
of the object as seen through the lens to the size of the y
θ0 = from Figure 55b
object as you would see it without a magnifier. By size, 25 cm
we mean the angle the object subtends at your eye. This which gives
is often called the angular magnification.
angular y/f 25 cm
The problem with this definition of magnification is = = (38)
magnification y/25 cm f
that different people, would hold the object at different
distances in order to look at it without a magnifier. For Thus if our magnifier lens has a focal length of 5 cm,
example, us nearsighted people would hold it a lot the angular magnification is 5× . Supposedly the
closer than a person with normal vision. To avoid this object will look five times bigger using the magnifier
ambiguity, we can choose some standard distance like than without it.
25 cm, a standard near point, at which a person would
normally hold an object when looking at it. Then the
angular magnification of the magnifier is the ratio of
the angle θ m subtended by the object when using the
parallel
magnifier, as shown in Figure (55a), to the angle θ 0 rays
subtended by the object held at a distance of 25 cm, as
shown in Figure (55b). y θm
angular θm angles defined f
(a)
= (37)
magnification θ0 in Figure 55

y θ0
(b)
25 cm

Figure 55
The angles used in defining angular magnification.
Optics-40

TELESCOPES As with the magnifier, we define the magnification of

The basic design of a telescope is to have a large lens the telescope as the ratio of the size of (angle subtended
or parabolic mirror to create a bright real image, and by) the object as seen through the object to the size of
then use an eyepiece to view the image. If we use a large (angle subtended by) the object seen by the unaided
lens, that lens is called an objective lens, and the eye. In Figure (56) we see that the constellation
telescope is called a refracting telescope. If we use a subtends an angle θ 0 as viewed by the unaided eye, and
parabolic mirror, then we have a reflecting telescope. an angle θ i when seen through the telescope. Thus we
define the magnification of the telescope as
The basic design of a refracting telescope is shown in θi magnification
Figure (56). Suppose, as shown in Figure (56a), we are m = (39)
looking at a constellation of stars that subtend an angle θ0 of telescope

θ 0 as viewed by the unaided eye. The eye is directed To calculate this ratio, we note from Figure (56c) that,
just below the bottom star and light from the top star using the small angle approximation sinθ ≈ θ , we
enters at an angle θ 0 . In Figure (56b), the lens system have
from the telescope is placed in front of the eye, and we yi yi
are following the path of the light from the top star in θ0 = ; θi = (40)
f0 fe
the constellation.
where f 0 and f e are the focal lengths of the objective
The parallel rays from the top star are focused at the and eyepiece lens respectively. In the ratio, the image
focal length f 0 of the objective lens. We adjust the height y i cancels and we get
eyepiece so that the image produced by the objective θi y i/f e
lens is at the focal point of the eyepiece lens, so that light m = =
θ0 y i/f 0
from the image will emerge from the eyepiece as
parallel rays that the eye can easily focus. f0
m =
fe (41)

Figure 56a
The unaided eye looking at a constellation of
stars that subtend an angle θ o .
θ0
constellation (a)

Figure 56b
Looking at the same constellation through a para
the llel ligh inverted parallel
simple refracting telescope. The objective lens the top sta t from image rays
con r
produces an inverted image which is viewed by stell in
atio
n θ0 θi
the eyepiece acting as a magnifier. Note that
the parallel light from the star focuses at the θ0 (b)
focal point of the objective lens. With the image
at the focal point of the eyepiece lens, light from
the image emerges as parallel rays that are fe
f0
easily focused by the eye.

θ0 θi (c)
Figure 56c yi
Relationship between the angles
θ0 , θi , and the focal lengths. f0 fe
objective eyepiece
lens lens
 Optics-41

The same formula also applies to a reflecting telescope Exercise 8

with f 0 the focal length of the parabolic mirror. Note To build your own refracting telescope, you purchase a
that there is no arbitrary number like 25 cm in the 3 inch diameter objective lens with a focal length of 50
formula for the magnification of a telescope because cm. You want the telescope to have a magnification
telescopes are designed to look at distant objects where m = 25 × .
the angle θ 0 the object subtends to the unaided eye is
the same for everyone. (a) What will be the f number of your telescope? (1 inch
= 2.54 cm).
The first and the last of the important refracting tele-
(b) What should the focal length of your eyepiece lens
scopes are shown in Figures (57). The telescope was be?
invented in Holland in 1608 by Hans Lippershy. Shortly
after that, Galileo constructed a more powerful instru- (c) How far behind the objective lens should the eye-
ment and was the first to use it effectively in astronomy. piece lens be located?
With a telescope like the one shown in Figure (57a), he
(d) Someone give you an eyepiece with a focal length
discovered the moons of Jupiter, a result that provided of 10 mm. Using this eyepiece, what magnification do
an explicit demonstration that heavenly bodies could you get with your telescope?
orbit around something other than the earth. This
countered the long held idea that the earth was at the (e) You notice that your new eyepiece is not in focus at
center of everything and provided support for the the same place as your old eyepiece. Did you have to
Copernican sun centered picture of the solar system. move the new eyepiece toward or away from the
objective lens, and by how much?
When it comes to building large refracting telescopes,
(f) Still later, you decide to take pictures with your
the huge amount of glass in the objective lens becomes
telescope. To do this you replace the eyepiece with a
a problem. The 1 meter diameter refracting telescope
film holder. Where do you place the film, and why did
at the Yerkes Observatory, shown in Figure (57b), is you remove the eyepiece?
the largest refracting telescope ever constructed. That
was built back in 1897. The largest reflecting telescope
is the new 10 meter telescope at the Keck Observatory
at the summit of the inactive volcano Mauna Kea in
Hawaii. Since the area and light gathering power of a
telescope is proportional to the area or the square of the
diameter of the mirror or objective lens, the 10 meter
Keck telescope is 100 times more powerful than the 1
meter Yerkes telescope.

Figure 57b
The Yerkes telescope is the world’s largest refracting
telescope, was finished in 1897. Since then all larger
telescopes have been reflectors.

Figure 57a
Galileo’s telescope. With such an instrument
Galileo discovered the moons of Jupiter.
Optics-42

Reflecting telescopes However, problems with keeping the reflecting surface

In several ways, the reflecting telescope is similar to the shinny, and the development of lens combinations that
refracting telescope. As we saw back in our discussion eliminated chromatic aberration, made refracting tele-
of parabolic mirrors, the mirror produces an image in scopes more popular until the late 1800’s. The inven-
the focal plane when the light comes from a distant tion of the durable silver and aluminum coatings on
object. This is shown in Figure (58a) which is similar glass brought reflecting telescopes into prominence in
to our old Figure (4). If you want to look at the image the twentieth century.
with an eyepiece, you have the problem that the image
θ0 parabolic reflector
is in front of the mirror where, for a small telescope,
your head would block the light coming into the scope.
Issac Newton, who invented the reflecting telescope, inverted
solved that problem by placing a small, flat, 45° image
reflecting surface inside the telescope tube to deflect
the image outside the tube as shown in Figure (58b).
There the image can easily be viewed using an eye-
piece. Newton’s own telescope is shown in Figure f0
(58d). Another technique, used in larger telescopes, is Figure 58a
to reflect the beam back through a hole in the mirror as A parabolic reflector focuses the parallel rays from a
shown in Figure (58c). distant object, forming an image a distance f0 in
front of the mirror.
The reason Newton invented the reflecting telescope
was to avoid an effect called chromatic aberration.
eyepiece to
When white light passes through a simple lens, differ- look at image
ent wavelengths or colors focus at different distances
behind the lens. For example if the yellow light is in
focus the red and blue images will be out of focus. In
contrast, all wavelengths focus at the same point using
a parabolic mirror.

parabolic reflector
Figure 58b
Issac Newton’s solution to viewing the image was to
deflect the beam using a 45° reflecting surface so that
the eyepiece could be outside the telescope tube.

secondary photographic film

mirror or eyepiece

Figure 58d
Issac Newton’s reflecting telescope. parabolic reflector
with hole in center
Figure 58c
For large telescopes, it is common to reflect the
beam back through a hole in the center of the
primary mirror. This arrangement is known as
the Cassegrain design.
 Optics-43

Large Reflecting Telescopes.

The first person to build a really large reflecting telescope
was William Hershel, who started with a two inch reflector
in 1774 and by 1789 had constructed the four foot diameter
telescope shown in Figure (59a). Among Hershel’s accom-
plishments was the discovery of the planet Uranus, and the
first observation a distant nebula. It would be another 130
years before Edwin Hubble, using the 100 inch telescope on
Mt. Wilson would conclusively demonstrate that such
nebula were in fact galaxies like our own milky way. This
also led hubble to discover the expansion of the universe.

During most of the second half of the twentieth century, the

largest telescope has been the 200 inch (5 meter) telescope
on Mt. Palomar, shown in Figure (59b). This was the first
telescope large enough that a person could work at the prime
focus, without using a secondary mirror. Hubbel himself is Figures 59b,c
seen in the observing cage at the prime focus in Figure (59c). The Mt. Palomar
200 inch telescope.
Recently it has become possible to construct mirrors larger Below is Edwin
than 5 meters in diameter. One of the tricks is to cast the Hubble in the
molten glass in a rotating container and keep the container observing cage.
rotating while the glass cools. A rotating liquid has a
parabolic surface. The faster the rotation the deeper the
parabola. Thus by choosing the right rotation speed, one can
cast a mirror blank that has the correct parabola built in. The
surface is still a bit rough, and has to be polished smooth, but
the grinding out oh large amounts of glass is avoided. The
6.5 meter mirror, shown in Figures (59d and e), being
installed on top of Mt. Hopkins in Arizona, was built this
way. Seventeen tons of glass would have to have been
ground out if the parabola had not been cast into the mirror
blank.

Figures 59d,e
The 6.5 meter MMT
telescope atop Mt.
Hopkins. Above, the
mirror has not been
silvered yet. The blue
Figure 59a is a temporary
William Hershel’s 4 ft diameter, 40 ft long protective coating.
reflecting telescope which he completed in 1789. Below, the mirror is
being hoisted into
the telescope frame.
Optics-44

Hubbel Space Telescope

An important limit to telescopes on earth, in their
ability to distinguish fine detail, is turbulence in the
atmosphere. Blobs of air above the telescope move
around causing the star image to move, blurring the
picture. This motion, on a time scale of about 1/60
second, is what causes stars to appear to twinkle.

The effects of turbulance, and any distortion caused by

the atmosphere, are eliminated by placing the telescope
in orbit above the atmosphere. The largest telescope in
orbit is the famous Hubble telescope with its 1.5 meter
diameter mirror, seen in Figure (60). After initial
problems with its optics were fixed, the Hubble tele-
scope has produced fantastic images like that of the
Figure 60a
Eagle nebula seen in Figure (7-17) reproduced here.
The Hubble telescope mirror. How
is that for a shaving mirror?
With a modern telescope like the Keck (see next page),
the effects of atmospheric turbulance can mostly be
eliminated by having a computer can track the image of
a bright star. The telescopes mirror is flexible enough
that the shape of the mirror can then be be modified
rapidly and by a tiny amount to keep the image steady.

Figure 60b
Hubble telescope before launch.

Figure 7-17
The eagle nebula, birthplace of stars. This Hubble
photograph, which apeared on the cover of Time
magazine, is perhaps the most famous.

Figure 60c
Hubble telescope being deployed.
 Optics-45

World’s Largest Optical Telescope Diffraction also affects the ability of telescopes to form
As of 1999, the largest optical telescope in the world is the sharp images. The bigger the diameter of the telescope,
Keck telescope located atop the Mauna Kea volcano in compared to the light wavelength, the less important
Hawaii, seen in Figure (61a). Actually there are two diffraction is and the sharper the image that can be
identical Keck telescopes as seen in the close-up, Figure formed. By combining the output from the two Keck
(61b). The primary mirror in each telescope consists of 36 telescopes, one creates a telescope whose effective
hexagonal mirrors fitted together as seen in Figure (61c) diameter, for handling diffraction effects, is equal to the
to form a mirror 10 meters in diameter. This is twice the 90 meter separation of the telescopes rather than just
diameter of the Mt. palomar mirrorwe discussed earlier. the 10 meter diameter of one telescope. The great
improvement in the image sharpness that results is seen
The reason for building two Keck telescopes has to do
in Figure (61d). On the left is the best possible image of
with the wave nature of light. As we mentioned in the
a star, taken using one telescope alone. When the two
introduction to this chapter, geometrical optics works
well when the objects we are studying are large compared
telescopes are combined, they get the much sharper
to the wavelength of light. This is illustrated by the ripple image on the right.
tank photographs of Figures (33-3) and (33-8) repro-
duced here. In the left hand figure, we see we see a wave
passing through a gap that is considerably wider than the
wave’s wavelength. On the other side of the gap there is
a well defined beam with a distinct shadow. This is what
we assume light waves do in geometrical optics.

In contrast, when the water waves encounter a gap whose

width is comparable to a wavelength,as in the right hand
figure, the waves spread out on the far side. This is a
phenomenon called diffraction. We can even see some
diffraction at the edges of the beam emerging from the
wide gap.
Figures 61 c
The 36 mirrors forming Keck’s primary mirror. We
have emphasized the outline of the upper 4 mirrors.

Figures 33-3,8
Unless the gap
is wide in
comparison to a
wavelength,
diffraction
effects are
important.

Figures 62
Figures 61 a,b Same star, photographed on the left using one scope,
The Keck telescopes atop Mauna Kea volcano in Hawaii on the right with the two Keck telescopes combined.
Optics-46

Infrared Telescopes Where does the infrared light come from? If you have
Among the spectacular images in astronomy are the studied Chapter 35 on the Bohr theory of hydrogen,
large dust clouds like the ones that form the Eagle you will recall that hydrogen atoms can radiate many
nebula photographed by the Hubble telescope, and the different wavelengths of light. The only visible wave-
famous Horsehead nebula shown in Figure (63a). But lengths are the three longest wavelengths in the Balmer
a problem is that astronomers would like to see through series. The rest of the Balmer series and all of the
the dust, to see what is going on inside the clouds and Lyman series consist of short wavelength ultraviolet
what lies beyond. light. But all the other wavelengths radiated by hydro-
gen are infrared, like the Paschen series where the
While visible light is blocked by the dust, other electron ends up in the third energy level. The infrared
wavelength’s of electromagnetic radiation can pen- wavelengths are longer than those of visible light.
etrate these clouds. Figure (63b) is a photograph of the Since hydrogen is the major constituent of almost all
same patch of sky as the Horsehead nebula in (63a), but stars, it should not be surprising that stars radiate
observed using infrared light whose wavelengths are infrared as well as visible light.
about 3 times longer than the wavelengths of visible
light. First notice that the brightest stars are at the same A telescope designed for looking at infrared light is
positions in both photographs. But then notice that the essentially the same as a visible light telescope, except
black cloud, thought to resemble a horses head, is for the camera. Figure (64) shows the infrared tele-
missing in the infrared photograph. The stars in and scope on Mt. Hopkins used to take the infrared image
behind the cloud shine through; their infrared light is of the Horsehead nebula. We enlarged the interior
not blocked by the dust. photograph to show the infrared camera which is
cooled by a jacket of liquid nitrogen (essentially a large
a) thermos bottle surrounding the camera).
Visible light
photograph

b)
Infrared light
photograph
Figure 64
Infrared telescope on
Mt. Hopkins. Note that
the infrared camera,
seen in the blowup, is in
a container cooled by
liquid nitrogen. You do
not want the walls of the
camera to be “infrared
Figure 63
hot” which would fog
The horsehead nebula photographed in visible
the image.
(a) and infrared light (b). The infrared light
passes through the dust cloud.
 Optics-47

You might wonder why you have to cool an infrared of the galaxy. But in an infrared photograph, the huge
camera and not a visible light camera. The answer is concentration of stars in the plane of the galaxy show
that warm bodies emit infrared radiation. The hotter the up clearly.
object, the shorter the wavelength of the radiation. If an
object is hot enough, it begins to glow in visible light, At the center of our galaxy is a gigantic black hole, with
and we say that the object is red hot, or white hot. Since a mass of millions of suns. For a visible light telescope,
you do not want the infrared detector in the camera the galactic center is completely obscured by dust. But
seeing camera walls glowing “infrared hot”, the cam- the center can be clearly seen in the infrared photograph
era has to be cooled. of Figure (67), taken by the Mt. Hopkins telescope of
Figure 64. This is not a single exposure, instead it is a
Not all infrared radiation can make it down through the composite of thousands of images in that region of the
earth’s atmosphere. Water vapor, for example is very sky. Three different infrared wavelengths were re-
good at absorbing certain infrared wavelengths. To corded, and the color photograph was created by
observe the wavelengths that do not make it through, displaying the longest wavelength image as red, the
infrared telescopes have been placed in orbit. Figure 65 middle wavelength as green, and the shortest wave-
is an artist’s drawing of the Infrared Astronomical length as blue. In this photograph, you not only see the
Satellite (IRAS) which was used to make the infrared intense radiation from the region of the black hole at the
map of the entire sky seen in Figure (66). The map is center, but also the enormous density of stars at the
oriented so that the Milky Way, our own galaxy, lies center of our galaxy. (You do not see radiation from the
along the center horizontal plane. In visible light pho- black hole itself, but from nearby stars that may be in
tographs, most of the stars in our own galaxy are the process of being captured by the black hole.)
obscured by the immense amount of dust in the plane

Figure 65
Artist’s drawing of the infrared
telescope IRAS in orbit.

Figure 66
Map of the entire sky made by
IRAS. The center of the Milky
Way is in the center of the map.
This is essentially a view of our Figure 67
galaxy seen from the inside. Center of our galaxy, where an enormous black hole resides. Not only is
the galactic center rich in stars, but also in dust which prevents viewing
this region in visible light.
Optics-48

Radio Telescopes it’s enormous size, and has been used to make significant
The earth’s atmosphere allows not only visible and discoveries, it has the resolving ability of an optical tele-
some infrared light from stars to pass through, but also scope about 3 centimeters in diameter, or a good set of
radio waves in the wavelength range from a few millimeters binoculars .
to a good fraction of a meter. To study the radio waves As we saw with the Keck telescope, there is a great
emitted by stars and galaxies, a number of radio telescopes improvement in resolving power if the images of two or
have been constructed. more telescopes are combined. The effective resolving
For a telescope reflector to produce a sharp image, the power is related to the separation of the telescopes rather
surface of the reflector should be smooth and accurate to than to the diameter of the individual telescopes. Figure (69)
within about a fifth of a wavelength of the radiation being shows the Very Large Array (VLA) consisting of twenty
studied. For example, the surface of a mirror for a visible seven 25 meter diameter radio telescopes located in south-
wavelength telescope should be accurate to within about ern New Mexico. The dishes are mounted on tracks, and can
10 – 4 millimeters since the wavelength of visible light is be spread out to cover an area 36 kilometers in diameter. At
–4
centered around 5 × 10 millimeters. Radio telescopes this spacing, the resolving power is nearly comparable to a
that are to work with 5 millimeter wavelength radio waves, 5 meter optical telescope at Mt. Palomar.
need surfaces accurate only to about a millimeter. Tele-
scopes designed to study the important 21 cm wavelength
radiation emitted by hydrogen, can have a rougher surface
yet. As a result, radio telescopes can use sheet metal or even
wire mesh rather than polished glass for the reflecting
surface.

This is a good thing, because radio telescopes have to be

much bigger than optical telescopes to order to achieve
comparable images. The sharpness of an image, due to
diffraction effects, is related to the ratio of the reflector
diameter to the radiation wavelength. Since the radio wave-
4
lengths are at least 10 times larger than those for visible
4
light, a radio telescope has to be 10 times larger than an
optical telescope to achieve the same resolution.
Figures 69
The worlds largest radio telescope dish, shown in Figure The “Very Large Array” (VLA) of radio telescopes.
The twenty seven telescopes can be spread out to a
(68), is the 305 meter dish at the Arecibo Observatory in
diameter of 36 kilometers.
Perto Rico. While this dish can see faint objects because of

Figure 68 Figures 69b

Arecibo radio telescope. While the world’s largest Radio galaxy image from the VLA. Studying the
telescope dish remains fixed in the earth, the focal radio waves emitted by a galaxy often gives a
point can be moved to track a star. very different picture than visible light.
 Optics-49

The Very Long Baseline Array (VLBA)

To obtain significantly greater resolving power, the
Very Long Baseline Array (VLBA) was set up in the
f early 1990’s. It consists of ten 25 meter diameter radio
j telescopes placed around the earth as shown in Figure
g i (70). When the images of these telescopes are com-
h d a
c b bined, the resolving power is comparable to an optical
a b telescope 1000 meters in diameter (or an array of
e optical telescopes spread over an area one kilometer
across).

The data from each telescope is recorded on a high

speed digital tape with a time track created by a
c d hydrogen maser atomic clock. The tapes are brought to
a single location in Socorrow New Mexico where a
high speed computer uses the accurate time tracks to
combine the data from all the telescopes into a single
image. To do this, the computer has to correct, for
example, for the time difference of the arrival of the
radio waves at the different telescope locations.

e f Because of it’s high resolution, the VLBA can be used

to study the structure of individual stars. In Figure 72
we see two time snapshots of the radio emission from
the stellar atmosphere of a star 1000 light years away.
With any of the current optical telescopes, the image of
this star is only a point.

g h

i j

Figure 71
Very Long Baseline Array (VLBA) radio images of the
variable star TX Cam which is located 1000 light years
away. The approximate size of the star as it would be
Figure 70 seen in visible light is indicated by the circle. The spots
The Very Long Baseline Array of radio antennas. They are silicon Monoxide (SiO) gas in the star’s extended
are located at a) Hancock New Hampshire b) Ft. Davis atmosphere. Motion of the these spots trace the periodic
Texas c) Kitt Peak Arizona d) North Liberty Iowa e) St. changes in the atmosphere of the star.
Croix Virgin Islands f) Brewster Washington g) Mauna (Credit P.J. Diamond & A.J. Kembal, National Radio
Kea Hawaii h) Pie Town New Mexico i) Los Alamos Astronomy, Associated Universities, Inc.)
New Mexico j) Owen’s Valley California.
Optics-50

MICROSCOPES Equating the values of β in Equations 29 and 30 and

Optically, microscopes like the one seen in Figure (72), solving for y i gives
are telescopes designed to focus on nearby objects. y0 yi L
Figure (73) shows the ray diagram for a simple micro- β = = ; yi = y0 (45)
f0 L f0
scope, where the objective lens forms an inverted
The eyepiece is placed so that the image of the objective
image which is viewed by an eyepiece.
is in the focal plane of the eyepiece lens, producing
To calculate the magnification of a simple microscope, parallel rays that the eye can focus. Thus the distance
note that if an object of height y 0 were viewed unaided EG equals the focal length f 0 of the eyepiece. From
at a distance of 25 cm, it would subtend an angle θ 0 triangle EFG we find that the angle θ i that image
given by subtends as seen by the eye is
y0 yi
θ0 = θi =
angle subtended
(46)
25 cm (42) fe by image
where throughout this discussion we will use the small Substituting Equation 45 for y i in Equation 46 gives
angle approximation sinθ ≈ tanθ ≈ θ .
L y0
θi = (47)
A ray from the tip of the object (point A in Figure 73b), f 0 fe
parallel to the axis, will cross the axis at point D, the Finally, the magnification m of the microscope is equal
focal point of the objective lens. Thus the height BC is to the ratio of the angle θ i subtended by the image in
equal to the height y 0 of the object, and the distance BD the microscope, to the angle θ 0 the object subtends at
is the focal length f 0 of the objective, and the angle β a distance of 25 cm from the unaided eye.
is given by
y0 θi L y0 1
β =
from triangle m = = × (48)
f0 BCD (43) θ0 f 0 fe y 0 /25 cm
From triangle DEF, where the small angle at D is also where we used Equation 47 for θ i and Equation 42 for
β, we have θ 0 . The distance y 0 cancels in Equation 48 and we get
yi from triangle
β = DEF (44) m = L×
25 cm magnificationof a
(49)
L f0 fe simple microscope
where y i is the height of the image and the distance L
is called the tube length of the microscope. (We could have inserted a minus sign in the formula for
magnification to indicate that the image is inverted.)

y0 θ0
(a)
25 cm

f0 L fe
A C (tube length )
y0 β E G θi (b)
B D β θi
yi
l
alle
F parys
Figure 73 ra
Figure 72 Optics of a simple microscope.
Standard optical microscope, which my grandfather
purchased as a medical student in the 1890’s. Compare
this with a microscope constructed 100 years later, seen
in Figure (69) on the next page.
 Optics-51

Scanning Tunneling Microscope line by line, by a tiny probe whose operation is based on
Modern research microscopes bear less resemblance to the particle-wave nature of electrons. An image of the
the simple microscope described above than the Hubble surface is then reconstructed by computer and displayed
telescope does to Newton’s first reflector telescope. In on a computer screen. These microscopes work at a
the research microscopes that can view and manipulate scale of distance much smaller than the wavelength of
individual atoms, there are no lenses based on geometri- light, a distance scale where the approximations inher-
cal optics. Instead the surface to be studied is scanned, ent in geometrical optics do not apply.
sample holder

probe

a) Probe and sample holder.

b) Vacuum chamber
enclosing the probe and
sample holder. Photograph
taken in Geoff Nunes’ lab
at Dartmouth College.

Figure 74
Scanning Tunneling Microscope (STM). The tungsten
probe seen in (a) has a very sharp point, about one
atom across. With a couple of volts difference between
the probe and the silicon crystal in the sample holder,
an electric current begins to flow when the tip gets to
within about fifteen angstroms (less than fifteen atomic
diameters) of the surface. The current flows because
the wave nature of the electrons allows them to
“tunnel” through the few angstrom gap. The current
increases rapidly as the probe is brought still closer. By
moving the probe in a line sideways across the face of
the silicon, while moving the probe in and out to keep
the current constant, the tip of the probe travels at a
constant height above the silicon atoms. By recording
how much the probe was moved in and out, one gets a
recording of the shape of the surface along that line.
By scanning across many closely spaced lines, one gets
a map of the entire surface. The fine motions of the c) Surface (111 plane) of a silicon
tungsten probe are controlled by piezo crystals which crystal imaged by this microscope. We
expand or contract by tiny amounts when a voltage is see the individual silicon atoms in the
applied to them. The final image you see was created by surface
computer from the scanning data.
PHOTOGRAPH CREDITS Figure Optics-58d, p42; Newton’s telescope
Dorling Kindersley, Pockets Inventions
Figure 36-1, p1, p8; Scattered Wave Figure Optics-59a, p43; Hershel’s telescope
Education Development Center © Royal Astronomical Society Library
Figure 33-30, p10; Shock wave Figure Optics-59b,c, p43; Palomar telescope
Education Development Center Courtesy of The Archives, California Institute of
Figures Optics-1 p3; Mormon Tabrenacle Technology
The Church of Jesus Christ of Latter-day Saints, Figure Optics-59d,e, p43; MMT telescope
Historical Department Archives Lori Stiles, Universitu of Arizona News Service
Figure Optics-8b p7; Corner reflectors Figure 7-17, p44; Eagle nebula
NASA Space Telescope Science Institute/NASA
Figure Optics-9, p9; Wave through lens Figure Optics-60, p44; Hubble telescope
Nils Abramson NASA
Figure Optics-10, p11; Refraction, ripple tank Figure Optics-61a,b,c, p45; Keck telescope
Education Development Center © Richard Wainscoat
Figure Optics-11, p9; Refraction, glass Figure 33-3,8, p45; Wave through gap
©1990 Richard Megna/ Fundamental Photographs Education Development Center
Figure Optics-16b, p14; Glass Fiber Figure Optics-62, p45; Sharpened star image
Foto Forum Keck
Figure Optics-18, p15; Duodenum Figure Optics-63, p46; Horsehead nebula
Dr. Richard Rothstein Two Micron All Sky Survay (2MASS)
Figure Optics-19, p18; Halo Figure Optics-64, p46; Infrared telescope
© Robert Greenler Two Micron All Sky Survay (2MASS)
Figure Optics-25, p18; Zoom lens Figure Optics-65, p47; IRAS drawing
Nikon Space Infrared Telescope Facility
Figure Optics-28, p22; Hubble Scope Figure Optics-66, p47; Iras Milky Way
NASA Space Infrared Telescope Facility
Figure Optics-42, p31; Human eye Figure Optics-67, p47; Center of galaxy
© Lennart Nilsson Two Micron All Sky Survay (2MASS)
Figure Optics-43, p31; Rods & cones Figure Optics-68, p48; Arecibo telescope
© Lennart Nilsson National Astronomy and Ionosphere Center
Figure Optics-47b, p33; Single lens reflex Figure Optics-69a, p48; VLA radio telescopes
Adapted from Nikon drawing Photo courtesy of NRAO/AUI.
Figure Optics-57a, p41; Galileo’s telescope Figure Optics-69b, p48; Radio galaxy
© Institute and Museum of History of Science of Photo courtesy of NRAO/AUI.
Florence Italy
Figure Optics-70, p49; Radio image of star
Figure Optics-57b, p41; Yerkes telescope Photo courtesy of NRAO/AUI.
University of Chicago

i
 Intro-1

Calculus 2000
A Physics Based Calculus Text

CALCULUS 2000
A PHYSICS BASED
CALCULUS TEXT

The Physics 2000 (P2000) text uses certain calculus five more theorems and proofs. The mathematical
concepts that are taught in Chapter 1 of this Calculus structure may be clear, but the usefulness of the math-
2000 (C2000) text. For students whose calculus is ematics is obscured.
rusty, or who have not had calculus, they should start
studying Chapter 1 of C2000 while studying Chapter 4 In Calculus 2000, the emphasis is on the intuitive use
of P2000 on the use of calculus in Physics. By the time of the mathematics. We do not introduce a new
the student reaches Part 2 of P2000, Chapter 23, the mathematical concept or technique until the founda-
ideas contained in Chapter 1 of C2000 should be well tion has been developed in the Physics 2000 text. In
understood. That is because the chapters on electric Chapter 3 of P2000, for example, we use strobe photo-
and magnetic fields make extensive use of these basic graphs to introduce the concepts of velocity, accelera-
calculus ideas. tion, and the limiting process. The calculus limit at ∆t
goes to zero is represented physically by the idea of
The remaining chapters in Calculus 2000 start from turning the strobe flashing rate all the way up. We point
physical concepts built up in P2000 and introduce the out, however, that because of the uncertainty principle
student to advanced mathematical techniques. These we reach a point where further increase of the flashing
include concepts such as gradient, divergence, and curl rate affects the behavior of the object being studied.
which are essential tools for further study of physics Before the student works with calculus formulas, we
and engineering. Later chapters in C2000 will include set both the intuitive basis for the calculus concepts and
such topics as an introduction to complex variables, the discuss the limitation of their applicability.
Lorentz transformation and 4 vector notation, and two
chapters on fluid dynamics. As we mentioned, Calculus 2000 provides the calculus
background necessary to complete all of the Physics
In the standard calculus text, there is an advantage to 2000 text. All further calculus and advanced math-
presenting a theorem in its most general form so that the ematics concepts used in P2000 are developed in the
theorem can be used effectively in later proofs. The physics text. Chapter 2 of P2000 discusses vectors and
emphasis is on the logical structure of the mathematics. their scalar and vector products. The scalar dot product
The problem the physics student often encounters is is applied in Chapter 8 to the discussion of work and
that before the intuitive implications of one theorem energy, and the vector cross product is applied in the
can be worked out, the instructor has plowed through discussion of torques and gyroscopes in Chapter 12.
Intro-2 Calculus 2000

In Chapter 14 on oscillations, we introduce the student After Chapter 1, the remaining chapters in Calculus
to some techniques of solving differential equations. 2000 use physical concepts developed in P2000 to
The idea is to guess an answer and plug the guess into introduce advanced calculus techniques needed by
the equation to see if it works. The important point, physicists and engineers. The concept of a second
however, is that we use experimental results to make an derivative, and the differential form of the wave equa-
informed guess. In Chapter 16 the student is introduced tion follows from our discussion of one dimensional
to Fourier analysis. We wrote the MacScope™ com- waves in Chapter 12 of P2000. The student sees how
puter program so that the student could easily use much easier it is to determine the speed of a wave from
Fourier analysis to study experimental data. One of our the differential wave equation than it is to go through
standard laboratory examples is to use Fourier analysis the non calculus arguments we used in Chapter 15 of
to determine the normal modes of oscillation of a the P2000 text.
system of coupled air carts.
Once the student is familiar with the field plotting
Part II of Physics 2000, particularly Chapters 23 through models and mapping techniques discussed in Chapter
32 on Electric and Magnetic fields, are more math- 25 of P2000, she or he is ready for the discussion of the
ematically based than the other chapters. Chapter 23 on gradient function ∇ϕ described in Chapter 3 of C2000.
fluids is used to introduce the concept of a vector field. In Chapter 4 of C2000 the divergence and curl opera-
It is easier to visualize the velocity field of a fluid with tion are presented as the differential form of the surface
its streamlines, than the more abstract electric and and line integrals discussed in Chapter 29 of P2000.
magnetic field lines. We also use the velocity field to Maxwell's equations discussed in Chapter 32 of P2000
introduce the concepts involved in Gauss' law. are presented in differential form in Chapter 5 of
C2000, and from them we derive the differential form
In Chapter 25 we use contour maps to introduce equi- of the wave equation for electric and magnetic fields.
potential lines and the concept of electrical voltage. Again we see that it is easier to determine the speed of
Chapter 29 formalizes Gauss' law as an example of a a wave from the differential wave equation than from
surface integral, and introduces the closed line integral the fairly complex derivation we carried out using the
in the discussion of Ampere's law. By pointing out that integral equation. Chapter 5 ends with the introduction
equations for both the surface integral and the line of the vector potential to simplify the wave equation
integral are required to uniquely determine a vector when source terms are involved.
field, the student sees why four equations are needed to
specify both the electric field and the magnetic field.
These are the four integral equations that form Maxwell's
equations studied in Chapter 32. Our main mathemati-
cal achievement in Chapter 32 is to show that a pulse of
crossed electric and magnetic fields travels through
space at the speed of light.

Chapters 33 through 40 of P2000 are not as mathemati-

cally focused because we concentrate on developing an
intuitive picture of the particle/wave nature of matter.
Because we have used special relativity throughout the
text, it is easy to introduce the zero rest mass photons as
the particle nature of light waves. The wave nature of
the electron is introduced with de Broglie's hypothesis
and an electron diffraction experiment that is similar to
the laser diffraction experiments of Chapter 33. Fourier
analysis plays a basic role in showing how the particle/
wave nature of matter leads to the uncertainty principle
discussed in Chapter 40.
Cal 1-2 Calculus 2000

0 0 1
V0
V0

∆t = 0.4 Sec ∆t = 0.1 Sec

(a) (c)

V0 V
~i
01

∆t = 0.025 Sec instantaneous velocity

(b) (d)

Figure 1
Transition to instantaneous velocity.
 Introduction to Calculus Cal 1-3

Calculus Chapter 1
Introduction to Calculus

CHAPTER 1 INTRODUCTION TO
CALCULUS

This chapter, which replaces Chapter 4 in Physics LIMITING PROCESS

2000, is intended for students who have not had calcu- In our discussion of instantaneous velocity we concep-
lus, or as a calculus review for those whose calculus is tually turned the strobe all the way up as illustrated in
not well remembered. If, after reading part way Figures (2-22a) through (2-22d), redrawn here in Fig-
through this chapter, you feel your calculus back- ure (1). In these figures, we initially see a fairly large
ground is not so bad after all, go back to Chapter 4 in change in v0 as the strobe rate is increased and ∆t
Physics 2000, study the derivation of the constant reduced. But the change becomes smaller and it looks
acceleration formulas beginning on page 4-8, and as if we are approaching some final value of v0 that
work the projectile motion problems in the appendix to does not depend on the size of ∆t , provided ∆t is small
Chapter 4. Those who study all of this introduction to enough. It looks as if we have come close to the final
calculus should then proceed to the projectile motion value in Figure (1c).
problems in the appendix to Chapter 4 of the physics
text. The progression seen in Figure (1) is called a limiting
process. The idea is that there really is some true value
In Chapter 3 of Physics 2000, we used strobe photo- of v0 which we have called the instantaneous velocity,
graphs to define velocity and acceleration vectors. The and that we approach this true value for sufficiently
basic approach was to turn up the strobe flashing rate small values of ∆t . This is a calculus concept, and in
as we did in going from Figure (3-3) to (3-4) until all the language of calculus, we are taking the limit as ∆ t
the kinks are clearly visible and the successive dis- goes to zero.
placement vectors give a reasonable description of the
motion. We did not turn the flashing rate too high, for
the practical reason that the displacement vectors The Uncertainty Principle
became too short for accurate work. For over 200 years, from the invention of calculus by
Newton and Leibnitz until 1924, the limiting process
and the resulting concept of instantaneous velocity was
one of the cornerstones of physics. Then in 1924
Werner Heisenberg discovered what he called the
uncertainty principle which places a limit on the
accuracy of experimental measurements.
Cal 1-4 Calculus 2000

Heisenberg discovered something very new and unex- When we study the uncertainty principle in Chapter 40
pected. He found that the act of making an experimen- of the physics text, we will see that a measurement
tal measurement unavoidably affects the results of an which is accurate enough to show that Position (2) is
experiment. This had not been known previously below Position (1), could disturb the electron enough to
because the effect on large objects like golf balls is reverse its direction of motion. The next position
undetectable. But on an atomic scale where we study measurement could find the electron over where we
small systems like electrons moving inside an atom, the drew Position (3), or back where we drew Position (0),
effect is not only observable, it can dominate our study or anywhere in the region in between. As a result we
of the system. could not even determine what direction the electron is
moving. This uncertainty would not be the result of a
One particular consequence of the uncertainly prin- sloppy experiment, it is the best we can do with the most
ciple is that the more accurately we measure the accurate and delicate measurements possible.
position of an object, the more we disturb the motion of
the object. This has an immediate impact on the The uncertainty principle has had a significant impact
concept of instantaneous velocity. If we turn the strobe on the way physicists think about motion. Because we
all the way up, reduce ∆t to zero, we are in effect trying now know that the measuring process affects the results
to measure the position of the object with infinite of the measurement, we see that it is essential to provide
precision. The consequence would be an infinitely big experimental definitions to any physical quantity we
disturbance of the motion of the object we are studying. wish to study. A conceptual definition, like turning the
If we actually could turn the strobe all the way up, we strobe all the way up to define instantaneous velocity,
would destroy the object we were trying to study. can lead to fundamental inconsistencies.
It turns out that the uncertainty principle can have a Even an experimental definition like our strobe defini-
significant impact on a larger scale of distance than the tion of velocity can lead to inconsistent results when
atomic scale. Suppose, for example, that we con- applied to something like the electron in Figure (2). But
structed a chamber 1 cm on a side, and wished to study these inconsistencies are real. Their existence is telling
the projectile motion of an electron inside. Using us that the very concept of velocity is beginning to lose
Galileo’s idea that objects of different mass fall at the meaning for these small objects.
same rate, we would expect that the motion of the
electron projectile should be the same as more massive On the other hand, the idea of the limiting process and
objects. If we took a strobe photograph of the electron’s instantaneous velocity is very convenient when ap-
motion, we would expect get results like those shown plied to larger objects where the effects of the uncer-
in Figure (2). This figure represents projectile motion tainty principle are not detectable. In this case we can
with an acceleration g = 980 cm/sec2 and ∆t = .01sec, apply all the mathematical tools of calculus developed
as the reader can easily check. over the past 250 years. The status of instantaneous
velocity has changed from a basic concept to a useful
0 1 mathematical tool. Those problems for which this
-1 2 mathematical tool works are called problems in classi-
v1
cal physics; and those problems for which the uncer-
3
1 centimeter

tainty principle is important, are in the realm of what we

call quantum physics.

1 centimeter
Figure 2
Hypothetical electron projectile motion experiment.
 Introduction to Calculus Cal 1-5

CALCULUS DEFINITION OF VELOCITY The velocity vector vi is now given by

With the above perspective on the physical limita- ∆R i
vi = (2)
tions on the limiting process, we can now return to ∆t
the main topic of this chapter—the use of calculus in This is just our old strobe definition vi = s i /∆t , but
defining and working with velocity and accelera- using a notation which emphasizes that the displace-
tion. ment s i = ∆R i is the change in position that occurs
during the time ∆t. The Greek letter ∆ (delta) is used
In discussing the limiting process in calculus, one
both to represent the idea that the quantity ∆R i or ∆t
traditionally uses a special set of symbols which we can
is small, and to emphasize that both of these quantities
understand if we adopt the notation shown in Figure
change as we change the strobe rate.
(3). In that figure we have drawn the coordinate vectors
R i and R i+1 for the i th and (i + 1) th positions of the The limiting process in Figure (1) can be written in the
object. We are now using the symbol ∆R i to represent form
limit ∆R i
the displacement of the ball during the i to i+1 interval.
The vector equation for ∆R i is vi ≡ ∆t→0 (3)
∆t
∆R i = R i+1 – R i (1) where the word “limit” with ∆t→0 underneath, is to be
read as “limit as ∆t goes to zero”. For example we
In words, Equation (1) tells us that ∆R i is the change,
would read Equation (3) as “the instantaneous veloc-
during the time ∆t, of the position vector R describing
ity vi at position i is the limit, as ∆t goes to zero, of
the location of the ball.
the ratio ∆R i /∆t . ”

i For two reasons, Equation (3) is not quite yet in

∆R i standard calculus notation. One is that in calculus, only
i +1 the limiting value, in this case, the instantaneous veloc-
ity, is considered to be important. Our strobe definition
vi = ∆R i /∆t is only a step in the limiting process.
Therefore when we see the vector vi , we should
Ri assume that it is the limiting value, and no special
symbol like the underline is used. For this reason we
R i +1 will drop the underline and write
limit ∆R i
vi = ∆t→0 (3a)
∆t
∆R i = R i +1 – R i

V i = ∆R i
∆t

Figure 3
Definitions of ∆Ri and vi .
Cal 1-6 Calculus 2000

The second change deals with the fact that when ∆t notation for the limiting process we have been describ-
goes to zero we need an infinite number to time steps ing. But to a physicist, there is a different, more
to get through our strobe photograph, and thus it is not practical meaning. Think of dt as a short ∆t , short
possible to locate a position by counting time steps. enough so that the limiting process has essentially
Instead we measure the time t that has elapsed since the occurred, but not too short to see what is going on. In
beginning of the photograph, and use that time to tell us Figure (1), a value of dt less than .025 seconds is
where we are, as illustrated in Figure (4). Thus instead probably good enough.
of using vi to represent the velocity at position i, we
write v(t) to represent the velocity at time t. Equation If dt is small but finite, then we know exactly what the
(3) now becomes dR(t) is. It is the small but finite displacement vector
at the time t. It is our old strobe definition of velocity,
limit ∆R(t)
v(t) = ∆t→0 (3b) with the added condition that dt is such a short time
∆t interval that the limiting process has occurred. From
where we also replaced ∆R i by its value ∆R(t) at this point of view, dt is a real time interval and dR(t) a
time t. real vector, which we can work with in a normal way.
The only thing special about these quantities is that
Although Equation (3b) is in more or less standard when we see the letter d instead of ∆ , we must
calculus notation, the notation is clumsy. It is a pain to remember that a limiting process is involved. In this
keep writing the word “limit” with a ∆t→0 under- notation, the calculus definition of velocity is
neath. To streamline the notation, we replace the Greek
letter ∆ with the English letter d as follows
dR(t)
v(t) =
dt
limit ∆R(t) ≡ dR(t) (5)
∆t→0 ∆t dt (4) where R(t) and v(t) are the particle’s coordinate vector
and velocity vector respectively, as shown in Figure
(The symbol ≡ means defined equal to.) To a (5). Remember that this is just fancy shorthand nota-
mathematician, the symbol dR(t)/dt is just shorthand tion for the limiting process we have been describing.
t = .1sec t = .2sec

t = 0sec t = .3sec
V(t)

t = .4sec
R(t) at t = .3 sec R(t)

t = .5sec

Figure 4 Figure 5
Rather than counting individual images, we can Instantaneous position and velocity at time t.
locate a position by measuring the elapsed time t.
In this figure, we have drawn the displacement
vector R(t) at time t = .3 sec.
 Introduction to Calculus Cal 1-7

ACCELERATION Now go through the limiting process, turning the strobe

In the analysis of strobe photographs, we defined both up, reducing ∆t until the value of a(t) settles down to
a velocity vector v and an acceleration vector a. The its limiting value. We have

limit v t + ∆t – v t
definition of a , shown in Figure (2-12) reproduced calculus
here in Figure (6) was a(t) definition = ∆t→0
∆t
v –v (9)
a i ≡ i+1 i (6) ∆v(t)
∆t limit
= ∆t→0
In our graphical work we replaced vi by s i /∆t so that ∆t
we could work directly with the displacement vectors Finally use the shorthand notation d/dt for the limiting
s i and experimentally determine the behavior of the process:
acceleration vector for several kinds of motion.
dv(t) (10)
a(t) =
Let us now change this graphical definition of accelera- dt
tion over to a calculus definition, using the ideas just
applied to the velocity vector. First, assume that the ball Equation (10) does not make sense unless you remem-
reached position i at time t as shown in Figure (6). ber that it is notation for all the ideas expressed above.
Then we can write Again, physicists think of dt as a short but finite time
interval, and dv(t) as the small but finite change in the
vi = v(t)
velocity vector during the time interval dt. It’s our
vi+1 = v(t+∆t) strobe definition of acceleration with the added re-
quirement that ∆t is short enough that the limiting
to change the time dependence from a count of strobe
process has already occurred.
flashes to the continuous variable t. Next, define the
vector ∆v(t) by
Components
∆v(t) ≡ v(t+∆t) – v(t) = vi+1 – vi (7) Even if you have studied calculus, you may not recall
We see that ∆v(t) is the change in the velocity vector as encountering formulas for the derivatives of vectors,
the time advances from t to t+∆t . The strobe like dR(t)/∆t and dv(t)/∆t which appear in Equations
definition of a i can now be written (5) and (10). To bring these equations into a more
familiar form where you can apply standard calculus
strobe v(t + ∆t) – v(t) ∆v(t) formulas, we will break the vector Equations (5) and
a(t) definition = ≡ (8)
∆t ∆t (10) down into component equations.

In the chapter on vectors, we saw that any vector

equation like
position at
time t position at A = B+C (11)
time t + ∆t
Vi is equivalent to the three component equations
A x = Bx + Cx
( Vi+1–Vi ) Vi+1 A y = By + Cy
(12)
–Vi A z = Bz + Cz
a i = ( Vi+1–Vi )
∆t
The advantage of the component equations was that
they are simply numerical equations and no graphical
work or trigonometry is required.

Figure 6
Experimental definition of the acceleration vector.
Cal 1-8 Calculus 2000

The limiting process in calculus does not affect the INTEGRATION

decomposition of a vector into components, thus Equa- When we worked with strobe photographs, the photo-
tion (5) for v(t) and Equation (10) for a(t) become graph told us the position R(t) of the ball as time
passed. Knowing the position, we can then use Equa-
v(t) = dR(t)/dt (5) tion (5) to calculate the ball's velocity v(t) and then
Equation (10) to determine the acceleration a(t) . In
vx(t) = dR x(t)/dt (5a)
general, however, we want to go the other way, and
vy(t) = dR y(t)/dt (5b) predict the motion from a knowledge of the accelera-
vz(t) = dR z(t)/dt (5c) tion. For example, imagine that you were in Galileo's
and position, hired by a prince to predict the motion of
a(t) = dv(t)/dt (10) cannonballs. You know that a cannonball should not
be much affected by air resistance, thus the acceleration
a x(t) = dvx(t)/dt (10a) throughout its trajectory should be the constant gravi-
a y(t) = dvy(t)/dt (10b) tational acceleration g . You know that a(t) = g ; how
then do you use that knowledge in Equations (5) and
a z(t) = dvz(t)/dt (10c)
(10) to predict the motion of the ball?
Often we use the letter x for the x coordinate of the
vector R and we use y for R y and z for R z . With this The answer is that you cannot with the equations in
notation, Equation (5) assumes the shorter and perhaps their present form. The equations tell you how to go
more familiar form from R(t) to a(t), while to predict motion you need to
y go the other way, from a(t) to R(t) . The topic of this
vx(t) = dx(t)/dt R
(5a’) section is to see how to reverse the directions in which
vy(t) = dy(t)/dt (5b’) we use our calculus equations. Equations (5) and (10)
vz(t) = dz(t)/dt x (5c’) involve the process called differentiation. We will see
Figure 7 that when we go the other way the reverse of differen-
At this point the notation has become deceptively short. tiation is a process called integration. We will see that
You now have to remember that x(t) stands for the x integration is a simple concept, but a process that is
coordinate of the particle at a time t. sometimes hard to perform without the aid of a com-
puter.
We have finally boiled the notation down to the point
where it would be familiar from any calculus course. If
we restrict our attention to one dimensional motion
along the x axis. Then all we have to concern ourselves
with are the x component equations

dx(t)
vx(t) =
dt
(10a)
dvx(t)
a x(t) =
dt
 Introduction to Calculus Cal 1-9

Prediction of Motion ties we want for a calculus discussion. In Figure (9) we

In our earlier discussion, we have used strobe photo- improved the situation by cutting ∆t to 1/4 of its
graphs to analyze motion. Let us see what we can learn previous value, giving us four times as many images
from such a photograph for predicting motion. Figure and more accurate velocities vi .
(8) is our familiar projectile motion photograph show-
ing the displacement s of a ball during the time the ball We see that the displacement s is now the sum of 16
traveled from a position labeled (0) to the position vectors
labeled (4). If the ball is now at position (0) and each s = s 1 + s 2 + s 3 + ... + s 15 + s 16 (13)
of the images is .1 seconds apart, then the vector s tells
Expressing this in terms of the velocity vectors v1 to
us where the ball will be at a time of .4 seconds from
v16 we have
now. If we can predict s , we can predict the motion of
the ball. The general problem of predicting the motion s = v1 ∆t + v2 ∆t + v3 ∆t + ... + v15 ∆t + v16 ∆t (14)
of the ball is to be able to calculate s(t) for any time t.
or using our more compact notation
From Figure (8) we see that s is the vector sum of the 16
individual displacement vectors s 1 , s 2 , s 3 and s 4 s = Σ vi ∆t
i=1
(14a)

s = s1 + s2 + s3 + s4 (11) While Equation (14) for s looks quite different than

We can then use the fact that s 1 = v1∆t , s 2 = v2∆t , Equation (12), the sum of sixteen vectors instead of
etc. to get four, the displacement vectors s in the two cases are
exactly the same. Adding more intermediate images
s = v1∆t + v2∆t + v3∆t + v4∆t (12) did not change where the ball was located at the time of
Rather than writing out each term, we can use the t = .4 seconds. In going from Equation (12) to (14),
summation sign Σ to write what has changed in shortening the time step ∆t , is that
4 the individual velocity vectors vi become more nearly
s = Σ vi∆t
i=1
(12a) equal to the instantaneous velocity of the ball at each
image.
Equation (12) is approximate in that the vi are approxi-
mate (strobe) velocities, not the instantaneous veloci-
0 S1 1 S1 0 1 2 3 4
S2
t=0 2 t=0 8
S3

3 12
S S

S4
S 16
S = S 1 + S 2 + ... + S 16 S = S 1 + S 2 + ... + S 16
4 16
t=.4 sec t=.4 sec

Figure 8 Figure 9
To predict the total displacement With a shorter time interval, we
s , we add up the individual add up more displacement vectors
displacements s i . to get the total displacement s .
Cal 1-10 Calculus 2000

If we reduced ∆t again by another factor of 1/4, so that Thus the displacement s has x and y components
we had 64 images in the interval t = 0 to t = .4 sec, the
s x = x(t f) – x(t i)
formula for s would become
64 s y = y(t f) – y(t i)
s = Σ
i=1
vi ∆t (15a)
Breaking Equation (17) into component equations
where now the vi are still closer to representing the gives
ball's instantaneous velocity. The more we reduce ∆t , tf
the more images we include, the closer each vi comes s x = x(t f) – x(t i) = vx(t)dt (18a)
ti
to the instantaneous velocity v t . While adding more
images gives us more vectors we have to add up to get
tf
the total displacement s , there is very little change in s y = y(t f) – y(t i) = vy(t)dt
our formula for s . If we had a million images, we ti (18b)
would simply write
1000000 Here we will introduce one more piece of notation
s = Σ
i=1
vi ∆t (16a) often used in calculus courses. On the left hand side of
Equation (18a) we have x(t f) – x(t i) which we can
In this case the vi would be physically indistinguish- think of as the variable x(t) evaluated over the interval
able from the instantaneous velocity v(t) . We have of time from t i to t f . We will often deal with variables
essentially reached a calculus limit, but we have prob- evaluated over some interval and have a special nota-
lems with the notation. It is clearly inconvenient to tion for that. We will write
label each vi and then count the images. Instead we
tf
would like notation that involves the instantaneous x(t f) – x(t i) ≡ x(t)
velocity v(t) and expresses the beginning and end ti
(19)
points in terms of the initial time t 1 and final time t f ,
rather than the initial and final image numbers i. You are to read the symbol x(t) ttfi as "x of t evaluated
from t i to t f ". We write the initial time t i at the bottom
In the calculus notation, we replace the summation sign of the vertical bar, the final time t f at the top.
Σ by something that looks almost like the summation
sign, namely the integral sign . (The French word i
yi
for integration is the same as their word for summa-
tion.) Next we replaced the individual vi by the
continuous variable v(t) and finally express the end
points by the initial time t i and the final time t f . The
(y – yi )

result is
S
f

n as the number tf
s = Σ vi ∆t →
i=1
n becomes
infinitely
ti
v(t)dt (17)
large

yf f
Calculus notation is more easily handled, or is at least
more familiar, if we break vector equations up into
component equations. Assume that the ball started at (x f – xi )
position i which has components x i = x(t i) [read x(t i) xi xf
as “ x at time t i ” ] and y i = y(t i) as shown in Figure x(t i ) x(t f)
(10). The final position f is at x f = x(t f) and y f = y(t f) . Figure 10
Breaking the vector s into components.
 Introduction to Calculus Cal 1-11

We use similar notation for any kind of variable, for To get this interpretation, let us start with the simple
example case of a ball moving in a straight line, for instance, the
x2 x direction, at a constant velocity vx . A strobe picture
f(x) ≡ f(x 2) – f(x 1) (19a) of this motion would look like that shown in Figure
x1
(11a).
Remember to subtract the variable when evaluated at
the value at the bottom of the vertical bar. Figure (11b) is a graph of the ball's velocity vx(t) as a
function of the time t. The vertical axis is the value of
With this notation, our Equation (18) can be written vx , the horizontal axis is the time t. Since the ball is
traveling at constant velocity, vx has a constant value
tf tf
s x = x(t) = vx(t)dt (18 a′) and is thus represented by a straight horizontal line. In
ti ti order to calculate the distance that the ball has traveled
during the time interval from t i to t f , we need to
tf tf evaluate the integral
s y = y(t) = vy(t)dt (18 b′)
ti ti distance ball
tf travels in
Calculating Integrals sx = vx(t)dt
ti time interval (18a)
Equation (20) is nice and compact, but how do you use t i to t f
it? How do you calculate integrals? The key is to To actually evaluate the integral, we will go back to our
remember that an integral is just a fancy notation for a summation notation
sum of terms, where we make the time step ∆(t) very i final
small. Keeping this in mind, we will see that there is a s x = Σ vxi ∆t (20)
very easy way to interpret an integral. i initial

ti tf and show individual time steps ∆t in the graph of vx

versus t, as in Figure (11c).

x We see that each term in Equation (20) is represented

Figure 11a
in Figure (11c) by a rectangle whose height is vx and
Strobe photograph of ball moving at whose width is ∆t . We have shaded in the rectangle
constant velocity in x direction. representing the 7th term vx 7∆t . We see that vx 7∆t is
vx (t) just the area of the shaded rectangle, and it is clear that
the sum of all the areas of the individual rectangles is the
vx total area under the curve, starting at time t i and ending
at time t f . Here we are beginning to see that the process
of integration is equivalent to finding the area under a
t curve.
ti tf
Figure 11b With a simple curve like the constant velocity vx(t) in
Graph of vx(t) versus t for the ball of Figure 11a. Figure (11c), we see by inspection that the total area
from t i to t f is just the area of the complete rectangle
vx (t)
of height vx and width (t f – t i) . Thus
vx7
vx
s x = vx × (t f – t i) (21)
This is the expected result for constant velocity, namely
∆t
t
ti tf distance = velocity × time for
traveled
constant (21a)
Figure 11c velocity
Each vx ∆ t is the area of a rectangle.
Cal 1-12 Calculus 2000

To see that you are not restricted to the case of constant Thus we can interpret the integral of a curve as the area
velocity, suppose you drove on a freeway due east (the under the curve even when the curve is not constant or
x direction) starting at 9:00 AM and stopping for lunch flat. Mathematicians concern themselves with curves
at 12 noon. Every minute during your trip you wrote that are so wild that it is difficult or impossible to
down the speedometer reading so that you had an determine the area under them. Such curves seldom
accurate plot of vx(t) for the entire morning, a plot like appear in physics problems.
that shown in Figure (12). From such a plot, could you
determine the distance s x that you had travelled? While the basic idea of integration is simple—just
finding the area under a curve–in practice it can be quite
Your best answer is to multiply each value vi of your difficult to calculate the area. Much of an introductory
velocity by the time ∆t to calculate the average dis- calculus course is devoted to finding the formulas for
tance traveled each minute. Summing these up from the areas under various curves. There are also books
the initial time t i = 9:00AM to the final time t f = noon , called tables of integrals where you look up the for-
you have as your estimate mula for a curve and the table tells you the formula for
the area under that curve.
sx ≈ Σi vxi ∆t
In Chapter 16 of the physics text, we will discuss a
(The symbol ≈ means approximately equal.) mathematical technique called Fourier analysis. This
To get a more accurate value for the distance traveled, is a technique in which we can describe the shape of any
you should measure your velocity at shorter time continuous curve in terms of a sum of sin waves. (Why
intervals ∆t and add up the larger number of smaller we want to do that will become clear then.) The process
rectangles. The precise answer should be obtained in of Fourier analysis involves finding the area under
the limit as ∆t goes to zero some very complex curves, curves often involving
experimental data for which we have no formula, only
tf
s x = limit Σ v ∆t
∆t → 0 i xi
=
ti
vx(t)dt (22)
graphs. Such curves cannot be integrated by using a
table of integrals, with the result that Fourier analysis
was not widely used until the advent of the modern
This limit is just the area under the curve that is digital computer.
supposed to represent the instantaneous velocity vx(t) .
vx(t) The computer made a difference, because we can find
the area under almost any curve by breaking the curve
vx7 into short pieces of length ∆t , calculating the area vi∆t
of each narrow rectangle, and adding up the area of the
rectangles to get the total area. If the curve is so wild
t that we have to break it into a million segments to get
9am ∆t noon an accurate answer, that might be too hard to do by
Figure 12 hand, but it usually a very simple and rapid job for a
Plot of vx(t) for a trip starting at 9:00 AM and computer. Computers can be much more efficient than
finishing at noon. The distance traveled is the people at integration.
area under the curve.
 Introduction to Calculus Cal 1-13

The Process of Integrating Since (t f – t i ) = t t f dt , we can replace (t f – t i ) in

i
There is a language for the process of integration which Equation (24) by the integral to get
we will now take you through. In each case we will
tf tf
check that the results are what we would expect from vx dt = vx dt vx a constant (25)
our summation definition, or the idea that an integral is ti ti
the area under a curve. and we see that a constant like vx can be taken outside
The simplest integral we will encounter in the calcula- the integral sign.
tion of the area under a curve of unit height as shown in Let us try the simplest case we can think of where vx
Figure (13). We have the area of a rectangle of height is not constant. Suppose vx starts at zero at time t i = 0
1 and length (t f – t i) and increases linearly according to the formula
tf tf
1 dt = dt = (t f – t i ) vx = at (26)
ti ti (22)
vx
1 at f
area = 1(t f – t i) at
vx =
t
ti tf t
Figure 13 0 tf
Area under a curve of unit height. Figure 15

When we get up to the time t f the velocity will be (at f)

We will use some special language to describe this as shown in Figure (15). The area under the curve
integration. We will say that the integral of dt is simply vx = at is a triangle whose base is of length t f and
the time t, and that the integral of dt from t i to t f is height is at f . The area of this triangle is one half the
equal to t evaluated from t i to t f . In symbols this is base times the height, thus we get for the distance s x
written as traveled by an object moving with this velocity
tf tf tf
dt = t = (t f – t i ) (23) sx = vx dt = 1 (base) × (altitude)
ti ti 0 2
Recall that the vertical line after a variable means to (27)
= 1 (t f)(at f) = 1 at f 2
evaluate that variable at the final position t f (upper 2 2
value), minus that variable evaluated at the initial Now let us repeat the same calculation using the
position t i (lower value). Notice that this prescription language one would find in a calculus book. We have
gives the correct answer.
tf tf
The next simplest integral is the integral of a constant, sx = vx dt = (at)dt (28)
0 0
like a constant velocity vx over the interval t i to t f
tf The constant (a) can come outside, and we know that
vx dt = vx (t f – t i ) (24) the answer is 1/2at f 2 , thus we can write
ti
tf
vx sx = a tdt = 1 at f 2 (29)
0 2
area = vx(t f – t i)
In Equation (29) we can cancel the a's to get the result
t tx
ti tf tdt = 1 t f 2 (30)
0 2
Figure 14
Area under the constant vx curve.
Cal 1-14 Calculus 2000

In a calculus text, you would find the statement that the n=3 t3
integral tdt is equal to t 2/2 and that the integral 4
should be evaluated as follows t 3dt = t
4
t
tf t2 t f tf2 0 t2 (33a,b,c,d)
tdt = = – = f (31)
0 2 2 2 2 Looking at the way these integrals are turning out, we
0
suspect that the general rule is
Indefinite Integrals n+1
When we want to measure an actual area under a curve, t n dt = t (34)
n+1
we have to know where to start and stop. When we put
these limits on the integral sign, like t i and t f , we have It turns out that Equation (34) is a general result for any
what is called a definite integral. However there are value of n except n = –1. If n = –1, then you would
times where we just want to know what the form of the have division by zero, which cannot be the answer.
integral is, with the idea that we will put in the limits (We will shortly discuss the special case where n = –1.)
later. In this case we have what is called an indefinite
integral, such as As long as we stay away from the n = –1 case, the
2 formula works for negative numbers. For example
tdt = t indefinite integral (32)
2 t – 2dt = dt = t (– 2 +1) = t –1
t2 –2 + 1 (–1)
The difference between our definite integral in Equa-
tion (31) and the indefinite one in Equation (32) is that dt = – 1
we have not chosen the limits yet in Equation (32). If t (35)
t2
possible, a table of integrals will give you a formula for
the indefinite integral and let you put in whatever limits In our discussion of gravitational and electrical poten-
you want. tial energy, we will encounter integrals of the form seen
in Equation (35).
Integration Formulas
Exercise 1
For some sets of curves, there are simple formulas for
the area under them. One example is the set of curves Using Equation (34) and the fact that constants can
come outside the integral, evaluate the following inte-
of the form t n . We have already considered the cases
grals:
where n = 0 and n = 1.
n=0 1 it does not matter whether
(a) xdx we call the variable t or x

t 0dt = dt = t
t x=2
(b) x5dx also sketch the area
x=1 being evaluated
n=1 t
2 t =2
t 1dt = tdt = t (c) dt Show that you get
2 t =1 t2 a positivearea.
t
Some results we will prove later are
(d) GmMdr where G, m, and M
n=2 t2 r2 are constants
3
t 2dt = t
3 a dy
t
(e) (a) is a constant
y3 / 2
 Introduction to Calculus Cal 1-15

NEW FUNCTIONS Two of the important but peculiar features of the

Logarithms natural logarithm are
We have seen that when we integrate a curve or ln(ab) = ln(a) + ln(b) (37)
function like t 2 , we get a new function t 3/3 . The
functions t 2 and t 3 appear to be fairly similar; the ln( 1a ) = – ln(a) (38)
integration did not create something radically differ-
ent. However, the process of integration can lead to Thus we get, for example
some curves with entirely different behavior. This
ln(t f ) – ln(t i ) = ln(t f ) + ln 1
happens, for example, in that special case n = –1 when ti
we try to do the integral of t – 1 .
tf (39)
It is certainly not hard to plot t – 1 , the result is shown = ln
ti
in Figure (16). Also there is nothing fundamentally
difficult or peculiar about measuring the area under the Thus the area under the curve in Figure (16) is
t – 1 curve from some t i to t f , as long as we stay away tf
from the origin t = 0 where t – 1 blows up. The formula dt = ln t f
ti t ti (40)
for this area turns out, however, to be the new function
called the natural logarithm, abbreviated by the sym- While the natural logarithm has some rather peculiar
bol ln. The area in Figure (16) is given by the formula properties it is easy to evaluate because it is available on
all scientific calculators. For example, if t i = .5 seconds
tf
1 dt = ln(t ) – ln(t ) (36) and t f = 4 seconds, then we have
t f i
ti
t
ln tf = ln 4 = ln (8) (41)
i .5
t –1
Entering the number 8 on a scientific calculator and
pressing the button labeled ln, gives

curve 1t ln (8) = 2.079 (42)

which is the answer.

t Exercise 2
ti tf Evaluate the integrals
Figure 16
Plot of t – 1. The area under this curve 1000
dx 1
dx
is the natural logarithm ln. .001 x .000001 x

Why are the answers the same?

Cal 1-16 Calculus 2000

The Exponential Function The inverse of the exponent to the base 10 is the
We have just seen that, while the logarithm function function called logarithm to the base 10 which is
may have some peculiar properties, it is easy to evalu- denoted by the key labeled log on a scientific calcula-
ate using a scientific calculator. The question we now tor. Formally this means that
want to consider is whether there is some function that
undoes the logarithm. When we enter the number 8 log 10 y = y (46)
into the calculator and press ln, we get the number
2.079. Now we are asking if, when we enter the number Check this out on your scientific calculator. For
2.079, can we press some key and get back the number example, enter the number 1,000,000 and press the log
8? The answer is, you press the key labeled e x . The button and see if you get the number 6. Try several
e x key performs the exponential function which examples so that you are confident of the result.
undoes the logarithm function. We say that the expo-
nential function e x is the inverse of the logarithm The Exponential Function yx
function ln.
Another key on your scientific calculator is labeled y x .
This allows you to determine the value of any number
Exponents to the Base 10 y raised to the power (or exponent) x. For example,
You are already familiar with exponents to the base 10, enter the number y = 10, and press the y x key. Then
as in the following examples enter the number x = 6 and press the = key. You should
see the answer
10 0 = 1
10 1 = 10 10 – 1 = 1/10 = .1 y x = 10 6 = 1000000
10 2 = 100 10 – 2 = 1/100 = .01 (43)
⋅⋅⋅⋅⋅⋅⋅ ⋅⋅⋅⋅⋅⋅⋅ It is quite clear that all exponents obey the same rules
10 6 = 1,000,000 10 – 6 = .000001 we saw for powers of 10, namely
The exponent, the number written above the 10, tells us
ya × yb = ya + b (47)
how many factors of 10 are involved. A minus sign
means how many factors of 10 we divide by. From this
alone we deduce the following rules for the exponent to (Example y 2 × y 3 = y × y y × y × y = y 5 .)
the base 10.
And as before
10 –a
= 1a (44)
10 y – a ≡ 1a (48)
y

10 a × 10 b = 10 a + b (45)

(Example 10 2 × 10 3 = 100 × 1000 = 100,000 .)

 Introduction to Calculus Cal 1-17

Exercise 3 Euler's Number e = 2.7183. . .

Use your scientific calculator to evaluate the following We have seen that the function log on the scientific
quantities. (You should get the answers shown.) calculator undoes, is the inverse of, powers of 10. For
example, we saw that
(a) 106 (1000000)
(b) 2 3 (8) log 10 x = x (46 repeated)
0
(c) 23 (1)
(d) 10 –1
(.1) Example: log 10 6 = 6
(To do this calculation, enter 10, then press yx . Then
Earlier we saw that the exponential function e x was the
enter 1, then press the +/– key to change it to –1, then
inverse of the natural logarithm ln. This means that
press = to get the answer .1)
(e) 2– .5 (1/ 2= .707) ln e x = x (49)
(f) log (10) (1)
(g) ln (2.7183) (1) (very close to 1) The difference between the logarithm log and the
natural logarithm ln, is that log undoes exponents of
Try some other examples on your own to become
completely familiar with the yx key. (You should note the number 10, while ln undoes exponents of the
that any positive number raised to the 0 power is 1. Also, number e. This special number e, one of the fundamen-
some calculators, in particular the one I am using, tal mathematical constants like π , is known as Euler's
cannot handle any negative values of y, not even (– 2)2 number, and is always denoted by the letter e.
which is +4)
You can find the numerical value of Euler's number e
on your calculator by evaluating
e1 = e (50)
To do this, enter 1 into your calculator, press the e x
key, and you should see the result
e 1 = e = 2.718281828 (51)
We will run into this number throughout the course.
You should remember that e is about 2.7, or you might
even remember 2.718. (Only remembering e as 2.7 is
as klutzy as remembering π as 3.1)

The terminology in math courses is that the function

log, which undoes exponents of the number 10, is the
logarithm to the base 10. The function ln, what we
have called the natural logarithm, which undoes
exponents of the number e, is the logarithm to the
base e. You can have logarithms to any base you
want, but in practice we only use base 10 (because
we have 10 fingers) and the base e. The base e is
special, in part because that is the logarithm that
naturally arises when we integrate the function 1/x.
We will see shortly that the functions ln and e x
have several more, very special features.
Cal 1-18 Calculus 2000

DIFFERENTIATION AND INTEGRATION While Equation (53) looks like it is applied to the
The scientific calculator is a good tool for seeing how explicit case of the strobe photograph of projectile
the functions like ln and e x are inverse of each other. motion, it is easily extended to cover any process of
Another example of inverse operations is integration differentiation. Whatever function we have [we had
and differentiation. We have seen that integration R(t), suppose it is now f(t)], evaluate it at two closely
allows us to go the other way from differentiation spaced times, subtract the older value from the newer
[finding x(t) from v(t), rather than v(t) from x(t)]. one, and divide by the time difference ∆t. Taking the
However it is not so obvious that integration and limit as ∆t becomes very small gives us the derivative
differentiation are inverse operations when you think
of integration as finding the area under a curve, and d f(t) limit f(t + ∆t) – f(t)
≡ ∆t→0 (54)
differentiation as finding limits of ∆x/∆t as ∆t goes to dt ∆t
zero. It is time now to make this relationship clear. The variable with which we are differentiating does not
have to be time t. It can be any variable that we can
First, let us review our concept of a derivative. Going divide into small segments, such as x;
back to our strobe photograph of Figure (3), replacing
R i by R(t) and R i+1 by R(t+∆t) , as shown in Figure
(3a), our strobe velocity was then given by d f(x) ≡ limit f(x + ∆x) – f(x) (55)
dx ∆x→ 0 ∆x
R(t+∆t) – R(t)
v(t) = (52)
∆t
Let us see how the operation defined in Equation (55)
The calculus definition of the velocity is obtained by is the inverse of finding the area under a curve.
reducing the strobe time interval ∆t until we obtain the
instantaneous velocity v . Suppose we have a curve, like our old vx(t) graphed as
a function of time, as shown in Figure (17). To find out
how far we traveled in a time interval from t i to some
limit R(t + ∆t) – R(t)
vcalculus = ∆t→ (53) later time T, we would do the integral
0 ∆t
T
x(T) = vx(t) dt (56)
ti
i
∆R = R(t+∆t) – R(t) The integral in Equation (56) tells us how far we have
i +1 gone at any time T during the trip. The quantity x(T)
is a function of this time T.

vx (t)
R(t)

R(t+∆t)
x(T)
t
ti T
Figure 17
R(t+∆t) – R(t)
V(t) = ∆R = The distance traveled by the time T is the area under
∆t ∆t
the velocity curve up to the time T.

Figure 3a
Defining the strobe velocity.
 Introduction to Calculus Cal 1-19

Now let us differentiate the function x(T) with respect The rectangle has a height approximately v(T) and a
to the variable T. By our definition of differentiation width ∆T for an area
we have
x(T + ∆T) – x(T) = vx(T)∆t (58)
d x(T) = limit x(T + ∆T) – x(T) (57)
dT ∆T→ 0 ∆T Dividing through by ∆T gives
Figure (17) shows us the function x(T). It is the area x(T + ∆T) – x(T)
vx(T) = (59)
under the curve v(t) starting at t i and going up to time ∆T
t = T. Figure (18) shows us the function x(T + ∆T) . It
The only approximation in Equation (59) is at the top
is the area under the same curve, starting at t i but going
of the rectangle. If the curve is not flat, vx(T + ∆T)
up to t = T + ∆t . When we subtract these two areas, all
will be different from vx(T) and the area of the sliver
we have left is the area of the slender rectangle shown
will have a value somewhere between vx(T)∆t and
in Figure (19).
vx(T + ∆t)∆t . But if we take the limit as ∆T goes to
vx (t) zero, the value of vx(T + ∆T) must approach vx(T) ,
and we end up with the exact result

limit x(T + ∆t) – x(T)

vx(T) = ∆t→0 (60)
x(T) ∆t
t This is just the derivative dx(t)/dt evaluated at t = T.
ti T
Figure 17 repeated dx(t)
The distance x(T) traveled by the time T vx(T) = (61a)
dt t=T

vx (t) where we started from

T
x(T) = vx(t) dt (61b)
x(T+∆t) ti

t
ti T T+∆t Equations (61a) and (61b) demonstrate explicitly how
Figure 18 differentiation and integration are inverse operations.
The distance x (T+ ∆ t ) traveled by the time T+ ∆ t . The derivative allowed us to go from x(t) to vx(t) while
the integral took us from vx(t) to x(t). This inverse is
not as simple as pushing a button on a calculator to go
vx (t)
from ln to e x . Here we have to deal with limits on the
vx (T) integration and a shift of variables from t to T. But
these two processes do allow us to go back and forth.
vx (T)∆t

t
Figure 18 T T+∆t
The distance x (T+ ∆ t ) – x(T)
traveled during the time ∆ t .
Cal 1-20 Calculus 2000

A Fast Way to go Back and Forth however, we say that we started our trip at x(t i) = 0 ,
We introduced our discussion of integration by point- then we get the result
ing out that equations T
x(T) = vx(t)dt (67)
dx(t) dvx(t) ti
vx(t) = ; a x(t) = (62a,b)
dt dt
representing the distance traveled since the start of the
went the wrong way in that we were more likely to trip.
know the acceleration a x(t) and from that want to
calculate the velocity vx(t) and distance traveled x(t).
Constant Acceleration Formulas
After many steps, we found that integration was what
we needed. The constant acceleration formulas, so well known
from high school physics courses, are an excellent
We do not want to repeat all those steps. Instead we application of the procedures we have just described.
would like a quick and simple way to go the other way
around. Here is how you do it. Think of the dt in (62a) We will begin with motion in one dimension. Suppose
as a small but finite time interval. That means we can a car is traveling due east, in the x direction, and for a
treat it like any other number and multiply both sides of while has a constant acceleration a x . The car passes us
Equation (62a) through by it. at a time t i = 0 , traveling at a speed vx0 . At some later
time T, if the acceleration a x remains constant, how far
dx(t) away from us will the car be?
vx(t) =
dt
dx(t) = vx(t)dt (63) We start with the equation
Now integrate both sides of Equation (63) from some dvx (t)
a x (t) = (68)
initial time t i to a final time T. (If you do the same thing dt
to both sides of an equation, both sides should still be Multiplying through by dt to get
equal to each other.) dvx(t) = a x(t)dt
T T then integrating from time t i = 0 to time tf = T, we get
dx(t) = vx(t) dt (64)
ti ti T T
dvx (t) = a x (t)dt (69)
If dt is to be thought of as a small but finite time step, 0 0
then dx(t) is the small but finite distance we moved in
Since the integral dvx (t) = vx (t) , we have
the time dt. The integral on the left side of Equation
(64) is just the sum of all these short distances moved, T T
dvx (t) = vx (t) = vx (T) – vx (0) (70)
which is just the total distance moved during the time 0 0
from t i to T.
where vx (0) is the velocity vx0 of the car when it
T T passed us at time t = 0.
dx(t) = x(t) = x(T) – x (ti) (65)
ti ti
While we can always do the left hand integral in
Thus we end up with the result Equation (69), we cannot do the right hand integral
T T until we know a x (t) . For the constant acceleration
x(t) = vx(t)dt
ti ti (66) problem, however, we know that a x (t) = a x is con-
stant, and we have
Equation (66) is a little more general than (62b) for it T T
allows for the fact that x(t i) might not be zero. If, a x (t)dt = a x dt (71)
0 0
 Introduction to Calculus Cal 1-21

Since constants can come outside the integral sign, we One of the results of integration that you should prove
get for yourself (just sketch the areas) is the rule
T T T f f f
a x dt = a x dt = a x t = axT (72) a(x) + b(x) dx = a(x)dx + b(x)dx (78)
0 0 0 i i i
where we used dt = t . Substituting Equations (70) thus we get
and (72) in (69) gives
T T T
(vx0 + a xt)dt = vx0 dt + a xt dt (79)
vx T – vx0 = a x T (73) 0 0 0

Since Equation (73) applies for any time T, we can Since constants can come outside the integrals, this is
replace T by t to get the well known result equal to
T T T
vx(t) = vx0 + a xt (a x constant) (74) (vx0 + a xt)dt = vx0 dt + a x t dt (80)
0 0 0

Equation (74) tells us the speed of the car at any time t Earlier we saw that
after it passed us, as long as the acceleration remains T T
constant. dt = t = T–0 = T (23)
0 0
To find out how far away the car is, we start with the
equation T
T t2 T2 T2
dx(t) tdt = = –0 = (30)
vx(t) = (62a) 0 2 2 2
0
dt
Multiplying through by dt to get Thus we get
dx(t) = vx(t) dt T 1
(vx0 + a xt)dt = vx0T + a xT2 (81)
then integrating from time t = 0 to time t = T gives 0 2
(as we saw earlier) Using Equations (76) and (81) in (75) gives
T T
dx(t) = vx(t) dt (75) 1
0 0 x(T) – x 0 = vx0T + a xT2
2
The left hand side is Taking x 0 = 0 and replacing T by t gives the other
constant acceleration formula
T T
dx(t) = x(t) = x(T) – x(0) (76) 1
0 0
x(t) = vx0t + a xt2 (a x constant) (82)
2
If we measure along the x axis, starting from where we
are (where the car was at t = 0) then x(0) = 0.
You can now see that the factor of t 2/2 in the constant
In order to do the right hand integral in Equation (75), acceleration formulas comes from the integral tdt .
we have to know what the function vx(t) is. But for
constant acceleration, we have from Equation (74)
vx(t) = vx0 + a xt , thus
T T
vx(t) dt = (vx 0 + a xt) dt (77)
0 0
Cal 1-22 Calculus 2000

Exercise 4 Then repeat, for each pair of equations, the steps that led
Find the formula for the velocity v(t) and position x(t) for
to the constant acceleration formulas for motion in the
a car moving with constant acceleration ax , that was x direction. The results will be
located at position xi at some initial time ti .
x(t) = vx0t + 1 a xt 2 vx(t) = vx0 + a xt
2
Start your calculation from the equations
y(t) = vy0t + 1 a yt 2 vy(t) = vy0 + a yt (84)
vx(t) =
dx(t) 2
dt 1
z(t) = vz0t + a zt 2 vz(t) = vz0 + a zt
dvx(t) 2
ax(t) =
dt The final step is to combine these six equations into the
and go through all the steps that we did to get Equations two vector equations
(74) and (82). See if you can do this without looking at
the text. x(t) = v0t + 1 at 2 ; v(t) = v0 + at (85)
2
If you have to look back to see what some steps are, then
finish the derivation looking at the text. Then a day or so These are the equations we analyzed graphically in
later, clean off your desk, get out a blank sheet of paper, Chapter 3 of the physics text, in Figure (3-34) and
write down this problem, put the book away and do the Exercise (3-9). (There we wrote s instead of x(t) , and
derivation. Keep doing this until you can do the deriva- vi rather than v0 .)
tion of the constant acceleration formulas without look-
ing at the text. In many introductory physics courses, considerable
emphasis is placed on solving constant acceleration
problems. You can spend weeks practicing on solving
Constant Acceleration Formulas these problems, and become very good at it. However,
in Three Dimensions when you have done this, you have not learned very
To handle the case of motion with constant accelera- much physics because most forms of motion are not
tion in three dimensions, you start with the separate with constant acceleration, and thus the formulas do
equations not apply. The formulas were important historically,
for they were the first to allow the accurate prediction
dx(t) dvx(t) of motion (of cannonballs). But if too much emphasis
vx(t) = a x(t) =
dt dt is placed on these problems, students tend to use them
dy(t) dvy(t) where they do not apply. For this reason we have
vy(t) = a y(t) = placed the exercises using the constant acceleration
dt dt
equations in an appendix at the end of chapter 4 of the
dz(t) dvz(t)
vz(t) = a z(t) = (83) physics text. There are plenty of problems there for all
dt dt the practice you will need with these equations. Doing
these exercises requires only algebra, there is no prac-
tice with calculus. To get some experience with
calculus, be sure that you can confidently do Exercise
4.
 Introduction to Calculus Cal 1-23

MORE ON DIFFERENTIATION When α is a number much smaller than 1 (α < < 1) ,

In our discussion of integration, we saw that the basic we can neglect α 2 compared to α (if α = .01,
idea was that the integral of some curve or function f(t) α 2 = .0001 ), with the result that we can accurately
was equal to the area under that curve. That is an easy approximate (x + α) n by
enough concept. The problems arose when we actu-
(x + α) n ≈ x n + nαx n–1 α << 1 (87)
ally tried to find the formulas for the areas under various
curves. The only areas we actually calculated were the Equation (87) gives us all the approximation formulas
rectangular area under f(t) = constant and the triangular found in Equations (1-20) through (1-25) on page 1-28
area under f(t) = at. It was perhaps a surprise that the of the physics text.
area under the simple curve 1/t should turn out to be a
logarithm. As an example of Equation (87), just to see that it
works, let us take x = 5, n = 7 and α = .01 to calculate
For differentiation, the basic idea of the process is given (5.01) 7 . From the calculator we get
by the formula
(5.01) 7 = 79225.3344 (88)
df(t) limit f t + ∆t – f(t)
= ∆t→0 (54 repeated)
dt ∆t (To do this enter 5.01, press the y x button, then enter
7 and press the = button.) Let us now see how this result
Equation (54) is short hand notation for a whole series compares with
of steps which we introduced through the use of strobe
photographs. The basic idea of differentiation is more (x + α) n ≈ x n + nαx n – 1
complex than integration, but, as we will now see, it is (89)
(5 + .01) 7 ≈ 5 7 + 7(.01)5 6
often a lot easier to find the derivative of a curve than
its integral. We have
5 7 = 78125 (90)
Series Expansions
7 × .01 × 5 6 = 7 × .01 × 15625 = 1093.75 (91)
An easy way to find the formula for the derivative of a
curve is to use a series expansion. We will illustrate the Adding the numbers in (90) and (91) together gives
process by using the binomial expansion to calculate
the derivative of the function x n where n is any 5 7 + 7(.01)5 6 = 79218.75 (92)
constant.
Thus we end up with 79218 instead of 79225, which is
We used the binomial expansion, or at least the first two not too bad a result. The smaller α is compared to one,
terms, in Chapter 1 of the physics text. That was during the better the approximation.
our discussion of the approximation formulas that are
useful in relativistic calculations. As we mentioned in
Exercise (1-5), the binomial expansion is
n(n – 1) 2 n – 2
(x + α) n = x n + nαx n – 1 + α x ⋅⋅⋅
2!
(86)
Cal 1-24 Calculus 2000

Derivative of the Function x n In our discussion of integration, we saw that a constant

could come outside the integral. The same thing
We are now ready to use our approximation formula
happens with a derivative. Consider, for example,
(87) to calculate the derivative of the function x n .
From the definition of the derivative we have d a f(x + ∆x) – a f(x)
a f(x) = limit
∆x→0
n
d(x ) limit (x + ∆x) – x
n n dx ∆x
= (93)
dx ∆x→0 ∆x
Since the constant a has nothing to do with the limiting
Since ∆x is to become infinitesimally small, we can process, this can be written
use our approximation formula for ( x + α )n . We get
d f(x + ∆x) – f(x)
af(x) = a limit
∆x→0
n n
( x + α ) ≈ x + n(α)x n–1
(α << 1) dx ∆x

( x + ∆x)n ≈ x n + n(∆x)x n–1 (∆x << 1) (94) (99)

df(x)
= a
dx
Using this in Equation (93) gives
Exercise 5
d(x n) x n + n(∆x)x n–1 – x n Calculate the derivative with respect to x (i.e., d/dx) of
limit
= ∆x→0 (95)
dx ∆x the following functions. (When negative powers of x are
involved, assume x is not equal to zero.)
We used an equal sign rather than an approximately
equal sign in Equation (95) because our approximation (a) x
formula (94) becomes exact when ∆x becomes infini-
tesimally small. (b) x2

In Equation (95), the terms x n cancel and we are left (c) x3

with (d) 5x2 – 3x
n–1
d(x n) limit n(∆x)x (Before you do part (d), use the definition of the deriva-
= (96)
dx ∆x→0 ∆x tive to prove that d f(x) + g(x) =
df(x) dg(x)
+ )
dx dx dx
At this point, the factors ∆x cancel and we have (e) x– 1
d(x n)
= limit
∆x→0
nx n–1 (97) (f) x– 2
dx
Since no ∆x's remain in our formula, we end up with
(g) x
the exact result
(h) 1/ x
d(x n)
= nx n–1 (98)
dx (i) 3x.73
Equation (98) is the general formula for the derivative
(j) 7x – .2
of the function x n .
(k) 1

(In part (k) first show that this should be zero from the
definition of the derivative. Then write 1 = x0 and show
that Equation (98) also works, as long as x is not zero.)

(l) 5
 Cal 1-25

The Chain Rule Using (104) and (105) in the chain rule (100) gives
There is a simple trick called the chain rule that makes
df(y) df dy
it easy to differentiate a wide variety of functions. The = × = ny n–1 × 2x
dx dy dx
rule is
= 2ny n–1x
df y(x) df(y) dy
= chain rule (100) n–1
dx dy dx = 2n x 2 x
To see how this rule works, consider the function
= 2n x 2(n–1) x
n
f(x) = x 2 (101)
= 2n x (2n–2) x (107)
We know that this is just f(x) = x 2n , and the derivative
is
= 2n x (2n– 2) + 1
df(x) d 2n
= x = 2nx 2n– 1 (102)
dx dx = 2nx 2n– 1
But suppose that we did not know this trick, and which is the answer we expect.
therefore did not know how to differentiate (x 2) n . We
do, however, know how to differentiate powers like x 2 In our example, using the chain rule was more difficult
and y n . The chain rule allows us to use this knowledge than differentiating directly because we already knew
in order to figure out how to differentiate the more how to differentiate x 2n . But we will shortly encounter
complex function (x 2) n . examples of new functions that we do not know how to
differentiate directly, but which can be written in the
We begin by defining y(x) as form f[y(x)]; and where we know df/dy and dy/dx. We
can then use the chain rule to evaluate the derivative
y(x) = x 2 (103)
df/dx. We will give you practice with the chain rule
Then our function f(x) = (x 2) n can be written in terms when we encounter these functions.
of y as follows
n n Remembering The Chain Rule
f(x) = (x 2) = y(x) = (y)n = f(y)
The chain rule can be remembered by thinking of the
f(y) = (y)n (104) dy's as cancelling as shown.

Differentiating (103) and (104) gives df(y) dy df(y) remembering

= the chain rule
(108)
dy(x) d 2 dy dx dx
= x = 2x (105)
dx dx
df(y)
= d y n = ny n–1 (106)
dy dy
Cal 1-26 Calculus 2000

Partial Proof of the Chain Rule (optional) (We call this a partial proof for the following reason.
The proof of the chain rule is closely related to cancel- For some functions y(x), the quantity
lation we showed in Equation (108). A partial proof of ∆y = y x + ∆x – y(x) may be identically zero for a
the rule proceeds as follows. small range of ∆x . In that case we would be dividing
by zero (the 1/∆y ) even before we took the limit as ∆x
Suppose we have some function f(y) where y is a goes to zero. A more complete proof handles the
function of the variable x. As a result f[y(x)] is itself a special cases separately. The resulting chain rule still
function of x and can be differentiated with respect to works however, even for these special cases.)
x.
Since ∆y = y(x + ∆x) – y(x) goes to zero as ∆x
d f y(x) = limit f y(x + ∆x) – f y(x) (123) goes to zero, we can write Equation (127) as
dx ∆x→0 ∆x
d f y(x)
Now define the quantity ∆y by dx
∆y ≡ y(x + ∆x) – y(x) (124) limit f(y + ∆y) – f(y)
= ∆y→0
∆y
so that
y(x + ∆x) = y(x) + ∆y limit y(x + ∆x) – y(x)
× ∆x→0
∆x
f[ y(x + ∆x)] = f(y + ∆y)
and Equation (123) becomes df(y) dy
= (100repeated)
dy dx
d f y(x) = limit f(y + ∆y) – f(y) (125)
dx ∆x→0 ∆x This rule works as long as the derivatives df/dy and
Now multiply (125) through by dy/dx are meaningful, i.e., we stay away from kinks or
∆y y(x + ∆x) – y(x) discontinuities in f and y.
1 = = (126)
∆y ∆y
to get

d
f y(x)
dx

limit f(y + ∆y) – f(y) y(x + ∆x) – y(x)

= ×
∆x→0 ∆x ∆y

limit f(y + ∆y) – f(y) y(x + ∆x) – y(x)

= ×
∆x→0 ∆y ∆x

(127)
whereweinterchanged ∆x and ∆yinthedenominator.
 Cal 1-27

INTEGRATION FORMULAS Dividing through by (n+1) gives

Knowing the formula for the derivative of the function T T
x n , and knowing that integration undoes differentia- t ndt = 1 t n+1 (133)
tion, we can now use Equation (98) ti n+1 ti

dx n = nx n – 1 (98repeated)
dx If we choose t i = 0 , we get the simpler result
to find the integral of the function x n . We will see that T n+1
this trick works for all cases except the special case t ndt = T (134)
0 n+1
where n = –1, i.e., the special case where the integral is
a natural logarithm. and the indefinite integral can be written
To integrate x n, let us go back to our calculation of the n+1
distance s x or x(t) traveled by an object moving in the t n dt = t (135)(also34)
n+1
x direction at a velocity vx . This was given by
Equations (19) or (56) as This is the general rule we stated without proof back in
T T Equation (34). Note that this formula says nothing
x(t) = vx(t) dt (128) about the case n = –1, i.e., when we integrate t – 1 = 1/t ,
ti ti
because n +1 = –1 +1 = 0 and we end up with division
where the instantaneous velocity vx(t) is defined as by zero. But for all other values of n, we now have
dx(t) derived a general formula for finding the area under any
vx(t) = (129)
dt curve of the form x n (or t n ). This is a rather powerful
Suppose x(t) had the special form result considering the problems one encounters actu-
ally finding areas under curves. (If you did not do
x(t) = t n + 1 (a special case) (130) Exercise 1, the integration exercises on page 14, or had
then we know from our derivative formulas that difficulty with them, go back and do them now.)
dx(t) (n+1)
v(t) = = dt = (n+1)t n (131)
dt dt
Substituting x(t) = t n + 1 and v(t) = (n+1)t n into
Equation (128) gives
T T
x(t) = vx(t) dt (128)
ti ti

T T
tn + 1 = (n +1)t ndt
ti ti
T (132)
= (n +1) t ndt
ti
Cal 1-28 Calculus 2000

Derivative of the Exponential Function Let us now see how to use the series 136 for calculating
The previous work shows us that if we have a series the derivative of e x . We have, from the definition of
expansion for a function, it is easy to obtain a formula a derivative,
for the derivative of the function. We will now apply d f(x) ≡ limit f(x + ∆x) – f(x) (56repeat)
this technique to calculate the derivative and integral of dx ∆x→ 0 ∆x
the exponential function e x .
If f(x) = e x , we get
There is a series expansion for the function e x that
works for any value of α in the range –1 to +1.
d(e x)
= limit e x + ∆x – e x (138)
dx ∆x→0 ∆x
eα ≈ 1 + α + α + α + ⋅ ⋅ ⋅
2 3
(136) To do this calculation, we have to evaluate the quantity
2! 3!
where 2! = 2 ×1 , 3! = 3 × 2 × 1 = 6 , etc. (The quanti- e x + ∆x . First, we use the fact that for exponentials
ties 2!, 3! are called factorials. For example 3! is called ea + b = ea eb
three factorial.)
(Remember that 10 2 + 3 = 10 2 × 10 3 = 10 5 .) Thus
To see how well the series (136) works, consider the
case α = .01 . From the series we have, up to the α 3 e x +∆x = e x e ∆x (139)
term
Now use the approximation formula (136), setting
α = .01 α = ∆x and throwing out the α 2 and α 3 and higher
α 2 = .0001 ; α 2/2 = .00005 terms because we are going to let ∆x go to zero
α 3 = .000001 ; α 3/ 6 = .000000167 e ∆x ≈ 1 + ∆x (140)
Giving us the approximate value Substituting (140) in (139) gives
1 + α + α + α = 1.010050167
2 3
(137) e x+∆x ≈ e x (1 + ∆x)
2! 3!
When we enter .01 into a scientific calculator and press = e x + e x∆x (141)
the e x button, we get exactly the same result. Thus the Next use (141) in (138) to get
calculator is no more accurate than including the α 3
term in the series, for values of α equal to .01 or less.
d ex e x + e x ∆x – e x
= limit (142)
dx ∆x→0 ∆x
The e x terms cancel and we are left with
d ex e x∆x = limit e x
= limit (143)
dx ∆x→0 ∆x ∆x→0
Since the ∆x′s cancelled, we are left with the exact
result

d ex
= ex (144)
dx

We see that the exponential function e x has the special

property that it is its own derivative.
 Cal 1-29

We will often want to know the derivative, not just of Integral of the Exponential Function
the function e x but of the slightly more general result To calculate the integral of e ax , we will use the same
e ax where a is a constant. That is, we want to find trick as we used for the integral of x n , but we will be
d e ax a bit more formal this time. Let us start with Equation
(a = constant) (145) (128) relating position x(t) and velocity v(t) = dx(t)/dt
dx
Solving this problem provides us with our first mean- go get
ingful application of the chain rule tf tf tf dx(t)
x(t) = vx(t) dt = dt (128)
df(y) df(y) dy ti ti ti dt
= (100repeated)
dx dy dx Since Equation (128) holds for any function x(t) [we
If we set did not put any restrictions on x(t)], we can write
Equation (128) in a more abstract way relating any
y=ax (146) function f(x) to its derivative df(x)/dx;
then we have
de ax = de y dy (147) xf xf df(x)
dx dy dx f(x) = dx (151)
xi xi dx
Now
de y = e y (148) To calculate the integral of e ax , we set f(x) = e ax and
dy df(x)/dx = ae ax to get
dy d dx xf xf
= (ax) = a = a×1 = a (149) e ax = a e ax dx (152)
dx dx dx xi xi
Using (148) and (149) in (147) gives Dividing (157) through by (a) gives us the definite
de ax integral
= e y (a) = e ax (a) = ae ax
dx xf xf
e ax dx = 1a e ax (a = constant) (153)
Thus we have xi xi

d ax The corresponding indefinite integral is

e = ae ax (150)
dx
e ax
e axdx = a (a = constant) (154)
This result will be used so often it is worth memorizing.

Exercise 6 Exercise 7
For further practice with the chain rule, show that The natural logarithm is defined by the equation
2 ln (x) = 1 dx (see Equations 33-40)
deax 2 x
= 2axeax
dx
Use Equation (151) to show that
2
Do this by choosing y = ax , and then do it again by d (ln x) = 1
choosing y = x2 . dx x (155)

(Hint—integrate both sides of Equation (155) with re-

spect to x.)
Cal 1-30 Calculus 2000

DERIVATIVE AS THE SLOPE OF A CURVE Figure (20a) is a blowup of the curve in the region
Up to now, we have emphasized the idea that the between x and x + ∆x . If the distance ∆x is suffi-
derivative of a function f(x) is given by the limiting ciently small, the curve between x and x + ∆x should
process be approximately a straight line and that part of the
curve should be approximately the hypotenuse of the
df(x) limit f(x + ∆x) – f(x) (55 repeated)
= ∆x→0 right triangle abc seen in Figure (20a). Since the side
dx ∆x opposite to the angle θ * is f(x + ∆x) – f(x) , and the
We saw that this form was convenient when we had an adjacent side is ∆x , we have the result that the tangent
explicit way of calculating f(x + ∆x) , as we did by of the angle θ * is
using a series expansion. However, a lot of words are f(x + ∆x) – f(x)
tan θ * = (156)
required to explain the steps involved in doing the ∆x
limiting process indicated in Equation (55). In contrast, When we make ∆x smaller and smaller, take the limit
the idea of an integral as being the area under a curve as ∆x → 0 , we see that the angle θ * becomes more
is much easier to state and visualize. Now we will nearly equal to the angle θ shown in Figure (21), the
provide an easy way to state and interpret the derivative angle of the curve when it passes through the point x.
of a curve. Thus
Consider the function f(x) graphed in Figure (20). At limit f(x + ∆x) – f(x)
tan θ = ∆x→0 (157)
a distance x down the x axis, the curve had a height ∆x
f(x) as shown. Slightly farther down the x axis, at The tangent of the angle at which the curve passes
x + ∆x , the curve has risen to a height f(x + ∆x) . through the point x is called the slope of the curve at the
f(x) point x. Thus from Equation (157) we see that the slope
f(x+∆x) of the curve is equal to the derivative of the curve at that
point. We now have the interpretation that the deriva-
f(x) tive of a curve at some point is equal to the slope of the
curve at that point, while the integral of a curve is equal
to the area under the curve up to that point.
f(x)
∆x θ
x
x x+∆x
Figure 20
Two points on a curve, a distance ∆ x apart.

c
f(x+∆x) x

f(x)
a θ*
}
b
f(x+∆x)
– f(x)
Figure 21
The tangent of the angle θ at which the curve
∆x passes through the point x is called the slope
of the curve at that point.
Figure 20a
At this point, the curve is tilted
by approximately an angle θ *.
 Cal 1-31

Negative Slope Exercise 8

In Figure (22) we compare the slopes of a rising and a Estimate the numerical value of the slope of the curve
falling curve. In (22a), where the curve is rising, the shown in Figure (23) at points (a), (b), (c), (d) and (e). In
quantity f(x + ∆x) is greater than f(x) and the deriva- each case do a sketch of f(x + ∆x) – f(x) for a small ∆x ,
tive or slope and let the slope be the ratio of f(x + ∆x) – f(x) to ∆x .
Your answers should be roughly 1, 0, –1, + ∞ , – ∞ .
df(x) limit f(x + ∆x) – f(x)
= ∆x→0
dx ∆x
f(x)
is a positive number. b
c e
In contrast, for the downward curve of Figure (22b), d
f(x + ∆x) is less than f(x) and the slope is negative. a
For a curve headed downward, we have x
Figure 23
df(x) downward heading
= – tan(θ) curve (158) Estimate the slope at the
dx
various points indicated.
(For this case you can think of θ as a negative angle,
so that tan(θ) would automatically come out negative.
However it is easier simply to remember that the slope
of an upward directed curve is positive and that of a
downward directed cure is negative.)

pe
e slo f(x+∆x)
itiv θ
os f(x)
p

f(x+∆x) – f(x) is positive

∆x
x x+∆x

f(x+∆x) – f(x) is negative

∆x
ne

ga f(x)
tiv θ
es
lo p f(x+∆x)
e

x x+∆x
Figure 22
Going uphill is a positive slope,
downhill is a negative slope.
Cal 1-32 Calculus 2000

THE EXPONENTIAL DECAY onds, only 27 remained. The decay of these muons is
A curve that we will encounter several times during the an example of an exponential decay of the form
course is the function e – ax shown in Figure (24), number of number of
which we call an exponential decay. Since exponents surviving = muons at × e – t/t0 (161)
muons time t = 0
always have to be dimensionless numbers, we are
writing the constant (a) in the form 1/x 0 so that the where t 0 is the time it takes for the number of muons
exponent x/x 0 is more obviously dimensionless. remaining to drop by a factor of 1/e = 1/2.7. That time
is called the muon lifetime.
The function e – x/x0 has several very special proper-
ties. At x = 0, it has the numerical value 1 (e 0 = 1) . We can use Equation (161) to estimate the muon
When we get up to x = x 0 , the curve has dropped to a lifetime t 0 . In the movie, the number of mesons at the
value top of the graph, reproduced in Figure (25), is 648. That
is at time t = 0. Down at time t = 6 microseconds, the
e – x/x0 = e – 1 = 1e (at x = x 0) number surviving is 27. Putting these numbers into
(159) Equation (161) gives
≈ 1
2.7
surviving
muons × e
= 648 initial
27 muons – 6/t0
When we go out to x = 2x o , the curve has dropped to

e – 2x0 /x0 = e – 2 = 12 (160) e – 6/t0 = 27 = .042 (162)

e 648
Out at x = 3x 0 , the curve has dropped by another Take the natural logarithm ln of both sides of Equation
factor of e to (1/e)(1/e)(1/e). This decrease continues (162), [remembering that ln e x = x ] gives
indefinitely. It is the characteristic feature of an expo-
nential decay. ln e – 6/t0 = –t 6 = ln .042 = – 3.17
0
where we entered .042 on a scientific calculator and
Muon Lifetime pressed the ln key. Solving for t 0 we get
In the muon lifetime experiment, we saw that the
number of muons surviving decreased with time. At t 0 = 6 = 1.9 microseconds (163)
3.17
the end of two microseconds, more than half of the This is close to the accepted value of t 0 = 2.2 0
original 648 muons were still present. By 6 microsec- microseconds which has been determined from the
study of many thousands of muon decays.

e–x/x0

1/e

1/e 2
x
0 x0 2x 0 3x 0
Figure 25
Figure 24 The lifetime of each detected muon is represented
As we go out an additional distance by the length of a vertical line. We can see that
x 0 , the exponential curve drops by many muons live as long as 2 microseconds (2µs),
another factor of 1/e. but few live as long as 6 microseconds.
 Cal 1-33

Half Life To help illustrate the nature of exponential decays,

The exponential decay curve e – t/t0 decays to suppose that you started with a million muons. How
long would you expect to wait before there was, on the
1/e = 1/2.7 of its value at time t 0 . While 1/e is a very
average, only one left?
convenient number from a mathematical point of view,
it is easier to think of the time t 1/2 it takes for half of the To solve this problem, you would want the number
muons to decay. This time t 1/2 is called the half life of e – t/t0 to be down by a factor of 1 million
the particle.
e – t/t0 = 1 × 10 – 6
From Figure (26) we can see that the half life t 1/2 is Taking the natural logarithm of both sides gives
slightly shorter than the lifetime t 0 . To calculate the
half life from t 0 , we have ln e – t/t0 = –t
t 0 = ln 1×10
–6
= –13.8 (166)
e – t/t0 = e – t1/2/t0 = 1 (164)
t = t 1/2 2 (To calculate ln 1×10 – 6 , enter 1, then press the exp
Again taking the natural logarithm of both sides of key and enter 6, then press the +/– key to change it to
Equation (164) gives –6. Finally press = to get the answer –13.8.)
–t Solving Equation (166) for t gives
ln e – t1/2/t0 = t 1/2 = ln 1 = – .693
0 2
t = 13.8 t 0 = 13.8 × 2.2 µsec
t 1/2 = .693 t 0 (165)
t = 30 microseconds (167)
From Equation (165) you can see that a half life t 1/2 is That is the nature of an exponential decay. While you
about .7 of the lifetime t 0 . If the muon lifetime is have nearly half a million left after around 2 microsec-
2.2 µsec (we will abbreviate microseconds as µsec ), onds, they are essentially all gone by 30 microseconds.
and you start with a large number of muons, you would
expect about half to decay in a time of Exercise 9
t 1/2 muon = .693 × 2.2µsec = 1.5 µsec How many factors of 1/2 do you have to multiply
together to get approximately 1/1,000,000? Multiply this
The basic feature of the exponential decay curve e – t/t0 number by the muon half-life to see if you get about
is that for every time t 0 that passes, the curve decreases 30 microseconds.
by another factor of 1/e. The same applies to the half
life t 1/2 . After one half life, e – t/t0 has decreased to half 1
its value. After a second half life, the curve is down to
1/4 = 1/2 x 1/2. After 3 half lives it is down to 1/8 = e–t /t 0
1/2 x 1/2 x 1/2 as shown in Figure (27).
1 1/2

1/4
e–t /t 0 1/8
1/2
0 t 1/2 2t1/2 3t 1/2
1/e Figure 27
After each half-life, the curve
decreases by another factor of 1/2.
t
0 t0
t 1/2
Figure 26
Comparison of the lifetime t0 and the half-life t1/2 .
Cal 1-34 Calculus 2000

Measuring the Time The height (y) of the point where we drew the tangent
Constant from a Graph curve is just the value of the function e – T/t0 . The
The idea that the derivative of a curve is the slope of the tangent of the angle θ is the opposite side (y) divided
curve, leads to an easy way to estimate a lifetime t 0 by the adjacent side (x)
from an exponential decay curve e – t/t0 . y e – T/t0
tanθ = x = x (169)
The formula for the derivative of an exponential curve
Equating the two magnitudes of tanθ in Equations
is
(169) in (168a) gives us
de at = ae at (150 repeated)
dt 1 e – T/t0 = 1 e – T/t0
t0 x
Setting a = – 1/t 0 gives
which requires that
d e – t/t0 = – 1 e – t/t0 (168)
dt t0 x = t0 (170)
Since the derivative of a curve is the slope of the curve, Equation (170) tells us that the distance (x), the distance
we set the derivative equal to the tangent of the angle down the axis where the tangent lines intersect the axis,
the curve makes with the horizontal axis. is simply the time constant t 0 .
d e – t/t0 = – 1 e – t/t0 = tanθ (168a) The result gives us a very quick way of determining the
dt t0 time constant t 0 of an exponential decay curve. As
The minus sign tells us that the curve is headed down. illustrated in Figure (29), choose any point on the
curve, draw a tangent to the curve at that point and
In Figure (28), we have drawn a line tangent to the measure the distance down the axis where the tangent
curve at the point t = T. This line intersects the (t) axis line intersects the axis. That distance will be the time
(the axis where e – t/t0 goes to zero) at a distance (x) constant t 0 . We will use this technique in several
down the t axis. laboratory exercises later in the course.

e–t /t 0 e–t/t 0

x
e –T/t 0 θ
y

t t
x
Figure 28
T Figure 29 t0
A line, drawn tangent to the exponential decay A quick way to estimate the time constant t0
curve at some point T, intersects the axis a for an exponential decay curve is to draw the
distance x down the axis. We show that this tangent line as shown.
distance x is equal to the time constant t0 . This
is true no matter what point T we start with.
 Cal 1-35

THE SINE AND COSINE FUNCTIONS Radian Measure

The final topic in our introduction to calculus will be the We are brought up to measure angles in degrees, but
functions sinθ and cosθ and their derivatives and physicists and mathematicians usually measure angles
integrals. We will need these functions when we come in radians. The angle θ measured in radians is defined
to rotational motion and wave motion. as the arc length subtended by the angle θ on a circle
of unit radius, as shown in Figure (32).
The definition of sinθ and cosθ , which should be
arc length subtended
familiar from trigonometry, are θ radians = by θ on a unit circle (173)
opposite
sinθ = ac hypotenuse (171a) (If we had a circle of radius c, then we would define
θ radians = /c , a dimensionless ratio. In the special case
c = 1, this reduces to θ radians = .)
adjacent
cosθ = bc hypotenuse (171b)
Since the circumference of a unit circle is 2π , we see
that θ for a complete circle is 2π radians, which is the
c same as 360 degrees. This tells us how to convert from
a
degrees to radians. We have the conversion factor
θ Figure 30
b 360 degrees degrees
= 57.3 (174)
where θ is an angle of a right triangle as sown in Figure 2π radians radian
(30), (a) is the length of the side opposite to θ , (b) the As an example of using this conversion factor, suppose
side adjacent to θ and (c) the hypotenuse. we want to convert 30 degrees to radians. We would
have
The formulas are simplified if we consider a right
30 degrees
triangle whose hypotenuse is of length c = 1 as in = .52 radians (175)
Figure (31). Then we have 57.3 degrees/radian
To decide whether to divide by or multiply by a
sinθ = a (172a)
conversion factor, use the dimensions of the conver-
cosθ = b (172b) sion factor. For example, if we had multiplied 30
degrees by our conversion factor, we would have
1 gotten
a
θ degrees degrees 2
Figure 31 30 degrees × 57.3 = 1719
b radian radian
We can then fit our right triangle inside a circle of radius This answer may be correct, but it is useless.
1 as shown in Figure (32).
The numbers to remember in using radians are the
following:
90° = π/2 radians
180° = π radians
(176)
270° = 3π/2 radians
1
a 360° = 2π radians
θ
b The other values you can work out as you need them.

Figure 32
Fitting our right triangle inside a unit radius circle.
Cal 1-36 Calculus 2000

The Sine Function Our next step is to construct a graph in which θ is

In Figure (33) we have started with a circle of radius 1 shown along the horizontal axis, and we plot the value
and, in a somewhat random way, labeled 10 points of sinθ = (a) on the vertical axis. The result is shown
around the circle. The arc length up to each of these in Figure (34). The eleven points, representing the
points is equal to the angle, in radian measure, sub- heights a 0 to a 10 at θ 0 to θ 10 are shown as large dots
tended by that point. The special values are: in Figure (34). We have also sketched in a smooth
curve through these points, it is the curve we would get
θ 0 = 0 radians if we had plotted the value of (a) for every value of θ
θ 4 = π/2 radians (90°) from θ = 0 to θ = 2π . The smooth curve is a graph of
θ 6 = π radians (180°) the function sin θ .
θ 8 = 3π/2 radians (270°)
θ 10 = 2π radians (360°) Exercise 10
Using the fact that the cosine function is defined as
In each case the sinθ is equal to the height (a) at that cos θ = b (b is defined in Figures 31, 32)
point. For example
plot the values of b0, b1, ⋅ ⋅ ⋅ , b10 on a graph similar to
sinθ 1 = a 1 Figure (34), and show that the cosine function cos θ
sinθ 2 = a 2 looks like the curve shown in Figure (35).
⋅⋅⋅⋅⋅⋅
sinθ 10 = a 10
We see that the height (a) starts out at a 0 = 0 for θ 0 ,
increases up to a 4 = 1 at the top of the circle, drops
back down to a 6 = 0 at θ 6 = π , goes negative, down
to a 8 = – 1 at θ 8 = 3π/2 , and returns to a 10 = 0 at
θ 10 = 2π .
θ4 θ3 1
θ5
θ2
a4 a2
a3 θ2
a5 a2 θ1
a1
a1 θ1
π 2π
θ6 θ0
θ 10 0 θ
π 3π
a7 a9 2 2

a8
a7 θ7
θ9
θ7

θ8 –1
Figure 33 Figure 34
The heights a i at various points around a unit circle. Graph of the function sin θ .
 Cal 1-37

There is nothing that says we have to stop measuring Amplitude of a Sine Wave
the angle θ after we have gone around once. On the A graph of the function y(θ) = c sinθ looks just like
second trip around, θ increases from 2π up to 4π , and the curve in Figure (36), except the curve goes up to a
the curve sinθ repeats itself. If we go around several height c and down to –c as shown in Figure (38). We
times, we get a result like that shown in Figure (36). would get the curve of Figure (38) by plotting points
around a circle as in Figure (33), but using a circle of
Several cycles of the curve cosθ are shown in Figure radius c. We call this factor c the amplitude of the sine
(37). You can see that the only difference between a wave. The function sinθ has an amplitude 1, while the
sine and a cosine curve is where you set θ = 0 . If you sine wave in Figure (38) has an amplitude c (its values
move the origin of the cosine axis back (to the left) 90° range from +c to –c).
(π/2) , you get a sine wave.
c sin θ

1 c

0 θ 0
π π 3π 2π
2 2

–1 –c
Figure 35 Figure 38
The cosine function. A sine wave of amplitude c.

0 θ
π 2π 3π 4π 5π 6π

–1
Figure 36
Several cycles of the curve sin θ .

0 θ
π 2π 3π 4π 5π 6π

–1
Figure 37
Several cycles of the curve cos θ .
Cal 1-38 Calculus 2000

Derivative of the Sine Function Now draw a line vertically down from point (c) and
Since the sine and cosine functions are smooth curves, horizontally over from point (b) to form the triangle bcd
we should be able to calculate the derivatives and shown in Figure (40). The important point is that the
integrals of them. We will do this by first calculating angle at point (c) in this tiny triangle is the same as the
the derivative, and then turning the process around to angle θ at point (a). To prove this, consider the sketch
find the integral, just as we did for the functions x n and in Figure (41). A line bf is drawn tangent to the circle
ex . at point (b), so that the angle abf is a right angle. That
means the other two angles in the triangle add up to 90°,
The derivative of the function sinθ is defined as usual the total angle in any triangle being 180°
by
θ + ϕ = 90° (178)
limit sin θ +∆θ – sinθ
d sinθ
= ∆θ→0 (177) Since the angle at (e) in triangle bef is also a right angle,
dθ ∆θ the other two angles in the triangle bef, must also add
where ∆θ is a small change in the angle θ . up to 90°.
α + ϕ = 90° (179)
The easiest way to evaluate this limit is to go back to the
unit circle of Figure (25) and construct both sinθ and For both Equations (178) and (179) to be true, we must
sin θ +∆θ as shown in Figure (39). We see that sinθ have α = θ . c
is the height of the triangle with an angle θ, while
sin θ +∆θ is the height of the triangle whose center θ

∆θ
angle is θ +∆θ . What we have to do is calculate the
difference in heights of these two triangles. c
d b
In Figure (40) we start by focusing our attention on the d

∆θ
slender triangle abc with an angle ∆θ at (a) and long b
sides of length 1 (since we have a unit circle). Since the ∆θ θ
angle ∆θ is small, the short side of this triangle is
essentially equal to the arc length along the circle from
point (b) to point (c). And since we are using radian a
θ
measure, this arc length is equal to the angle ∆θ . Figure 40 r=1
The difference between sinθ and
sin θ + ∆θ is equal to the height
of the side cd of the triangle cdb.

sin(θ+∆θ) b
sin(θ)
∆θ
α
θ
θ φ
a
e f
r=1 θ + φ = 90°
α + φ = 90°
∴α = θ
Figure 41
Figure 39 Demonstration that the angle α equals the angle θ .
Triangles for the sinθ and the sin θ + ∆θ .
 Cal 1-39

The final step is to note that when ∆θ in Figure (40) is Exercise 11

very small, the side cb of the very small triangle is Using a similar derivation, show that
essentially tangent to the circle, and thus parallel to the
side bf in Figure (41). As a result the angle between cb d (cosθ) = – sinθ (182)
and the vertical is also the same angle θ . dθ

Because the tiny triangle, shown again in Figure (42)

Exercise 12
has a hypotenuse ∆θ and a top angle θ , the vertical
Using the chain rule for differentiation, show that
side, which is equal to the difference between sinθ and
sin θ +∆θ has a height (cos θ)∆θ . Thus we have d (sinaθ) = a cosaθ
dθ
sin θ +∆θ – sinθ = (cosθ)∆θ (180) a = constant (183)
d (cosaθ) = – a sinaθ
Equation (180) becomes exact when ∆θ becomes an dθ
infinitesimal angle.
(Hint—if you need to, look at Equation (145) through
We can now evaluate the derivative (150).

limit sin θ +∆θ – sinθ

d sinθ Exercise 13
= ∆θ→0 Using the fact that integration reverses differentiation,
dθ ∆θ
as we did in integrating the function e x (Equations (151)
through (154), show that
cosθ ∆θ
= limit
∆θ→0 ∆θ
θf (184a)
1 sinaθ θf
(cosaθ)dθ = a
= limit
∆θ→0
cosθ θi θi (a = constant)
θf (184b)
Thus we get the exact result 1 cosaθ θf
(sinaθ)dθ = – a
θi θi
d (sinθ) = cosθ (181)
dθ
Use sketches of the integrals from θi = 0 to θf = π /2 to
show that Equations (184a) and (184b) have the correct
numerical sign. (Explicitly explain the minus sign in
∆θ cosθ

(184b).
θ
∆θ

sin(θ+∆θ)
∆θ

sin(θ)
∆θ

a
θ
r=1
Figure 42
The difference between sinθ and
sin θ + ∆θ is equal to ∆θ cos θ .
 Index-1

Adiabatic expansion
P2000 Index Calculation of work 18-26
In Carnot cycle 18-11
Introduction to 18-9
Air cart
Analysis of coupled carts 16-12
Symbols
Construction of 6-2
µ zero, definition of 31-19
In impulse experiments 11-9
π meson 40-23 In recoil experiments 6-2
∆E∆t>=h, Uncertainty principle 40-19 Oscillating cart 14-5
∆x∆p>=h, Uncertainty principle 40-15 Speed detector 30-5. See also Experiments II: - 6-
10 dimensions, String theory int-16 Faraday's law air cart speed detector
100 billion to one, Matter over anti matter int-30 Air Resistance
13.6 eV, hydrogen spectrum 35-4 Calculus analysis for projectile motion 4-12
1836 times, proton/electron mass ratio int-17 Computer analysis for projectile motion 5-24, 8-3
1987 supernova 6-14, 20-14 Strobe analysis for projectile motion 3-22
2.74 degrees, cosmic background radiation int-29 Airplane wing, Bernoulli's equation 23-13
2D or 3D? equipotential plotting experiment 25-8 Allowed orbits, Bohr theory int-8, 35-1
4 dimensions int-16 Allowed projections, spin 39-3
Allowed standing wave patterns 37-1
A Alpha particles 20-8
Aberration Amount of sin(3t) present in a wave 16-28
Astigmatism Optics-21 Ampere
Chromatic Optics-21 Definition of 27- 2
Newton's reflecting telescope Optics-22 MKS units 24-2
Spherical Optics-21 Ampere's law
In Hubble telescope mirror Optics-22
Applied to a solenoid 29-15
Absolute zero 17-9, 17-21
Chapter on 29-1
Abundance of the elements 34-24 Derivation of line integral 29-7
AC voltage generator 30-21 Field of straight wire 29-11
Magnetic flux in 30-21 Final result 29-11
Accelerating field in electron gun 26-10 Maxwell’s correction to 32- 4
Acceleration Amplitude
Angular 12-3 And intensity, Fourier analysis lecture 16-33
Angular analogy 12-3 And phase
Calculus definition of 4-5, Cal 1-7 Fourier analysis lecture 16-31
Component equations Cal 1-8 Wave motion 15-17
Vector equation Cal 1-7 Diffraction pattern by Fourier analysis 16-33
Constant acceleration formulas Fourier coefficients 16-32
Calculus derivation 4-9, Cal 1-20 Of a sine wave Cal 1-37
In three dimensions 4-11, Cal 1-22 Analysis
Definition of 3-13 Fourier 16-6
Due to gravity 3-21 Of coupled air carts 16-12
From a strobe photograph 3-15 Of path 1 for electromagnetic pulse 32- 14
Intuitive discussion 3-20 Of path 2 for electromagnetic pulse 32- 16
On inclined plane 9-11
Analytic solution
Radial 12-5
Of the RC circuit 27- 22
Tangential 12-4
Oscillation of mass on spring 14-7
Uniform circular motion
Projectile motion with air resistance 4-12
Direction of 3-18
Magnitude of 3-18 Anderson, C., positrons int-13
Vector, definition of 3-15 Andromeda galaxy int-2, int-3, 1-22
Acceleration versus time graphs 4-7 Angle of reflection (scattering of light) 36-3, Optics-
Accelerators, particle int-1, 28- 22 1
Accurate values of Fourier coefficients 16-32 Angles of incidence and reflection Optics-3
Adding sines and cosines in Fourier analysis 16-28 Angular acceleration 12-3
Addition of charge 19-10 Angular analogy 12-3
Addition of forces 9-2 For Newton’s second law 12-14
Torque (angular force) 12-15
Index-2

Angular frequency Applications of Faraday’s law 30-21

Definition of 14-4 AC voltage generator 30-21
Wave motion 15-14 Gaussmeter 30-23
Angular magnification of magnifier Optics-39 Applications of Newton’s Second Law 9-1
Angular mass Applications of the second law of thermodynam-
Moment of inertia 12-7 ics 18-17
Rotational kinetic energy 12-22 Arbitrary wave, Fourier analysis 16-28
Angular momentum Area
As a Vector 7- 14, 12-7 As a vector 24-22
Movie 7- 15 Negative or positive 16-29
Bohr model 35-1, 35-8 Related to integration Cal 1-11
Planck's constant 35-8 Under the curve Cal 1-12
Conservation of 7- 9, 12-16 Arecibo radio telescope int-15, Optics-48
Derivation from F = ma 12-16 Arithmetic of vectors. See also Vector
X. See also Experiments I: - 4- Conservation of
Addition 2-3
angular momentum
Associative law 2-4
Definition of 7- 10
Commutative law 2-4
Definition of, more general 7- 12
Multiplication by number 2-5
Definition of, still more general 12-6
Negative of 2-5
Definition of, cross product 12-11
Scalar or dot product 2-12, 10-13
Formation of planets 7- 17
Subtraction of 2-5
Gyroscopes 12-18
Vector cross product 2-15, 12-9
Kepler's second law 8-32
Aspirator, Bernoulli's equation 23-16
Magnetic moment 31-24
Associative law, Exercise 2-7
Movie on vector nature 7- 15, 12-6, 12-17
Of bicycle wheel 12-6 Astigmatism Optics-21
Projections of, classical 7- 14 Astronomy
Projections of electron spin 39-3 1987 supernova int-19, 6-14, 20-14
Quantized int-9 Abundance of the elements 34-24
Quantized projections 38-5 Big bang model of universe 33-25, 34-26
Quantum number 38-7 Binary stars int-2
Angular velocity Black dwarf star int-19
As a vector 12-7 Black holes 10-29
Introduction to int-19
Definition of 12-2
Mass on spring 14-9 Blackbody radiation, color of stars 34-2
Oscillating cart 14-5 Copernicus 8-25
Crab nebula 20-16
Annihilation of antimatter 34-17
Decoupling of light and matter 34-31
Antielectron int-13
Doppler effect 33-23
Antielectron type neutrino int-22
Eagle nebula 7- 18, Optics-44
Antimatter 34-16 Early universe int-27, 34-29
Annihilation of 34-17 Escape velocity 10-28
Excess of matter over, in early universe 34-17, 34-29 Evolution of the universe 34-21
Introduction to int-12, 34-16 Excess of matter over antimatter 34-29
Neutrino int-22 Expanding universe, Hubble int-3
Neutron int-13 Formation of planets 7- 17
Positron int-13 Galaxy
Positron electron pair 34-17 Andromeda int-2
Proton int-13 Introduction to int-2
Wave equation for 15-2 Most distant int-3
Antiparticle int-13 Sombrero int-2
Created by photon 34-17 General relativity 8-29
Applications of Bernoulli’s equation 23-12 Globular cluster 11-2
Airplane wing 23-13 Gravitational lensing 34-20
Aspirator 23-16 Helium abundance in universe 34-26
Hydrostatics 23-12 Helium core of massive star 20-15
Leaky tank 23-12 Hubble rule for expanding universe int-3
Sailboat 23-14 Iron core of massive star 20-15
Venturi meter 23-15 Kepler's laws 8-24
 Index-3

Astronomy Continued Atoms

Light years int-2 Angular momentum quantum number 38-7
Magnetic field of the earth 28- 11 Atomic nucleus, chapter on 20-1
Models of the universe 34-23 Atomic processes 17-4
Neutrino 6-14, 11-21 Avogadro’s law 17-24
Neutron star BASIC program, hydrogen molecule ion 19-24
And black holes 20-18 Beryllium in periodic table 38-13
In Crab nebula 20-17 Bohr model int-8
Introduction to int-19 Boron in periodic table 38-13
Nuclear fusion and stellar evolution 20-12 Brownian motion 17-7
Orion nebula 7- 17 Chapter on 17-1, 38-1
Penzias and Wilson, cosmic radiation 34-27 Classical hydrogen atom 35-2
Powering the sun 34-23 Effective nuclear charge 38-12
Ptolemy, epicycle in Greek astronomy 8-25 Electron binding energy 19-20, 38-11
Quantum fluctuations in space 40-25 Electron energy in hydrogen molecule ion 19-21
Quasar, gravitational lens 34-20 Electron spin 38-9
Radio galaxy Optics-48 Equipartition of energy 17-28
Radio images of variable star Optics-49 Expanded energy level diagram 38-8
Radio telescope. See Radio telescope Failure of classical physics 17-31
Radio telescope, three degree radiation int-30, 34- Freezing out of degrees of freedom 17-32
27 Heat capacity 17-26
Radio telescopes Optics-48 Hydrogen molecule 19-16
Red shift and expanding universe int-3, 33-24, 34- Introductory view of int-16
21 Ionic bonding 38-15
Red supergiant star 20-15 L= 0 Patterns in hydrogen 38-4
Retrograde motion of Mars 8-24 Lithium 38-12
Space travel and time dilation 1-22 Model atom 37-4
Star, blackbody spectrum 34-3 Molecular and atomic processes 17-1
Steady state model of the universe 34-25 Molecular forces 19-15
Stellar evolution int-19 Multi electron 38-9
Telescopes Optics-40 Nuclear matter, chapter on 20-1
Arecibo radio telescope Optics-48 Nucleus. See also Nuclear
Galileo's Optics-41 Discovery of 11-19
Hubbel Space Telescope Optics-44 Particle-wave nature of matter int-10
Issac Newton’s Optics-42 Pauli exclusion principle 38-9
Mt. Hopkins Optics-43 Periodic table 38-10
Mt. Palomar Optics-43
Potassium to krypton 38-14
Very Large Array, radio telescopes Optics-48
Precession of, in magnetic field 39-15
Very Long Baseline Array (VLBA) Optics-49
Quantized projections of angular momentum 38-5
William Hershel’s Optics-43
World’s Largest Optical, Keck Optics-45 Schrödinger’s equation for hydrogen 38-2
Yerkes Optics-41 Silicon, surface (111 plane) of Optics-51
Thermal equilibrium of the universe 34-28 Sodium to argon 38-13
Three degree cosmic radiation int-29, 34-27 Standing wave patterns in hydrogen 38-3
Tycho Brahe 8-25 Table of 19-5
Van Allen radiation belts 28- 32 Thermal motion of 17-6
Visible universe int-3 Up to neon 38-13
White dwarf star 20-15 Xenon, photograph of 17-1
Atmospheric pressure 17-23 Atwood’s machine 9-16
Atomic Avogadro’s law 17-24
And molecular forces, electric interaction 19-1 Avogadro's number, the mole 17-24
Clocks 1-21
Microscopes 17-1
Scanning Tunneling Microscope Optics-51
Processes 17-4
Spectra 33-16
Structure 19-3
Units 19-22
Index-4

B Beryllium
Balancing weights, equilibrium 13-2 Binding energy of last electron 38-12
Ball Spring Pendulum. See Pendulum: Spring In periodic table 38-13
Balmer series Beta decay
Energy level diagram for 35-6 And energy conservation int-21
Formula from Bohr theory 35-5 Neutrinos 20-6
Hydrogen spectrum 35-4 Neutrons 20-7
Introduction to, hydrogen star 33-19 Protons 20-7
Barometer, mercury, pressure measurement 17-22 Recoil experiment 6-6
Basic electric circuits 27- 1 Beta, Hans, proton cycle, energy from sun 34-23
BASIC program. See also Computer Beta ray int-21
Calculating circle 5-6 Betatron 30-16
Calculational loop for satellite motion 8-19 Bi-concave lens Optics-27
Comment lines in 5-7 Bi-convex lens Optics-27
Computer time step 5-14 Bicycle wheel
Conservation of angular momentum 8-32 As a collection of masses 12-5
Conservation of energy 8-35 As a gyroscope 12-18
DO LOOP 5-4 Right hand rule for rotation 12-11
For drawing circle 5-11 Vector nature of angular momentum 7- 14, 12-
For hydrogen molecule ion 19-24 6, 12-17
For oscillating cart 14-32 Big bang model of universe int-4, 33-25, 34-26
For oscillatory motion 14-21, 14-30 Binary stars int-2
For projectile motion 5-18, 5-19, 5-21, 8-21 Binding energy
For projectile motion with air resistance 5-22 Hydrogen molecule ion 19-23
For satellite motion 8-21 Molecular forces 17-13
For spring pendulum 9-20 Nuclear 20-9
Kepler's first law 8-26 Nuclear stability 20-10
Kepler's second law 8-27 Of inner electrons 38-12
Kepler's third law 8-28 Binomial expansion 1-31, Cal 1-23
LET Statement 5-5 Black dwarf star int-19
Modified gravity 8-29 Black holes
Multiplication 5-6 And neutron stars 20-18
New calculational loop 8-17 Critical radius for sun mass 10-30
Orbit-1 program 8-21 Introduction to int-19
Perihelion, precession of 8-30 Stellar evolution int-20
Plotting a point 5-6 Theory of 10-29
Plotting window 5-7 Blackbody radiation
Prediction of satellite orbits 8-16 Electromagnetic spectrum 32- 22
Satellite motion laboratory 8-23 Photon picture of 34-22
Selected printing (MOD command) 5-10 Planck's formula 34-4
Sine wave products 16-29 Theory of 34-2
Unit vectors 8-18 Wein's displacement law 34-2
Variable names 5-6 Blood flow, fluid dynamics 23-23
Bathtub vortex 23-2 Bohr magneton
Baud rate, for fiber optics Optics-14 Dirac wave equation 39-5
Bell Telephone Lab, electron waves 35-12 Unit of magnetic moment 39-4
Berkeley synchrotron 28- 22 Bohr model int-8
Bernoulli’s equation Allowed orbits 35-1
Applications of Angular momentum 35-1, 35-8
Airplane Wing 23-13 Chapter on 35-1
Aspirator 23-16 De Broglie explanation 35-1
Leaky tank 23-12 Derivation of 35-8
Sailboat 23-14 Energy levels 35-4
Venturi meter 23-15 Introduction to int-8
Applies along a streamline 23-11 Planck's constant 35-1, 35-8
Care in applying 23-16 Quantum mechanics 35-1
Derivation of 23-9 Rydberg constant 35-9
Formula for 23-11 Bohr orbits, radii of 35-7
Hydrodynamic voltage 23-17
 Index-5

Boltzman Capacitance
Constant 17-11 Electrical 27- 16
Formula for entropy 18-24 Introduction to 27- 14
Bonding Capacitor
Covalent 19-15 Electrolytic 27- 17
Ionic 38-15 Energy storage in 27- 18
Born interpretation of particle waves 40-6 Examples of 27- 17
Boron In circuits. See also Circuits
Binding energy of last electron 38-12 As circuit elements 27- 20
In periodic table 38-13 Hydrodynamic analogy 27- 14
Bottom quark int-24 LC circuit 31-10
Bragg reflection 36-4 Parallel connection 27- 20
RC circuit 27- 22
Brahe, Tycho 8-25
Series connection 27- 21
Brownian motion Introduction to 27- 14
Discussion 17-7 Magnetic field in 32- 6
Movie 17-7 Parallel plate
Bubble chambers 28- 26 Capacitance of 27- 16
Bulk modulus 15-8 Deflection plates 26-16
Button labeled ø on MacScope 16-32 Introduction to 26-14
Voltage in 26-15
C Carbon
c (speed of light) int-2, 1-12. See also Speed of Burning in oxygen 17-5
light Graphite crystal, electron diffraction 36-8
Calculating Fourier coefficients 16-28 Carnot cycle
Calculational loop 5-17 As thought experiment 18-4
For projectile motion 5-19, 8-17 Efficiency of
For projectile motion with air resistance 5-24 Calculation of 18-28
Satellite Motion 8-19 Discussion 18-12
Calculations Formula for 18-13, 18-29
Computer, step-by-step 5-1 Reversible engines 18-18
Of flux 24-22 Energy flow diagrams 18-15
Of integrals Cal 1-11 Entropy 18-22
Calculus Introduction to 18-11
And the uncertainty principle 4-1, Cal 1-3 Maximally efficient engines 18-15
Calculating integrals Cal 1-11 Refrigerator, energy flow diagrams 18-15
Calculus in physics 4-1, Cal 1-3 Reversible engines 18-13
Chain rule Cal 1-25 Reversiblility 18-17
Definition of acceleration 4-5, Cal 1-7 Second law of thermodynamics 18-4
Component equations Cal 1-8 Cassegrain telescope Optics-42
Vector equation Cal 1-7 Cavendish experiment 8-7
Definition of velocity 4-3, Cal 1-5 Center of mass
Component equations Cal 1-8 Diver movie 11-1
Vector equation Cal 1-6 Dynamics of 11-4
Derivation, electric force of charged rod 24-6 Formula for 11-3
Derivation of constant acceleration formulas 4- Gravitational force acting on 13-4
9, Cal 1-20 Introduction to 11-2
In three dimensions 4-11, Cal 1-22 Center of our galaxy Optics-47
Limiting process 4-1, Cal 1-3, Cal 1-5 Cerenkov radiation Optics-10
Vector equation for Cal 1-5 CERN
Line integral 29-5 Electroweak theory int-26
Special chapter on Cal 1-3 Proton synchrotron at 28- 24
Surface integral 29-2 CGS units Back cover-1
Calculus in Physics 4-1 Classical hydrogen atom 35-2
Calibration of force detector 11-10 Coulomb's law 24-2
Camera Definition of electric charge 19-8
Depth of field Optics-34 Chain rule Cal 1-25
Pinhole Optics-35 Proving it (almost) Cal 1-26
Single lens reflex Optics-33 Remembering it Cal 1-25
Index-6

Chaos 23-1 Classical physics int-7

Charge Clock
Addition of 19-10 Atomic clocks 1-21
Conservation of int-21 Lack of simultaneity 1-32
Density, created by Lorentz contraction 28- 6 Light pulse clock 1-14
Discussion of int-6 Muon clock 1-20
Electric, definition of (CGS units) 19-8 Time dilation 1-22
Fractional (quarks) int-24, 19-15 Cluster, globular 11-2
Magnetic moment for circular orbit 31-24 Cochlea (inside of the ear) 16-34
On electron, Millikan oil drop experiment 26-17 Coefficient of friction 9-13
Positive and negative int-6, 19-10 Coefficients, Fourier (Fourier analysis lecture) 16-28
Quantization of electric 19-14 Coil
Surface 26-2 As a circuit element 31-7
Unit test 24-11 Field of a solenoid 28- 17, 29-14
Charges, static, line integral for 30-2 Inductance of 31-5
Charm quark int-24 Magnetic field of Helmholtz coils 28- 19
Chemistry. See Atoms: Angular momentum quantum Primary 30-26
number Toroidal 31-6
Cholera molecule 17-2 In LC circuit 31-11
Chromatic aberration Optics-21 Torroidal 29-17
Newton's reflecting telescope Optics-22 Coil, primary 30-26
Ciliary muscle, eye Optics-31 Collisions
Circuits Discovery of the atomic nucleus 11-19
Basic 27- 1 Energy loss 11-14
Grounding 26-8 Experiments on momentum conservation 7- 4
Inductor as a circuit element 31-7 Force detector 11-10
Kirchoff’s law 27- 10 Impulse 11-9
LC circuit Introduction to 11-9
Experiment 31-13 Momentum conservation during 11-13
Fourier analysis of 31-31 Subatomic 7- 7
Introduction to 31-10 That conserve momentum and energy (elastic) 11-
Ringing like a bell 31-36 16
LR circuit X. See Experiments I: - 8- Collisions
Exponential decay 31-9 Color force 19-15
Introduction to 31-8 Colors
Neon oscillator circuit 27- 29 And Fourier analysis 16-28
Power in 27- 9 Blackbody radiation 32- 22
RC circuit Color of stars 34-2
Exponential decay 27- 23 Electromagnetic Spectrum 32- 20
Exponential rise 27- 26 Glass prism and rainbow of colors Optics-15
Initial slope 27- 25
Comment lines, computer 5-7
Introduction to 27- 22
Measuring time constant 27- 25 Commutative law, exercise on 2-7
Time constant 27- 24 Compass needles, direction of magnetic field 28-
X. See Experiments II: - 3- The RC Circuit 12
Short 27- 9 Component sine wave, Fourier analysis 16-28
Simple 27- 8 Components, vector Cal 1-7
The voltage divider 27- 13 Formula for cross product 2-17
Circular electric field 30-13 Introduction to 2-8
Line integral for 30-13 Compton scattering, photon momentum 34-15
Circular motion
Force causing 8-2
Particles in magnetic field 28- 20
Uniform
Introduction to 3-17
Magnitude of acceleration 3-18
Circular orbit, classical hydrogen atom 35-2
Circular wave patterns, superposition of 33-2
Classical hydrogen atom 35-2
 Index-7

Computer. See also BASIC program Conservation of

BASIC. See also BASIC program Angular momentum 8-32
Calculations Derivation from F = ma 12-16
Introduction to 5-2 Introduction to 7- 9
Step-by-step 5-1 Electric charge 19-13
Commands Energy int-11, 8-35
Comment lines 5-7 Feynman's introduction to 10-2
DO LOOP 5-4 Mass on spring 14-11
LET Statement 5-5 Uncertainty principle 40-24
Multiplication notation 5-6 Work Energy Theorem 10-20
Selected Printing (MOD command) 5-10 X. See Experiments I: - 9- Conservation of energy
Variable names 5-6 Energy and momentum, elastic collisions 11-16
English program Linear and angular momentum, chapter on 7- 1
For projectile motion 5-16, 5-19 Linear momentum 7- 2, 11-7
For satellite motion 8-19 during collisions 11-13
Plot of electric fields Conservative force 25-5
Field plot model 25-12 And non-conservative force 10-21
In electron gun 26-13 Definition of 29-6
Of various charge distributions 24-19 Conserved field lines, flux tubes 24-17
Plotting Constant acceleration formulas
A point 5-6
Angular analogy 12-3
Crosses 5-11
Calculus derivation 4-9, Cal 1-20
Window 5-7
In three dimensions 4-11, Cal 1-22
Prediction of motion 5-12
Chapter on 5-1 Constant, integral of Cal 1-13
Satellite orbits 8-16 Constant voltage source 27- 15
Satellite with modified gravity 8-30 Continuity equation
Program for For electric fields 24-14
Air resistance 5-24 For fluids 23-5
Damped harmonic motion 14-34 Continuous creation theory int-4
Harmonic motion 14-30 Contour map 25-1
Hydrogen molecule ion 19-24 Contraction, Lorentz relativistic 1-24
Plotting a circle 5-2, 5-4, 5-11 Conversion factors Back cover-1
Projectile motion, final one 5-21 Cook’s Bay, Moorea, rainbow over Optics-17
Projectile motion, styrofoam projectile 5-28
Coordinate system, right handed 2-18
Projectile motion with air resistance 5-22
Satellite motion 8-21 Coordinate vector
Programming, introduction to 5-4 Definition of 3-11
Satellite motion calculational loop 8-19 In computer predictions 5-12, 8-17
Time Step and Initial Conditions 5-14 In definition of velocity vector 3-13
Computer analysis of satellite motion. See Experi- Cornea Optics-31
ments I: - 5- Computer analysis of satellite motion Corner reflector
Computer prediction of projectile motion. See How it works Optics-7
Experiments I: - 2- Computer prediction of On the surface of the moon Optics-7
projectile motion Cosine function
Amplitude of Cal 1-37
Computers
Definition of Cal 1-35
Why they are so good at integration Cal 1-12
Derivative of Cal 1-38
Conductors
Cosine waves
And electric fields, chapter on 26-1
Derivative of 14-8
Electric field in hollow metal sphere 26-4
Fourier analysis lecture 16-28
Electric field inside of 26-1
Phase of 14-6
Surface charge density 26-3
Cosmic background neutrinos int-30
Cones, nerve fibers in eye Optics-31
Cosmic background radiation int-30, 34-27
Conical pendulum 9-18
Cosmic radiation int-30
And simple pendulum 14-17
Cosmic rays int-13
Index-8

Coulomb's law D
And Gauss' law, chapter on 24-1 Damped harmonic motion
Classical hydrogen atom 35-2 Computer program for 14-34
For hydrogen atom 24-4 Differential equation for 14-21
For two charges 24-3 Damping, critical 14-23
Units, CGS 24-2 Davisson & Germer, electron waves 35-12
Units, MKS 24-2 De Broglie
Coupled air cart system, analysis of 16-12 Electron waves int-10, 35-11
Covalent bonding 19-15 Formula for momentum 35-11
Cp = Cv + R, specific heats 18-7 Hypothesis 35-10
Cp and Cv, specific heats 18-6 Introduction to wave motion 15-1
Crab nebula 20-16 Key to quantum mechanics 35-1
Creation of antimatter, positron-electron pairs 34-17 Wavelength, formula for 35-11
Critical damping 14-23 Waves, movie of standing wave model 35-11
Cross product Debye, on electron waves 37-1, 38-2
Angular momentum 12-11 Decay
Component formula for 2-17 Exponential decay Cal 1-32
Discussion of 2-15 Decoupling of light and matter in early universe 34-
Magnitude of 2-17 31
Review of 12-9 Definite integral
Right hand rule 12-10 Compared to indefinite integrals Cal 1-14
Crystal Defining new functions Cal 1-15
Diffraction by Thin 36-6 Introduction to Cal 1-11
Graphite, electron diffraction by 36-8 Of velocity Cal 1-11
Structures Process of integrating Cal 1-13
Graphite 36-8 Deflection plates in electron gun 26-16
Ice, snowflake 17-4 Degrees of freedom
X ray diffraction 36-5 Freezing out of 17-32
Crystalline lens, eye Optics-31 Theory of 17-28
Current and voltage Depth of field, camera Optics-34
Fluid analogy 27- 6 Derivative
Ohm’s law 27- 7 As a limiting process Cal 1-6, Cal 1-18, Cal 1-
Resistors 27- 6 23, Cal 1-28, Cal 1-30
Current, electric Constants come outside Cal 1-24
Inertia of (inductance) 31-12 Negative slope Cal 1-31
Introduction to 27- 2 Of exponential function e to the x Cal 1-28
Magnetic force on 31-18 Of exponential function e to the ax Cal 1-29
Positive and negative 27- 3 Of function x to the n'th power Cal 1-24
Current loop Of sine function Cal 1-38
Magnetic energy of 31-22 Derivative as the Slope of a Curve Cal 1-30
Torque on 31-20 Descartes, explanation of rainbow Optics-16
Currents Description of motion 3-3
Magnetic force between 28- 14, 31-19
Detector for radiated magnetic field 32- 26
Curve
Diagrams, PV (pressure, volume) 18-8
Area under, integral of Cal 1-12
Differential equation
Slope as derivative Cal 1-30
For adiabatic expansion 18-27
That increases linearly, integral of Cal 1-13
For damped harmonic motion 14-21
Velocity, area under Cal 1-12
For forced harmonic motion 14-25, 14-28
Curved surfaces, reflection from Optics-3
For LC circuit 31-10
Cycle, Carnot 18-11
For LR circuit 31-9
For oscillating mass 14-8
Introduction to 4-14
Differentiation. See also Derivative
Chain rule Cal 1-25
More on Cal 1-23
 Index-9

Differentiation and integration Disorder

As inverse operations Cal 1-18 Direction of time 18-25
Velocity and position Cal 1-18, Cal 1-19 Entropy and the second law of thermodynamics 18-
Fast way to go back and forth Cal 1-20 4
Position as integral of velocity Cal 1-20 Formula for entropy 18-24
Velocity as derivative of position Cal 1-20 Displacement vectors
Diffraction From strobe photos 3-5
By thin crystals 36-6 Introduction to 2-2
Electron diffraction tube 36-9 Distance, tangential 12-4
Of water waves 33-5 Distant galaxies, Hubble photograph int-3
X Ray 36-4 Dive movie, time reversed 18-1
Diffraction, electron. See Experiments II: -11- Elec- Diver, movie of 11-1
tron diffraction experiment Diverging lenses Optics-26
Diffraction grating 33-12. See also Experiments II: - DO LOOP, computer 5-4
10- Diffraction grating and hydrogen spectrum Doppler effect
Diffraction limit, telescopes Optics-45 Astronomer's Z factor 33-23
Diffraction pattern 16-33, 33-5 For light 33-22
Analysis of 36-11 In Astronomy 33-23
By strand of hair 36-14 Introduction to 33-20
Electron 36-10 Relativistic formulas 33-22
For x rays 36-5 Stationary source, moving observer 33-21
Of human hair 36-14 Universe, evolution of 34-21
Recording 33-28 Dot product
Single slit 33-27 Definition of 2-12
Student projects 36-13 Interpretation 2-14
Two-slit 33-6 Work and energy 10-13
Dimensional analysis Down quark int-24
For predicting the speed of light 15-9 Drums, standing waves on 16-22
For predicting the speed of sound 15-6 Duodenum, medical imaging Optics-15
Dimensions
Period and frequency 14-4 E
Using, for remembering formulas 14-4 e - charge on an electron 19-9
Dimensions of E = hf, photoelectric effect formula 34-7
Capacitance Front cover-2 E = mc2, mass energy int-11, 10-3
Electric charge Front cover-2 E.dl meter 30-18
Electric potential Front cover-2 Eagle nebula
Electric resistance Front cover-2 Big photo of 7- 18
Energy Front cover-2 Hubble telescope photo Optics-44
Force Front cover-2 Planet formation 7- 16
Frequency Front cover-2 Ear, human
Inductance Front cover-2 Inside of cochlea 16-34
Magnetic field Front cover-2 Structure of 16-15
Magnetic flux Front cover-2 Early universe. See Universe, early
Power Front cover-2
Earth
Pressure Front cover-2
Gravitational field inside of 24-24
Dirac equation
Mass of 8-8
Antimatter int-13, 15-2, 34-16
Earth tides 8-12
Electron spin 39-3
Eclipse expedition, Eddington 34-19
Bohr magneton 39-5
Dirac, P. A. M., prediction of antiparticles int-13 Edit window for Fourier transform data 16-32
Direction an induced electric field 31-3 Effective nuclear charge, periodic table 38-12
Direction of time Efficiency
And strobe photographs 3-27 Of Carnot cycle, calculation of 18-26
Dive movie 18-1 Of electric cars 18-19
Entropy 18-25 Of heat pump 18-19
Neutral K meson 18-25 Of reversible engines 18-18
Rising water droplets 18-3
Discovery of the atomic nucleus 11-19
Index-10

Einstein Electric field continued

General relativity int-15, 8-29 Introduction to 24-10
Mass formula 6-10 Line integral of 30-14
Photoelectric effect int-8 Lines 24-12
Photoelectric effect formula 34-7 Mapping 24-12
Principle of relativity, chapter on 1-12 Mapping convention for 24-17
Einstein cross, gravitational lens 34-20 Of a line charge
Elastic collisions, conservation, energy, momen- Using calculus 24-6
tum 11-16 Using Gauss' law 24-21
Elasticity of rubber 17-35 Of electromagnet 30-15
Electric and weak interactions unified 19-3 Of static charges, conservative field 30-2, 30-16
Electric cars, efficiency of 18-19 Radiation by line charge 32- 28
Radiation by point charge 32- 30
Electric charge
Van de Graaff generator 26-6
Conservation of 19-13
Electric force
Definition (CGS units) 19-8
Definition (MKS units) 31-19 Between garden peas 19-12
Quantization of 19-14 Produced by a line charge 24-6
Produced by a short rod 24-9
Electric circuits
Electric force law
Basic 27- 1
Grounding 26-8 Four basic interactions 19-2
Kirchoff’s law 27- 10 In CGS units 19-8
LC circuit, oscillation of 31-10 Introduction to 19-7
LR circuit, exponential decay of 31-9 Lorentz force law, electric and magnetic forces 28-
Power in 27- 9 15
RC circuit Electric force or interaction int-6, int-13
Equations for 27- 22 Atomic & molecular forces 19-1
Exponential decay of 27- 23 Electroweak theory int-26
Half life 27- 25 Strength of
Initial slope 27- 25 Between garden peas 28- 2
Time constant 27- 24 Comparison to gravity int-6, 19-8
The voltage divider 27- 13 Comparison to nuclear force int-18, 20-2
Electric current Origin of magnetic forces 28- 6
Inertia of, due to inductance 31-12 Electric potential
Introduction to 27- 2 Contour map 25-1
Positive and negative 27- 3 Field plots 25-1
Electric discharge of Van de Graaff generator 26-7 Of a point charge 25-5
Electric field Plotting experiment 25-7
And conductors 26-1 Electric potential energy. See Potential energy:
And light int-7, 32- 20 Electric
Circular electric field Electric voltage. See also Voltage
Introduction to 30-13 Introduction to 25-6
Line integral for 30-13 Van de Graaff generator 26-6
The betatron 30-16 Electrical capacitance 27- 16. See also Capacitor
Computer plot, -3,+5 charges 24-19 Electrically neutral int-6
Computer plotting programs 25-12 Electromagnet 31-28
Continuity equation for 24-14 Electromagnetic radiation
Contour map 25-1 Energy radiated by classical H atom 35-3
Created by changing magnetic flux 31-2 Observed by telescopes
Created by moving magnetic field, Lorentz force 30- Infrared Optics-46
9 Radio Optics-48
Direction of, when created by magnetic flux 31-3 Visible Optics-42
Energy density in 27- 19 Pulse 32- 10
Equipotential lines 25-3 Analysis of path 1 32- 14
Flux, definition of 24-15 Analysis of path 2 32- 16
Gauss’ law 24-20 Calculation of speed 32- 14
In electromagnetic waves 32- 18 Electromagnetic spectrum int-7, 32- 20, 34-11
Inside a conductor 26-1 Photon energies 34-11
Integral of E.dl meter 30-20 Electromagnetic waves 32- 18
Probability wave for photons 40-7
 Index-11

Electron Elementary particles

Beam, magnetic deflection 28- 9 A confusing picture int-22
Charge on 19-8 Short lived 40-23
Diffraction Pattern 36-10 Elements
In classical hydrogen atom 35-2 Abundance of 34-24
Lepton family int-22 Creation of int-4
Mass in beta decay 6-7 Table of 19-5
Motion of in a magnetic field. See Experiments II: - 5- Ellipse
Motion of electrons in a magnetic field Becoming a parabola Optics-4
Radius 39-3 Drawing one 8-26, Optics-3
Rest energy in electron volts 26-12 Focus of 8-26, Optics-3
Spin Empty space, quantum fluctuations 40-25
And hydrogen wave patterns 38-9 Energy
Chapter on 39-1 Bernoulli's equation 23-10
Spin resonance Black holes 10-29
Details of experiment 39-9 Capacitors, storage in 27- 18
Introduction to 39-5
Chapter on 10-1
X. See Experiments II: -12- Electron spin resonance
Conservation of energy
Stability of 19-14
And the uncertainty principle 40-24
Two slit experiment for 40-3 Conservative and non-conservative forces 10-21
Electron binding energy Derivation from work theorem 10-20
A classical approach 19-21 Feynman story 10-2
And the periodic table 38-11 In collisions 11-14
In classical hydrogen molecule ion 19-23 In satellite motion 8-35
Electron diffraction experiment 36-8 Mass on spring 14-11
Diffraction tube 36-9 Neutrinos in beta decay 11-20
X. See Experiments II: -11- Electron diffraction Overview int-11
experiment Work energy theorem 10-20
Electron gun X. See Experiments I: - 9- Conservation of energy
Accelerating field 26-10 E = Mc2 10-3
Electron volt 26-12 Electric potential energy
Equipotential plot 26-11, 26-13 And molecular force 17-12
Filament 26-9 Contour map of 25-1
In classical hydrogen atom 35-3
In magnetic field
In hydrogen atom int-11
Bend beam in circle 28- 20
In hydrogen molecule ion 19-21
Magnetic focusing 28- 29
In nuclear fission int-18, 20-5
Introduction to 26-8
Negative and positive 25-4
X. See Experiments II: - 2- The Electron Gun
Of a point charge 25-5
Electron positron pair 34-17 Plotting 25-7. See also Experiments II: - 1- Potential
Electron scattering plotting
Chapter on scattering 36-1 Storage in capacitors 27- 18
First experiment on wave nature 35-12 Electron binding and the periodic table 38-11
Electron screening, periodic table 38-10 Electron, in the hydrogen molecule ion 19-21
Electron type neutrino int-22 Electron volt as a unit of energy 19-21, 26-12
Electron volt Energy density in an electric field 27- 19
As a Unit of Energy 19-21, 26-12 Energy level 35-1
Electron gun used to define 26-12 Energy loss during collisions 11-14
Electron waves Equipartition of energy 17-28
Davisson & Germer experiment 35-12 Failure of classical physics 17-31
De Broglie picture 35-11 Freezing out of degrees of freedom 17-32
In hydrogen 38-1 Real molecules 17-30
Scattering of 35-12 From nuclear fission 20-4
Wavelength of 36-9 From sun int-18
Electroweak interaction 19-3
Theory of int-26
Weak interaction int-26
Z and W mesons int-26
Electroweak interactions 19-3
Index-12

Energy Continued Total energy

Gravitational potential energy int-11, 8-35 Classical H atom 35-3
Bernoulli's equation 23-10 Escape velocity 10-28
Black holes 10-29 Satellite motion 8-36, 10-26
In a room 10-25 Uncertainty principle
In satellite motion 8-36, 10-26 Energy conservation 40-24
In stellar evolution 20-13 Energy-time form of 40-19
Introduction to 10-8 Fourier transform 40-20
Modified 8-37 Voltage as energy per unit charge 25-6
On a large scale 10-22 Work
Zero of 10-22 Conservation of energy int-11
Joules and Ergs 10-4 Definition of 10-12
Kinetic energy int-8 Vector dot product 10-13
Always positive 8-35 Work energy theorem 10-18
Bohr model of hydrogen 35-3 X Ray photons, energy of 36-4
Classical hydrogen atom 35-3 Zero of potential energy 10-22
Electron diffraction apparatus 36-9 Zero point energy 37-7
Equipartition of energy 17-28 Chapter on 37-1
Escape velocity 10-28 Energy flow diagrams for reversible engines 18-15
Hydrogen molecule ion 19-21 Energy from sun, proton cycle 34-24
Ideal gas law 17-18 Energy, kinetic, in terms of momentum 37-5
In collisions 11-14
Energy level diagram
In model atom 37-5
Nuclear fusion 20-12 Balmer series 35-6
Origin of 10-5 Bohr theory 35-4
Oscillating mass 14-11 Expanded 38-8
Overview int-8 Lyman series 35-6
Pendulum 10-10 Model atom 37-4
Relativistic definition of 10-5 Paschen series 35-6
Rotational 12-22 Photon in laser 37-4
Satellite motion 8-36, 10-26 Energy-time form of the uncertainty principle 40-19
Slowly moving particles 10-6, 10-29 Engines
Temperature scale 17-11 Internal combustion 18-21
Theorem on center of mass 12-26 Maximally efficient 18-15
Thermal motion 17-6 Reversible, efficiency of 18-18
Translation and rotation 12-24 English program
Work energy theorem 10-18
For oscillatory motion 14-31
Magnetic energy of current loop 31-22
For projectile motion 5-16
Mass energy int-18, 10-3
For satellite motion 8-19
Negative and positive potential energy 25-4
English program for projectile motion 5-16
Neutron mass energy int-22
Entropy
Nuclear potential energy
Fusion int-18, 20-12 Boltzman's formula for 18-24
Nuclear binding 20-9 Definition of 18-22
Nuclear energy well. 20-10 Number of ways to hang tools 18-23
Nuclear structure int-22 Second law of thermodynamics 18-1
Pendulum motion, energy in 10-10 Epicycle, in Greek astronomy 8-25
Photon energy 34-9 Equations, differential. See Differential equation
Photon pulse, uncertainty principle 40-21 Equations, vector
Potential. See also Potential energy Components with derivatives Cal 1-7
Powering the sun 34-23 In component form 2-10
Rest energy of electron and proton 26-12 Equilibrium
Rotational kinetic energy 12-26 Balancing weights 13-2
Spin magnetic energy Chapter on 13-1
Dirac equation 39-15 Equations for 13-2
Magnetic moment 39-4 Example - bridge problem 13-9
Magnetic potential 39-1 Example - wheel and curb 13-5
Spring potential energy int-11, 10-16, 14-11 How to solve equilibrium problems 13-5
Thermal energy int-8, 17-7 Thermal equilibrium 17-8
Working with rope 13-10
 Index-13

Equipartition of energy Experiments II

Failure of classical physics 17-31 - 1- Potential plotting 25-7
Freezing out of degrees of freedom 17-32 - 2- The electron gun 26-8
Normal modes 17-28 - 3- The RC circuit 27- 22
Real molecules 17-30 - 4- The neon bulb oscillator 27- 28
Theory of 17-28 - 5- Motion of electrons in a magnetic field 28- 19
Equipotential lines 25-3 - 5a- Magnetic focusing, space physics 28- 30
Model 25-10 - 6- Faraday's law air cart speed detector 30-5
Plotting experiment, 2D or 3D? 25-8 - 7- Magnetic field mapping using Faraday's law 30-
Equipotential plot for electron gun 26-11 24
Ergs and joules 10-4 - 8- Measuring the speed of light with LC circuit 31-
Ergs per second, power in CGS units 24-2 15
Escape velocity 10-28 - 9- LC circuit and Fourier analysis 31-31
Euler's number e = 2.7183. . . Cal 1-17 -10- Diffraction grating and hydrogen spectrum 33-
Evaporation 17
Of water 17-5 -11- Electron diffraction experiment 36-8
Surface tension 17-14 -12- Electron spin resonance 39-9
Even harmonics in square wave 16-28 -13- Fourier analysis and uncertainty principle 40-21
Evolution Exponential decay Cal 1-32
Of stars int-19, 17-17, 20-13 In LR circuits 31-9
Neutrinos role in 20-15 In RC circuits 27- 23
Of the universe 34-21 Exponential function
Excess of matter over antimatter 34-29 Derivative of Cal 1-28
Exclusion principle 38-1, 38-9 Exponential decay Cal 1-32
Indefinite integral of Cal 1-29
Exercises, finding them. See under x in this index
Integral of Cal 1-29
Expanding gas, work done by 18-5
Introduction to Cal 1-16
Expanding universe
Inverse of the logarithm Cal 1-16
Hubble, Edwin int-3 Series expansion Cal 1-28
Hubble rule for int-3 y to the x power Cal 1-16
Red shift 33-24, 34-19, 34-21
Eye glasses experiment Optics-36
Expansion
Eye, human
Adiabatic
Ciliary muscle Optics-31
Carnot cycle 18-11
Cornea Optics-31
Equation for 18-26
PV Diagrams 18-9 Crystalline lens Optics-31
Reversible engines 18-13 Farsightedness Optics-32
Isothermal Focusing Optics-32
Carnot cycle 18-11 Introduction Optics-31
Equation for 18-26 Iris Optics-31
PV Diagrams 18-8 Nearsightedness Optics-32
Reversible engines 18-13 Nerve fibers Optics-31
Thermal 17-33 Cones Optics-31
Uniform, of the universe int-3 Rods Optics-31
Expansion, binomial 1-31, Cal 1-23 Eyepiece Optics-37
Experimental diffraction pattern 16-33
Experiments I
- 1- Graphical analysis of projectile motion 3-17
- 2- Computer prediction of projectile motion 5-21
- 3- Conservation of linear momentum 7- 4
- 4- Conservation of angular momentum 7- 10
- 5- Computer analysis of satellite motion 8-23
- 6- Spring pendulum 9-4
- 7- Conservation of energy
Check for, in all previous experiments 10-26
- 8- Collisions 11-9
- 9- The gyroscope 12-18
-10- Oscillatory motion of various kinds 14-2
-11- Normal modes of oscillation 16-4
-12- Fourier analysis of sound waves 16-18
Index-14

F Field continued
F = ma. See also Newton's second law Electromagnetic field 32- 18
Applied to Newton’s law of gravity 8-5 Flux, introduction of concept 24-15
Applied to satellite motion 8-8 Gauss’ law 24-20
For Atwood’s machine 9-16 Gravitational field
For inclined plane 9-10 Definition of 23-3
For spring pendulum 9-7 Inside the earth 24-24
Of point mass 24-23
For string forces 9-15
Of spherical mass 24-24
Introduction to 8-4
Magnetic field
Vector addition of forces 9-6
Between capacitor plates 32- 6
f number Detector, radio waves 32- 26
For camera Optics-33 Direction of, north pole 28- 11
For parabolic mirror Optics-5 Gauss's law for (magnetic monopole) 32- 2
F(t) = (1)sin(t) + (1/3)sin(3t) + ... In coils 28- 17
Fourier analysis of square wave 16-28 In Helmholtz coils 28- 18
Failure of classical physics 17-31 Interaction with Spin 39-4
Faraday's law Introduction to 28- 10
AC voltage generator 30-21 Of a solenoid 29-14
Applications of 30-15 Of a toroid 29-17
Chapter on 30-1 Of straight wire 29-11
Derivation of 30-11 Surface integral 32- 2
Field mapping experiment 30-24 Thought experiment on radiated field 32- 11
Uniform 28- 16
Gaussmeter 30-23
Visualizing using compass needles 28- 12
Induced Voltage 31-4
Visualizing using iron filings 28- 12
Line integral 30-15
Plotting experiment 25-7
One form of 30-12
Vector field
Right hand rule for 30-15
Definition of 23-3
The betatron 30-16 Two kinds of 30-18
Velocity detector 30-25 Velocity field
Voltage transformer 30-26 Introduction to 23-2
X. See Experiments II: - 6- Faraday's law air cart Of a line source 23-7
speed detector; Experiments II: - 7- Magnetic field Of a point source 23-6
mapping using Faraday's law Field lines
Farsightedness Optics-32 Computer plots, programs for 25-12
Fermi Lab accelerator 28- 23 Electric
Feynman, R. P. int-14 Definition of 24-12
FFT Data button 16-32 Drawing them 24-13
Fiber optics Three dimensional model 25-10
Introduction to Optics-14 Field mapping
Medical imaging Optics-15 Magnetic field of Helmholtz coils 30-24
Field Magnetic field of solenoid 30-24
Conserved lines, fluid and electric 24-17 Field plots and electric potential, chapter on 25-1
Electric Filament, electron gun 26-9
Circular, line integral for 30-13 First maxima of two-slit pattern 33-8
Computer plot of (–3,+5) 24-19 Fission, nuclear 20-3
Continuity equation for 24-14 Fitch, Val, K mesons and the direction of time 18-27
Created by changing magnetic flux 31-2 Fluctuations, quantum, in empty space 40-25
Direction of circular or induced 31-3
Fluid dynamics, chapter on 23-1
Inside a conductor 26-1
Inside hollow metal sphere 26-4 Fluid flow, viscous effects 23-19
Integral of - E.dl meter 30-20 Fluorescence and reflection 40-8
Introduction to 24-10 Flux
Line integral of 30-14 Calculations, introduction to 24-22
Mapping convention 24-17 Definition of 24-15
Mapping with lines 24-12 Of magnetic field 30-11
Of electromagnet (turned on or off) 30-15 Of velocity and electric fields 24-15
Of line charge 24-21 Of velocity field 23-8
Of static charges 30-2, 30-16 Tubes of flux, definition of 24-17
Radiation by line charge 32- 28
Radiation by point charge 32- 30
Van de Graaff generator 26-6
 Index-15

Flux, magnetic Force detector 11-10

AC voltage generator 30-21 Forced harmonic motion, differential equation
Definition of 30-11 for 14-25, 14-28
Faraday's law Forces, addition of 9-2
Line integral form 30-15 Formation of planets 7- 17
Voltage form 30-12 Four basic interactions int-25, 19-1
Field mapping experiment 30-24 Fourier analysis
Gaussmeter 30-23 Amplitude and intensity 16-33
In the betatron 30-16 Amplitude and phase 16-31
Integral E.dl meter 30-19 And repeated wave forms 16-11
Magnetic field detector 32- 26 Calculating Fourier coefficients 16-28
Maxwell's equations 32- 8 Energy-time form of the uncertainty principle 40-20
Velocity detector 30-25 Formation of pulse from sine waves 40-27
Voltage transformer 30-26 In the human ear 16-16
Focal length Introduction to 16-6
For parabolic mirror Optics-5 Lecture on Fourier analysis 16-28
Negative, diverging lenses Optics-26 Normal modes and sound 16-1
Of a spherical surface Optics-20 Of a sine wave 16-7
Two lenses together Optics-29 Of a square wave 16-9, 16-28
Focus Of coupled air carts, normal modes 16-12
Eye, human Optics-32 Of LC circuit 31-31
Of a parabolic mirror Optics-4 Of slits forming a diffraction pattern 16-33
Of an ellipse 8-26, Optics-3 Of sound waves. See Experiments I: -12- Fourier
Focusing, magnetic 28- 29 analysis of sound waves
Focusing of sound waves 8-26, Optics-3 Of violin, acoustic vs electric 16-19
Force 8-2 X. See Experiments II: -13- Fourier analysis & the
Color force 19-15 uncertainty principle
Conservative and non-conservative 10-21 Fourier coefficients
Conservative forces 25-5 Accurate values of 16-32
Electric force Calculating 16-31
Classical hydrogen atom 35-2 Lecture on 16-28
Introduction to 19-7 Fourier, Jean Baptiste 16-2
Produced by a line charge 24-6
Fractional charge int-24
Strength of (garden peas) 28- 2
Freezing out of degrees of freedom 17-32
Four basic forces or interactions 19-1
Introduction to force 8-2 Frequencies (Fourier analysis) 16-28
Lorentz force law 32- 8 Frequency
Magnetic force Angular 15-14
Between currents 31-19 Of oscillation of LC circuit 31-10
On a current 31-18 Photon energy E=hf 34-7
Origin of 28- 10 Spacial frequency 15-14
Magnetic force law Frequency, period, and wavelength 15-13
Derivation of 28- 10 Friction
Vector form 28- 14 Coefficient of 9-13
Molecular force Inclined plane 9-12
A classical analysis 19-19 Functions obtained from integration Cal 1-15
Analogous to spring force 14-20 Logarithms Cal 1-15
Introduction to 19-15 Fusion, nuclear int-18, 20-12
Potential energy for 17-12
Non linear restoring force 14-19
Nuclear force int-18
Introduction to 20-2
Range of 20-3
Particle nature of forces int-13
Pressure force 17-16
Spring force
As molecular force 14-20, 17-12
Hook's law 9-3
String force
Atwood’s machine 9-16
Tension 9-15
Index-16

G Graphite crystal
Galaxy Electron diffraction experiment 36-8
Andromeda int-2 Electron scattering 36-1
Center of our galaxy Optics-47 Structure of 36-8
Introduction to int-2 Grating
Most distant int-3 Diffraction 33-12
Sombrero int-2 Multiple slit
Space travel 1-22 Fourier analysis of 16-33
Galileo Interference patterns for 33-12
Falling objects (Galileo Was Right!) 8-6 Three slit 16-33
Inclined plane 9-10 Gravitational field
Portrait of 9-11 An abstract concept 23-3
Galileo’s inclined plane 9-11 Gauss' law for 24-23
Galileo’s telescope Optics-41 Inside the earth 24-24
Of point mass 24-23
Gallbladder operation, medical image Optics-15
Of spherical mass 24-24
Gamma = Cp/Cv, specific heats 18-7
Gravitational force. See Gravity
Gamma rays 32- 20, 32- 22
Gravitational lens, Einstein cross 34-20
Photon energies 34-11
Gravitational mass 6-5
Wavelength of 32- 20
Gamov, George, big bang theory int-4 Gravitational potential energy
Energy conservation int-11
Garden peas
Graviton int-15
Electric force between 19-12
Garden peas, electric forces between 28- 2 Gravity int-15
Acceleration due to 3-21
Gas constant R 17-25
And satellite motion 8-8
Gas, expanding, work done by 18-5
Black hole int-20, 10-29
Gas law, ideal 17-18
Cavendish experiment 8-7
Gaudsmit and Uhlenbeck, spin 39-1 Deflection of photons 34-19
Gauss' law Earth tides 8-12
Electric field of line charge 24-21 Einstein's general relativity int-15
For gravitational fields 24-23 Four basic interactions 19-1
For magnetic fields 32- 2 Gravitational force acting at center of mass 13-4
Introduction to 24-20 Gravitational potential energy
Solving problems 24-26 Bernoulli's equation 23-10
Surface integral 29-3 Black holes 10-29
Gauss, tesla, magnetic field dimensions 28- 16 Conservation of Energy 8-35
Gaussmeter 30-23 Energy conservation int-11
Gell-Mann In a room 10-25
Quarks int-24, 19-14 In satellite motion 8-36
General relativity int-15, 8-29 Introduction to 10-8
Modified gravity 8-29 Modified 8-37
Geometrical optics On a large scale 10-22
Zero of 10-22
Chapter on Optics-1
Inertial and gravitational mass 8-8
Definition of Optics-2
Interaction with photons 34-18
Glass prism Optics-13
Modified, general relativity 8-29
Globular cluster 11-2
Newton's universal law int-15, 8-5
Gluons, strong nuclear force int-25 Potential energy. See also Potential energy: Gravita-
Graph paper tional
For graphical analysis 3-33 Introduction to 10-8
For projectile motion 3-29 On a Large Scale 10-22
Graphical analysis Quantum theory of int-16
Of instantaneous velocity 3-26 Strength, comparison to electricity int-6, 19-8
Of projectile motion 3-17 Weakness & strength int-20
Of projectile motion with air resistance 3-22 "Weighing” the Earth 8-8
X. See Experiments I: - 1- Graphical analysis of Weight 8-11
projectile motion Green flash Optics-17
Grounding, electrical circuits 26-8
 Index-17

Guitar string Helmholtz coils 28- 17, 28- 18

Sound produced by 15-22 100 turn search coil 39-12
Waves 15-20 Electron spin resonance apparatus 39-11
Waves, frequency of 15-21 Field mapping experiment 30-24. See also Experi-
Gun, electron. See Electron gun ments II: - 7- Magnetic field mapping using
Gyromagnetic ratio for electron spin 39-14 Faraday's law
Gyroscopes Motion of electrons in 28- 20, 28- 29
Atomic scale 39-15 Uniform magnetic field inside 28- 17
Movie 12-18 Hertz, Heinrich, radio waves 34-1
Precession formula 12-21 Hexagonal array
Precession of 12-19 Graphite crystal and diffraction pattern 36-9
Theory of 12-18 Homework exercises, finding them. See X-Ch (chap-
X. See Experiments I: - 9- The gyroscope ter number): Exercise number
Hooke’s law 9-4
H In dimensional analysis 15-7
h bar, Planck's constant 35-9 Horsehead nebula in visible & infrared light Optics-
Hair, strand of, diffraction pattern of 36-14 46
Half-life Hot early universe int-4
In exponential decay Cal 1-33 Hubbel space telescope Optics-44
In RC circuit 27- 25 Hubble, Edwin, expanding universe int-3
Of muons (as clock) 1-20 Hubble photograph of most distant galaxies int-3
Of muons, exponential decay Cal 1-33 Hubble rule for expanding universe int-3, 33-24
Halos around sun Optics-18 Hubble telescope mirror Optics-44
Harmonic motion Spherical aberration in Optics-22
Computer program 14-30 Human ear
Damped Description of 16-15
Computer program 14-34
Inside of cochlea 16-34
Critical damping 14-23
Differential equation for 14-21 Human Eye Optics-31
Forced Huygens
Analytic solution 14-28 Wave nature of light 34-1, Optics-1
Differential equation for 14-25, 14-28 Huygens' principle 33-4
Harmonic oscillator 14-12 Preliminary discussion of 15-2
Differential equation for 14-14 Hydrodynamic voltage
Harmonic series 16-3 Bernoulli's equation 23-17
Harmonics and Fourier coefficients 16-28 Resistance 27- 7
Hays, Tobias Town water supply 23-18
Dive Movie, center of mass 11-1 Hydrogen atom
Dive movie, time reversed 18-1 Angular momentum quantum number 38-7
Heat capacity 17-26 Big bang theory int-4
Molar 17-26 Binding energy of electron 38-12
Heat pump, efficiency of 18-19 Bohr theory 35-1
Heat, specific. See Specific heat Classical 35-2
Coulomb's law 24-4
Heisenberg, Werner 4-1, Cal 1-3
Expanded energy level diagram 38-8
Hele-Shaw cell, streamlines 23-4
Quantized projections of angular momentum 38-5
Helium
Solution of Schrödinger’s equation 38-2
Abundance in early universe 34-26 Standing wave patterns in 38-3
And electron spin 38-9 The L = 0 Patterns 38-4
And the Pauli exclusion principle 38-9 The L ≠ 0 Patterns 38-5
Binding energy of last electron 38-12
Hydrogen atom, classical
Creation of in universe int-4
Failure of Newtonian mechanics 35-3
Energy to ionize 38-9
Hydrogen bomb int-18, 20-13
In periodic table 38-11
Hydrogen molecule
Isotopes helium 3 and 4 int-17
Formation of 19-16
Hydrogen molecule ion
Binding energy and electron clouds 19-23
Computer program for 19-24
Formation of 19-16
Index-18

Hydrogen nucleus int-6, 19-3 Inductor

Isotopes of 19-6 As a circuit element 31-7
Hydrogen spectrum Definition of 31-2
Balmer series 33-19, 35-4 Iron core 31-29
Bohr model int-8 LC Circuit 31-10
Experiment on 33-17 LR circuit 31-8
Lyman series 35-6 Toroidal coil 31-6
Of star 35-4 Inertia
Paschen series 35-6 Inertial mass 6-5
X. See Experiments II: -10- Diffraction grating and Moment of (Angular mass) 12-7
hydrogen spectrum Of a massive object 31-12
Hydrogen wave patterns Of an electric current 31-12
Intensity at the origin 38-5 Infrared light
L= 0 patterns 38-4 Ability to penetrate interstellar dust Optics-46
Lowest energy ones 38-3 Center of our galaxy Optics-47
Schrödinger’s Equation 38-2 Horsehead nebula in visible & infrared Optics-46
Hydrogen-Deuterium molecule, NMR experi- In the electromagnetic spectrum int-7
ment 39-12 Paschen series, hydrogen spectra 35-6
Hydrostatics, from Bernoulli's equation 23-12 Wavelength of 32- 20
Infrared Telescopes Optics-46
I Infrared camera Optics-46
IBM Labs, atomic microscopes 17-1 IRAS satellite Optics-47
Ice crystal 17-4 Map of the entire sky Optics-47
Ideal gas law 17-18 Mt. Hopkins 2Mass telescope Optics-46
Chemist's form 17-25 Viewing center of our galaxy Optics-47
Ideal gas thermometer 17-20 Initial conditions in a computer program 5-14
Absolute zero 17-21 Initial slope in RC circuit 27- 25
Image Inside the cochlea 16-34
Image distance Instantaneous velocity
Lens equation Optics-24 And the uncertainty principle 4-2, Cal 1-4
Negative Optics-26 Calculus definition of 4-3, Cal 1-5
In focal plane of telescope mirror Optics-5 Definition of 3-24
Medical Optics-15 From strobe photograph 3-26
Gallbladder operation Optics-15 Instruments
Of duodenum Optics-15 Percussion 16-22
Impulse Stringed 16-18
Change in momentum 11-12 Violin, acoustic vs electric 16-19
Experiment on 11-9 Wind 16-20
Measurement 11-11 Integral
Inclined plane 9-10 As a sum Cal 1-10
Galileo’s inclined plane, photo of 9-11 Calculating them Cal 1-11
Objects rolling down 12-25 Definite, introduction to Cal 1-11
With friction 9-12 Formula for integrating x to n'th power Cal 1-14, Cal
Indefinite integral 1-27
Definition of Cal 1-14 Indefinite, definition of Cal 1-14
Of exponential function Cal 1-29 Of 1/x, the logarithm Cal 1-15
Index of refraction Of a constant Cal 1-13
Definition of Optics-9 Of a curve that increases linearly Cal 1-13
Glass prism and rainbow of colors Optics-15 Of a velocity curve Cal 1-12
Introduction to Optics-2 Of exponential function e to the ax Cal 1-29
Of gas of supercooled sodium atoms Optics-9 Of the velocity vector Cal 1-10
Table of some values Optics-9 As area under curve Cal 1-12
Induced voltage Of x to n'th power
In moving loop of wire 30-4 Indefinite integral Cal 1-27
Line integral for 31-4
Inductance
Chapter on 31-1
Derivation of formulas 31-5
 Index-19

Integral, line Iron 56, most tightly bound nucleus int-18, 20-11
Ampere’s law 29-7 Iron core inductor 31-29
Conservative force 29-6 Iron core of massive star 20-15
Evaluation for solenoid 29-15 Iron fillings, direction of magnetic field 28- 12
Evaluation for toroid 29-17 Iron magnets 31-26
Faraday's law 30-15 Isothermal expansion
For circular electric field 30-13 Calculation of work 18-26
For static charges 30-2 PV Diagrams 18-8
In Maxwell's equations 32- 8 Isotopes of nuclei int-17, 19-6
Introduction to 29-5 Stability of int-22
Two kinds of fields 30-18
Integral of E.dl meter 30-18, 30-20 J
Integral sign Cal 1-10 Jeweler using magnifier Optics-38
Integral, surface 29-2 Joules and Ergs 10-4
For magnetic fields 32- 2
Formal introduction 29-2 K
Gauss’ law 29-3 K meson and direction of time int-23, 40-23
In Maxwell's equations 32- 8 Karman vortex street 14-25
Two kinds of fields 30-18 Kepler's laws
Integration Conservation of angular momentum 8-32
Equivalent to finding area Cal 1-11 First law 8-26
Introduction to Cal 1-8 Introduction to 8-24
Introduction to finding areas under curves Cal 1-13 Second law 8-27
Why computers do it so well Cal 1-12 Third law 8-28
Integration and differentiation Kilobaud, fiber optics communication Optics-14
As inverse operations Cal 1-18 Kinetic energy
Velocity and position Cal 1-19 Always positive 8-35
Fast way to go back and forth Cal 1-20 Bohr model of hydrogen 35-3
Position as integral of velocity Cal 1-20 Classical hydrogen atom 35-3
Velocity as derivative of position Cal 1-20 Electron diffraction apparatus 36-9
Integration formulas Cal 1-27 Equipartition of energy 17-28
Intensity Escape velocity 10-28
And amplitude, Fourier analysis lecture 16-33 Hydrogen molecule ion 19-21
Of diffraction pattern 16-33 Ideal gas law 17-18
Of harmonics in Fourier analysis of light pulse 40-22 In collisions 11-14
Of probability wave 40-22 In model atom 37-5
Sound intensity, bells and decibels 16-24 In terms of momentum 37-5
Sound intensity, speaker curves 16-27 Nonrelativistic 10-6
Interactions. See also the individual forces Nuclear fusion 20-12
Electric int-14 Origin of 10-5
Four basic 19-1 Oscillating mass 14-11
Gravitational int-15 Overview int-8
Nuclear int-14 Pendulum 10-10
Photons and gravity 34-18 Relativistic definition of 10-5
Weak int-14 Rotational 12-22
Interactions, four basic 19-1 Satellite motion 8-36, 10-26
Interference patterns Slowly moving particles 10-6, 10-29
A closer look at 33-26 Temperature scale 17-11
Introduction to 33-3 Theorem on center of mass 12-26
Two-slit Thermal motion 17-6
Light waves 33-10 Translation and rotation 12-24
Probability waves 40-9 Work energy theorem 10-18
Water waves 33-6 Kirchoff’s law
Internal combustion engine 18-21 Applications of 27- 11
Internal reflection Optics-13 Introduction to 27- 10
Interval, evaluating variables over Cal 1-10
Ionic Bonding 38-15
Iris Optics-31
Index-20

L Lenses, transmitted waves 36-3

L = 0 Patterns, hydrogen standing waves 38-4 Lensing, gravitational 34-20
Lack of simultaneity 1-32 Lepton family
Lambda max. See Blackbody radiation: Wein's Electron int-22
displacement law Electron type neutrino int-22
Lambda(1520), short lived elementary particle 40- Muon int-22
23 Muon type neutrino int-22
Tau int-22
Landé g factor 39-14
Tau type neutrino int-22
Largest scale of distance int-25
Leptons, conservation of int-22
Laser
LET statement, computer 5-5
Chapter on 37-1
Lifetime
Diffraction patterns, Fourier analysis 16-33
Pulse in gas of supercooled sodium atoms Optics-9 Muon, exponential decay Cal 1-32
Standing light waves 37-2 Lifting weights and muscle injuries 13-11
LC circuit Light int-7
Experiment on 31-13 Atomic spectra 33-16
Fourier analysis of 31-31 Balmer series 33-19
Introduction to 31-10 Blackbody radiation 34-2
Another View of 34-22
Ringing like a bell 31-36
Chapter on 33-1
LC oscillation, intuitive picture of 31-12
Decoupling from matter in early universe 34-31
Left hand rule, as mirror image Optics-6
Diffraction of light
Leibnitz 4-1, Cal 1-3 By thin crystals 36-6
Lens Fourier analysis of slits 16-33
Crystalline, eye Optics-31 Grating for 33-12
Diverging Optics-26 Pattern, by strand of hair 36-14
Eye glasses experiment Optics-36 Patterns, student projects 36-13
Eyepiece Optics-37 Doppler effect
Introduction to theory Optics-18 In astronomy 33-23
Lens equation Introduction to 33-20
Derivation Optics-25 Relativistic 33-22
Introduction Optics-24 Electromagnetic spectrum int-7, 32- 20
Multiple lens systems Optics-28 Photon energies 34-11
Negative focal length, diverging lens Optics-26 Electromagnetic waves 32- 18
Negative image distance Optics-26 Gravitational lensing of 34-20
Negative object distance Optics-27 Hydrogen spectrum. See Experiments II: -10- Diffrac-
The lens equation itself Optics-25 tion grating and hydrogen spectrum
Two lenses together Optics-29 Balmer formula 35-5
Magnification of lenses Optics-30 Bohr model. See Bohr Model
Magnifier Optics-38 Lab experiment 33-17
Jeweler using Optics-38 Spectrum of star 33-19
Magnification of Optics-39 Infra red. See Infrared light
Magnifying glass Optics-37 Interaction with gravity 34-18
Multiple lens systems Optics-28 Interference patterns for various slits 33-12
Optical properties due to slowing of light Optics-9 Laser pulse in gas of supercooled sodium
Simple microscope Optics-50 atoms Optics-9
Spherical surface Lasers, chapter on 37-1
Grinding one Optics-19 Light pulse clock 1-14
Optical properties of Optics-19 Maxwell’s theory of int-7
Thin lens Optics-23 Microwaves. See Microwaves
Two lenses together Optics-29 Mirror images Optics-6
Zoom lens Optics-18 Motion through a Medium Optics-8
Lens, various types of Particle nature of 34-1
Bi-concave Optics-27 Photoelectric effect 34-5
Bi-convex Optics-27
Meniscus-concave Optics-27
Meniscus-convex Optics-27
Planar-concave Optics-27
Planar-convex Optics-27
 Index-21

Photon Lithium
Chapter on 34-1 And the Pauli exclusion principle 38-9
Creates antiparticle 34-17 Atom 38-12
Energies in short laser pulse 40-21 Binding energy of last electron 38-12
In electron spin resonance experiment 39-9 In the periodic table 38-11
Introduction to int-8 Nucleus int-6
Mass 34-12 Logarithms
Momentum 34-13
Integral of 1/x Cal 1-15
Thermal, 2.74 degrees int-29, 34-27, 34-31
Introduction to Cal 1-15
Polarization of light 32- 23
Inverse of exponential function Cal 1-16
Polarizers 32- 25
Lorentz contraction 1-24
Prism, analogy to Fourier analysis 16-28
Radiated electric fields 32- 28 Charge density created by 28- 6
Radiation pressure of 34-14 Lorentz force law
Radiation pressure, red supergiant stars 20-15 And Maxwell's equations 32- 8
Radio waves. See Radio waves Electric and magnetic forces 28- 15
Rays Optics-1 Relativity experiment 30-9
Red shift and the expanding universe 33-24, 34-21 Lorenz, chaos 23-1
Reflection 36-3, Optics-1 LR circuit 31-8
Reflection and fluorescence 40-8 Exponential decay time constant 31-9
Reflection from curved surfaces Optics-4 LRC circuit, ringing like a bell 31-36
Spectral lines, hydrogen int-7 Lyman series, energy level diagram 35-6
Bohr theory 35-4
Speed in a medium Optics-8 M
Speed of light Magnetic bottle 28- 31
Electromagnetic pulse 32- 14 Magnetic constant (µ zero), definition of 28- 11
Experiment to measure 1-9, 31-15 Magnetic energy
Same to all observers 1-12 Of current loop 31-22
Structure of electromagnetic wave 32- 19 Of spin 39-1
Thermal, 2.74 degrees 34-27 Of spin, semi classical formula 39-14
Three degree radiation 34-27 Magnetic field 28- 10
Two-slit interference pattern for 33-10 Between capacitor plates 32- 6
Ultraviolet. See Ultraviolet light Detector 32- 26
Visible. See Visible light Dimensions of, tesla and gauss 28- 16
Visible spectrum of 33-15 Direction of
Wave equation for int-7 Compass needles 28- 12
Waves, chapter on 33-1 Definition 28- 11
X ray diffraction 36-4 Iron fillings 28- 12
X rays. See X-rays Gauss' law for 32- 2
Light-hour int-2 Helmholtz coils 28- 17, 28- 18
Light-minute int-2 100 turn search coil 39-12
Light-second int-2 Electron spin resonance apparatus 39-11
Light-year int-2 Field mapping experiment 30-24
Motion of electrons in 28- 20, 28- 29
Limiting process 4-1, Cal 1-3
In electromagnetic waves 32- 18
Definition of derivative Cal 1-30
In light wave int-7
In calculus Cal 1-5
Interaction with spin 39-4
Introduction to derivative Cal 1-6
Mapping. See Experiments II: - 7- Magnetic field
With strobe photographs Cal 1-2
mapping using Faraday's law
Line charge, electric field of
Mapping experiment with Helmholtz coils 30-24
Calculated using calculus 24-6
Motion of charged particles in 28- 19
Calculated using Gauss' law 24-21
Motion of electrons in. See Experiments II: - 5- Motion
Line integral. See Integral, line
of electrons in a magnetic field
Linear and nonlinear wave motion 15-10 Oersted, Hans Christian 28- 12
Linear momentum. See Momentum Of a solenoid 28- 17, 29-14
Lines, equipotential 25-3 Of a straight wire 29-11
Of a toroid 29-17
Of permanent magnet, experiment to measure 30-25
Radiated, a thought experiment 32- 11
Index-22

Magnetic field Continued Magnets

Right-hand rule for current 28- 13 Electromagnet 31-28
Right-hand rule for solenoids 29-14 Iron 31-26
Surface integral of 32- 2 Superconducting 31-30
Uniform 28- 16 Magnification
Between Helmholtz coils 28- 17 Definition Optics-30
Between pole pieces 28- 16 Negative (inverted image) Optics-30
Inside coils 28- 17 Of Magnifier Optics-39
Magnetic flux Of two lenses, equation for Optics-30
AC voltage generator 30-21 Magnifier Optics-38
Definition of 30-11 Jeweler using Optics-38
Faraday's law Magnification of Optics-39
Line integral form 30-15 Magnifying glass Optics-30, Optics-37
Voltage form 30-12 Magnitude of a Vector 2-6
Field mapping experiment 30-24
Map, contour 25-1
Gaussmeter 30-23
Mapping convention for electric fields 24-17
In the betatron 30-16
Mars, retrograde motion of 8-24
Integral E.dl meter 30-19
Magnetic field detector 32- 26 Mass
Maxwell's equations 32- 8 Addition of 6-4
Velocity detector 30-25 Angular
Moment of inertia 12-7
Voltage transformer 30-26
Center of mass
Magnetic focusing 28- 29. See also Experiments II:
Diver movie 11-1
- 5a- Magnetic focusing and space physics Dynamics of 11-4
Movie 28- 30 Formula 11-3
Magnetic force Introduction to 11-2
Between currents 31-19 Chapter on mass 6-1
Deflection of electron beam 28- 9 Definition of mass
Movie 28- 9 Newton's second law 8-3
On a current 31-18 Recoil experiments 6-2
On electrons in a wire 30-3 Electron mass in relativistic beta decay 6-7
Origin of 28- 8 Energy
Parallel currents attract 28- 14 In nuclei int-18
Relativity experiment (Faraday's law) 30-9 Introduction to 10-3
Thought experiment (on origin of) 28- 7 Of neutron int-22
Magnetic force law Gravitational force on int-6
Lorentz force law, electric and magnetic forces 28- Gravitational mass 6-5
15 Inertial mass 6-5
Magnetic force law, derivation of Measuring mass 6-4
F = qvB 28- 10 Of a moving object 6-5
Vector form 28- 14 Of a neutrino 6-13
Magnetic moment Of a photon 34-12
And angular momentum 31-24 Properties of 6-3
Bohr magneton 39-4 Relativistic formula for 6-10
Definition of 31-21 Relativistic mass
Nuclear 39-6 Beta decay 6-6
Of charge in circular orbit 31-24 Beta decay of Plutonium 246 6-8
Of electron 39-4 Beta decay of Protactinium 236 6-9
Of neutron 39-6 Intuitive discussion 6-6
Of proton 39-6 Rest mass int-11, 6-10, 10-5
Summary of equations 31-24 Role in mechanics 8-3
Magnetic resonance Standard mass 6-3
Classical picture of 39-8, 39-14 Zero rest mass 6-11
Precession of atom 39-15 Mass on a spring, analytic solution 14-7
Electron spin resonance experiment 39-5 Mass, oscillating, differential equation for 14-8
X. See Experiments II: -12- Electron spin resonance Mass spectrometer 28- 28
Magnetism Mathematical prism, Fourier analysis 16-28
Chapter on 28- 1 Matter over antimatter, in early universe int-30, 34-
Thought experiment to introduce 28- 4 17
Matter, stability of 19-14
 Index-23

Mauna Kea, Hawaii Optics-45 MOD command, computer 5-10

Maxima, first, of two-slit pattern 33-8 Model Atom 37-4
Maximally efficient engines 18-15 Chapter on 37-1
Maxwell’s correction to Ampere’s law 32- 4 Energy levels in 37-4
Maxwell's equations 32- 8 Model showing equipotential and field lines 25-10
Chapter on 32- 1 Modulus
Failure of Bulk 15-8
In classical hydrogen atom 35-2 Definition of 15-8
In photoelectric effect 34-6 Young’s 15-8
In empty space 32- 10 Molar heat capacity 17-26
Probability wave for photons 40-7 Molar specific heat of helium gas 17-27
Symmetry of 32- 9 Mole
Maxwell’s theory of light int-7, 1-9, 32- 2 Avogadro's number 17-24
Maxwell's wave equation int-7, 15-1, 32- 18 Volume of 17-25
Measurement limitation Molecular forces int-6, 14-20, 17-12
Due to photon momentum 40-11 A classical analysis 19-18
Due to uncertainty principle 4-2, Cal 1-4 The bonding region 19-19
Two slit thought experiment 40-9 A more quantitative look 19-18
Using waves 40-10 Binding energy 17-13
Measuring short times using uncertainty prin- Electric interaction 19-1
ciple 40-22 Four basic interactions 19-2
Measuring time constant from graph Cal 1-34 Introduction to 19-15
Mechanics Represented by springs 17-13
Newtonian Molecular weight 17-24
Chapter on 8-1 Molecules 17-2
Classical H atom 35-3 Cholera 17-2
Newton's second law 8-4 Hydrogen, electric forces in 19-16
Newton’s Third Law 11-6 Hydrogen molecule ion 19-16
Photon mechanics 34-12 Myoglobin 17-3
Relativistic int-12 Water 17-2
The role of mass in 8-3 Moment, magnetic
Medical imaging Optics-15 And angular momentum 31-24
Megabit, fiber optics communication Optics-14 Definition of 31-21
Meiners, Harry, electron scattering apparatus 36-1 Of charge in circular orbit 31-24
Meniscus-concave lens Optics-27 Summary of equations 31-24
Meniscus-convex lens Optics-27 Moment of inertia
Mercury barometer, pressure measurement 17-22 Angular mass 12-7
Meter, definition of int-2 Calculating 12-8
Microscope int-1, Optics-50 Rotational kinetic energy 12-22
Atomic 17-1 Momentum
Scanning tunneling microscope Optics-51 Angular. See Angular momentum
Surface (111 plane) of a silicon Optics-51 Collisions and impulse 11-9
Simple microscope Optics-50 Conservation of 7- 2
Microwaves Derivation from Newton's second law 11-7
Electromagnetic spectrum int-7, 32- 20 During collisions 11-13
General discussion 7- 1
Microwave polarizer 32- 24
In collision experiments 7- 4
Photon energies 34-11
In subatomic collisions 7- 7
Milky Way int-2 X. See also Experiments I: - 3- Conservation of linear
Center of our galaxy Optics-47 momentum
Millikan oil drop experiment 26-17 De Broglie formula for momentum 35-11
Mirror images Kinetic energy in terms of momentum 37-5
General discussion Optics-6 Linear momentum, chapter on 7- 1
Reversing front to back Optics-7 Momentum of photon
Right-hand rule Optics-6 Compton scattering. 34-15
Mirror, parabolic Formula for 34-13
Focusing properties of Optics-4, Optics-42 Momentum version of Newton’s second law 11-8
MKS units Uncertainty principle, position-momentum form 40-
Ampere, volt, watt 24-2 15
Coulomb's law in 24-2
Index-24

Mormon Tabernacle, ellipse Optics-3 Muon

Mormon Tabernacle, focusing of sound waves 8- And Mt. Washington, Lorentz contraction 1-29
26, Optics-3 Discovery of int-23
Motion Half life used as clock 1-20
Angular analogy 12-3 Lepton family int-22
Damped harmonic motion 14-21 Lifetime, exponential decay Cal 1-32
Differential equation for 14-21 Movie on lifetime 1-21
Description of, chapter on 3-3 Cerenkov radiation Optics-10
Forced harmonic motion Muon type neutrino int-21
Differential equation for 14-25, 14-28 Muscle, ciliary, in the eye Optics-31
Harmonic motion 14-12 Muscle injuries lifting weights 13-11
Computer program for 14-30 Myoglobin molecule 17-3
Of charged particles Myopia, nearsightedness Optics-32
In magnetic fields 28- 19
In radiation belts 28- 32 N
Of electrons in a magnetic field. See Experiments II: - Nature’s speed limit int-12, 6-11
5- Motion of electrons in a magnetic field Nearsightedness Optics-32
Of light through a medium Optics-8 Nebula
Oscillatory motion 14-2. See also Experiments I: - Crab, neutron star 20-16
10- Oscillatory motion of various kinds Eagle 7- 16, 7- 18
Prediction of motion 5-12 Orion 7- 17
Projectile. See Projectile Motion Negative and positive charge 19-10
Resonance 14-24 Negative focal length, diverging lenses Optics-26
Tacoma Narrows bridge 14-24
Negative image distance Optics-26
Rotational motion 12-1
Angular acceleration 12-3 Negative object distance Optics-27
Angular velocity 12-2 Negative slope Cal 1-31
Radian measure 12-2 Neon bulb oscillator 27- 28
Satellite. See Satellite motion Experimental setup 27- 31
Thermal motion 17-6 X. See Experiments II: - 4- The Neon Bulb Oscillator
Translation and rotation 12-24 Neon, up to, periodic table 38-13
Uniform circular motion 3-17, 8-2 Nerve fibers, human eye Optics-31
Particles in magnetic field 28- 20 Cones Optics-31
Wave motion, amplitude and phase 15-17 Rods Optics-31
Movie Net area (Fourier analysis) 16-29
Angular momentum as a Vector 7- 15 Neutrino astronomy 6-14, 11-21
Brownian motion 17-7 Neutrinos 6-13, 11-20
Circular motion of particles in magnetic field 28- 20 1987 Supernova 6-14
Diver 11-1 Beta (ß) decay reactions 20-6
Magnetic deflection 28- 9 Cosmic background int-30
Magnetic focusing 28- 30 Created in the weak interaction 20-6
Muon Lifetime 1-21 Electron type int-21
Standing De Broglie like waves 35-11 From the sun 6-13, 11-21
Time reversed dive, second law of thermodynam- In nuclear structure 20-7
ics 18-1 In stellar evolution 20-15
Mt. Hopkins telescope Optics-43 In supernova explosions 20-16
Mt. Palomar telescope Optics-43 Muon type int-21
Mu (µ) zero, definition of 31-19 Passing through matter int-22
Multi Electron Atoms 38-9 Pauli's prediction of int-21, 20-6
Multiple lens systems Optics-28 Rest mass 20-6
Multiple slit grating 16-33 Stability of 19-14
Multiple slit interference patterns 33-12 Tau type int-21
Multiplication notation, computer 5-6
Multiplication of vectors
By a number 2-5
Scalar or dot product 2-12
Vector cross product 2-15, 12-9
 Index-25

Neutron Nuclear
Decay Binding energy 20-9
Beta decay reactions 20-7 Charge, effective, in periodic table 38-12
Charge conservation in int-21 Energy well, binding energies 20-10
Energy problems int-21 Fission int-18, 20-3
Weak interaction 20-6 Energy from 20-4
Formation of deuterium in early universe 34-30 Force int-18, 20-2
In alpha particles 20-8 Four basic interactions 19-2
In isotopes int-17, 19-6 Meson, Yukawa theory int-22
In nuclear matter 20-1 Range of 20-3
Neutron proton balance in early universe 34-30 Range vs. electric force int-18
Neutron proton mass difference in early universe 34- Fusion int-18, 20-12
30 Binding energies, stellar evolution 20-13
Nuclear binding energies 20-9 Nuclear interaction int-14
Quark structure of int-24 Alpha particles 20-8
Rest mass energy int-22, 20-9 Binding energy 20-9
Role in nuclear structure 20-7 Neutron stars 20-17
Neutron star Nuclear magnetic moment 39-6
And black holes int-20, 20-18 Nuclear matter, chapter on 20-1
Binary int-15 Nuclear reactions, element creation int-4
In Crab nebula 20-17 Nuclear stability, binding energy 20-10
Pulsars 20-17 Nuclear structure int-22, 20-7
Stellar evolution int-19 Nucleon int-17
New functions, obtained from integration Cal 1-15 Nucleus int-17
Newton Discovery of, Rutherford 35-1
Particle nature of light 34-1, Optics-1 Hydrogen int-6
Newtonian mechanics 8-1 Isotopes of 19-6
Classical H atom 35-3 Large int-22
Failure of Lithium int-6
In specific heats 17-31 Most tightly bound (Iron 56) 20-11
In the classical hydrogen atom 35-3
The role of mass 8-3
O
Newton’s laws Odd harmonics in a square wave 16-28
Chapter on 8-1 Odor of violets 17-6
Classical physics int-7 Oersted, Hans Christian 28- 12
Gravity int-15, 8-5 Off axis rays, parabolic mirror Optics-4
Second law 8-4 Ohm’s law 27- 7
And Newton’s law of gravity 8-5 One cycle of a square wave 16-28
Angular analogy 12-14 One dimensional wave motion 15-1
Applications of, chapter on 9-1 Optical properties
Atwood’s machine 9-16 Parabolic reflectors Optics-4
Inclined plane 9-10 Spherical surface Optics-19
Momentum version of 11-8 Optics, fiber
Satellite motion 8-8
Introduction to Optics-14
String forces 9-15
Medical imaging Optics-15
Vector addition of forces 9-6
Optics, geometrical
Third law 11-6
Chapter on Optics-1
Newton’s reflecting telescope Optics-22, Optics-42
Definition of Optics-2
NMR experiment, the hydrogen-deuterium mol-
Optics of a simple microscope Optics-50
ecule 39-12
Orbitals 19-15. See also Hydrogen atom: Standing
Nonlinear restoring forces 14-19
wave patterns in
Nonrelativistic wave equation int-12, 15-2, 34-16
Orbits
Normal modes
Allowed, Bohr theory int-8
Degrees of freedom 17-29
Bohr, radii of 35-7
Fourier analysis of coupled air cart system 16-12
Classical hydrogen atom 35-2
Modes of oscillation 16-4
Orbit-1 program 8-21
X. See Experiments I: -11- Normal modes of oscilla-
Precession of, general relativity 8-30
tion
Satellite. See Satellite motion
Index-26

Order and disorder Particle accelerators 28- 22

Direction of time 18-25 Particle decays and four basic interactions int-23
Entropy and the second law of thermodynamics 18- Particle nature of light 34-1
4 Photoelectric effect 34-5
Formula for entropy 18-24 Particle-wave nature
Orion nebula 7- 17 Born’s interpretation 40-6
Oscillation 14-1 De Broglie picture int-10, 35-10
Critical damping 14-23 Energy level diagrams resulting from 37-4
Damped Of electromagnetic spectrum 34-11
Computer approach 14-21 Of electrons
Differential equation for 14-21 Davisson and Germer experiment 35-12
LC circuit 31-10 De Broglie picture int-10, 35-10
Intuitive picture of 31-12 Electron diffraction experiment 36-8
Mass on a spring Electron waves in hydrogen 38-2
Analytic solution 14-7 Pauli exclusion principle 38-9
Computer solution 14-30 Of forces int-13
Differential equation for 14-8 Of light
Non linear restoring forces 14-19 Electromagnetic spectrum 34-11
Normal modes 16-4. See also Experiments I: -11- Photoelectric effect 34-5
Normal modes of oscillation Photon mass 34-12
Period of 14-4 Photon momentum 34-13
Phase of 14-6 Photon waves 40-6
Resonance Photons, chapter on 34-1
Analytic solution for 14-26, 14-28 Of matter int-10, 34-11
Differential equation for 14-25 Probability interpretation of 40-6
Introduction to 14-24 Fourier harmonics in a laser pulse 40-22
Small oscillation Reflection and fluorescence 40-8
Molecular forces 14-20 Quantum mechanics, chapter on 40-1
Simple pendulum 14-16 Two slit experiment from a particle point of view
Torsion pendulum 14-12 Probability interpretation 40-9
Transients 14-27 The experiment 40-3
Oscillator Uncertainty principle 40-14
Harmonic oscillator 14-12 Energy conservation 40-24
Differential equation for 14-14 Position-momentum form of 40-15
Forced 14-28 Quantum fluctuations 40-25
Neon bulb oscillator 27- 28 Time-energy form of 40-19
Period of oscillation 27- 30 Paschen series
X. See also Experiments II: - 4- The neon bulb Energy level diagram 35-6
oscillator Hydrogen spectra 35-6
Oscillatory motion 14-2. See also Experiments I: - Pauli exclusion principle 38-1, 38-9
10- Oscillatory motion of various kinds Pauli, W., neutrinos int-21, 20-6
Osmotic pressure 17-34 Peebles, radiation from early universe 34-27
Overview of physics int-1 Pendulum
Conical 9-18
P Energy conservation 10-10
Parabola Simple and conical 14-17
How to make one Optics-4 Simple pendulum 14-15
Parabolic mirror Spring pendulum 9-4
f number Optics-5 Ball spring program 9-20
Focusing properties of Optics-4, Optics-42 Computer analysis of 9-8
Off axis rays Optics-4 F = ma 9-7
Parallel currents attract 28- 14 X. See Experiments I: - 6- Spring pendulum
Parallel plate capacitor 26-14 Torsion pendulum 14-12
Differential equation for 14-14
Capacitance of 27- 16
Penzias and Wilson, cosmic background
voltage in 26-15
Parallel resistors 27- 12 radiation int-29, 34-27
Particle Percussion instruments 16-22
Point size int-14 Period of oscillation 14-4
Systems of particles 11-1 Neon bulb circuit 27- 30
 Index-27

Period, wavelength, and frequency 15-13 Planck's constant

Periodic table 38-1, 38-10 And blackbody radiation 34-4, 34-22
Beryllium 38-13 Angular momentum, Bohr model 35-8
Boron 38-13 Bohr theory 35-1
Effective nuclear charge 38-12 In Bohr magneton formula 39-4
Electron binding energies 38-11 In de Broglie wavelength formula 35-11
Electron screening 38-10 In photon mass formula 34-12
Lithium 38-12 In photon momentum formula 34-13
Potassium to krypton 38-14 In the photoelectric effect 34-7
Sodium to argon 38-13 In the uncertainty principle 40-15, 40-19
Summary 38-14 Introduction to 34-8
Phase and amplitude Spin angular momentum 39-3
Fourier analysis lecture 16-31 Plane, inclined 9-10
Wave motion 15-17 Planetary units 8-14
Phase of an oscillation 14-6 Planets
Phase transition, electroweak theory int-26 Formation of 7- 17
Phases of Fourier coefficients 16-32 Plates, electron gun deflection 26-16
Photoelectric effect int-8 Plotting
Einstein's formula 34-7 A point by computer 5-6
Introduction to 34-5 Experiment, electric potential 25-7
Maxwell's theory, failure of 34-6 Potentials and fields. See Experiments II: - 1- Poten-
Planck's constant 34-8 tial plotting
Photon int-8 Window, computer 5-7
Blackbody radiation 34-22 Plutonium 246 6-8
Chapter on 34-1 Point
Creates antiparticle 34-17 Mass, gravitational field of 24-23
Electric interaction int-23 Particle int-14
Energies in short laser pulse 40-21 Source, velocity field of 23-6
Energy 34-9 Polarization of light waves 32- 23
Energy levels in laser 37-4 Polarizer
Uncertainty principle, Fourier transform 40-20 Light 32- 25
Gravitational deflection of 34-19 Microwave 32- 24
Hydrogen spectrum 35-5 Polaroid, light polarizer 32- 25
In electron spin resonance experiment 39-9 Position measurement, uncertainty principle 40-15
Interaction with gravity 34-18 Positive and negative
Mass 34-12 Charge 19-10
Mechanics 34-12 Electric current 27- 3
Momentum 34-13 Positive area in Fourier analysis 16-29
Compton scattering 34-15
Positron (antimatter) int-13, 34-17
Measurement limitation 40-11
Probability wave 40-7 Positronium, annihilation into photons 34-17
Rest mass int-12 Potassium to krypton, periodic table 38-14
Stability of 19-14 Potential, electric
Standing waves 37-3 Contour map 25-1
Thermal, 2.74 degrees int-29, 34-27, 34-31 Of a point charge 25-5
Photon pulse
Photon energy in 40-21
Probability Interpretation of 40-22
Photon waves, probability interpretation 40-6
Physical constants
In CGS units Back cover-1
Pi mesons int-23
Piela, electron clouds and binding energy 19-23
Pinhole camera Optics-35
Planar-concave lens Optics-27
Planar-convex lens Optics-27
Planck, M., blackbody radiation law 34-4
Index-28

Potential energy Pressure

Conservative forces 25-5 Atmospheric 17-23
Electric potential energy Bernoulli's equation 23-9, 23-11
Contour map of 25-1 Airplane wing 23-13
In classical hydrogen atom 35-3 Care in applying 23-16
In hydrogen atom int-11 Sailboats 23-14
In hydrogen molecule ion 19-21 Superfluid helium 23-17
In molecules 17-12 Ideal gas law 17-18
In nuclear fission int-18, 20-5 In stellar evolution 17-17
Negative and positive 25-4 Measurement, using mercury barometer 17-22
Of a point charge 25-5 Osmotic pressure 17-34
Potential plotting 25-7 Pressure in fluids
Storage in capacitors 27- 18 Aspirator 23-16
Energy conservation int-11, 8-35, 10-20 Definition of 23-10
Conservative and non-conservative forces 10-21 Hydrodynamic voltage 23-17
Mass on spring 14-11 Hydrostatics 23-12
Uncertainty principle 40-24 Venturi meter 23-15
Equipartition of energy 17-28 Viscous effects 23-19
Gravitational potential energy int-11, 8-35 Pressure of a gas 17-16
Bernoulli's equation 23-10 Pressure of light
Black holes 10-29 Nichols and Hull experiment 34-14
In a room 10-25 Red giant stars 20-15, 34-15
In satellite motion 8-36, 10-26 PV diagrams 18-8
In stellar evolution 20-13 Adiabatic expansion 18-9, 18-26
Introduction to 10-8 Internal combustion engine 18-21
Modified 8-37 Isothermal expansion 18-8, 18-26
On a large scale 10-22 Reversible engines 18-13
Zero of 10-22 The Carnot cycle 18-11, 18-26, 18-28
In collisions 11-14 Work dome by pressure, Bernoulli's equation 23-10
Negative and positive 25-4 Primary coil 30-26
Nuclear potential energy int-18 Principle of relativity. See also Relativistic physics
Binding energies 20-9
A statement of 1-4
Fusion 20-12
And the speed of light 1-11
Spin potential energy
As a basic law of physics 1-4
Magnetic 39-1
Chapter on 1-1
Magnetic moment 39-4
Spring potential energy int-11, 10-16, 14-11 Einstein’s theory of 1-12
Work int-11, 10-15 Introduction to 1-2
Work energy theorem 10-18 Special theory of 1-13
A consistent theory 1-32
Potential plotting. See Experiments II: - 1- Potential
Causality 1-36
plotting Lack of simultaneity 1-32
Power Light pulse clock 1-14
1 horsepower = 746 watts 18-20 Lorentz contraction 1-24
Definition of watt 10-31 Mass energy 10-3
Efficiency of a power plant 18-18 Nature’s speed limit 6-11
In electric circuits 27- 9 Origin of magnetic forces 28- 8
Of the sun 34-23 Photon mass 34-12
Sound intensity 16-24 Photon momentum 34-13
Powers of 10, names of Front cover-2 Relativistic energy and momenta 28- 24
Practical system of units 10-31 Relativistic mass 6-6
Precession Relativity experiment leading to Faraday's law 30-9
Time dilation 1-16, 1-22
Of atom, magnetic interaction 39-15
Zero rest mass particles 6-11
Of orbit, modified gravity 8-30
Principle of superposition
Prediction of motion
For 1 dimensional waves 15-11
Using a computer 5-12
For 2 dimensional waves 33-2
Using calculus Cal 1-9
Preliminary discussion of 15-2, 33-1
 Index-29

Prism Proton int-17

Atmospheric, the green flash Optics-17 In alpha particles 20-8
Glass Optics-13 In atomic structure 19-4
Glass, rainbow of colors Optics-15 In hydrogen molecule 19-16
Prism, mathematical (Fourier analysis) 16-28 In nuclear structure 20-2
Probability interpretation In the weak interaction 20-6
And the uncertainty principle 40-14 Quark structure int-24
Of electron diffraction pattern 40-6 Rest energy in electron volts 26-12
Of particle waves 40-6 Stability of 19-14
Of photon pulse 40-22 Proton cycle, energy from sun 34-24
Of two slit experiment 40-2 Proton synchrotron at CERN 28- 24
Probability wave Proton-neutron mass difference, early universe 34-
For photons 40-7 30
Intensity of 40-22 Ptolemy, epicycle, in Greek astronomy 8-25
Reflection and Fluorescence 40-8 Pulleys
Probe for scanning tunneling microscope Optics-51 Atwood’s machine 9-16
Problem solving. See Solving problems Working with 9-16
Gauss' law problems 24-26 Pulsars, neutron stars 20-17
How to go about it 24-29 Pulse
Projectile motion problems 4-16 Formation from sine waves 40-27
Program, BASIC. See BASIC program Of electromagnetic radiation 32- 10
Program, English PV = NRT 17-25
For oscillatory motion 14-31 PV diagrams
For projectile motion 5-16 Adiabatic expansion 18-9
For satellite motion 8-19 Carnot cycle 18-11
Project suggestion on wave speed 15-8 Isothermal expansion 18-8
Projectile motion
Analysis of Q
Calculus 4-9 Quantization of electric charge 19-14
Computer 5-16 Quantized angular momentum int-9
Graphical 3-16 Angular momentum quantum number 38-7
And the uncertainty principle 4-2, Cal 1-4 Electron spin 38-9
BASIC program for 5-19 Chapter on 39-1
Calculus definition of velocity 4-3, Cal 1-5 Concept of spin 39-3
Computer program for 8-21 In Bohr theory 35-9
Constant acceleration formulas In de Broglie's hypothesis 35-10
Calculus derivation 4-9 In hydrogen wave patterns 38-3
Graphical analysis of 3-26 Quantized projections 38-5
Determining acceleration for 3-16 Quantized vortices in superfluids 23-22
English program for 5-16 Quantum electrodynamics int-14
Graph paper tear out pages 3-29 Quantum fluctuations in empty space 40-25
Gravitational force 8-2 Quantum mechanics. See also Particle-wave
Instantaneous velocity 3-24
nature; Schrödinger’s equation
Solving problems 4-16
Bohr theory of hydrogen 35-1
Strobe photograph of 3-7
Chapter on 40-1
Styrofoam projectile 5-28
Concept of velocity 4-2, Cal 1-4
With air resistance
Electron and nuclear spin 39-1
Calculus analysis 4-12
Model atom 37-4
Computer calculation 5-22
Graphical analysis 3-22 Schrodinger's equation applied to atoms 38-1
X. See Experiments I: - 2- Computer prediction of Uncertainty principle 40-14
projectile motion Zero point energy 37-7
Projections of angular momentum Quantum number, angular momentum 38-7
Classical 7- 14 Quantum theory of gravity int-16, int-20
Electron spin 39-3 Quark confinement 19-15
Quantized 38-5 Quarks int-24
Protactinium 236, recoil definition of mass 6-9 Quantization of electric charge 19-14
Quasars
Gravitational lens, Einstein cross 34-20
Size of 34-19
Index-30

R Reactive metal, lithium int-6, 38-9

R, gas constant 17-25 Recoil experiments, definition of mass 6-2
Radar waves Red shift and the expanding universe
Photon energies in 34-11 Doppler effect 34-21
Wavelength of 32- 20 Evolution of universe 34-21
Radial acceleration 12-5 Uniform expansion 33-24
Radian measure 12-2, Cal 1-35 Red supergiant star 20-15
Radiated electromagnetic pulse 32- 10 Reflecting telescope Optics-42
Radiated magnetic field thought experiment 32- 11 Cassegrain design Optics-42
Radiation Diffraction limit Optics-45
Blackbody 32- 22 Hubbel space telescope Optics-44
Photon picture of 34-22 Keck, world’s largest optical Optics-45
Theory of 34-2 Mt. Hopkins Optics-43
Wein's displacement law 34-2 Mt. Palomar Optics-43
Cerenkov Optics-10 Newton's Optics-22
Electromagnetic field Newton’s Optics-42
Analysis of path 1 32- 14 Secondary mirror Optics-42
Analysis of path 2 32- 16 William Hershel’s Optics-43
Calculation of speed 32- 14 Reflection
Spectrum of 32- 20 And fluorescence, probability interpretation 40-8
Radiated electric fields 32- 28 Bragg reflection 36-4
Radiated energy and the classical H atom 35-3 From curved surfaces Optics-3
Radiated field of point charge 32- 30 Internal Optics-13
Three degree cosmic radiation int-30, 34-27 Of light 36-3, Optics-1
UV, X Rays, and Gamma Rays 32- 22 Refracting telescopes Optics-40
Radiation belts, Van Allen 28- 32 Galileo's Optics-41
Radiation pressure Yerkes Optics-41
In red supergiant stars 20-15 Refraction, index of
Of light 34-14 Definition of Optics-9
Radio galaxy Optics-48 Glass prism and rainbow of colors Optics-15
Radio images of variable star Optics-49 Introduction to Optics-2
Radio telescope int-15 Of gas of supercooled sodium atoms Optics-9
Arecibo int-15 Table of some values Optics-9
Three degree radiation 34-27 Relativistic mass. See Relativistic physics: Relativis-
Radio telescopes Optics-48 tic mass
Arecibo Optics-48 Relativistic physics. See also Principle of Relativity
Radio galaxy image Optics-48 A consistent theory 1-32
Radio images of variable star Optics-49 Antimatter int-12, 34-16
Very Large Array Optics-48 Black holes 10-29
Very Long Baseline Array Optics-49 Blackbody radiation 34-22
Radio waves Causality 1-36
Hertz, Heinrich 34-1 Chapter on 1-1
In the electromagnetic spectrum int-7, 32- 20 Clock
Photon energies 34-11 Light pulse 1-14
Predicted from the classical hydrogen atom 35-2 Moving 1-13
Wavelength of 32- 20 Muon 1-20
Radius of electron 39-3 Muon lifetime movie 1-21
Rainbow Other kinds 1-18
Real ones 1-20
Glass prism Optics-15
Creation of positron-electron pair 34-17
Photograph of Optics-16
Definition of mass 6-2
Range of nuclear force 20-3
Doppler effect for light 33-22
RC circuit 27- 22
Einstein mass formula 6-10
Exponential decay 27- 23 Electric or magnetic field: depends on viewpoint 30-
Exponential rise 27- 26 10
Half-lives 27- 25 Electric or magnetic force: depends on view-
Initial slope 27- 25 point 28- 10
Measuring time constant 27- 25 Electromagnetic radiation, structure of 32- 19
X. See Experiments II: - 3- The RC Circuit
 Index-31

Relativistic physics Continued Renormalization int-14, int-17

Electron mass in beta decay 6-7 Repeated wave forms in Fourier analysis 16-11
Gravitational lensing 34-20 Resistors
Interaction of photons and gravity 34-18 In parallel 27- 12
Kinetic energy 10-5 In series 27- 11
Slowly moving particles 10-6 Introduction to 27- 6
Lack of simultaneity 1-32 LR circuit 31-8
Longer seconds 1-16 Ohm's law 27- 7
Lorentz contraction 1-24 Resonance
Origin of magnetic forces 28- 8 Electron spin
Thought experiment on currents 28- 4 Classical picture of 39-14
Lorentz force law 28- 15 Experiment 39-9
Mass energy 10-3 Introduction to 39-5
Mass-energy relationship int-11 X. See Experiments II: -12- Electron spin resonance
Maxwell’s equations 32- 8 Introduction to 14-24
Motion of charged particles in magnetic fields 28- Phenomena 14-26
19 Tacoma Narrows bridge 14-24
Muon lifetime movie 1-21 Vortex street 14-25
Nature's speed limit int-2, int-12, 6-11 Transients 14-27
Neutrino astronomy 6-14 Rest energy of proton and electron in eV 26-12
Neutrinos 6-13 Rest mass int-11
Origin of magnetic forces 28- 8 And kinetic energy 10-5
Particle accelerators 28- 22 Einstein formula 6-10
Photon mass 34-12 Restoring forces
Photon rest mass 6-12 Linear 14-7
Photon momentum 34-13 Non linear 14-19
Compton scattering 34-15 Retrograde motion of Mars 8-24
Principle of relativity 1-2 Reversible engines
As a basic law 1-4
As thought experiment 18-13
Radiated electric fields 32- 28
Carnot cycle 18-17
Red shift and the expansion of the universe 34-21
Efficiency of 18-18
Relativistic calculations 1-28
Rifle and Bullet, recoil 7- 7
Approximation formulas 1-30
Muons and Mt. Washington 1-29 Right handed coordinate system 2-18
Slow speeds 1-29 Right-hand rule
Relativistic energy and momenta 28- 24 For cross products 12-10
Relativistic mass 6-6 For Faraday's law 30-15
Formula for 6-10 For magnetic field of a current 28- 13
In beta decay 6-6 For magnetic field of a solenoid 29-14
Relativistic mechanics int-12 For surfaces 29-16
Relativistic speed limit 6-11 Mirror images of Optics-6
Relativistic wave equation int-12 Rods, nerve fibers in eye Optics-31
Relativity experiment for Faraday's law 30-9 Rope, working with 13-10
Short lived elementary particles 40-23 Rotational motion
Space travel 1-22 Angular acceleration 12-3
Special theory of relativity 1-13 Angular analogy 12-3
Speed of light, measurement of 1-9 Angular velocity 12-2
Speed of light wave 32- 17 Bicycle wheel as a collection of masses 12-5
Spiraling electron in bubble chamber 28- 27 Chapter on 12-1
The betatron 30-16 Radian measure 12-2
The early universe 34-29 Rolling down inclined plane 12-25
Thought experiment on expanding magnetic Rotational kinetic energy 12-22
field 32- 11 Proof of theorem 12-26
Time dilation 1-22 Translation and rotation 12-24
Time-energy form of the uncertainty principle 40- Rubber, elasticity of 17-35
19, 40-23 Rutherford and the nucleus 35-1
Zero rest mass int-12 Rydberg constant, in Bohr theory 35-9
Zero rest mass particles 6-11
Relativity, general int-15, 8-29
Index-32

S Second law of thermodynamics. See also Carnot

Sailboats, Bernoulli's equation 23-14 cycle; Thermal energy
Salt Applications of 18-17
Dissolving 17-6 Chapter on 18-1
Ionic bonding 38-15 Statement of 18-4
Satellite motion 8-8 Time reversed movie 18-1
Calculational loop for 8-19 Second, unit of time, definition of int-2
Classical hydrogen atom 35-2 Secondary mirror, in telescope Optics-42
Compare with projectile, Newton's sketch 8-10 Semi major axis, Kepler's laws 8-28
Computer lab 8-23 Series expansions Cal 1-23
Computer prediction of 8-16 Binomial Cal 1-23
Conservation of angular momentum 8-32 Exponential function e to the x Cal 1-28
Conservation of energy 8-35 Series, harmonic 16-3
Earth tides 8-12 Series wiring
Gravitational potential energy 10-22 Capacitors 27- 21
Kepler's laws 8-24 Resistors 27- 11
Kepler's first law 8-26 Set Window, BASIC computer command 5-7
Kepler's second law 8-27 Short circuit 27- 9
Kepler's third law 8-28 Short rod, electric force exerted by 24-9
Modified gravity 8-29 Silicon, surface (111 plane) of Optics-51
Moon 8-8 Simple electric circuit 27- 8
Orbit, circular 10-27 Simple pendulum
Orbit, elliptical 10-27 Simple and conical pendulums 14-17
Orbit, hyperbolic 10-27 Theory of 14-15
Orbit, parabolic 10-27 Simultaneity, lack of 1-32
Planetary units for 8-14 Sine function
Program for (Orbit 1) 8-21 Amplitude of Cal 1-37
Total energy 10-26 Definition of Cal 1-35, Cal 1-36
X. See Experiments I: - 5- Computer analysis of Derivative of, derivation Cal 1-38
satellite motion
Sine waves
Scalar dot product 2-12
AC voltage generator 30-22
Definition of work 10-13 Amplitude of 14-10
Scanning tunneling microscope Optics-51 As solution of differential equation 14-9
Surface (111 plane) of a silicon Optics-51 Definition of 14-3
Scattering of waves Derivative of 14-8
By graphite crystal, electron waves 36-8 Formation of pulse from 40-27
By myoglobin molecule 36-5 Fourier analysis lecture 16-28
By small object 36-2 Fourier analysis of 16-7
By thin crystals 36-6 Harmonic series 16-3
Chapter on 36-1 Normal modes 16-4
Davisson-Germer experiment 35-12 Phase of 14-6
Reflection of light 36-3 Pulse in air 15-3
Two slit thought experiment 40-10 Sinusoidal waves motion 15-12
X ray diffraction 36-4 Standing waves on a guitar string 15-20
Schmidt, Maarten, quasars 34-19 Traveling wave 15-16
Schrödinger wave equation Traveling waves add to standing wave 15-20
Felix Block story on 37-1 Single slit diffraction 33-26
Introduction to wave motion 15-1 Analysis of pattern 33-27
Particle-wave nature of matter int-10 Application to uncertainty principle 40-16
Solution for hydrogen atom 38-2 Huygens principle 33-4
Standing waves in fuzzy walled box 38-1 Recording patterns 33-28
Schrödinger, Erwin int-10, 37-1 Slit pattern, Fourier transform of 16-33
Schwinger, J., quantum electrodynamics int-14 Slope of a curve
Search coil As derivative Cal 1-30
For magnetic field mapping experiment 30-24 Formula for Cal 1-30
Inside Helmholtz coils 39-12 Negative slope Cal 1-31
Second law, Newton's. See Newton’s laws: Second
law
 Index-33

Small angle approximation Specific heat

Simple pendulum 14-16 Cp and Cv 18-6
Snell’s law Optics-19 Definition of 17-26
Small oscillations Failure of Newtonian mechanics 17-31
For non linear restoring forces 14-19 Gamma = Cp/Cv 18-7
Molecular forces act like springs 14-20 Spectral lines
Of simple pendulum 14-16 Atomic spectra 33-16
Smallest scale of distance, physics at int-25 Bohr's explanation of int-9
Snell’s law Hydrogen
Applied to spherical surfaces Optics-19 Bohr theory of 35-4
Derivation of Optics-12 Colors of int-7
For small angles Optics-19 Experiment to measure 33-17
Introduction to Optics-2, Optics-11 The Balmer Series 33-19
Sodium to argon, periodic table 38-13 Introduction to int-7
Solar neutrinos 6-13. See also Neutrinos Spectrometer, mass 28- 28
Solenoid Spectrum
Ampere's law applied to 29-15 Electromagnetic
Photon energies 34-11
Magnetic field of 28- 17, 29-14
Visible spectrum 33-15
Right hand rule for 29-14
Wavelengths of 32- 20
Toroidal, magnetic field of 29-17
Hydrogen
Solving problems Balmer series 33-19
Gauss' law problems 24-26 Bohr theory of 35-4
How to go about it 24-29 Experiment to measure 33-17
Projectile motion problems 4-16 Lyman series, ultraviolet 35-6
Sombrero galaxy int-2 Paschen series, infrared 35-6
Sound Hydrogen star 33-19, 35-4
Focusing by an ellipse 8-26, Optics-3 Spectrum.
Fourier analysis, and normal modes 16-1 Electromagnetic
Fourier analysis of violin notes 16-18 Photon energies 34-11
Intensity Speed and mass increase int-11
Bells and decibels 16-24 Speed detector, air cart 30-5. See also Experi-
Definition of 16-24 ments II: - 6- Faraday's law air cart speed
Speaker curves 16-27 detector
Percussion instruments 16-22
Speed limit, nature’s int-12, 6-11
Sound meters 16-26
Speed of an electromagnetic pulse 32- 14
Sound produced by guitar string 15-22
Speed of light
Stringed instruments 16-18
The human ear 16-16 Absolute speed limit int-2, int-12
Wind instruments 16-20 Calculation of speed of light
Analysis of path 1 32- 14
X. See Experiments I: -12- Fourier analysis of sound
Analysis of path 2 32- 16
waves
Using Maxwell's Equations 32- 14
Sound waves, speed of Dimensional analysis 15-9
Calculation of 15-8 Experiment to measure 1-9, 31-15
Formula for 15-9 In a medium Optics-8
In various materials 15-9 Same to all observers 1-12
Space Speed of sound
And time int-2, 1-1 Formula for and values of 15-9
Quantum fluctuations in 40-25 Theory of 15-8
The Lorentz contraction 1-24 Speed of wave pulses
Travel 1-22 Dimensional analysis 15-6
Space physics 28- 31. See also Experiments II: - On rope 15-4
5a- Magnetic focusing and space physics Sphere
Space telescope Area of 23-6
Hubble Optics-44 Electric field inside of 26-4
Infrared (IRAS) Optics-47 Spherical aberration Optics-21
Spacial frequency k 15-14 In Hubble telescope mirror Optics-22
Speaker curves 16-27 Spherical lens surface Optics-19
Formula for focal length Optics-20
Index-34

Spherical mass, gravitational field of 24-24 Star

Spin Binary int-2
Allowed projections 39-3 Black hole 20-19
Chapter on 39-1 Black dwarf int-19
Concept of 39-3 Black hole int-20, 20-18
Dirac equation 39-3 Blackbody spectrum of 34-3
Electron, introduction to 38-1 Hydrogen spectrum of 33-19, 35-4
Electron, periodic table 38-9 Neutron int-20, 20-17
Electron Spin Resonance Red supergiant 20-15
Experiment 39-5, 39-9 Stellar evolution 17-17
Gyroscope like 39-15 White dwarf 20-15
Interaction with magnetic field 39-4 Statamps, CGS units 24-2
Magnetic energy 39-1 Static charges
Semi classical formula 39-14 Electric field of 30-16
Uhlenbeck and Gaudsmit 39-1 Line integral for 30-2
X. See Experiments II: -12- Electron spin resonance Stationary source, moving observer, Doppler ef-
Spin flip energy, Dirac equation 39-15 fect 33-21
Spring Statvolts, CGS units of charge 24-2
Constant 9-3 Steady state model of the universe 34-25
Forces 9-3 Stellar evolution
Mass on a spring General discussion int-19
Computer analysis of 14-30 Role of neutrinos 20-15
Differential equation for 14-8 Role of the four basic interactions 20-13
Theory 14-7
Role of thermal energy 17-17
Oscillating cart 14-5
Step-By-Step Calculations 5-1
Spring model of molecular force 17-13
Stomach, medical image Optics-15
Spring pendulum
Ball spring program 9-20 Strain, definition of 15-8
Computer analysis of 9-8 Strange quark int-24
F = ma 9-7 Streamlines
Introduction to 9-4 And electric field lines 24-14, 24-17
Spring potential energy int-11 Around airplane wing 23-13
Square of amplitude, intensity 16-33 Around sailboat sail 23-14
Square wave Bernoulli’s equation 23-11
Fourier analysis of 16-9, 16-28 Bounding flux tubes 23-8
Stability of matter 19-14 Definition of 23-4
Standard model of basic interactions int-25 Hele-Shaw cell 23-4
Standing waves In blood flow experiment 23-23
Allowed standing waves in hydrogen 37-1 In superfluid helium venturi meter 23-17
De Broglie waves In venturi meter 23-15
Movie 35-11 Strength of the electric interaction 19-8
Formulas for 15-20 And magnetic forces 28- 6
Hydrogen, L= 0 patterns 38-4 In comparison to gravity int-6
Introduction to 15-18 In comparison to nuclear force int-18, 20-2
Light waves in laser 37-2 Two garden peas 28- 2
Made from traveling waves 15-19 Stress, definition of 15-8
On a guitar string 15-20 String forces
Frequency of 15-21 Atwoods machine 9-16
On drums 16-22 Conical pendulum 9-18
On violin backplate 16-23 Introduction to 9-15
Particle wave nature int-13 Solving pulley problems 9-16
Patterns in hydrogen 38-3 Working with rope 13-10
Photons in laser 37-3 String theory int-16
The L= 0 patterns in hydrogen 38-4 Stringed instruments 16-18
Two dimensional 37-8 Violin, acoustic vs electric 16-19
Electrons on copper crystal 37-9
On drumhead 37-8
 Index-35

Strobe photographs T
Analyzing 3-8, 3-11 Table of elements 19-5
And the uncertainty principle 4-2, Cal 1-4 Tacoma narrows bridge 14-24
Defining the acceleration vector 3-15 Tangential distance, velocity and acceleration 12-4
Defining the velocity vector 3-11 Tau particle, lepton family int-22
Taking 3-7
Tau type neutrino int-22
Structure, nuclear 20-7
Taylor, Joe, binary neutron stars int-15
Styrofoam projectile 5-28
Telescope, parabolic reflector Optics-4
SU3 symmetry 19-15
Telescopes int-1
Gell-Mann and Neuman int-24
Infrared
Subtraction of vectors 2-7 IRAS satellite Optics-47
Summation Mt. Hopkins, 2Mass Optics-46
Becoming an integral Cal 1-10 Radio Optics-48
Of velocity vectors Cal 1-10 Arecibo Optics-48
Sun Arecibo, binary neutron stars int-15
Age of 34-23 Holmdel, three degree radiation int-30
And neutron stars 20-17 Radio galaxy image Optics-48
Energy source int-18 Radio images of variable star Optics-49
Halos around Optics-18 Very Large Array Optics-48
Kepler's laws 8-24 Very Long Baseline Array Optics-49
Neutrinos from 6-13, 11-21 Reflecting Optics-42
Stellar evolution 17-17, 20-13 Cassegrain design Optics-42
Sun dogs Optics-18 Diffraction limit Optics-45
Superconducting magnets 31-30 Hubbel space telescope Optics-44
Image plane Optics-5
Superfluids
Keck, world’s largest optical Optics-45
Quantized vortices in 23-22
Mt. Hopkins Optics-43
Superfluid helium venturi meter 23-15 Mt. Palomar Optics-43
Supernova, 1987 Newton’s Optics-42
In stellar evolution 20-14 Secondary mirror Optics-42
Neutrino astronomy 6-14 William Hershel’s Optics-43
Neutrinos from 20-16 Refracting Optics-40
Photograph of int-19, 6-14 Galileo's Optics-41
Superposition of waves Yerkes Optics-41
Circular waves 33-2 Television waves
Principle of 15-11 Photon energies 34-11
Two slit experiment 40-2 Wavelength of 32- 20
Surface (111 plane) of silicon Optics-51 Temperature
Surface charges 26-2 Absolute zero 17-9
Surface integral. See also Integral And zero point energy 37-8
Definition of 29-2 Boltzman’s constant 17-11
For magnetic fields 32- 3 Heated hydrogen int-9
Gauss’ law 29-3 Ideal gas thermometer 17-20
In Maxwell's equations 32- 8 Introduction to 17-9
Two kinds of vector fields 30-19 Temperature scales 17-10
Surface tension 17-14 Tesla and Gauss, magnetic field dimensions 28- 16
Symmetry Test charge, unit size 24-11
SU3 19-15 Text file for FFT data 16-32
Gell-Mann and Neuman int-24 Thermal efficiency of Carnot cycle. See Carnot cycle,
Symmetry of Maxwell’s equations 32- 9 efficiency of
Synchrotron 28- 22 Thermal energy
Systems of particles, chapter on 11-1 Dollar value of 18-1
In a bottle of hydrogen int-8
Time reversed movie of dive 18-1
Thermal equilibrium
And temperature 17-9
Introduction to 17-8
Of the universe 34-28
Index-36

Thermal expansion Time, direction of

Adiabatic expansion 18-9 And strobe photographs 3-27
Derivation of formula 18-26 Time reversal
Formula for 18-10 Dive movie 18-1
In stellar evolution 17-17, 20-13 Water droplets 18-2
Isothermal expansion 18-8 Time step and initial conditions 5-14
Derivation of formula for 18-26 Time-energy form of the uncertainty principle 40-19
Molecular theory of 17-33 Top quark
Of gas in balloon 17-16 Mass of int-24
Work done by an expanding gas 18-5 Quark family int-24
Thermal motion Toroid
Boltzman’s constant 17-11 Inductor 31-6
Brownian motion 17-6 In LC experiment 31-11
Movie 17-7 In resonance 31-14
Thermal photons. See Three degree radiation Speed of light measuerment 1-9, 31-15
In blackbody radiation 34-22 Magnetic field of 29-17
Thermometer Torque
Ideal gas 17-20 As angular force 12-15
Mercury or alcohol 17-9 In torsion pendulum 14-12
Temperature scales 17-10 On a current loop 31-20
Thin lenses Optics-23 Torsion pendulum 14-12. See also Cavendish
Thought experiments experiment
Carnot cycle 18-4 Total energy
Causality 1-36 Classical hydrogen atom 35-3
Lack of simultaneity 1-32 Escape velocity 10-28
Light pulse clock 1-13 Satellite motion 8-36
Lorentz contraction and space travel 1-22 Town water supply, hydrodynamic voltage 23-18
Magnetic force and Faraday's law 30-3 Transients 14-27
No width contraction 1-28 Translation and rotation 12-24
Origin of magnetic forces 28- 8
Transmitted wave and lenses 36-3
Two slit experiment and the uncertainty principle 40-
Traveling waves, formula for 15-16
9
Tritium, a hydrogen isotope int-17, 19-7
Three degree cosmic radiation int-30, 34-27
True BASIC. See BASIC program
Penzias and Wilson 34-27
Tides, two a day 8-12 Tubes of flux 24-17
Time Two dimensional standing waves 37-8
Age of sun 34-23 Two kinds of vector fields 30-18
And the speed of light int-2 Two lenses together Optics-29
Behavior of 1-1 Two slit experiment
Dilation 1-22 And the uncertainty principle 40-9
Dilation formula 1-16 Measurement limitations 40-9
Direction of One particle at a time 40-3
Entropy 18-26 One slit experiment 40-2
Neutral K meson 18-26 Particle point of view 40-3
In early universe int-27, 34-29 Particle/wave nature 40-2
Measuring short times using uncertainty prin- Using electrons 40-3
ciple 40-22 Two-slit interference patterns 33-6
Moving clocks 1-13 A closer look at 33-26
Muon lifetime movie 1-21 First maxima 33-8
On light pulse clock 1-14 For light 33-10
On other clocks 1-18 Tycho Brahe’s apparatus 8-25
On real clocks 1-20
Steady state model of universe 34-25 U
Time constant Uhlenbeck and Gaudsmit, electron spin 39-1
For RC circuit 27- 24 Ultraviolet light
Measuring from a graph 27- 25, Cal 1-34 Electromagnetic spectrum int-7, 32- 20, 32- 22
Photon energies 34-11
Wavelength of 32- 20
 Index-37

Uncertainty principle Cal 1-3 Universe, early int-27, 34-29

∆x∆p>=h 40-15 10 to the 10 degrees int-27
And definition of velocity 4-1, Cal 1-3 10 to the 13 degrees int-27
And strobe photographs 4-2, Cal 1-4 10 to the 14 degrees int-27
Applied to projectile motion 4-2, Cal 1-4 13.8 seconds int-28
Elementary particles, short lived 40-23 24% neutrons int-28
Energy conservation 40-24 38% neutrons int-27
Fourier transform 40-20 At .7 million years, decoupling int-29
Introduction to 40-14 At various short times 34-30
Particle/wave nature int-10 Big bang model 34-26
Position-momentum form 40-15 Books on
Single slit experiment 40-16 Coming of Age in the Milky Way 34-32
Time-energy form 40-19 The First Three Minutes 34-32
Used as clock 40-22 Decoupling (700,000 years) 34-31
X. See Experiments II: -13- Fourier analysis & the Deuterium bottleneck int-28
uncertainty principle Excess of Matter over Antimatter 34-29
Uniform circular motion. See Circular motion Frame #2 (.11 seconds) 34-30
Uniform expansion of universe int-3, 33-24 Frame #3 (1.09 seconds) 34-30
Uniform magnetic fields 28- 16 Frame #4 (13.82 seconds) 34-30
Unit of angular momentum int-9 Frame #5 (3 minutes and 2 seconds) 34-30
Electron spin 39-3 Helium abundance 34-26
In Bohr theory 35-9 Helium created int-28
Unit test charge 24-11 Hot int-4
Unit vectors 8-18 Matter particles survive int-27, 34-29
Neutrinos escape at one second int-28
Units
Overview int-27, 34-29
Atomic units 19-22
Positrons annihilated int-28
CGS
Thermal equilibrium of 34-28
centimeter, gram, second Back cover-1
Coulomb's law 24-2 Thermal photons int-29, 34-28
Statamp, statvolt, ergs per second 24-2 Three degree cosmic radiation int-30, 34-27, 34-32
Checking MKS calculations 24-3 Transparent universe int-29
MKS Why it is hot 34-29
Ampere, volt, watt 24-2 Up quark int-24
Coulomb's law 24-2 Uranium
Meter, kilogram, second Front cover-2 Binding energy per nucleon 20-10
Planetary units 8-14 Nuclear fission 20-4
Practical System of units (MKS) 10-31 Nuclear structure int-17, 20-2
Universe
Age of int-3 V
As a laboratory int-1 Van Allen radiation belts 28- 32
Becomes transparent, decoupling 34-31 Van de Graaff generator 26-6
Big bang model int-4, 33-25, 34-26 Variable names, computer 5-6
Continuous creation theory int-4, 34-25 Variables
Decoupling (700,000 years) 34-31 Evaluated over interval Cal 1-10
Early. See Universe, early
Evolution of, Doppler effect 34-21
Excess of matter over antimatter int-27, 34-29
Expanding int-3
Quasars and gravitational lensing 34-19
Red shift 33-24
Helium abundance in 34-26
Models of 34-23
Steady state model of 34-25
The First Three Minutes 34-32
Thermal equilibrium of 34-28
Three degree cosmic radiation int-30, 34-27
Visible int-3, 34-32
Index-38

Vector Velocity detector

Addition 2-3 Air cart 30-5
Addition by components 2-9 Magnetic flux 30-25
Angular momentum 7- 14, 12-7 Velocity field 23-2
Angular velocity 12-7 Flux of 23-8, 24-15
Area 24-22 Of a line source 23-7
Components 2-8 Of a point source 23-6
Cross Product 12-9 Velocity vector from coordinate vector 3-13
Definition of acceleration Cal 1-7 Venturi meter
Component equations Cal 1-8 Bernoulli's equation 23-15
Definition of velocity Cal 1-6 With superfluid helium 23-17
Component equations Cal 1-8 Very Large Array, radio telescopes Optics-48
Dot product 10-13 Very Long Baseline Array , radio telescopes Optics-
Equations 49
Components with derivatives 4-6, Cal 1-7
Violets, odor of 17-6
Constant acceleration formulas 4-11
Exercise on 2-7, 2-8 Violin
In component form 2-10 Acoustic vs electric 16-19
Magnitude of 2-6 Back, standing waves on 16-23
Measuring length of 3-9 Viscous effects in fluid flow 23-19
Multiplication 2-11 Visible light int-7, 32- 20
Multiplication, cross product 2-15, 12-9 Photon energies 34-11
Formula for 2-17 Spectrum of 33-15
Magnitude of 2-17 Wavelength of 32- 20
Right hand rule 12-10 Visible universe int-3, 34-32
Multiplication, scalar or dot product 2-12, 10-13 Volt
Interpretation of 2-14 Electron 26-12
Velocity from coordinate vector 3-13 MKS units 24-2
Vector fields 23-3 Voltage
Electric field 24-10 Air cart speed detector 30-5
Magnetic field 28- 10 Divider, circuit for 27- 13
Two kinds of 30-18 Electric 25-6
Velocity field 23-2 Fluid analogy
Vectors 2-2 Hydrodynamic voltage 23-17
Arithmetic of 2-3 Resistance 27- 7
Associative law 2-4 Town water supply 23-18
Commutative law 2-4 Induced 31-4
Coordinate 3-11 Induced in a moving loop 30-4
Displacement 2-2 Voltage and current 27- 6
from Strobe Photos 3-5 Resistors 27- 6
Graphical addition and subtraction 3-10 Ohm’s law 27- 7
Multiplication by number 2-5 Voltage source, constant 27- 15
Negative of 2-5 Voltage transformer and magnetic flux 30-26
Subtraction of 2-5 Volume of mole of gas 17-25
Unit 8-18 Vortex
Velocity 3-11 Bathtub 23-2, 23-20
And the uncertainty principle 4-2, Cal 1-4 Hurricane 23-20
Angular 12-2, 14-5 Tornado 23-20
Angular analogy 12-3 Vortex street, Karman 14-25
Calculus definition of 4-3, Cal 1-5 Vortices 23-20
Component equations Cal 1-8 Quantized, in superfluids 23-22
Curve, area under Cal 1-12
Definite integral of Cal 1-11
Instantaneous 3-24
From strobe photograph 3-26
Integral of Cal 1-10
Of escape 10-28
Tangential 12-4
Using strobe photos 3-11
 Index-39

W Wave motion
W and Z mesons, electroweak interaction int-26 Amplitude and phase 15-17
Water Linear and nonlinear 15-10
Evaporating 17-5 One dimensional 15-1
Molecule 17-2 Principle of superposition 15-11
Water droplets, time reversal 18-2 Wave nature of light
Watt, MKS units 24-2 Young, Thomas 34-1, Optics-1
Wave Wave patterns
Circular water waves 33-2 Hydrogen
Intensity at the Origin 38-5
Cosine waves 16-28
Schrödinger’s equation 38-2
De Broglie, standing wave movie 35-11
Standing waves 38-3
Diffraction pattern 33-5
Wave/particle nature. See also Particle-wave nature
Electromagnetic waves 32- 18
Born Interpretation 40-6
Probability wave for photons 40-7
Electron waves, de Broglie picture 35-11 Probability interpretation 40-6
Electron waves, in hydrogen 38-1 Two slit experiment 40-2
Forms, repeated, in Fourier analysis 16-11 Wavelength
Fourier analysis De Broglie 35-11
Of a sine wave 16-7 Electron 36-9
Of a square wave 16-9 Fourier analysis 16-28
Huygens' principle, introduction to 33-4 Laser beam 16-33
Interference patterns 33-3 Period, and frequency 15-13
Light waves Weak interaction int-14, int-20, 19-2
Chapter on 33-1 Creation of neutrinos 20-6
Polarization of 32- 23 Electroweak theory int-26
Patterns, superposition of 33-2 Four basic interactions 19-2
Photon wave 40-6 Neutron decay 20-6
Probability wave Range of int-21
Intensity of 40-22 Strength of int-21
Reflection and Fluorescence 40-8 Weighing the earth 8-8
Scattering, measurement limitation 40-10 Weight 8-11
Single slit experiment, uncertainty principle 40-16 Lifting 13-11
Sinusoidal waves 15-12 Weinberg, S., book "The first Three Minutes" 34-32
Time dependent 14-3 Wein's displacement law for blackbody radia-
Speed of waves tion 34-2
Dimensional analysis 15-6 White dwarf star 20-15
Project suggestion 15-8
White light (Fourier analysis) 16-28
Standing waves
Allowed 37-1 William Hershel’s telescope Optics-43
Formulas for 15-20 Wilson, Robert, 3 degree radiation int-29, 34-27
Frequency of 15-21 Wind instruments 16-20
Hydrogen L= 0 Patterns 38-4 Work int-11
Introduction to 15-18 And potential energy 10-14
On a guitar string 15-20 Bernoulli's equation 23-10
Two dimensional 37-8 Calculation of in adiabatic expansion 18-26
Transmitted waves 36-3 Calculation of in isothermal expansion 18-26
Traveling waves, formula for 15-16 Definition of 10-12
Wave pulses 15-3 Done by an expanding gas 18-5
Speed of 15-4 Integral formula for 10-14, 10-15
Wave equation int-7, 15-1 Non-constant forces 10-14
Dirac's 15-2 Vector dot product 10-13
Bohr magneton 39-5 Work energy theorem 10-18
Spin 39-3 World’s largest optical telescope Optics-45
For light int-7
Maxwell's 15-1
Nonrelativistic int-12, 34-16
Relativistic int-12, 15-2
Schrödinger's 15-1
Discovery of 37-1
Index-40

X Exercise 7 5-20
x-Cal 1 Exercise 8 Changing the time step 5-20
Exercise 1 Cal 1-14 Exercise 9 Numerical Output 5-20
Exercise 2 Cal 1-15 Exercise 10 Attempt to reduce output 5-20
Exercise 3 Cal 1-17 Exercise 11 Reducing numerical output 5-20
Exercise 4 Cal 1-22 Exercise 12 Plotting crosses 5-21
Exercise 5 Cal 1-24 Exercise 13 Graphical analysis 5-25
Exercise 6 Cal 1-29 Exercise 14 Computer prediction 5-26
Exercise 7 Cal 1-29 Exercise 15 Viscous fluid 5-26
Exercise 8 Cal 1-31 Exercise 16 Nonlinear air resistance (optional) 5-27
Exercise 9 Cal 1-33 Exercise 17 Fan Added 5-27
Exercise 10 Cal 1-36 X-Ch 6
Exercise 11 Cal 1-39 Exercise 1 6-4
Exercise 12 Cal 1-39 Exercise 2 Decay of Plutonium 246 6-8
Exercise 13 Cal 1-39 Exercise 3 Protactinium 236 decay. 6-9
X-Ch 1 Exercise 4 Increase in Electron Mass. 6-9
Exercise 1 1-3 Exercise 5 A Thought Experiment. 6-9
Exercise 2 1-4 Exercise 6 6-10
Exercise 3 1-11 Exercise 7 6-10
Exercise 4 1-16 Exercise 8 6-10
Exercise 5 1-31 X-Ch 7
Exercise 6 1-31 Exercise 1 7- 6
Exercise 7 1-31 Exercise 2 7- 8
X-Ch 2 Exercise 3 Frictionless Ice 7- 8
Exercise 1 2-7 Exercise 4 Bullet and Block 7- 8
Exercise 2 2-7 Exercise 5 Two Skaters Throwing Ball 7- 8
Exercise 3 2-7 Exercise 6 Rocket 7- 8
Exercise 4 2-7 Exercise 7 7- 10
Exercise 5 2-8 Exercise 8 7- 10
Exercise 6 2-10 Exercise 9 7- 10
Exercise 7 2-12 Exercise 10 7- 13
Exercise 8 2-13 Exercise 11 7- 14
Exercise 9 2-15 Exercise 12 7- 16
Exercise 10 2-16 Exercise 13 7- 16
Exercise 11 2-17 Exercise 14 7- 18
Exercise 12 2-17 X-Ch 8
Exercise 13 2-18 Exercise 1 8-5
X-Ch 3 Exercise 2 8-8
Exercise 1 3-10 Exercise 3 8-9
Exercise 2 3-10 Exercise 4 8-11
Exercise 3 3-10 Exercise 5 8-12
Exercise 4 3-12 Exercise 6 8-15
Exercise 5 3-17 Exercise 7 8-15
Exercise 6 3-18 Exercise 8 8-16
Exercise 7 3-22 Exercise 9 8-16
Exercise 8 3-27 Exercise 10 8-22
Exercise 9 3-27 Exercise 11 8-23
Exercise 10 3-27 Exercise 12 8-23
X-Ch 4 Exercise 13 8-27
Exercises 1-7 4-19 Exercise 14 8-27
X-Ch 5 Exercise 15 8-28
Exercise 1 5-3 Exercise 16 8-28
Exercise 2 A running program 5-8 Exercise 17 8-28
Exercise 3 Plotting a circular line 5-8 Exercise 18 8-31
Exercise 4 Labels and axes 5-9 Exercise 19 8-34
Exercise 5a Numerical output 5-9 Exercise 20 8-34
Exercise 5b 5-10 Exercise 21 8-36
Exercise 6 Plotting crosses 5-11 Exercise 22 8-36
Exercise 23 8-37
 Index-41

X-Ch 9 Exercise 8 Working with rope 13-10

Exercise 1 9-7 Exercise 9 13-11
Exercise 2 9-11 Exercise 10 13-12
Exercise 3 9-17 Exercise 11 13-12
Exercise 4 9-17 X-Ch14
Exercise 5 Conical Pendulum 9-19 Exercise 1 14-4
Exercise 6 9-19 Exercise 2 14-5
X-Ch10 Exercise 3 14-5
Exercise 1 10-4 Exercise 4 14-7
Exercise 2 10-6 Exercise 5 14-8
Exercise 3 10-8 Exercise 6 14-9
Exercise 4 10-9 Exercise 7 14-10
Exercise 5 10-9 Exercise 8 14-10
Exercise 6 10-9 Exercise 9 14-10
Exercise 7 10-11 Exercise 10 14-11
Exercise 8 10-12 Exercise 11 14-11
Exercise 9 10-13 Exercise 12 14-14
Exercise 10 10-17 Exercise 13 14-14
Exercise 11 10-17 Exercise 14 14-15
Exercise 12 10-28 Exercise 15 14-16
Exercise 13 10-28 Exercise 16 14-16
Exercise 14 10-28 Exercise 17 14-17
Exercise 15 10-31 Exercise 18 14-18
X-Ch11 Exercise 19 Physical pendulum 14-18
Exercise 1 11-4 Exercise 20 Damped harmonic motion 14-23
Exercise 2 11-4 Exercise 21 14-27
Exercise 3 11-4 Exercise 22 14-33
Exercise 4 11-6 Exercise 23 14-34
Exercise 5 11-8 X-Ch15
Exercise 6 11-12 Exercise 1 15-4
Exercise 7 11-13 Exercise 2 15-9
Exercise 8 11-13 Exercise 3 15-9
Exercise 9 11-13 Exercise 4 15-9
Exercise 10 11-14 Exercise 5 15-14
Exercise 11 11-14 Exercise 6 15-15
Exercise 12 11-15 Exercise 7 15-16
Exercise 13 11-16 Exercise 8 15-22
X-Ch12 Exercise 9 15-22
Exercise 1 12-2 Exercise 10 15-22
Exercise 2 12-3 X-Ch16
Exercise 3 12-5 Exercise 1 16-21
Exercise 4 12-9 Exercise 2 16-27
Exercise 5 12-10 Exercise 3 16-27
Exercise 6 12-11 X-Ch17
Exercise 7 12-13 Exercise 1 17-8
Exercise 8 12-15 Exercise 2 17-11
Exercise 9 12-21 Exercise 3 17-11
Exercise A1 12-24 Exercise 4 17-11
Exercise A2 12-26 Exercise 5 17-14
Exercise A3 potential lab experiment 12-26 Exercise 6 17-24
X-Ch13 Exercise 7 17-25
Exercise 1 13-3 Exercise 8 17-26
Exercise 2 13-3
Exercise 3 13-4
Exercise 4 13-6
Exercise 5 13-8
Exercise 6 Ladder problem 13-8
Exercise 7 13-9
Index-42

X-Ch18 X-Ch27
Exercise 1 18-5 Exercise 1 27- 10
Exercise 2 18-7 Exercise 2 27- 13
Exercise 3 18-10 Exercise 3 The voltage divider 27- 13
Exercise 4 18-13 Exercise 4 - Electrolytic capacitor 27- 17
Exercise 5 18-13 Exercise 5 27- 19
Exercise 6 18-14 Exercise 6 27- 21
Exercise 7 18-16 Exercise 7 27- 24
Exercise 8 18-19 Exercise 8 27- 26
Exercise 9 18-20 Exercise 9 27- 28
Exercise 10 18-20 Exercise 10 27- 30
X-Ch19 Exercise 11 27- 32
Exercise 1 19-9 X-Ch28
Exercise 2 19-11 Exercise 1 28- 3
Exercise 3 19-11 Exercise 2 28- 9
Exercise 4 19-12 Exercise 3 28- 15
Exercise 5 19-12 Exercise 4 28- 21
X-Ch20 Exercise 5 28- 21
Exercise 1 20-9 Exercise 6 28- 25
X-Ch23 Exercise 7 28- 25
Exercise 1 23-12 Exercise 8 28- 25
Exercise 2 23-12 Exercise 9 28- 27
Exercise 3 23-15 Exercise 10 28- 28
Exercise 4 23-15 X-Ch29
Exercise 5 23-22 Exercise 1 29-12
Exercise 6 23-22 Exercise 2 29-12
X-Ch24 Exercise 3 29-12
Exercise 1 24-5 Exercise 4 29-12
Exercise 2 24-5 Exercise 5 29-13
Exercise 3 24-5 Exercise 6 29-13
Exercise 4 24-5 Exercise 7 29-16
Exercise 5 24-8 Exercise 8 29-18
Exercise 6 24-9 Exercise 9 29-18
Exercise 7 24-12 X-Ch30
Exercise 8 24-25 Exercise 1 30-10
Exercise 9 24-27 Exercise 2 30-12
Exercise 10 24-27 Exercise 3 30-18
Exercise 11 24-27 Exercise 4 30-20
Exercise 12 24-27 Exercise 5 30-22
Exercise 13 24-28 Exercise 6 30-22
Exercise 14 24-28 Exercise 7 30-25
X-Ch25 Exercise 8 30-26
Exercise 1 25-5 X-Ch31
Exercise 2 25-9 Exercise 1 31-3
Exercise 3 25-12 Exercise 2 31-9
Exercise 4 25-12 Exercise 3 31-10
X-Ch26 Exercise 4 31-11
Exercise 1 26-5 Exercise 5 31-15
Exercise 2 26-5 Exercise 6 31-15
Exercise 3 26-5 Exercise 7 31-16
Exercise 4 26-5 X-Ch32
Exercise 5 26-11 Exercise 1 32- 7
Exercise 6 26-13 Exercise 2 32- 8
Exercise 7 26-13 Exercise 3 32- 9
Exercise 8 26-16 Exercise 4 32- 17
Exercise 9 26-17 Exercise 5 32- 17
Exercise 10 Millikan oil drop experiment 26-17 Exercise 6 32- 32
 Index-43

X-Ch33 x-Ch39
Exercise 1 33-4 Exercise 1 39-5
Exercise 2 33-8 Exercise 2 39-5
Exercise 3 33-10 Exercise 3 39-5
Exercise 4 33-10 Exercise 4 39-7
Exercise 5 33-13 Exercise 5 39-7
Exercise 6 33-13 Exercise 6 39-12
Exercise 7 33-16 x-Ch40
Exercise 8 33-18 Exercise 1 40-6
Exercise 9 33-19 Exercise 2 40-7
Exercise 10 33-21 Exercise 3 40-18
Exercise 11 33-23 X-Optics
Exercise 12 33-23 Exercise 1a Optics-10
Exercise 13 33-23 Exercise 1b Optics-10
Exercise 14 33-28 Exercise 2 Optics-12
Exercise 15 33-30 Exercise 3 Optics-13
X-Ch34 Exercise 4 Optics-17
Exercise 1 34-3 Exercise 5 Optics-21
Exercise 2 34-9 Exercise 6 Optics-21
Exercise 3 34-9 Exercise 7 Optics-21
Exercise 4 34-10 Exercise 8 Optics-23
Exercise 5 34-10 Exercise 9 Optics-24
Exercise 6 34-10 Exercise 10 Optics-27
Exercise 7 34-10 Exercise 11 Optics-29
Exercise 8 34-10 Exercise 12 Optics-30
Exercise 9 34-10 Exercise 13 Optics-30
Exercise 10 34-10 X-rays
Exercise 11 34-17 Diffraction 36-4
Exercise 12 34-18 Diffraction pattern 36-5
x-Ch35 Electromagnetic spectrum int-7, 32- 20, 32- 22
Exercise 1 35-5 Photon energies 34-11, 36-4
Exercise 2 35-5 Wavelength of 32- 20
Exercise 3 35-6
Exercise 4 35-6 Y
Exercise 5 35-7 Yerkes telescope Optics-41
Exercise 6 35-9 Young, Thomas
Exercise 7 35-9 Wave nature of light 34-1, Optics-1
Exercise 8 35-9 Young’s modulus 15-8
Exercise 9 35-10 Yukawa, H. int-22
Exercise 10 35-10 Yukawa's theory int-22
x-Ch36
Exercise 1 36-3 Z
Exercise 2 36-7 Z and W mesons, electroweak interaction int-26
Exercise 3 36-7 Z, astronomer's Z factor for Doppler effect 33-23
Exercise 4 36-9 Zero, absolute 17-9, 17-21
Exercise 5 36-12 Zero of potential energy 10-22
Exercise 6 36-12 Zero point energy 37-7
x-Ch37 And temperature 37-8
Exercise 1 37-4 Chapter on 37-1
Exercise 2 37-6 Zero rest mass particles 6-11
Exercise 3 37-6 Zoom lens Optics-18
Exercise 4 37-7 Zweig, George, quarks 19-15
x-Ch38
Exercise 1 38-4
Exercise 2 38-7
Physical Constants in CGS Units (link to MKS units)
speed of light c = 3 × 10 10cm/ sec = 1000 ft / µsec = 1 ft / nanosecond
acceleration due to gravity
at the surface of the earth g = 980 cm/ sec2 = 32 ft/ sec2
gravitational constant G = 6.67 × 10 – 8cm3 / (gm sec2)
charge on an electron e = 4.8 × 10– 10esu
Planck's constant h = 6.62 × 10 – 27 erg sec (gm cm2/sec )
Planck constant / 2 π h = 1.06 × 10 – 27erg sec (gm cm2 / sec )
Bohr radius a0 = .529 × 10 – 8cm
rest mass of electron me = 0.911×10 – 27gm
rest mass of proton Mp = 1.67 × 10 – 24gm
rest energy of electron m ec 2 = 0.51 MeV ( ≈ 1 / 2 MeV)
rest energy of proton Mpc 2 = 0.938 BeV ( ≈ 1 BeV)
proton radius rp = 1.0× 10 – 13cm
Boltzmann's constant k = 1.38 × 10 – 16ergs/ kelvin
Avogadro's number N 0 = 6.02 × 10 23molecules/ mole

absolute zero = 0°K = – 273°C

3
density of mercury = 13.6 gm / cm
mass of earth = 5.98 × 10 27gm
mass of the moon = 7.35 × 10 25gm
mass of the sun = 1.97 × 10 33gm
earth radius = 6.38 × 10 8cm = 3960 mi
moon radius = 1.74 × 10 8cm = 1080 mi
mean distance to moon = 3.84 × 10 10cm
mean distance to sun = 1.50 × 10 13cm
mean earth velocity in orbit about sun = 29.77 km / sec

Conversion Factors
1 meter = 100 cm (100 cm/meter)
1 in. = 2.54 cm (2.54 cm/in.)
1 mi = 5280 ft (5280 ft/mi)
5 5
1 km (kilometer) = 10 cm (10 cm / km)
1 mi = 1.61 km = 1.61 × 10 cm (1.61× 10 5cm/ mi)
5

1 A° (angstrom ) = 10 – 8cm (10 – 8cm / A° )

1 day = 86,000 sec ( 8.6 × 10 4sec / day )
1 year = 3.16 × 10 7sec (3.16× 10 7sec/ year)
1 µ sec (microsecond ) = 10 – 6sec (10 – 6sec / µ sec )
– 9
1 nanosecond = 10 sec (10 – 9sec /nanosecond )
1 mi/hr = 44.7 cm/sec
60 mi/hr = 88 ft/sec
3 3
1 kg (kilogram) = 10 gm (10 gm / kg)
1 coulomb = 3 × 109esu (3 × 10 9esu/coulomb)
1 ampere = 3 × 109statamps (3 × 109statamps/ ampere)
1 statvolt = 300 volts (300 volts/statvolt)
7 7
1 joule = 10 ergs (10 ergs / joule )
7 7
1 W (watt) = 10 ergs / sec (10 erg / W)
1 eV = 1.6 × 10– 12ergs (1.6 × 10– 12ergs/ eV)
6 6
1 MeV = 10 eV (10 eV /MeV)
9 9
1 BeV = 10 eV (10 eV /BeV)
1 µ (micron ) pressure = 1.33 dynes / cm
2

1 cm Hg pressure = 10 µ
4

1 atm = 76 cm Hg = 1.01×10 6dynes/ cm2

Moose Mountain
Digital Press
Calculus 2000
E. R. Huggins
Dartmouth College

physics2000.com
Front cover

MKS Units
m = meters kg = kilograms s = seconds
N = newtons J = joules C = coulombs
T = tesla F = farads H = henrys Powers of 10
A = amperes K = kelvins mol = mole
Power Prefix Symbol
speed of light c 3.00 × 10 8 m / s
10 12 tera T
gravitational constant G 6.67 × 10 –11
N⋅m2 / kg 2 9
10 giga G
permittivity constant ε0 8.85 × 10 – 12F / m
10 6 mega M
permeability constant µ0 1.26 × 10 –6
H/m 10 3 kilo k
elementary charge e 1.60 × 10 –19C 10 2 hecto h
electron volt eV 1.60 × 10 –19 J 10 – 1 deci d
electron rest mass me 9.11 × 10 – 31kg 10 – 2 centi c
proton rest mass mp 1.67 × 10 – 27kg 10 – 3 milli m
Planck constant h 6.63 × 10 – 34 J⋅ s 10 – 6 micro µ
–9
Planck constant / 2π h 1.06 × 10 – 34 J⋅ s 10 nano n
5.29 × 10 – 11m 10 – 12 pico p
Bohr radius rb
– 15
10 femto f
Bohr magneton µb 9.27 × 10 – 24J / T
Boltzmann constant k 1.38 × 10 –23J / K
Avogadro constant NA 6.02 × 10 23mol – 1
universal gas constant R 8.31 J /mol⋅ K

Dimensions
Quantity Unit Equivalents
Force newton N J/m kg •m/ s2
Energy joule J N• m kg • m2/s2
Power watt W J/s kg • m2/s3
Pressure pascal Pa N/m 2 kg/m• s2
Frequency hertz Hz cycle/s s–1
Electric charge coulomb C A•s
Electric potential volt V J/C kg • m2/A • s3
Electric resistance ohm Ω V/A kg • m2/A2• s3
Capacitance farad F C/V A2• s4/kg • m2
Magnetic field tesla T N • s/C • m kg/A • s2
Magnetic flux weber Wb T• m2 kg • m2/A• s2
Inductance henry H V•s/A kg • m2/A2• s2

Copyright © Moose Mountain Digital Press

New Hampshire 03750
All rights reserved
 Calculus 2000 - Preface & Table of Contents Cal - i

Calculus 2000
A Physics-Based Calculus Text

When developing a physics curriculum, a major After the introductory courses, the standard physics
concern is the mathematical background of the curriculum repeatedly goes over the same topics at
student. The Physics 2000 text was developed successively higher mathematical levels. A typical
teaching premedical students who were supposed to example is the subject of electricity and magnetism
have had one semester of calculus. Because many which is taught using integral equations in the
of the students had taken calculus several years introductory course, using differential operators in
previously, and had forgotten much of it, the physics an upper level undergraduate course, and then
text used strobe photographs and the computer to taught all over again in a graduate level course. In
carefully introduce the calculus concepts such as each of the courses it takes a while for the student to
velocity, acceleration, and the limiting process. By realize that this is just the same old subject dressed
the time we got to electricity and magnetism in Part up in new math.
2 of Physics 2000 we relied on the student being
familiar with the basic steps of differentiation and With Chapters 2 through 13 of the Calculus 2000,
integration. we introduce a different approach. We take the
topics that we have already introduced in Physics
For students who have forgotten much of their 2000, and show how these topics can be handled in
calculus course, or those who have not had calculus progressively more sophisticated mathematical
but wish to study the Physics 2000 text, we have ways. Once we have introduced the mathematical
written Chapter 1 of Calculus 2000. This chapter concepts of gradient, divergence and curl in the
not only covers all the calculus needed for the calculus text, we can turn the integral form of
Physics 2000 text, but is also carefully integrated Maxwell's equation into a wave equation for elec-
with it. The chapter is much shorter than the typical tric and magnetic fields. With the introduction of
introductory calculus text because the basic calcu- the Laplacian and complex variables, we can study
lus concepts are discussed in the physics text and the Schrödinger's equation and begin to solve for the
calculus chapter only has to deal with the formal- hydrogen wave patterns discussed in Chapter 38 of
ism. the physics text.
Cal - ii Calculus 2000 - Preface & Table of Contents

Beyond seeing the same topics in a more sophisti- However our discussion of electric and magnetic
cated way, the student finds that new insights can fields, particularly in this calculus text, go way
result from the advanced mathematical approach. beyond the simple fluid flow topics we introduced in
Chapter 10 of the calculus text is a short chapter less the physics text. In the last two chapters of the
than two pages. But it is one of the most significant calculus text, we turn the tables and apply to fluid
chapters in the text. For there we see that Maxwell's theory the mathematical techniques we learned
equations for electric and magnetic fields require studying electricity and magnetism. In Chapter 12
that electric charge be conserved. This intimate we discuss the concept of vorticity which is the curl
connection between a conservation law and field of the velocity field. The focus is to develop an
theory becomes clear when we have sufficiently intuitive understanding of the nature of vorticity
powerful mathematical tools to handle the theory. and the role it plays in fluid flows, particularly
vortices and vortex rings.
The physics text began its discussion of vector fields
in Chapter 23, using the velocity field as its first Chapter 13 is an introduction to fluid dynamics.
example. We did that because it is much easier to The idea is to bring our discussion of the velocity
visualize the familiar flow of water than the abstract field up to the same level as our treatment of electric
concept of an electric field. We saw that the stream- and magnetic fields. We begin with a derivation of
lines in fluid flow went over to electric field lines, the Navier-Stokes equation which applies to con-
Gauss's law in fluid theory simply represented the stant density viscous fluids. This is then converted
incompressibility of the fluid, and Bernoulli's equa- into an equation for vortex dynamics from which we
tion provided an introduction to the concept of derive an extended form of the famous Helmholtz
voltage and potential. equation. We then use that to derive the well known
properties of vortex motion such as the so called
Magnus force, and discuss the experiment Rayfield
and Reif used to measure the circulation and core
diameter of quantized vortices in superfluid helium.
TABLE OF CONTENTS
Calculus 2000 - Preface & Table of Contents Cal - iii

Table of Contents
FRONT COVER CHAPTER 2 SECOND DERIVATIVES
MKS Units ................................... Front cover-2 AND THE ONE DIMENSIONAL WAVE
Dimensions ................................. Front cover-2 EQUATION
Powers of 10 ............................... Front cover-2 The Second Derivative ............................. Cal 2-2
Example 1 ............................................. Cal 2-2
TABLE OF CONTENTS Geometrical Interpretation .................... Cal 2-3
CHAPTER 1 INTRODUCTION TO Curvature .................................................. Cal 2-4
CALCULUS Curve Fitting and Boat Lofting .............. Cal 2-5
Limiting Process ....................................... Cal 1-2 The Binomial Expansion ........................... Cal 2-6
The Taylor Series Expansion ................ Cal 2-7
The Uncertainty Principle ......................... Cal 1-2 The Constant Acceleration Formulas ... Cal 2-9
Uncertainty Principle on a Larger ScaleCal 1-4
The Wave Equation ................................. Cal 2-10
Calculus Definition of Velocity .................. Cal 1-5 Waves on a Rope ................................ Cal 2-10
Acceleration .............................................. Cal 1-7 Partial Derivatives ............................... Cal 2-12
Components ......................................... Cal 1-7 The One Dimensional Wave Equation Cal 2-14
Integration ................................................. Cal 1-8 Compressional Waves on a Spring .... Cal 2-15
Prediction of Motion .............................. Cal 1-9 The Speed of Sound ........................... Cal 2-17
Calculating Integrals ........................... Cal 1-11
The Process of Integrating ................. Cal 1-13 CHAPTER 3 THE GRADIENT
Indefinite Integrals .............................. Cal 1-14 Two views of the gradient ......................... Cal 3-2
Integration Formulas ........................... Cal 1-14 Calculating the Electric Field .................... Cal 3-4
New Functions ........................................ Cal 1-15 Interpretation ......................................... Cal 3-6
Logarithms .......................................... Cal 1-15 The Gradient OPERATOR ......................... Cal 3-7
The Exponential Function ................... Cal 1-16
The Parallel Plate Capacitor ..................... Cal 3-8
Exponents to the Base 10 ................... Cal 1-16
Voltage Inside a Conductor .................. Cal 3-9
The Exponential Function yx .............. Cal 1-16
Euler's Number e = 2.7183. . . ........... Cal 1-17 Electric Field of a Point Charge .............. Cal 3-10
Differentiation and Integration ................ Cal 1-18 Gradient in Cartesian Coordinates ......... Cal 3-12
A Fast Way to go Back and Forth ....... Cal 1-20 Gradient in Cylindrical Coordinates ....... Cal 3-14
Constant Acceleration Formulas ........ Cal 1-20 Gradient in Spherical Coordinates ......... Cal 3-16
In Three Dimensions ........................... Cal 1-22
Summary of Gradient Formulas .............. Cal 3-18
More on Differentiation ........................... Cal 1-23 Cartesian Coordinates ........................ Cal 3-18
Series Expansions .............................. Cal 1-23 Cylindrical Coordinates ...................... Cal 3-18
Derivative of the Function xn .............. Cal 1-24 Spherical Coordinates ........................ Cal 3-18
The Chain Rule ................................... Cal 1-25
Examples ................................................ Cal 3-18
Remembering the Chain Rule ............ Cal 1-25
Electric Field of a Point Charge .......... Cal 3-18
Proof of the Chain Rule (optional) ....... Cal 1-26
Electric Field of a Line Charge ........... Cal 3-19
Integration Formulas ............................... Cal 1-27 The Coaxial Cable .............................. Cal 3-21
Derivative of the Exponential Function Cal 1-28
Integral of the Exponential Function ... Cal 1-29 CH 3 VIEW 2 THE GRADIENT FROM A
Derivative as the Slope of a Curve ......... Cal 1-30 GEOMETRICAL PERSPECTIVE
Negative Slope ................................... Cal 1-31 Slope in Two Dimensions ....................... Cal 3-23
The Exponential Decay ........................... Cal 1-32 The Gradient ........................................... Cal 3-25
Muon Lifetime ..................................... Cal 1-32 Gradient as a Vector Field .................. Cal 3-28
Half Life ............................................... Cal 1-33
Measuring the Time Constant ............. Cal 1-34 CH 3 VIEW 3
The Sine and Cosine Functions .............. Cal 1-35 Pressure Force as a Gradient ................ Cal 3-29
Radian Measure .................................. Cal 1-35
The Sine Function ............................... Cal 1-36
Amplitude of a Sine Wave .................. Cal 1-37
Derivative of the Sine Function ........... Cal 1-38
Cal - iv Calculus 2000 - Preface & Table of Contents

CHAPTER 6 INTRODUCTION TO
CHAPTER 4 THE OPERATOR ∇ 2 ≡ ∇ ⋅ ∇
SCHRÖDINGER'S EQUATION
Fluid theory ........................................... Cal 4-1
Schrödinger's Wave Equation .................. Cal 6-2
Schrödinger's Equation ........................ Cal 4-2
The Formulary ....................................... Cal 4-2 Potential Energy &
Schrödinger's Equation ........................... Cal 6-6
∇ in Cartesian Coordinates .................... Cal 4-3
2
∇2 in Spherical Polar Coordinates ....... Cal 4-3 The Hydrogen Atom ................................. Cal 6-6
Newtonian Fluids ...................................... Cal 4-4 Interpretation of Solutions to Schrödinger's
Equation .................................................. Cal 6-9
Viscous Force on a Fluid Element ............ Cal 4-5
Normalization ...................................... Cal 6-10
Viscous Force for 3D Flows ...................... Cal 4-6
The Dirac Equation ................................. Cal 6-12
Viscous Force in
Cylindrical Coordinates ...................... Cal 4-7 Appendix I – Evaluation of a Normalization
Measuring the Viscosity Coefficient ..... Cal 4-9 Integral .................................................. Cal 6-13
Appendix II – Schrödinger's Equation Applied to
APPENDIX THE OPERATOR ∇ 2 IN the Hydrogen Atom ............................... Cal 6-14
The Hydrogen Atom ............................ Cal 6-14
SPHERICAL POLAR COORDINATES The Second Energy Level .................. Cal 6-16
Spherical Polar Coordinates ................... Cal 4-12 Non Spherically Symmetric Solutions . Cal 6-18
Derivatives of r ................................... Cal 4-13
Derivatives of θ .................................. Cal 4-14 CHAPTER 7 DIVERGENCE
Derivatives of φ .................................. Cal 4-14 The Divergence ........................................ Cal 7-2
Summary of Derivatives ...................... Cal 4-16 Electric Field of a Point Charge ............ Cal 7-6
Calculation of ∇2f ............................... Cal 4-16
The δ Function .......................................... Cal 7-8
CHAPTER 5 INTRODUCTION TO Divergence Free Fields .......................... Cal 7-10
COMPLEX VARIABLES Appendix – Flux Equation Derivation ..... Cal 7-11

CHAPTER 8 CURL
A Road Map .............................................. Cal 5-1
ABOUT THE CURL ................................... Cal 8-1
Introduction ............................................... Cal 5-1
Introduction to the curl .............................. Cal 8-2
Imaginary Numbers .................................. Cal 5-2
Stokes' Law ............................................... Cal 8-4
Complex Numbers .................................... Cal 5-2
Ampere's Law ........................................... Cal 8-7
Exponential Form of Complex Number .... Cal 5-3
Small Angle Approximations ................ Cal 5-4 Curl of the Magnetic Field of a Wire ....... Cal 8-10
The Complex Conjugate Z* .................. Cal 5-6 Curl in Cylindrical Coordinates ............... Cal 8-11
Differential Equations for R, L, C Circuits . Cal 5-6 Curl of the Magnetic Field of a Wire ....... Cal 8-12
The RC Circuit ....................................... Cal 5-6 Outside the wire .................................. Cal 8-12
An Aside on Labeling Voltages ............ Cal 5-7 Inside the Wire .................................... Cal 8-13
Solving the RC Circuit Equation ........... Cal 5-8
The LC Circuit ....................................... Cal 5-8 CHAPTER 9 ELECTROMAGNETIC WAVES
A Faster Way to Find Real Solutions ...... Cal 5-10 Vector Identities ........................................ Cal 9-2
The RLC Circuit ...................................... Cal 5-11 Derivation of the Wave Equation .............. Cal 9-4
The Easy Way ..................................... Cal 5-12 Plane Wave Solution ................................. Cal 9-6
Impedance .............................................. Cal 5-15 Three Dimensional Wave Equation ......... Cal 9-7
Impedance Formulas .......................... Cal 5-18
Appendix – Order of partial differentiation Cal 9-8
The Driven RLC Circuit ....................... Cal 5-19
Transients ............................................... Cal 5-22 CHAPTER 10 CONSERVATION OF
Particular Solution ............................... Cal 5-22 ELECTRIC CHARGE
Transient Solutions .............................. Cal 5-22 The Continuity Equation.......................... Cal 10-2
Combined Solutions ............................ Cal 5-23
From Maxwell's Equations ...................... Cal 10-2
Solutions of the 1D Wave Equation ........ Cal 5-24 Integral Form of Continuity Equation .. Cal 10-3
 Calculus 2000 - Preface & Table of Contents Cal - v

CHAPTER 11 SCALAR AND VECTOR CHAPTER 13 INTRODUCTION TO

POTENTIALS FLUID DYNAMICS
The Vector Potential ................................ Cal 11-2 The Navier-Stokes Equation ................... Cal 13-2
Wave Equations for φ and A ................. Cal 11-4 Rate of Change of Momentum ............ Cal 13-2
Summary ............................................. Cal 11-6 Einstein Summation Convention ......... Cal 13-5
Mass Continuity Equation ................... Cal 13-5
CHAPTER 12 VORTICITY Rate of Change of Momentum
DIVERGENCE FREE FIELDS .................. Cal 12-2 when Mass is Conserved ................. Cal 13-6
Newton's Second Law ........................ Cal 13-6
The Vorticity Field ................................... Cal 12-2
Bernoulli's Equation ................................ Cal 13-8
Potential Flow .......................................... Cal 12-3 Applies Along a Streamline ................ Cal 13-9
Examples of Potential Flow ................. Cal 12-4 The Viscosity Term ............................ Cal 13-10
Potential Flow in a Sealed Container .. Cal 12-4
The Helmholtz Theorem ........................ Cal 13-11
Potential Flow in a Straight Pipe ......... Cal 12-5
Equation for Vorticity ......................... Cal 13-11
Superfluids .............................................. Cal 12-6 Non Potential Forces ......................... Cal 13-11
Vorticity as a Source of Fluid Motion ...... Cal 12-7 A Vector Identity for a Moving Circuit... Cal 13-12
Picturing Vorticity ................................ Cal 12-8 The Integral Form of the
Solid Body Rotation ................................ Cal 12-9 Vortex Dynamics Equation ............. Cal 13-15
Vortex Core ....................................... Cal 12-10 The Helmholtz Theorem .................... Cal 13-16
Stokes’ Law Revisited ........................... Cal 12-11 Extended Helmholtz Theorem .............. Cal 13-16
Total Circulation and The Rayfield-Reif Experiment ........... Cal 13-16
Density of Circulation ..................... Cal 12-11 Motion of Charged Vortex Rings ...... Cal 13-18
Velocity Field of a Rotating Shaft ...... Cal 12-12 Conservation of Energy .................... Cal 13-19
Wheel on Fixed Axle ......................... Cal 12-12 Measurement of the
A Conservation Law for Vorticity ....... Cal 12-13 Quantized Circulation κ = h / mHe ... Cal 13-20
Circulation of a Vortex .......................... Cal 12-14 The Magnus Equation ....................... Cal 13-20
Quantum Vortices ................................. Cal 12-15 Impulse of a Vortex Ring ...................... Cal 13-23
Circulation of a Quantum Vortex ....... Cal 12-15 The Airplane Wing ................................ Cal 13-24
Rotating Bucket of Superfluid HeliumCal 12-16 The Magnus Lift Force ...................... Cal 13-25
Bose-Einstein Condensates ............. Cal 12-18 The Magnus Force & Fluid Vortices . Cal 13-26
The Vorticity Field ................................ Cal 12-18
CH 13 APP 1 COMPONENT NOTATION
Helmholtz Theorem ............................... Cal 12-19
The Two Dimensional “Vortex Ring” . Cal 12-19 The Summation Convention............... Cal 13 A1-2
The Circular Vortex Ring ................... Cal 12-20 The Dot Product and δij .................... Cal 13 A1-2
Smoke Rings ..................................... Cal 12-20 The Cross Product and ε ijk .............. Cal 13 A1-3
Creating the Smoke Ring .................. Cal 12-21 Handling Multiple Cross Products . Cal 13 A1-5
Proof of the ε Identity .................... Cal 13 A1-6

CH 13 APP 2 VORTEX CURRENTS

Conserved Two
Dimensional Currents ....................... Cal 13 A2-2
Continuity Equation for Vorticity ........ Cal 13 A2-3
A Single Vortex Line .......................... Cal 13 A2-4
Center of Mass Motion ................... Cal 13 A2-5
Magnus Formula for Curved Vortices Cal 13 A2-7
Creation of Vorticity ........................... Cal 13 A2-9
Energy Dissipation in Fluid Flow ..... Cal 13 A2-10
Cal - vi Calculus 2000 - Preface & Table of Contents

FORMULARY
Cylindrical Coordinates ........................... Formulary-2
Divergence .......................................... Formulary-2
Gradient ............................................... Formulary-2
Curl ....................................................... Formulary-2
Laplacian ............................................. Formulary-2
Laplacian of a vector ........................... Formulary-2
Components of (A⋅∇)B ....................... Formulary-2
Spherical Polar Coordinates .................... Formulary-3
Divergence .......................................... Formulary-3
Gradient ............................................... Formulary-3
Curl ....................................................... Formulary.-3
Laplacian ............................................. Formulary-3
Laplacian of a vector ........................... Formulary-3
Components of (A⋅∇)B ....................... Formulary-3
Vector Identities ....................................... Formulary-4
Integral Formulas ..................................... Formulary-5
Working With Cross products .................. Formulary-6
The cross product ................................ Formulary-6
Product of ε 's ...................................... Formulary-6
Example of use .................................... Formulary-6
Tensor Formulas ...................................... Formulary-7
Definition .............................................. Formulary-7
Formulas .............................................. Formulary-7
Div. of Tensor (Cylindrical Coord.) ...... Formulary-7
Div. of Tensor (Spherical Coord.) ....... Formulary-7
Short Table of Integrals ........................... Formulary-8
Series Expansions ................................... Formulary-9
The binomial expansion ....................... Formulary-9
Taylor series expansion ....................... Formulary-9
Sine and cosine ................................... Formulary-9
Exponential .......................................... Formulary-9

INDEX

BACK COVER
Physical Constants in CGS Units.........Back cover-1
Conversion Factors ..............................Back cover-1
 Calculus 2000 - Index Calculus only Index-1

Besier curves Cal 2-6

Index Binomial expansion Cal 1-23
Derivation of Cal 2-6, Form.-9
Boat lofting Cal 2-5
Bohr atom
Symbols Quantum vortices Cal 12-15
Bohr radius Cal 6-7
∇× , del cross, curl chapter on Cal 8-1 Born, Max
Interpretation of solutions to Schrödinger's Eq. Cal
* ∇⋅ , divergence, chapter on Cal 7-1 6-9
∇ 2 , del squared, chapter on Cal 4-1 C
δ ij and ε ijk , appendix on Cal 13 A1-1 Calculation of integrals Cal 1-11
Calculus
∇f , gradient operation, chapter on Cal 3-1 And the uncertainty principle Cal 1-2
Calculating integrals Cal 1-11
A Calculus in physics Cal 1-1
Acceleration Chain rule Cal 1-25
Calculus definition of Cal 1-7 Definition of acceleration Cal 1-7
Component equations Cal 1-8 Component equations Cal 1-8
Vector equation Cal 1-7 Vector equation Cal 1-7
Constant acceleration formulas Definition of velocity Cal 1-5
Calculus derivation Cal 1-20 Component equations Cal 1-8
In three dimensions Cal 1-22 Vector equation Cal 1-6
Adiabatic expansion Derivation of constant acc. formulas Cal 1-20
Sound waves Cal 2-18 In three dimensions Cal 1-22
Adobe Illustrator™ Cal 2-6 Limiting process Cal 1-2, Cal 1-5
Airplane wing Vector equation for Cal 1-5
Magnus equation Cal 13-24 Special chapter on Cal 1-1
Magnus lift force Cal 13-24 Calculus in Physics i, Cal 2-1, Cal 3-1, Cal 4-1, Cal
Wing vortex Cal 13-24 5-1, Cal 6-1, Cal 7-1, Cal 8-1, Cal 9-1, Cal 10-
Allowed standing wave patterns, hydrogen Cal 6-7 1, Cal 11-1, Form.-1
Ampere's law Capacitance
In differential form Cal 8-3 Dimensions of Front cover-2
Derivation of Cal 8-7 Cartesian coordinates
Amplitude Del squared Cal 4-3
Of a sine wave Cal 1-37 Gradient Cal 3-12
Angular momentum Right hand rule Cal 3-12
Quantum vortices as giant Bohr atom Cal 12-15 Unit vectors Cal 3-12
Schrödinger's equation solutions Cal 6-18 Center of mass
Area Vortex line motion Cal 13 A2-4
Related to integration Cal 1-11 Center of mass motion
Under the curve Cal 1-12 Conserved vortex current Cal 13 A2-5
Atoms Vortex line Cal 13 A2-5
Standing wave patterns in hydrogen Cal 6-8 CGS units Back cover-1
Chain rule Cal 1-25
B Proving it (almost) Cal 1-26
Bernoulli’s equation Remembering it Cal 1-25
Applies along a streamline Cal 13-9 Charge
Derivation from Navier-Stokes equation Cal 13-8 Electric
For potential flow Cal 13-9 Conservation due to Maxwell's equations Cal 10-1
Gradient of hydrodynamic voltage Cal 13-9 Continuity equation for Cal 10-2
Hydrodynamic voltage Cal 13-9 Choice of gauge Cal 11-4
Index-2 Calculus 2000 - Index Calculus only

Circuits Complex variables

Driven LRC circuit Cal 5-19 Chapter on Cal 5-1
Resonance in Cal 5-21 Complex numbers Cal 5-2
Impedance Cal 5-15 Exponential function
Impedance formulas Cal 5-18 Series expansion Cal 5-4
LC circuit Fast way to find real solutions Cal 5-10
Ringing like a bell Cal 5-11 Imaginary numbers Cal 5-2
RC circuit Impedance Cal 5-15
Solving with complex numbers Cal 5-8 Driven LRC circuit Cal 5-19
RLC circuit Formulas for Cal 5-18
Decaying oscillation Cal 5-10 Why Schrödinger's equation is complex Cal 6-5
Differential equation for Cal 5-11 Component notation
Solution using complex variables Cal 5-12 Appendix on Cal 13 A1-1
Transient solutions Cal 5-22 Components, vector Cal 1-7
Circulation In rotated coordinate system Cal 3-26
Density of Cal 12-11 Compressible fluids
For roller bearings Cal 12-13 Continuity equation
For vortex sheet Cal 12-13
Integral form Cal 10-3
For wheel on fixed axle Cal 12-13
Compressional waves on spring Cal 2-15
Flux of vorticity in flow tube Cal 12-18
Computers
Measurement of quantized value Cal 13-20
Why they are so good at integration Cal 1-12
Of a quantum vortex Cal 12-15
Conjugate, complex Cal 5-6
Of a Vortex Cal 12-14
Conservation law
Unchanged by local non potential forces Cal 13-17
Total Cal 12-11 For vorticity Cal 12-13
Circulation, total Related field Cal 10-3
Of rotating shaft Cal 12-12 Conservation of
Classical theory of electromagnetism Cal 11-3 Electric charge
Chapter on Cal 10-1
Coaxial cable
Conservation of charge
Example of gradient in Cyl. Coord. Cal 3-21
Related to electric fields Cal 10-3
Voltage in Cal 3-21
Conservation of energy
Coefficient of viscosity Cal 4-4
In vortex ring motion Cal 13-19
Measuring Cal 4-9
Related to gravity Cal 10-3
Experimental formula for Cal 4-11
Second viscosity coefficient Cal 4-6 Conservative field
Complex analysis Fluid flow Cal 12-3
Of driven LRC circuit Cal 5-19 Conservative force
Complex conjugate Cal 5-6 And Faraday's law Cal 11-2
Of Schrödinger wave function Cal 6-9 Conserved current
Complex numbers Cal 5-2 Vortex, intuitive discussion of Cal 13 A2-2
Analogy to coordinate vector Cal 5-2 Constant acceleration formulas
As a complex exponential Cal 5-5 Calculus derivation Cal 1-20
In three dimensions Cal 1-22
Exponential form Cal 5-3
Second derivative Cal 2-9
Plotting Cal 5-2
Constant, integral of Cal 1-13
Real part, imaginary part Cal 5-2
Continuity equation
Solving the LC circuit Cal 5-9
Derivation from Maxwell's equations Cal 10-2
Solving the RLC circuit Cal 5-12
For compressible fluids
Transient solutions Cal 5-22
Differential form Cal 10-2
Integral form Cal 10-3
For electric charge and current Cal 10-2
For flow of mass Cal 13-5
Formula for Cal 13-6
For flow of vorticity Cal 13 A2-1
Derivation of Cal 13 A2-3
Conversion factors Back cover-1
 Calculus 2000 - Index Calculus only Index-3

Coordinate system Curvature, radius of

Cartesian Cal 3-12 Definition Cal 2-4
Del squared Cal 4-3 Rope waves Cal 2-10
Unit vectors Cal 3-12 Second derivative Cal 2-4
Cylindrical Cal 3-14 Curve
Unit vectors Cal 3-14 Area under, integral of Cal 1-12
Spherical Cal 3-16 Besier (Adobe Illustrator) Cal 2-6
Derivation of del squared in Cal 4-12 Slope as derivative Cal 1-30
Unit vectors Cal 3-16
That increases linearly, integral of Cal 1-13
Coordinate system rotated
Velocity, area under Cal 1-12
Components of a vector in Cal 3-26
Curve fitting Cal 2-5
Core
Cylindrical coordinates
Vortex core structure Cal 12-10
Curl in Cal 8-11
Analogy to magnetic field Cal 12-10
Cosine function Curl in solid body rotation Cal 12-9
Amplitude of Cal 1-37 Curl of magnetic field of wire Cal 8-12
As function of complex exponential Cal 5-5 Div, grad, curl, del squared, A dot del B Form.-2
Definition of Cal 1-35 Gradient in Cal 3-14
Radial component Cal 3-14
Derivative of Cal 1-38
Theta component Cal 3-15
Series expansion Cal 5-4 Unit vectors Cal 3-14
Coulomb potential in Schrödinger's equation Cal 6-6 Viscous force in Cal 4-7
Creating a smoke ring Cal 12-21
Creation of vorticity Cal 13 A2-9 D
Cross product De Broglie
Multiple, easy way to handle Cal 13 A1-5 Schrödinger's equation Cal 6-2
Relation to curl Cal 8-2 Debye, on electron waves Cal 6-1
Use of epsilon i,j,k Cal 13 A1-3 Decay
Working with Form.-6 Exponential decay Cal 1-32
Curl Decaying oscillation
Chapter on Cal 8-1 RLC circuit Cal 5-10
Definition of Cal 8-2 Definite integral
In cylindrical coordinates Cal 8-11 Compared to indefinite integrals Cal 1-14
Vorticity calculation Cal 12-9 Defining new functions Cal 1-15
Introduction to Cal 8-2 Introduction to Cal 1-11
Line integral shrunk down Cal 8-1, Cal 8-3 Of velocity Cal 1-11
Of a divergence = 0 Cal 11-1 Process of integrating Cal 1-13
Of magnetic field of wire Cal 8-10 Del
Calculating the curl Cal 8-12 Relation to curl Cal 8-2
Of solid body rotation velocity field Cal 12-10 Del - gradient operator Cal 3-7
Of vector potential Cal 11-4 Del cross; curl
Gauge invarience Cal 11-4
Chapter on Cal 8-1
Of vortex velocity field Cal 12-9
Del squared
Of vorticity
Chapter on Cal 4-1
Viscosity term in Navier-Stokes equation Cal 13-10
Curl and divergence In Cartesian coordinates Cal 4-3
Uniquely determined field Cal 12-2 In spherical polar coordinates
Derivation of Cal 4-12
Curl theorem
Spherical harmonics Cal 6-18
Called Stokes law Cal 8-3 Relation to curl Cal 8-2
Current, vortex Relation to potential flow Cal 12-4
Appendix on Cal 13 A2-1 Schrödinger's equation
Conserved, intuitive discussion of Cal 13 A2-2 Applied to hydrogen atom Cal 6-14
Continuity equation for Cal 13 A2-3 Schrödinger's equation Cal 4-2
Viscous force Cal 4-1
For 3D flows Cal 4-6
In cylindrical coordinates Cal 4-7
Index-4 Calculus 2000 - Index Calculus only

Delta function Dimensions of

Definition of Cal 7-8 Capacitance Front cover-2
In three dimensions Cal 7-8 Electric charge Front cover-2
Used in Gauss' law Cal 7-9 Electric potential Front cover-2
Delta i,j Electric resistance Front cover-2
Handling multiple cross products Cal 13 A1-5 Energy Front cover-2
Used in dot product Cal 13 A1-2 Force Front cover-2
Delta i,j and epsilon i,j,k Frequency Front cover-2
Appendix on Cal 13 A1-1 Inductance Front cover-2
Density of circulation Magnetic field Front cover-2
Stokes' law Cal 12-11 Magnetic flux Front cover-2
Derivative Power Front cover-2
As a limiting process Cal 1-6, Cal 1-18, Cal 1- Pressure Front cover-2
23, Cal 1-28, Cal 1-30 Dirac equation
As the Slope of a Curve Cal 1-30 Discussion of Cal 6-12
Constants come outside Cal 1-24 Divergence
Negative slope Cal 1-31 Chapter on Cal 7-1
Of exponential function e to the x Cal 1-28 Relation to curl Cal 8-2
Of exponential function e to the ax Cal 1-29 Shrinking the surface integral Cal 7-2
Of function x to the n'th power Cal 1-24 Theorem Cal 7-5
Of sine function Cal 1-38 Handling a point charge Cal 7-7
Partial Cal 5-24 Relation to curl Cal 8-3
Second Divergence and curl
Chapter on Cal 2-1 Surface & line integrals shrunken Cal 7-1, Cal 7-2
Constant acceleration formulas Cal 2-9 Uniquely determined field Cal 12-2
Third, boat lofting Cal 2-5 Divergence and gradient compared Cal 7-5
Derivative, partial Divergence free fields Cal 7-10
Order of, appendix on Cal 9-8 Dot product
Derivative, second Cal 2-2 Relation to curl Cal 8-2
Geometrical interpretation Cal 2-3 Use of delta i,j Cal 13 A1-2
Of a sine wave Cal 2-2 Driven LRC circuit Cal 5-19
Differential equation
Fast way to find real solutions Cal 5-10
For LC circuit
Solving with complex numbers Cal 5-8
For LRC circuit
Transient solutions Cal 5-22
For R, L, and C circuits Cal 5-6
Homogenous Cal 5-9
To integral equation Cal 3-4
Differentiation. See also Derivative
Chain rule Cal 1-25
More on Cal 1-23
Differentiation and integration
As inverse operations Cal 1-18
Velocity and position Cal 1-18
Fast way to go back and forth Cal 1-20
Position as integral of velocity Cal 1-20
Velocity as derivative of position Cal 1-20
 Calculus 2000 - Index Calculus only Index-5

E Equation
Einstein Continuity
Summation convention Cal 13-5 For electric charge and current Cal 10-2
Electric and magnetic fields Extended Helmholtz equation Cal 13-15
In terms of scalar and vector potentials Cal 11-3 Magnus Cal 13-21
Electric charge Airplane wing Cal 13-24
Conservation of Maxwell's
Derivation of the wave equation Cal 9-4
Consequence of Maxwell's equations Cal 10-1
Vector identities for Cal 9-2
Continuity equation for Cal 10-2
Navier-Stokes Cal 13-2
Dimensions of Front cover-2
Nonlinear effects Cal 13-7
Electric field Schrödinger's. See Schrödinger wave equation
Gradient of voltage Cal 3-3 Vector
Equation for Cal 3-7
Components with derivatives Cal 1-7
Field of point charge Cal 3-10
Vortex dynamics equation Cal 13-12
Interpretation Cal 3-6
In terms of scalar & vector potentials Cal 11-3 Wave, one dimensional Cal 2-1
General form of Cal 2-14
Of a line charge
Solutions using complex variables Cal 5-24
Using gradient in cylind. coord. Cal 3-19
Wave, relativistic
Of a point charge
Dirac's Cal 6-12
Using gradient in spherical coord. Cal 3-18
For zero rest mass particles Cal 6-2
Wave equation for Particles with rest mass Cal 6-3
With sources Cal 11-6 Schrödinger's Cal 6-3
Electric potential Euler's number e = 2.7183. . . Cal 1-17
Dimensions of Front cover-2 Expansion, binomial Cal 1-23
Plotting experiment Cal 3-2 Derivation of Cal 2-6, Form.-9
Related to fluid flows Cal 12-3 Expansion, series
Electric resistance Exponential function in complex variables Cal 5-4
Dimensions of Front cover-2 Sin and cosine Cal 5-4
Electromagnetic waves Taylor series Cal 2-7
Chapter on wave equation Cal 9-1 Experiments II
Electromagnetism Potential plotting Cal 3-2
Classical theory of Cal 11-3 Exponential decay Cal 1-32
Electron Exponential form complex number Cal 5-3
In Standard model of elementary particles Cal 7-7 Exponential function
Point particle? Cal 7-7 As function of sin and cos Cal 5-5
Energy Derivative of Cal 1-28
Dimensions of Front cover-2 Exponential decay Cal 1-32
Energy levels, hydrogen Indefinite integral of Cal 1-29
Calculation of lowest Cal 6-15 Integral of Cal 1-29
Lowest two from Schrödinger's equation Cal 6-7 Introduction to Cal 1-16
Epsilon i,j,k Inverse of the logarithm Cal 1-16
Use in cross product Cal 13 A1-3 Series expansion Cal 1-28
Handling multiple cross products Cal 13 A1-5 y to the x power Cal 1-16
Epsilon i,j,k and delta i,j Extended Helmholtz's theorem Cal 13-15
Appendix on Cal 13 A1-1 Discussion of Cal 13-16
Index-6 Calculus 2000 - Index Calculus only

F Functions delta i,j and epsilon i,j,k

Fall line. See Gradient: Of voltage: Interpretation Appendix on Cal 13 A1-1
As a field line Cal 3-23 Functions obtained from integration Cal 1-15
Faraday's law Logarithms Cal 1-15
In terms of the vector potential Cal 11-3
Non potential field Cal 12-3
G
Feynman Cal 7-7 Gamma
Quantized vortices Cal 12-16 Speed of sound Cal 2-18
Parabolic surface of rotating helium Cal 12-6 Gauge invariance
Field Choice of vector potential divergence Cal 11-4
Divergence free Cal 7-10 Gauge invariant theory Cal 11-4
Plotting experiment Cal 3-2 Gauss' law
Pressure field Cal 3-1 Derived from differential equation Cal 7-7
Scalar field Cal 3-7 Electric field of point charge
Uniquely determined, conditions for Cal 12-2 Using delta function Cal 7-9
Vector field Geometrical interpretation
Created by gradient Cal 3-1 Of Gradient Cal 3-4, Cal 3-22
Vorticity field Cal 12-18 Equations for Cal 3-25
Field lines Of second derivative Cal 2-3
And contour lines Cal 3-23 Geometry, fractal Cal 3-23
Two dimensional slope Cal 3-24 Gibbs, Willard; gradient notation Cal 3-7
Fluid dynamics Gradient
Introductory chapter on Cal 13-1 A summary of gradient formulas Cal 3-18
Vorticity Cal 12-1 As a vector field Cal 3-28
Fluids Chapter on Cal 3-1
Compressible From a Geometrical Perspective Cal 3-4, Cal 3-22
Equations for Cal 3-25
Continuity equation for Cal 10-3
Laminar flow Cal 4-8 In Cartesian coordinates Cal 3-12
Newtonian, definition of Cal 4-4 In cylindrical coordinates Cal 3-14
Coaxial cable Cal 3-21
Potential flow Cal 12-3
Electric field of line charge Cal 3-19
In a straight pipe Cal 12-5
Radial component Cal 3-14
Zero vorticity Cal 12-3
Theta component Cal 3-15
Solid body rotation Cal 12-9 In spherical coordinates
Viscous force on Cal 4-5 Phi component Cal 3-17
Vorticity as a source of fluid motion Cal 12-7 Theta component Cal 3-17
Flux Of pressure Cal 3-29
Of vorticity in flow tube Cal 12-18 Of voltage Cal 3-3
Rate of change of through moving circuit Cal 13- Field of point charge Cal 3-10
15 Interpretation Cal 3-6
Flux equation, derivation of Cal 7-11 Parallel plate capacitor Cal 3-8
Force Voltage inside conductor Cal 3-9
Conservative forces Operator "del" Cal 3-7
And Faraday's law Cal 11-2 Relation to curl Cal 8-2
Dimensions of Front cover-2 Vector and scalar fields Cal 3-1
Non potential Gradient vector
In Navier-Stokes equation Cal 13-11 In three dimensions Cal 3-28
Viscous Steepest slope Cal 3-25
In cylindrical coordinates Cal 4-7 Transformation of Cal 3-25
In pipe flow Cal 4-7 Gravity
On a fluid element Cal 4-5 Quantum theory of Cal 7-7
Formulary Form.-1 Gyroscope like behavior
Discussion of Cal 4-2 Of vortex line due to non potential force Cal 13-18
Fractal geometry Cal 3-23 Gyroscopes
Frequency Superfluid Cal 12-17
Dimensions of Front cover-2
 Calculus 2000 - Index Calculus only Index-7

H Integral
Half-life As a sum Cal 1-10
In exponential decay Cal 1-33 Calculating them Cal 1-11
Of muons, exponential decay Cal 1-33 Definite, introduction to Cal 1-11
Heisenberg, Werner Cal 1-2 Formula for integrating x to n'th power Cal 1-
Helium, superfluid Cal 12-6 14, Cal 1-27
Helmholtz theorem Indefinite, definition of Cal 1-14
Application to smoke rings Cal 12-21 Of 1/x, the logarithm Cal 1-15
Derivation from Navier-Stokes equation Cal 13-11 Of a constant Cal 1-13
From extended Helmholtz theorem Cal 13-16 Of a curve that increases linearly Cal 1-13
Introduction to Cal 12-19 Of a velocity curve Cal 1-12
Helmholtz theorem extended Of exponential function e to the ax Cal 1-29
Discussion of Cal 13-16 Of the velocity vector Cal 1-10
Including non potential forces Cal 13-15 As area under curve Cal 1-12
Homogeneous solution for RLC equation Cal 5-23 Of x to n'th power
Homogenous differential equation Cal 5-9, Cal 5-23 Indefinite integral Cal 1-27
Hydrodynamic voltage Integral formulas
Gradient of and Bernoulli's equation Cal 13-9 Many of them Form.-5
Hydrogen atom Integral, line
Bohr radius Cal 6-7 Becomes curl for infinitesimal paths Cal 8-3
Schrödinger's equation for Cal 6-6 Integral sign Cal 1-10
Schrödinger's equation solutions Cal 6-7 Integral, surface
Lowest two energy levels Cal 6-7 Shrinking for divergence Cal 7-2
Non spherically symmetric Cal 6-18 Integral to differential equations Cal 3-4
Spherical harmonics Cal 6-18 Integration
Standing wave patterns in Cal 6-8 Equivalent to finding area Cal 1-11
Hydrogen wave patterns Introduction to Cal 1-8
Lowest energy ones Cal 6-8 Introduction to finding areas under curves Cal 1-13
Why computers do it so well Cal 1-12
I Integration and differentiation
Illustrator™, Adobe Cal 2-6 As inverse operations Cal 1-18
Imaginary numbers Cal 5-2 Fast way to go back and forth Cal 1-20
Impedance Cal 5-15 Position as integral of velocity Cal 1-20
Formulas for Cal 5-18 Velocity as derivative of position Cal 1-20
Impulse Integration formulas Cal 1-27
Of a vortex ring Cal 13-23 Intensity
Impulse equation Cal 13-23 Of wave function Cal 6-9
Indefinite integral Interpretation of solutions to Schrödinger's Eq. Cal
Definition of Cal 1-14 6-9
Of exponential function Cal 1-29 Interval, evaluating variables over Cal 1-10
Inductance
Dimensions of Front cover-2
Infinities in the gravitational interaction
String theory Cal 7-7
Instantaneous velocity
And the uncertainty principle Cal 1-2
Calculus definition of Cal 1-5
Index-8 Calculus 2000 - Index Calculus only

L Magnus equation
Laminar flow Cal 4-8, Cal 7-10 Airplane wing Cal 13-24
Landau, Lev Relative motion of vortex line and fluid particles
Superfluid helium Cal 12-6 Cal 13-20
Landau's prediction for Cal 12-6 The equation Cal 13-21
Laplacian Magnus formula
Relation to curl Cal 8-2 Exact for curved vortices Cal 13 A2-1
Laplacian (del squared) Magnus lift force Cal 13-25
Chapter on Cal 4-1 On fluid core vortices - a pseudo force Cal 13-26
Relation to potential flow Cal 12-4 Mass
LC circuit Continuity equation for flow of Cal 13-5
Ringing like a bell Cal 5-11 Maxwell's equations
Leibnitz Cal 1-2 All forms of Cal 11-6
Leptons Conservation of electric charge Cal 10-1
Standard model of elementary particles Cal 7-7 Derivation of the wave equation Cal 9-4
Lifetime In differential form Cal 8-9
Muon, exponential decay Cal 1-32 In terms of scalar and vector potentials Cal 11-3
Light Introducing vector potential into Cal 11-3
Speed of light One dimensional wave equation
From one dimensional wave equation Cal 9-7 Gives speed of light Cal 9-7
Structure of electromagnetic wave Cal 9-1, Cal 9- Plane wave solution Cal 9-6
6 Relativistic wave equation for photons Cal 6-3
Limiting process Cal 1-2 Vector identities for Cal 9-2
Definition of derivative Cal 1-30 Vector potential in Cal 11-2
In calculus Cal 1-5 Measurement limitation
Introduction to derivative Cal 1-6 Due to uncertainty principle Cal 1-2
With strobe photographs Cal 1-3 Measurement of quantized circulation Cal 13-20
Line charge, electric field of Measuring time constant from graph Cal 1-34
Calculated using calculus MKS units Front cover-2
In cylindrical coordinates Cal 3-19 Modulus
Line integral Spring Cal 2-15
Becomes curl for infinitesimal paths Cal 8-3 Momentum of fluid particles
Localized non potential force Navier-Stokes equation Cal 13-2
Effect on vortex motion Cal 13-17 Motion
Lofting, boat Cal 2-5 Of charged vortex rings Cal 13-18
Logarithms Of vortex line, relative directions Cal 13-22
Integral of 1/x Cal 1-15 Moving circuit
Introduction to Cal 1-15 Vector Identity for Cal 13-12
Inverse of exponential function Cal 1-16 Multiple cross products
LRC circuit. See RLC circuit Easy way to handle Cal 13 A1-5
LRC circuit, ringing like a bell Cal 5-11 Muon
In Standard model of elementary particles Cal 7-7
M Lifetime, exponential decay Cal 1-32
Magnetic and electric fields
In terms of scalar and vector potentials Cal 11-3
Magnetic field
Analogous to vorticity in fluids Cal 12-7
Of a straight wire
Calculating curl of Cal 8-12
Curl of Cal 8-10
Wave equation for
With sources Cal 11-6
Magnetic flux
Dimensions of Front cover-2
 Calculus 2000 - Index Calculus only Index-9

N P
Navier-Stokes equation Cal 13-2 Parabolic profile, pipe flow Cal 4-8
As starting point for fluid theory Cal 13-7 Parabolic surface, rotating fluid
Bernoulli equation derivation Cal 13-8 Superfluid helium Cal 12-6
Derivation of Helmholtz's theorem from Cal 13-11 Telescope mirror Cal 12-6
Final equation! Cal 13-7 Parallel plate capacitor
Momentum of fluid particles Cal 13-2 Example of voltage gradient Cal 3-8
Newton’s second law for fluids Cal 13-2 Partial derivative Cal 5-24
Non potential forces in Cal 13-11 Order of
Nonlinear equation Cal 13-7 Appendix on Cal 9-8
Rate of change of momentum Cal 13-2 Partial derivative operator Cal 8-2
Role of viscosity Cal 13-7 Particular solution, driven RLC circuit Cal 5-22
Viscosity term in Cal 13-10 Perpendicular components of flow Cal 12-2
Curl of vorticity Cal 13-10 Phi component
Negative slope Cal 1-31 Gradient in spherical coordinates Cal 3-17
Neutrinos Photons
In Standard model of elementary particles Cal 7-7 Relativistic wave equation for Cal 6-3
New functions, obtained from integration Cal 1-15 Physical constants
Newtonian Fluids In CGS units Back cover-1
Definition of Cal 4-4 In MKS units Front cover-2
Newton’s laws Pipe flow
Second law Calculating viscous forces Cal 4-7
For fluids, the Navier-Stokes equation Cal 13-2 Measuring viscosity coefficient Cal 4-9
Non potential field Cal 12-3 Parabolic profile Cal 4-8
Non potential forces Potential flow in Cal 12-5
In extended Helmholtz theorem Cal 13-15 Pressure force Cal 4-9
In Navier-Stokes equation Cal 13-11 Viscous force formula Cal 4-8
Localized Plane, tangent Cal 3-23
Causing sideways motion Cal 13-17 Plane wave
Rayfield-Reif experiment Cal 13-16 Discussion of Cal 9-6
Nonlinear equation Solution for Maxwell's equations Cal 9-6
Navier-Stokes equation Cal 13-7 Plotting
Normalization of wave function Cal 6-10 Experiment, electric potential Cal 3-2
Plywood model. See Gradient: Of voltage: Interpre-
O tation
One dimensional wave equation Cal 2-1, Cal 2-14 Point charge
Maxwell's equations Divergence theorem Cal 7-7
Gives speed of light Cal 9-7
Quantum electrodynamics Cal 7-7
Solutions using complex variables Cal 5-24, Cal
Point particles
5-25
Delta function Cal 7-8
Order of partial derivative Cal 9-8
Problems with gravity theory Cal 7-7
Oscillation
Standard model Cal 7-7
Decaying Cal 5-10
Postscript™ language Cal 2-6
Index-10 Calculus 2000 - Index Calculus only

Potential, magnetic Q
Wave equation for Cal 11-4 Quantized angular momentum
Potential, electric In hydrogen wave patterns Cal 6-8
Wave equation for Cal 11-4 Quantized circulation
Potential energy Measurement of Cal 13-20
Electric potential energy Quantized vortex ring
Electric field as gradient of Cal 3-4 Rayfield-Reif experiment Cal 13-16
Schrödinger's Equation Cal 6-6 Quantum electrodynamics
Potential flow Feynman, Schwinger, and Tomonaga Cal 7-7
And the Laplacian (del squared) Cal 12-4 Point charges Cal 7-7
Bernoulli’s equation in Cal 13-9 Quantum mechanics
Definition of Cal 12-3 Concept of velocity Cal 1-4
Examples of Quantum theory
In a sealed container Cal 12-4 Vector potential needed in Cal 11-3
In a straight pipe Cal 12-5
Quantum theory of gravity Cal 7-7
Superfluids Cal 12-6
Quantum vortices Cal 12-15
Zero curl, no vorticity Cal 12-3
Core of Cal 12-16
Power
Giant Bohr atom Cal 12-15
Dimensions of Front cover-2
Number in rotating bucket Cal 12-16
Power series. See Series expansions
Quarks Cal 7-7
Powers of 10, names of Front cover-2
In Standard model of elementary particles Cal 7-7
Prediction of motion
Using calculus Cal 1-9 R
Pressure Radial component
Dimensions of Front cover-2 Gradient in cylindrical coordinates Cal 3-14
Pressure field Cal 3-1 Radian measure Cal 1-35
Pressure force Radians to degrees Cal 5-4
As gradient of pressure Cal 3-29 Radius of curvature
In pipe flow Cal 4-9 Definition Cal 2-4
Per unit volume Cal 3-30 Second derivative Cal 2-4
Probability wave, Schrödinger's Equation Cal 6-9 Rate of change of momentum
Projectile motion Of fluid particles, Navier-Stokes equation
And the uncertainty principle Cal 1-4 When mass is conserved Cal 13-6
Calculus definition of velocity Cal 1-5 Rayfield-Reif experiment Cal 13-16
Pulse Creation of vorticity Cal 13 A2-9
Formation of wave pulse Cal 2-14 Motion of charged vortex rings Cal 13-18
RC circuit
Differential equation for Cal 5-6
Solving with complex numbers Cal 5-8
Labeling voltages Cal 5-7
Real part of complex number Cal 5-2
Relativistic physics
Electromagnetic radiation, structure of Cal 9-
1, Cal 9-6
Relativistic wave equation
For zero rest mass particles Cal 6-2
Particles with rest mass Cal 6-3
Schrödinger's Cal 6-3
Resonance
In driven RLC circuits Cal 5-21
Rest mass
Non zero
Relativistic wave equation for Cal 6-3
Zero
Relativistic wave equation for Cal 6-2
 Calculus 2000 - Index Calculus only Index-11

Right-hand rule Schrödinger's equation

For Cartesian coordinates Cal 3-12 Allowed standing wave patterns, hydrogen Cal 6-
RLC circuit 7
Decaying oscillation Cal 5-10 Angular momentum in solutions Cal 6-18
Differential equation for Cal 5-11 Applied to the hydrogen atom Cal 6-14
Driven LRC circuit Cal 5-19 Bohr radius Cal 6-7
Impedance Cal 5-15 Calculation of lowest energy level Cal 6-15
Formulas for Cal 5-18 Chapter on Cal 6-1
Labeling voltages Cal 5-7 Complex conjugate of wave function Cal 6-9
Solution using complex variables Cal 5-12 Coulomb potential Cal 6-6
Transient solutions Cal 5-22 Del squared in Cal 4-2
Roller bearings Felix Block story on Cal 6-1
For wheel on fixed axle, Stokes' law Cal 12-13 For hydrogen atom Cal 6-6
Rope Full three dimensional form Cal 6-6
Wave equation for Cal 2-10 Hydrogen atom solution Cal 6-6, Cal 6-7
Rotated coordinate system Ideas that led to it Cal 6-2
Components of a vector in Cal 3-26 Intensity of wave function Cal 6-9, Cal 6-10
Rotating bucket of superfluid helium Interpretation of solutions Cal 6-9
Quantized vortices in Cal 12-16 Lowest two energy levels Cal 6-7
Rotating shaft Non spherically symmetric solutions Cal 6-18
Total circulation of Cal 12-12 Spherical harmonics Cal 6-18
Velocity field of Cal 12-12 Normalization of wave function Cal 6-10
Potential energy in Cal 6-6
S Probability interpretation Cal 6-9
Scalar and vector potentials Second energy level Cal 6-16
Chapter on Cal 11-1 Solutions of definite energy Cal 6-14
Scalar field Cal 3-7 Solved for hydrogen atom Cal 6-14
Gradient gives vector field Cal 3-29 Why it is complex Cal 6-5
Pressure Cal 3-29 Schrödinger's relativistic wave equation Cal 6-3
Scalar potential Two solutions Cal 6-4
And the electric field Cal 11-3 Schwinger Cal 7-7
Relation to vector potential Cal 11-2 Second derivative Cal 2-2
Wave equation for Cal 11-4 Constant acceleration formulas Cal 2-9
Coulomb gauge Cal 11-6 Geometrical interpretation Cal 2-3
Gauge invariant form Cal 11-4 Of a sine wave Cal 2-2
Wave gauge Cal 11-5
Radius of curvature Cal 2-4
Schrödinger, Erwin Cal 6-1
Second energy level, hydrogen Cal 6-16
Second viscosity coefficient Cal 4-6
Series expansions Cal 1-23
Binomial Cal 1-23
Exponential function
Complex variables Cal 5-4
Exponential function e to the x Cal 1-28
Sine and cosine Cal 5-4
Taylor Cal 2-7
Sideways motion of vortex line
Caused by localized non potential force Cal 13-17
Sine function
Amplitude of Cal 1-37
Definition of Cal 1-35, Cal 1-36
Derivative of, derivation Cal 1-38
Series expansion Cal 5-4
Sine waves
As function of complex exponential Cal 5-5
Second derivative Cal 2-2
Traveling wave Cal 2-14
Index-12 Calculus 2000 - Index Calculus only

Single vortex line Cal 13 A2-4 Spline fitting Cal 2-5

Singly connected surface, Stokes' law Cal 12-13 Spring
Slinky™ Wave equation for Cal 2-17
Compressional wave on Cal 2-15 Speed of wave Cal 2-17
Slope of a curve Spring modulus Cal 2-15
And contour maps Cal 3-23 Stability of smoke rings Cal 12-21
As derivative Cal 1-30, Cal 3-23 Standard model of elementary particles Cal 7-7
Formula for Cal 1-30 Leptons Cal 7-7
In two dimensions Cal 3-23 Electrons Cal 7-7
Negative slope Cal 1-31 Muons Cal 7-7
Steepest slope, gradient vector Cal 3-25 Neutrinos Cal 7-7
Tau particle Cal 7-7
Smoke rings Cal 12-20
Quarks Cal 7-7
Approaching each other Cal 12-21
Standing waves
Creating Cal 12-21
Patterns in hydrogen Cal 6-8
Role of viscosity in Cal 12-21
From Schrödinger's equation Cal 6-7
Prediction of Helmholtz’s theorem Cal 12-21
Wave equation Cal 2-14
Stability of Cal 12-21
Stokes' law
Titanium tetrachloride for Cal 12-20
Applied to wheel on fixed axle Cal 12-13
Soap film analogy
Convert line to surface integral Cal 8-4
Stokes' law Cal 8-7
Derivation of Cal 8-4
Solar neutrinos. See Neutrinos
Final result Cal 8-6
Solid body rotation Cal 12-9
Introduction to Cal 8-3
Curl of velocity field Cal 12-10
Revisited Cal 12-11
Sound
Roller bearings Cal 12-13
Speed, formula for Cal 2-20
Soap film analogy Cal 8-7
Speed of air molecules Cal 2-21
Total circulation and density of circulation Cal 12-
Speed of, calculating Cal 2-17
11
Wave equation for Cal 2-17, Cal 2-20
Strain, definition of
Adiabatic expansion Cal 2-18
Sound waves, speed of Wave equation Cal 2-15
Formula for Cal 2-21 Streamlines
Source of fields, conserved Cal 10-3 Bernoulli’s equation
Applies along a streamline, derivation of Cal 13-9
Source terms for wave equations Cal 11-6
Stress Cal 4-4
Speed of
Viscous Cal 4-4
Air molecules Cal 2-21
String theory Cal 7-7
Sound, formula for Cal 2-20
Vortex current tensor Cal 13 A2-1
Vortex rings
Strobe photographs
Circular rings Cal 12-20
Two dimensional rings Cal 12-19 And the uncertainty principle Cal 1-2
Wave pulses Substantive derivative Cal 13-2
On rope, calculus derivation Cal 2-13 Summation
Waves Becoming an integral Cal 1-10
One dimensional wave equation Cal 5-24 Of velocity vectors Cal 1-10
Spherical coordinates Summation convention, Einstein's Cal 13-5, Cal 13
Derivation of del squared in Cal 4-12 A1-2
Div, grad, curl, del squared, A dot del B Form.-3 Superfluid gyroscope Cal 12-17
Gradient in Cal 3-16 Superfluids
Phi component Cal 3-17 Potential flow in Cal 12-6
Theta component Cal 3-17 Superfluid helium Cal 12-6
Schrödinger's equation Cal 6-6 Feynman's prediction for Cal 12-6, Cal 12-16
Unit vectors Cal 3-16 Landau's prediction for Cal 12-6
Derivative of changing unit vectors Cal 4-12 Surface integral
Spherical harmonics Cal 6-18 Shrinking for divergence Cal 7-2
 Calculus 2000 - Index Calculus only Index-13

T V
Tangent line Cal 3-23 Variables
Tangent plane Cal 3-23 Evaluated over interval Cal 1-10
Tau particle Vector
In Standard model of elementary particles Cal 7-7 Components in rotated coordinate system Cal 3-
Taylor series expansion Cal 2-7 26
Constant acceleration formulas Cal 2-9 Definition of acceleration Cal 1-7
Tensor formulas Form.-7 Component equations Cal 1-8
Theta component Definition of velocity Cal 1-6
Gradient Component equations Cal 1-8
In cylindrical coordinates Cal 3-15 Dot product
In spherical coordinates Cal 3-17 Summation convention, Einstein's Cal 13-5
Third derivative, boat lofting Cal 2-5 Equations
Time constant Components with derivatives Cal 1-7
Measuring from a graph Cal 1-34 Gradient vector
Titanium tetrachloride for smoke rings Cal 12-20 Transformation of Cal 3-25
Tomonaga Cal 7-7 Vector fields
Total circulation As gradient of scalar field Cal 3-29
Of a vortex Cal 12-14 Created by gradient Cal 3-1
Of rotating shaft Cal 12-12 Gradient as a vector field Cal 3-28
Stokes' law and density of circulation Cal 12-11 Vector identities Cal 8-3
Total circulation and density of circulation For a moving circuit Cal 13-12
Stokes' law Cal 12-11 Rate of change of flux through Cal 13-15
Transients Cal 5-22 For use with Maxwell's equations Cal 9-2
Particular solution Cal 5-22 Many of them Form.-4
Transient solution Cal 5-23 Vector potential
Traveling wave And the electric field Cal 11-3
Sine wave Cal 2-14 Chapter on Cal 11-1
Tritton, fluid dynamics text Cal 4-9 Curl and divergence of Cal 11-4
Turbulence Cal 7-10 Divergent and solenoidal parts Cal 11-4
Gauge invarience Cal 11-4
Two dimensional conserved current
In Faraday's law Cal 11-3
Vortex, intuitive discussion of Cal 13 A2-2
Introducing into Maxwell's equations Cal 11-3
Two dimensional vortex ring Cal 12-19
Introduction to Cal 11-2
Speed of Cal 12-19
Needed in quantum theory Cal 11-3
U Unneeded in classical electromagnetism Cal 11-3
Uncertainty principle Cal 1-2 Wave equation for Cal 11-4
And definition of velocity Cal 1-2 Coulomb gauge Cal 11-6
Gauge invariant form Cal 11-5
And strobe photographs Cal 1-2
Wave gauge Cal 11-5
Applied to projectile motion Cal 1-4 Velocity
Uniquely determined field And the uncertainty principle Cal 1-4
Conditions for Cal 12-2 Calculus definition of Cal 1-5
Unit vectors Component equations Cal 1-8
Cylindrical coordinate system Cal 3-14 Curve, area under Cal 1-12
Derivative of changing unit vectors Cal 4-12 Definite integral of Cal 1-11
Spherical coordinate system Cal 3-16 Integral of Cal 1-10
Units Velocity field
CGS Derived from a potential Cal 12-3
Centimeter, gram, second Back cover-1 Of a rotating shaft
MKS Solid body rotation Cal 12-9
Meter, kilogram, second Front cover-2 Stokes' law applied to Cal 12-12
Index-14 Calculus 2000 - Index Calculus only

Viscosity Vortex line

Coefficient of Cal 4-4 Center of mass motion Cal 13 A2-4
Experimental formula for Cal 4-11 Gyroscope like behavior Cal 13-18
Measuring Cal 4-9 Magnus equation
None in superfluids Cal 12-6 Airplane wing Cal 13-24
Role in Navier-Stokes equation Cal 13-7 Relative motion of line and fluid Cal 13-20
Second viscosity coefficient Cal 4-6 Magnus formula for Cal 13 A2-7
Term in Navier-Stokes equation Cal 13-10 Magnus lift force Cal 13-25
Viscous stress Cal 4-4 On fluid core vortices - a pseudo force Cal 13-26
Viscous force Measurement of quantized circulation of Cal 13-20
Del squared Cal 4-1 Relative motion with fluid particles Cal 13-16
In cylindrical coordinates Cal 4-7 Magnus equation Cal 13-20
In pipe flow Cal 4-7 Sideways motion
Viscous force formula Cal 4-8 Caused by localized non potential force Cal 13-17
On a fluid element Cal 4-5 Single Cal 13 A2-4
For 1D flows Cal 4-5 Small unit flux tube of vorticity Cal 13-16
For 3D flows Cal 4-6 Vortex motion
Where it acts in a vortex Cal 13-10 Effect of localized non potential force Cal 13-17
Voltage Vortex ring
Electric field as gradient of Cal 3-4 Creation of Cal 13 A2-9
Gradient of Impulse of Cal 13-23
Equation for electric field Cal 3-7 Impulse equation Cal 13-23
Field of point charge Cal 3-10 Ring does not carry linear momentum Cal 13-23
Inside a conductor Cal 3-9 Motion of
Interpretation Cal 3-6 Conservation of energy in Cal 13-19
Parallel plate capacitor Cal 3-8 Motion of charged vortex ring Cal 13-18
Impedance formulas for Cal 5-18 Push forward to slow down Cal 13-19
In coaxial cable Cal 3-21 Rayfield-Reif experiment Cal 13-16
Voltage in circuit Smoke ring Cal 12-20
Driven LRC circuit Cal 5-19 Tubes of vorticity Cal 12-18
Transient solutions Cal 5-22 Two dimensional Cal 12-19
In R,L,C circuits Vortex sheet Cal 12-13
Differential Equations for Cal 5-6 Vortices
Using impedance formulas Cal 5-16
Quantized, in superfluids
Vortex
Circulation of Cal 12-15
Core, structure of Core of Cal 12-16
Analogy to magnetic field Cal 12-10 Number in rotating bucket Cal 12-16
Curl of velocity field Cal 12-9 Vortex sheet Cal 12-13
Quantum Cal 12-15 Vorticity
Where viscous forces act Cal 13-10 Analogous to magnetic field Cal 12-7
Vortex currents As a source of fluid motion Cal 12-7
Appendix on Cal 13 A2-1 Chapter on Cal 12-1
Center of mass motion Cal 13 A2-5 Conservation law for Cal 12-13
Conserved, intuitive discussion of Cal 13 A2-2 Continuity equation for Cal 13 A2-3
Continuity equation for Cal 13 A2-3 Creation of Cal 13 A2-9
Vortex dynamics Equation for Cal 13-12
Two ways to handle Cal 13 A2-1 In solid body rotation Cal 12-10
Vortex dynamics equation Cal 13-12 Vortex core structure Cal 12-10
Analogy to magnetic field Cal 12-10
Zero in potential flow Cal 12-3
Vorticity field
Divergence of Cal 12-18
Flux in flow tube of vorticity Cal 12-18
 Calculus 2000 - Index Calculus only Index-15

W Wave motion
Wave Sinusoidal, 1D Cal 5-24
Compressional wave on spring Cal 2-15 Wave patterns
Pulse, formation of Cal 2-14 Hydrogen
Rope Cal 2-10 Standing waves Cal 6-8
Speed of waves Wheel on fixed axle Cal 12-12
Calculus derivation, rope Cal 2-13 Stokes' law applied to Cal 12-13
One dimensional wave equation Cal 5-24 Wing vortex Cal 13-24
Spring compressional wave Cal 2-15
Wave equation Cal 2-1
Addition of waves Cal 2-14
Dirac's Cal 6-12
Electric field
With sources Cal 11-6
For scalar and vector potentials Cal 11-4
For waves on a spring Cal 2-17
Introduction to Cal 2-10
Magnetic field
With sources Cal 11-6
Maxwell's
Derivation of Cal 9-4
Plane wave solution Cal 9-6
One dimensional Cal 2-1
General form of Cal 2-14
Solutions summary Cal 5-25
Solutions using complex variables Cal 5-24
Relativistic
For zero rest mass particles Cal 6-2
Particles with rest mass Cal 6-3
Rope Cal 2-10
Scalar potential
Coulomb gauge Cal 11-6
Gauge invariant form Cal 11-4
Wave gauge Cal 11-5
Schrödinger's
Chapter on Cal 6-1
Discovery of Cal 6-1
Full three dimensional form Cal 6-6
Ideas that led to it Cal 6-2
Relativistic wave equation Cal 6-3
Sound Cal 2-17
The equation for Cal 2-20
Standing wave Cal 2-14
Three dimensional Cal 9-7
Vector potential
Coulomb gauge Cal 11-6
Gauge invariant form Cal 11-5
Wave gauge Cal 11-5
Index-16 Calculus 2000 - Index Calculus only

X X.-Cal Ch 6
X.-Cal Ch 1 Exercise 1 Cal 6-3
Exercise 1 Cal 1-14 Exercise 2 Cal 6-5
Exercise 2 Cal 1-15 Exercise 3 Cal 6-7
Exercise 3 Cal 1-17 Exercise 4 Cal 6-11
Exercise 4 Cal 1-22 Exercise 5 Cal 6-15
Exercise 5 Cal 1-24 Exercise 6 Cal 6-16
Exercise 6 Cal 1-29 X.-Cal Ch 7
Exercise 7 Cal 1-29 Exercise 1 Cal 7-4
Exercise 8 Cal 1-31 Exercise 2 Cal 7-6
Exercise 9 Cal 1-33 Exercise 3 Cal 7-9
Exercise 10 Cal 1-36 Exercise 4 Cal 7-9
Exercise 11 Cal 1-39 X.-Cal Ch 8
Exercise 12 Cal 1-39 Exercise 1 Cal 8-5
Exercise 13 Cal 1-39 Exercise 2 Cal 8-9
X.-Cal Ch 2 X.-Cal Ch 9
Exercise 1 Cal 2-7 Exercise 1 Cal 9-8
Exercise 2 Cal 2-8 X.-Cal Ch 11
Exercise 3 Cal 2-8 Exercise 1 Cal 11-6
Exercise 4 Cal 2-9 Exercise 2 Cal 11-6
X.-Cal Ch 3 X.-Cal Ch 12
Exercise 1 Cal 3-12 Exercise 1 Cal 12-9
Exercise 2 Cal 3-15 Exercise 2 Cal 12-10
Exercise 3 Cal 3-16 Exercise 3 Cal 12-17
Exercise 4 Cal 3-21 X.-Cal Ch 13
X.-Cal Ch 3 view 2 Exercise 1 Cal 13-18
Exercise 1 Cal 3-26 Exercise 2 Cal 13-18
Exercise 2 Cal 3-27 Exercise 3 Cal 13-19
X.-Cal Ch 4 Exercise 4 Cal 13-20
Exercise 1 Cal 4-11
Exercise 2 Cal 4-17
Z
Exercise 3 Cal 4-17 Zero rest mass particles
X.-Cal Ch 5 Relativistic wave equation for Cal 6-2
Photons Cal 6-3
Exercise 1 Cal 5-5
Exercise 2 Cal 5-10
Exercise 3 Cal 5-12
Exercise 4 Cal 5-14
 Calculus 2000 - Chapter 1 Introduction to Calculus Cal 1-1

Calculus Chapter 1
Introduction to Calculus
This first chapter covers all the calculus that is needed
for the Physics 2000 text. The remaining chapters allow
students to look at the physics from an advanced
mathematical point of view.

CHAPTER 1 INTRODUCTION TO
CALCULUS

This chapter, which replaces Chapter 4 in Physics

2000, is intended for students who have not had
calculus, or as a calculus review for those whose
calculus is not well remembered. If, after reading
part way through this chapter, you feel your calculus
background is not so bad after all, go back to
Chapter 4 in Physics 2000, study the derivation of
the constant acceleration formulas beginning on
page 4-8, and work the projectile motion problems
in the appendix to Chapter 4. Those who study all of
this introduction to calculus should then proceed to
the projectile motion problems in the appendix to
Chapter 4 of the Physics text.
Cal 1-2 Calculus 2000 - Chapter 1 Introduction to Calculus

LIMITING PROCESS THE UNCERTAINTY PRINCIPLE

In Chapter 3 of Physics 2000, we used strobe photo- For over 200 years, from the invention of calculus by
graphs to define velocity and acceleration vectors. Newton and Leibnitz until 1924, the limiting pro-
The basic approach was to turn up the strobe flashing cess and the resulting concept of instantaneous ve-
rate, as we did in going from Figure (3-3) to (3-4) locity was one of the cornerstones of physics. Then
shown below. We turned the rate up until all the in 1924 Werner Heisenberg discovered what he
kinks are clearly visible and the successive displace- called the uncertainty principle which places a limit
ment vectors give a reasonable description of the on the accuracy of experimental measurements.
motion. We did not turn the flashing rate too high,
for the practical reason that the displacement vectors Heisenberg discovered something very new and
became too short for accurate work. unexpected. He found that the act of making an
experimental measurement unavoidably affects the
In our discussion of instantaneous velocity we con- results of an experiment. This had not been known
ceptually turned the strobe all the way up as illus- previously because the effect on large objects like
trated in Figures (2-22a) through (2-22d), redrawn golf balls is undetectable. But on an atomic scale
here in Figure (1). In these figures, we initially see where we study small systems like electrons moving
a fairly large change in v0 as the strobe rate is inside an atom, the effect is not only observable, it
increased and ∆t reduced. But then the change can dominate our study of the system.
becomes smaller, and it looks as if we are approach-
ing some final value of v0 that does not depend on One particular consequence of the uncertainly prin-
the size of ∆t , provided ∆t is small enough. It looks ciple is that the more accurately we measure the
as if we have come close to the final value in Figure position of an object, the more we disturb the motion
(1c). of the object. This has an immediate impact on the
concept of instantaneous velocity. If we turn the
The progression seen in Figure (1) is called a limit- strobe all the way up, reduce ∆t to zero, we are in
ing process. The idea is that there really is some true effect trying to measure the position of the object
value of v0 which we have called the instantaneous with infinite precision. The consequence would be
velocity, and that we approach this true value for an infinitely big disturbance of the motion of the
sufficiently small values of ∆t . This is a calculus object we are studying. If we actually could turn the
concept, and in the language of calculus, we are strobe all the way up, we would destroy the object
taking the limit as ∆ t goes to zero. we were trying to study.

Figures 3-3 and 3-4 from Physics 2000

Strobe photographs of a moving object. In the first photograph,
the time between flashes is so long that the motion is difficult to
understand. In the second, the time between flashes was reduced
and the motion is more easily understood.
 Calculus 2000 - Chapter 1 Introduction to Calculus Cal 1-3

0 0 1
V0
V0

∆t = 0.4 Sec ∆t = 0.1 Sec

(a) (c)

V0 V
~i
01

∆t = 0.025 Sec instantaneous velocity

(b) (d)

Figure 1
Transition to instantaneous velocity. As we reduce ∆ t ,
there is less and less change in the vector V 0 . It looks
as if we are approaching an exact final value.
Cal 1-4 Calculus 2000 - Chapter 1 Introduction to Calculus

Uncertainty Principle on a Larger Scale The uncertainty principle has had a significant im-
It turns out that the uncertainty principle can have a pact on the way physicists think about motion.
significant impact on a larger scale of distance than Because we now know that the measuring process
the atomic scale. Suppose, for example, we con- affects the results of the measurement, we see that it
structed a chamber that is 1 cm on each side, and is essential to provide experimental definitions to
wished to study the projectile motion of an electron any physical quantity we wish to study. A concep-
inside. Using Galileo’s idea that objects of different tual definition, like turning the strobe all the way up
mass fall at the same rate, we would expect that the to define instantaneous velocity, can lead to funda-
motion of the electron projectile should be the same mental inconsistencies.
as more massive objects. If we took a strobe photo-
graph of the electron’s motion, we would expect to Even an experimental definition like our strobe
get results like those shown in Figure (2). This definition of velocity can lead to inconsistent results
figure represents projectile motion with an accelera- when applied to something like the electron in
tion g = 980 cm/sec2 and ∆t = .01sec, as the reader Figure (2). But these inconsistencies are real. Their
can easily check. existence is telling us that the very concept of
velocity is beginning to lose meaning for these small
When we study the uncertainty principle in Chapter objects.
40 of the Physics text, we will see that a measure-
ment which is accurate enough to show that position On the other hand, the idea of the limiting process
(2) is below position (1), could disturb the electron and instantaneous velocity is very convenient when
enough to reverse its direction of motion. The next applied to larger objects where the effects of the
position measurement could find the electron over uncertainty principle are not detectable. In this case
where we drew position (3), or back where we drew we can apply all the mathematical tools of calculus
position (0), or anywhere in the region in between. developed over the past 250 years. The status of
As a result we could not even determine what instantaneous velocity has changed from a basic
direction the electron is moving. This uncertainty concept to a useful mathematical tool. Those prob-
would not be the result of a sloppy experiment, it is lems for which this mathematical tool works are
the best we can do with the most accurate and called problems in classical physics; those problems
delicate measurements possible. for which the uncertainty principle is important, are
in the realm of what we call quantum physics.

0 1
-1 2
v1
3
1 centimeter

1 centimeter
Figure 2
Hypothetical electron projectile motion experiment.
The uncertainty principle tells us that such an
experiment cannot lead to predictable results.
 Calculus 2000 - Chapter 1 Introduction to Calculus Cal 1-5

CALCULUS DEFINITION OF VELOCITY The velocity vector vi is now given by

With the above perspective on the physical limita- ∆R i
vi = (2)
tions of the limiting process, we can now return to ∆t
the main topic of this chapter—the use of calculus in This is just our old strobe definition vi = s i /∆t , but
defining and working with velocity and accelera- using a notation which emphasizes that the displace-
tion. ment s i = ∆R i is the change in position that occurs
during the time ∆t . The Greek letter ∆ (delta) is
In discussing the limiting process in calculus, one
used both to represent the idea that the quantity ∆R i
traditionally uses a special set of symbols which we
or ∆t is small, and to emphasize that both of these
can understand if we adopt the notation shown in
quantities change as we change the strobe rate.
Figure (3). In that figure we have drawn the coordi-
nate vectors R i and R i+1 for the i th and (i + 1) The limiting process in Figure (1) can be written in
positions of the object. We are now using the the form
symbol ∆R i to represent the displacement of the ball
during the i to i+1 interval. The vector equation for limit ∆R i
vi ≡ ∆t→0
∆t (3)
∆R i is
where the word “limit” with ∆t→0 underneath, is to
∆R i = R i+1 – R i (1) be read as “limit as ∆t goes to zero”. For example
In words, Equation (1) tells us that ∆R i is the we would read Equation (3) as “the instantaneous
change, during the time ∆t, of the position vector R velocity vi at position i is the limit, as ∆t goes to
describing the location of the ball. zero, of the ratio ∆R i /∆t . ”

i For two reasons, Equation (3) is not quite yet in

∆R i standard calculus notation. One is that in calculus,
i +1 only the limiting value, in this case, the instanta-
neous velocity, is considered to be important. Our
strobe definition vi = ∆R i /∆t is only a step in the
limiting process. Therefore when we see the vector
vi , we should assume that it is the limiting value, and
Ri
no special symbol like the underline is used. For this
R i +1 reason we will drop the underline and write
limit ∆R i
vi = ∆t→0 (3a)
∆t

∆R i = R i +1 – R i

V i = ∆R i
∆t

Figure 3
Definitions of ∆Ri and vi .
Cal 1-6 Calculus 2000 - Chapter 1 Introduction to Calculus

The second change deals with the fact that when ∆t hand notation for the limiting process we have been
goes to zero we need an infinite number of time steps describing. But to a physicist, there is a different,
to get through our strobe photograph, and thus it is more practical meaning. Think of dt as a short ∆t ,
not possible to locate a position by counting time short enough so that the limiting process has essen-
steps. Instead we measure the time t that has elapsed tially occurred, but not too short to see what is going
since the beginning of the photograph, and use that on. In Figure (1), a value of dt less than .025 seconds
time to tell us where we are, as illustrated in Figure is probably good enough.
(4). Thus instead of using vi to represent the
velocity at position i, we write v(t) to represent the If dt is small but finite, then we know exactly what
velocity at time t. Equation (3) now becomes the dR(t) is. It is the small but finite displacement
vector at the time t. It is our old strobe definition of
limit ∆R(t)
v(t) = ∆t→0 (3b) velocity, with the added condition that dt is such a
∆t short time interval that the limiting process has
where we also replaced ∆R i by its value ∆R(t) at occurred. From this point of view, dt is a real time
time t. interval and dR(t) a real vector, which we can work
with in a normal way. The only thing special about
Although Equation (3b) is in more or less standard these quantities is that when we see the letter d
calculus notation, the notation is clumsy. It is a pain to instead of ∆ , we must remember that a limiting
keep writing the word “limit” with a ∆t→0 under- process is involved. In this notation, the calculus
neath. To streamline the notation, we replace the definition of velocity is
Greek letter ∆ with the English letter d as follows
dR(t)
limit ∆R(t) ≡ dR(t) v(t) =
dt
∆t→0 ∆t dt (5)
(4)
where R(t) and v(t) are the particle’s coordinate vec-
(The symbol ≡ means defined equal to.) To a tor and velocity vector respectively, as shown in Figure
mathematician, the symbol dR(t)/dt is just short- (5). Remember that this is just fancy shorthand nota-
tion for the limiting process we have been describing.
t = .1sec t = .2sec

t = 0sec t = .3sec
V(t)

t = .4sec
R(t) at t = .3 sec R(t)

t = .5sec

Figure 4 Figure 5
Rather than counting individual images, we can Instantaneous position and velocity at time t.
locate a position by measuring the elapsed time t.
In this figure, we have drawn the displacement
vector R(t) at time t = .3 sec.
 Calculus 2000 - Chapter 1 Introduction to Calculus Cal 1-7

ACCELERATION Now go through the limiting process, turning the

In the analysis of strobe photographs, we defined both strobe up, reducing ∆t until the value of a(t) settles
a velocity vector v and an acceleration vector a . The down to its limiting value. We have
definition of a , shown in Figure (2-12) reproduced calculus limit v(t + ∆t) – v(t)
a(t) definition = ∆t→0
here in Figure (6), was ∆t
v –v (9)
a i ≡ i+1 i (6) limit ∆v(t)
∆t = ∆t→0
∆t
In our graphical work we replaced vi by s i /∆t so that
Finally use the shorthand notation d/dt for the
we could work directly with the displacement vectors
limiting process:
s i and experimentally determine the behavior of the
acceleration vector for several kinds of motion. dv(t)
a(t) = (10)
Let us now change this graphical definition of accel- dt
eration over to a calculus definition, using the ideas
Equation (10) does not make sense unless you re-
just applied to the velocity vector. First, assume that
member that it is notation for all the ideas expressed
the ball reached position i at time t as shown in
above. Again, physicists think of dt as a short but
Figure (6). Then we can write
finite time interval, and dv(t) as the small but finite
vi = v(t) change in the velocity vector during the time interval
vi+1 = v(t+∆t) dt. It’s our strobe definition of acceleration with the
added requirement that ∆t is short enough that the
to change the time dependence from a count of limiting process has already occurred.
strobe flashes to the continuous variable t. Next,
define the vector ∆v(t) by
Components
∆v(t) ≡ v(t+∆t) – v(t) = vi+1 – vi (7) Even if you have studied calculus, you may not
recall encountering formulas for the derivatives of
We see that ∆v(t) is the change in the velocity vector vectors, like dR(t)/dt and dv(t)/dt which appear in
as the time advances from t to t+∆t . The strobe Equations (5) and (10). To bring these equations
definition of a i can now be written into a more familiar form where you can apply
strobe v(t + ∆t) – v(t) ∆v(t) standard calculus formulas, we will break the vector
a(t) definition = ≡ (8)
∆t ∆t Equations (5) and (10) down into component equa-
tions.

In the chapter on vectors, we saw that any vector

position at equation like
time t position at
time t + ∆t A = B+C (11)
Vi
is equivalent to the three component equations
( Vi+1–Vi ) A x = Bx + Cx
Vi+1
A y = By + Cy
(12)
–Vi A z = Bz + Cz
a i = ( Vi+1–Vi ) The advantage of the component equations was that
∆t
they are simply numerical equations and no graphi-
cal work or trigonometry is required.

Figure 6
Experimental definition of the acceleration vector.
Cal 1-8 Calculus 2000 - Chapter 1 Introduction to Calculus

The limiting process in calculus does not affect the INTEGRATION

decomposition of a vector into components, thus When we worked with strobe photographs, the pho-
Equation (5) for v(t) and Equation (10) for a(t) tograph told us the position R(t) of the ball as time
become passed. Knowing the position, we can then use
Equation (5) to calculate the ball's velocity v(t) and
v(t) = dR(t)/dt (5)
then Equation (10) to determine the acceleration
vx(t) = dR x(t)/dt (5a) a(t) . In general, however, we want to go the other
vy(t) = dR y(t)/dt (5b) way, and predict the motion from a knowledge of the
acceleration. For example, imagine that you were in
vz(t) = dR z(t)/dt (5c) Galileo's position, hired by a prince to predict the
and motion of cannonballs. You know that a cannonball
a(t) = dv(t)/dt (10) should not be much affected by air resistance, thus
the acceleration throughout its trajectory should be
a x(t) = dvx(t)/dt (10a) the constant gravitational acceleration g . You know
a y(t) = dvy(t)/dt (10b) that a(t) = g . How then do you use that knowledge
a z(t) = dvz(t)/dt (10c) in Equations (5) and (10) to predict the motion of the
ball?
Often we use the letter x for the x coordinate of the
vector R and we use y for R y and z for R z . With this The answer is that you cannot with the equations in
notation, Equation (5) assumes the shorter and perhaps their present form. The equations tell you how to go
more familiar form from R(t) to a(t), while to predict motion you need
y
to go the other way, from a(t) to R(t) . The topic of
vx(t) = dx(t)/dt R (5a’) this section is to see how to reverse the directions in
vy(t) = dy(t)/dt (5b’) which we use our calculus equations. Equations (5)
vz(t) = dz(t)/dt x (5c’) and (10) involve the process called differentiation.
Figure 7
We will see that when we go the other way the
At this point the notation has become deceptively short. reverse of differentiation is a process called integra-
You now have to remember that x(t) stands for the x tion. We will see that integration is a simple con-
coordinate of the particle at a time t. cept, but a process that is sometimes hard to perform
without the aid of a computer.
We have finally boiled the notation down to the point
where it would be familiar in any calculus course. If we
restrict our attention to one dimensional motion along
the x axis, then all we have to concern ourselves with
are the x component equations
dx(t)
vx(t) =
dt
dvx(t) (10a)
a x(t) =
dt
 Calculus 2000 - Chapter 1 Introduction to Calculus Cal 1-9

Prediction of Motion Equation (12) is approximate in that the vi are

In our earlier discussion, we have used strobe pho- approximate (strobe) velocities, not the instanta-
tographs to analyze motion. Let us see what we can neous velocities we want for a calculus discussion.
learn from such a photograph for predicting motion. In Figure (9) we improved the situation by cutting
Figure (8) is our familiar projectile motion photo- ∆t to 1/4 of its previous value, giving us four times
graph showing the displacement s of a ball during as many images and more accurate velocities vi .
the time the ball traveled from a position labeled (0)
to the position labeled (4). If the ball is now at We see that the displacement s is now the sum of 16
position (0) and each of the images is (.1) seconds vectors
apart, then the vector s tells us where the ball will be s = s 1 + s 2 + s 3 + ... + s 15 + s 16 (13)
at a time of (.4) seconds from now. If we can predict
Expressing this in terms of the velocity vectors v1 to
s , we can predict the motion of the ball. The general
v16 we have
problem of predicting the motion of the ball is to be
able to calculate s(t) for any time t. s = v1 ∆t + v2 ∆t + v3 ∆t + ... + v15 ∆t + v16 ∆t (14)

From Figure (8) we see that s is the vector sum of the or using our more compact notation
individual displacement vectors s 1 , s 2 , s 3 and s 4 16
s = Σ vi ∆t
i=1
(14a)
s = s1 + s2 + s3 + s4 (11)
We can then use the fact that s 1 = v1∆t , s 2 = v2∆t , While Equation (14) for s looks quite different than
etc. to get Equation (12)—the sum of sixteen vectors instead
of four—the displacement vectors s in the two cases
s = v1∆t + v2∆t + v3∆t + v4∆t (12) are exactly the same. Adding more intermediate
images did not change where the ball was located at
Rather than writing out each term, we can use the the time of t = .4 seconds. In going from Equation
summation sign Σ to write (12) to (14), what has changed as a result of shorten-
4 ing the time step ∆t , is that the individual velocity
s = Σ vi∆t
i=1
(12a)
vectors vi become more nearly equal to the instan-
taneous velocity of the ball at each image.

0 S1 1 1 2 3
S2 S1 0 4
t=0 2
t=0 8
S3

3
S 12
S
S4
S = S 1 + S 2 + ... + S 16 S 16
S = S 1 + S 2 + ... + S 16
4 16
t=.4 sec t=.4 sec

Figure 8 Figure 9
To predict the total displacement s , we With a shorter time interval, we add up more
add up the individual displacements s i . displacement vectors to get the total displacement s .
Cal 1-10 Calculus 2000 - Chapter 1 Introduction to Calculus

If we reduced ∆t again by another factor of 1/4, so Thus the displacement s has x and y components
that we had 64 images in the interval t = 0 to t = .4 sec, s x = x(t f) – x(t i)
the formula for s would become
s y = y(t f) – y(t i)
64
s = Σ
i=1
vi ∆t (15a) Breaking Equation (17) into component equations
gives
where now the vi are still closer to representing the
ball's instantaneous velocity. The more we reduce tf
∆t , the more images we include, the closer each vi s x = x(t f) – x(t i) =
ti
vx(t)dt (18a)
comes to the instantaneous velocity v(t) . While
adding more images gives us more vectors that we tf
have to add up to get the total displacement s , there s y = y(t f) – y(t i) = vy(t)dt
is very little change in our formula for s . If we had ti (18b)
a million images, we would simply write
1000000
Here we will introduce one more piece of notation
s = Σ
i=1
vi ∆t (16a) often used in calculus courses. On the left hand side
of Equation (18a) we have x(t f) – x(t i) which we
In this case the vi would be physically indistinguish- can think of as the variable x(t) evaluated over the
able from the instantaneous velocity v(t) . We have interval of time from t i to t f . We will often deal with
essentially reached a calculus limit, but we have variables evaluated over some interval and have a
problems with the notation. It is clearly inconve- special notation for that. We will write
nient to label each vi and then count the images. tf
Instead we would like notation that involves the x(t f) – x(t i) ≡ x(t) (19)
instantaneous velocity v(t) and expresses the begin- ti
ning and end points in terms of the initial time t i and
final time t f , rather than the initial and final image You are to read the symbol x(t) ttfi as "x of t evaluated
numbers i. from t i to t f ". We write the initial time t i at the bottom
of the vertical bar, the final time t f at the top.
In the calculus notation, we replace the summation
sign Σ by something that looks almost like the
summation sign, namely the integral sign . (The i
yi
French word for integration is the same as their word
for summation.) Next we replaced the individual vi
by the continuous variable v(t) and finally express
the end points by the initial time t i and the final time
(y – yi )

t f . The result is S
f

n as the number tf
s = Σ vi ∆t →
i=1
n becomes
infinitely
ti
v(t)dt (17)
large

yf f
Calculus notation is more easily handled, or is at
least more familiar, if we break vector equations up
(x f – xi )
into component equations. Assume that the ball
started at position i which has components x i = x(t i) xi xf
[read x(t i) as “x at time t i ”] and y i = y(t i) as shown x(t i ) x(t f)
in Figure (10). The final position f is at x f = x(t f) Figure 10
Breaking the vector s into components.
and y f = y(t f) .
We use similar notation for any kind of variable, for To get this interpretation, let us start with the simple
example case of a ball moving in a straight line, for instance,
x2 the x direction, at a constant velocity vx . A strobe
f(x) ≡ f(x 2) – f(x 1) (19a) picture of this motion would look like that shown in
x1
Figure (11a).
(Remember to subtract when the variable is evalu-
ated at the value at the bottom of the vertical bar.) Figure (11b) is a graph of the ball's velocity vx(t) as
a function of the time t. The vertical axis is the value
With this notation, our Equation (18) can be written of vx , the horizontal axis is the time t. Since the ball
is traveling at constant velocity, vx has a constant
tf tf
s x = x(t) = vx(t)dt (18 a′ ) value and is thus represented by a straight horizontal
ti ti line. In order to calculate the distance that the ball
has traveled during the time interval from t i to t f ,
tf tf we need to evaluate the integral
s y = y(t) = vy(t)dt (18 b′ )
ti ti distance ball
tf travels in
Calculating Integrals sx = vx(t)dt time interval
ti (18a)
Equation (18) is nice and compact, but how do you t i to t f
use it? How do you calculate integrals? The key is To actually evaluate the integral, we will go back to
to remember that an integral is just a fancy notation our summation notation
for a sum of terms, where we make the time step ∆t i final
very small. Keeping this in mind, we will see that s x = Σ vxi ∆t (20)
there is a very easy way to interpret an integral. i initial

ti tf and show individual time steps ∆t in the graph of vx

versus t, as in Figure (11c).

x We see that each term in Equation (20) is repre-

Figure 11a
sented in Figure (11c) by a rectangle whose height is
Strobe photograph of ball moving at vx and whose width is ∆t . We have shaded in the
constant velocity in x direction. rectangle representing the 7th term vx 7∆t . We see
that vx 7∆t is just the area of the shaded rectangle,
vx (t)
and it is clear that the sum of all the areas of the
vx individual rectangles is the total area under the
curve, starting at time t i and ending at time t f . Here
we are beginning to see that the process of integra-
t tion is equivalent to finding the area under a curve.
ti tf
Figure 11b With a simple curve like the constant velocity vx(t)
Graph of vx(t) versus t for the ball of Figure 11a. in Figure (11c), we see by inspection that the total
area from t i to t f is just the area of the complete
vx (t)
rectangle of height vx and width (t f – t i) . Thus
vx7
vx
s x = vx × (t f – t i) (21)
This is the expected result for constant velocity, namely
∆t
t
ti tf distance = velocity × time for
constant (21a)
Figure 11c traveled velocity
Each vx ∆ t is the area of a rectangle.
Cal 1-12 Calculus 2000 - Chapter 1 Introduction to Calculus

To see that you are not restricted to the case of Thus we can interpret the integral of a curve as the
constant velocity, suppose you drove on a freeway area under the curve even when the curve is not
due east (the x direction) starting at 9:00 AM and constant or flat. Mathematicians concern them-
stopping for lunch at 12 noon. Every minute during selves with curves that are so wild that it is difficult
your trip you wrote down the speedometer reading or impossible to determine the area under them.
so that you had an accurate plot of vx(t) for the entire Such curves seldom appear in physics problems.
morning, a plot like that shown in Figure (12). From
such a plot, could you determine the distance s x that While the basic idea of integration is simple—just
you had traveled? finding the area under a curve—in practice it can be
quite difficult to calculate the area. Much of an
Your best answer is to multiply each value vi of your introductory calculus course is devoted to finding
velocity by the time ∆t to calculate the average the formulas for the areas under various curves.
distance traveled each minute. Summing these up There are also books called tables of integrals where
from the initial time t i = 9:00 AM to the final time you look up the formula for a curve and the table tells
t f = noon , you have as your estimate you the formula for the area under that curve.

sx ≈ Σi vxi ∆t In Chapter 16 of the Physics text, we will discuss a

mathematical technique called Fourier analysis.
(The symbol ≈ means approximately equal.) This is a technique in which we can describe the
To get a more accurate value for the distance trav- shape of any continuous curve in terms of a sum of
eled, you should measure your velocity at shorter sine waves. (Why we want to do that will become
time intervals ∆t and add up the larger number of clear then.) The process of Fourier analysis involves
smaller rectangles. The precise answer should be finding the area under some very complex curves,
obtained in the limit as ∆t goes to zero curves often involving experimental data for which
we have no formula, only graphs. Such curves
tf
s x = limit Σ v ∆t
∆t → 0 i xi
=
ti
vx(t)dt (22)
cannot be integrated by using a table of integrals,
with the result that Fourier analysis was not widely
used until the advent of the modern digital computer.
This limit is just the area under the curve that is
supposed to represent the instantaneous velocity vx(t) . The computer made a difference, because we can
find the area under almost any curve by breaking the
vx(t)
curve into short pieces of length ∆t , calculating the
vx7 area vi∆t of each narrow rectangle, and adding up
the area of the rectangles to get the total area. If the
curve is so wild that we have to break it into a million
segments to get an accurate answer, that might be
t
9am ∆t noon too hard to do by hand, but it usually a very simple
and rapid job for a computer. Computers can be
Figure 12
Plot of vx(t) for a trip starting at 9:00 AM and much more efficient than people at integration.
finishing at noon. The distance traveled is the
area under the curve.
 Calculus 2000 - Chapter 1 Introduction to Calculus Cal 1-13

The Process of Integrating Since (t f – t i ) = t t f dt , we can replace (t f – t i ) in

i
There is a language for the process of integration Equation (24) by the integral to get
which we will now take you through. In each case tf tf
we will check that the results are what we would vx dt = vx dt vx a constant (25)
expect from our summation definition, or the idea ti ti

that an integral is the area under a curve. and we see that a constant like vx can be taken
outside the integral sign.
The simplest integral we will encounter is the calcu-
lation of the area under a curve of unit height as Let us try the simplest case we can think of where vx
shown in Figure (13). We have the area of a is not constant. Suppose vx starts at zero at time
rectangle of height 1 and length (t f – t i) t i = 0 and increases linearly according to the formula
tf tf vx = at (26)
1 dt = dt = (t f – t i )
ti ti (22)
vx
at f
1
area = 1(t f – t i) at
vx =
t
ti tf t
0 tf
Figure 13
Figure 15
Area under a curve of unit height.
When we get up to the time t f the velocity will be
We will use some special language to describe this (at f) as shown in Figure (15). The area under the
integration. We will say that the integral of dt is curve vx = at is a triangle whose base is of length t f
simply the time t, and that the integral of dt from t i and height is at f . The area of this triangle is one half
to t f is equal to t evaluated from t i to t f . In symbols the base times the height, thus we get for the distance
this is written as s x traveled by an object moving with this velocity
tf tf
tf
dt = t = (t f – t i ) (23) sx = vx dt = 1 (base) × (height)
ti ti 0 2
Recall that the vertical line after a variable means to (27)
= 1 (t f)(at f) = 1 at f 2
evaluate that variable at the final position t f (upper 2 2
value), minus that variable evaluated at the initial
Now let us repeat the same calculation using the
position t i (lower value). Notice that this prescrip-
language one would find in a calculus book. We have
tion gives the correct answer.
tf tf
The next simplest integral is the integral of a constant, sx = vx dt = (at)dt (28)
0 0
like a constant velocity vx over the interval t i to t f
tf The constant (a) can come outside, and we know that
vx dt = vx (t f – t i ) (24) the answer is 1/2at f 2 , thus we can write
ti
tf
vx sx = a tdt = 1 at f 2 (29)
0 2
area = vx(t f – t i)
In Equation (29) we can cancel the a's to get the result
t tx
ti tf tdt = 1 t f 2 (30)
0 2
Figure 14
Area under the constant vx curve.
Cal 1-14 Calculus 2000 - Chapter 1 Introduction to Calculus

In a calculus text, you would find the statement that t3

n=3
the integral tdt is equal to t 2/2 and that the integral 4
should be evaluated as follows t 3dt = t (33d)
4 t
tf
tf t2 tf2 0 t2 Looking at the way these integrals are turning out,
tdt = = – = f (31) we suspect that the general rule is
0 2 2 2 2
0
n+1
Indefinite Integrals t n dt = t (34)
n+1
When we want to measure an actual area under a
curve, we have to know where to start and stop. It turns out that Equation (34) is a general result for any
When we put these limits on the integral sign, like t i value of n except n = –1. If n = –1, then you would
and t f , we have what is called a definite integral. have division by zero, which cannot be the answer.
However there are times where we just want to know (We will shortly discuss the special case where n = –1.)
what the form of the integral is, with the idea that we
will put in the limits later. In this case we have what As long as we stay away from the n = –1 case, the
is called an indefinite integral, such as formula works for negative numbers. For example
2 t – 2dt = dt = t (– 2 +1) = t –1
tdt = t indefinite integral (32) t2 –2 + 1 (–1)
2
The difference between our definite integral in Equa- dt = – 1
t (35)
tion (31) and the indefinite one in Equation (32) is t2
that we have not chosen the limits yet in Equation
(32). If possible, a table of integrals will give you a In our discussion of gravitational and electrical
formula for the indefinite integral and let you put in potential energy, we will encounter integrals of the
whatever limits you want. form seen in Equation (35).

Exercise 1
Integration Formulas
Using Equation (34) and the fact that constants can
For some sets of curves, there are simple formulas come outside the integral, evaluate the following inte-
for the area under them. One example is the set of grals:
curves of the form t n . We have already considered
the cases where n = 0 and n = 1. it does not matter whether
(a) xdx we call the variable t or x
n=0 1
x=2
t 0dt = dt = t (33a) (b) x5dx also sketch the area
x=1 being evaluated
t

n=1 t t =2
dt show that you get
(c) a positive area
t2 t =1 t2
t 1dt = tdt = (33b)
2
t
Some results we will prove later are (d) GmMdr where G, m, and M
r2 are constants
n=2 t2
3 a dy
t 2dt = t (33c) (e)
y3 / 2
"a" is a constant
3
t
 Calculus 2000 - Chapter 1 Introduction to Calculus Cal 1-15

NEW FUNCTIONS Two of the important but peculiar features of the

We have seen that when we integrate a curve or natural logarithm are
function like t 2 , we get a new function t 3/3 . The
ln(ab) = ln(a) + ln(b) (37)
functions t 2 and t 3 appear to be fairly similar; the
integration did not create something radically differ- ln( 1a ) = – ln(a) (38)
ent. However, the process of integration can lead to
some curves with entirely different behavior. This Thus we get, for example
happens, for example, in that special case n = –1
ln(t f ) – ln(t i ) = ln(t f ) + ln 1
when we try to do the integral of t – 1 . ti
Logarithms tf (39)
It is certainly not hard to plot t – 1 , the result is shown = ln
ti
in Figure (16). Also there is nothing fundamentally
difficult or peculiar about measuring the area under Thus the area under the curve in Figure (16) is
the t – 1 curve from some t i to t f , as long as we stay tf
away from the origin t = 0 where t – 1 blows up. The dt = ln t f
ti t ti (40)
formula for this area turns out, however, to be the
new function called the natural logarithm, abbrevi- While the natural logarithm has some rather peculiar
ated by the symbol ln. The area in Figure (16) is properties it is easy to evaluate because it is available
given by the formula on all scientific calculators. For example, if t i = .5
seconds and t f = 4 seconds, then we have
tf
1 dt = ln(t ) – ln(t ) (36) t
ti t f i
ln tf = ln 4 = ln (8) (41)
i .5
Entering the number 8 on a scientific calculator and
t –1 pressing the button labeled ln, gives
ln (8) = 2.079 (42)
curve 1t which is the answer.

Exercise 2
Evaluate the integrals
t
ti tf 1000
dx 1
dx
Figure 16
–1 .001 x .000001 x
Plot of t . The area under this curve
is the natural logarithm ln. Why are the answers the same?
Cal 1-16 Calculus 2000 - Chapter 1 Introduction to Calculus

The Exponential Function The inverse of the exponent to the base 10 is the
We have just seen that, while the logarithm function function called logarithm to the base 10 which is
may have some peculiar properties, it is easy to denoted by the key labeled log on a scientific calcu-
evaluate using a scientific calculator. The question lator. Formally this means that
we now want to consider is whether there is some
function that undoes the logarithm. When we enter log (10 y ) = y (46)
the number 8 into the calculator and press ln, we get Check this out on your scientific calculator. For
the number 2.079. Now we are asking if, when we example, enter the number 1,000,000 and press the
enter the number 2.079, can we press some key and log button and see if you get the number 6. Try
get back the number 8? The answer is, you press the several examples so that you are confident of the
key labeled e x . The e x key performs the exponen- result.
tial function which undoes the logarithm function.
We say that the exponential function e x is the
inverse of the logarithm function ln. The Exponential Function yx
Another key on your scientific calculator is labeled
Exponents to the Base 10 y x . This allows you to determine the value of any
You are already familiar with exponents to the base number y raised to the power (or exponent) x. For
10, as in the following examples example, enter the number y = 10, and press the y x
key. Then enter the number x = 6 and press the =
10 0 = 1 key. You should see the answer
10 1 = 10 10 – 1 = 1/10 = .1
10 2 = 100 10 – 2 = 1/100 = .01 (43) y x = 10 6 = 1000000
⋅⋅⋅⋅⋅⋅⋅ ⋅⋅⋅⋅⋅⋅⋅
10 6 = 1,000,000 10 – 6 = .000001 It is quite clear that all exponents obey the same rules
we saw for powers of 10, namely
The exponent, the number written above the 10, tells
us how many factors of 10 are involved. A minus ya × yb = ya + b (47)
sign means how many factors of 10 we divide by.
From this alone we deduce the following rules for [Example y 2 × y 3 = (y × y)(y × y × y) = y 5 .]
the exponent to the base 10.
And as before
10 –a = 1 a (44)
10 y – a ≡ 1a (48)
y

10 a × 10 b = 10 a + b (45)

(Example 10 2 × 10 3 = 100 × 1000 = 100,000 .)

 Calculus 2000 - Chapter 1 Introduction to Calculus Cal 1-17

Exercise 3 Euler's Number e = 2.7183. . .

Use your scientific calculator to evaluate the follow- We have seen that the function log on the scientific
ing quantities. (You should get the answers shown.) calculator undoes, is the inverse of, powers of 10.
For example, we saw that
(a) 106 (1000000)
3 (8)
log (10 x ) = x (46) repeated
(b) 2
0
(c) 23 (1) Example: log (10 6 ) = 6
–1
(d) 10 (.1)
Earlier we saw that the exponential function e x was
(To do this calculation, enter 10, then press yx . Then
the inverse of the natural logarithm ln. This means that
enter 1, then press the +/– key to change it to –1, then
press = to get the answer .1) ln(e x ) = x (49)
– .5
(e) 2 (1/ 2= .707)
(f) log (10) (1) The difference between the logarithm log and the
(g) ln (2.7183) (1) (very close to 1) natural logarithm ln, is that log undoes exponents
of the number 10, while ln undoes exponents of the
Try some other examples on your own to become number e. This special number e, one of the
completely familiar with the yx key. (You should note fundamental mathematical constants like π , is known
that any positive number raised to the 0 power is 1. as Euler's number, and is always denoted by the
Also, some calculators, in particular the one I am letter e.
using, cannot handle any negative values of y, not
even (– 2)2 which is +4) You can find the numerical value of Euler's number
e on your calculator by evaluating
e1 = e (50)
To do this, enter 1 into your calculator, press the e x
key, and you should see the result
e 1 = e = 2.718281828 (51)
We will run into this number throughout the course.
You should remember that e is about 2.7, or you might
even remember 2.718. (Only remembering e as 2.7 is
as klutzy as remembering π as 3.1)

The terminology in math courses is that the function

log, which undoes exponents of the number 10, is the
logarithm to the base 10. The function ln, what we
have called the natural logarithm, which undoes
exponents of the number e, is the logarithm to the
base e. You can have logarithms to any base you
want, but in practice we only use base 10 (because
we have 10 fingers) and the base e. The base e is
special, in part because that is the logarithm that
naturally arises when we integrate the function 1/x.
We will see shortly that the functions ln and e x have
several more, very special features.
Cal 1-18 Calculus 2000 - Chapter 1 Introduction to Calculus

DIFFERENTIATION AND INTEGRATION While Equation (53) looks like it is applied to the
The scientific calculator is a good tool for seeing explicit case of the strobe photograph of projectile
how the functions like ln and e x are inverse of each motion, it is easily extended to cover any process of
other. Another example of inverse operations is differentiation. Whatever function we have [we had
integration and differentiation. We have seen that R(t), suppose it is now f(t)], evaluate it at two closely
integration allows us to go the other way from spaced times, subtract the older value from the
differentiation [finding x(t) from v(t), rather than newer one, and divide by the time difference ∆t.
v(t) from x(t)]. However it is not so obvious that Taking the limit as ∆t becomes very small gives us
integration and differentiation are inverse opera- the derivative
tions when you think of integration as finding the
area under a curve, and differentiation as finding d f(t) limit f(t + ∆t) – f(t)
≡ ∆t→0 (54)
limits of ∆x/∆t as ∆t goes to zero. It is time now to dt ∆t
make this relationship clear. The variable with which we are differentiating does
not have to be time t. It can be any variable that we
First, let us review our concept of a derivative. can divide into small segments, such as x
Going back to our strobe photograph of Figure (3),
replacing R i by R(t) and R i+1 by R(t+∆t) , as shown
in Figure (3a), our strobe velocity was then given by d f(x) ≡ limit f(x + ∆x) – f(x) (55)
dx ∆x→ 0 ∆x
R(t+∆t) – R(t)
v(t) = (52)
∆t
Let us see how the operation defined in Equation
The calculus definition of the velocity is obtained by (55) is the inverse of finding the area under a curve.
reducing the strobe time interval ∆t until we obtain
the instantaneous velocity v . Suppose we have a curve, like our old vx(t) graphed
as a function of time, as shown in Figure (17). To
find out how far we traveled in a time interval from
limit R(t + ∆t) – R(t)
vcalculus = ∆t→ (53) t i to some later time T, we would do the integral
0 ∆t
T
x(T) = vx(t) dt (56)
ti
i
∆R = R(t+∆t) – R(t) The integral in Equation (56) tells us how far we
have gone at any time T during the trip. The quantity
i +1 x(T) is a function of this time T.

vx (t)
R(t)

R(t+∆t) x(T)
t
ti T
Figure 17
The distance traveled by the time T is the area under
V(t) = ∆R =
R(t+∆t) – R(t)
∆t ∆t the velocity curve up to the time T.

Figure 3a
Defining the strobe velocity.
 Calculus 2000 - Chapter 1 Introduction to Calculus Cal 1-19

Now let us differentiate the function x(T) with The rectangle has a height approximately vx(T) and
respect to the variable T. By our definition of a width ∆t for an area
differentiation we have
x(T + ∆t) – x(T) = vx(T)∆t (58)
d x(T) = limit x(T + ∆t) – x(T) (57)
dT ∆t→ 0 ∆t Dividing through by ∆t gives

Figure (17) shows us the function x(T). It is the area x(T + ∆t) – x(T)
vx(T) = (59)
under the curve v(t) starting at t i and going up to ∆t
time t = T. Figure (18) shows us the function
The only approximation in Equation (59) is at the top
x(T + ∆t) . It is the area under the same curve,
of the rectangle. If the curve is not flat, vx(T + ∆t)
starting at t i but going up to t = T + ∆t . When we
will be different from vx(T) and the area of the sliver
subtract these two areas, all we have left is the area
will have a value somewhere between vx(T)∆t and
of the slender rectangle shown in Figure (19).
vx(T + ∆t)∆t . But if we take the limit as ∆t goes to
vx (t) zero, the value of vx(T + ∆t) must approach vx(T) ,
and we end up with the exact result

limit x(T + ∆t) – x(T)

vx(T) = ∆t→0 (60)
x(T) ∆t
t This is just the derivative dx(t)/dt evaluated at t = T.
ti T
Figure 17 repeated
dx(t)
The distance x(T) traveled by the time T vx(T) = (61a)
dt t = T
vx (t)
where we started from

T
x(T+∆t) x(T) = vx(t) dt (61b)
ti
t
ti T T+∆t
Figure 18 Equations (61a) and (61b) demonstrate explicitly how
The distance x (T+ ∆ t ) traveled by the time T+ ∆ t .
differentiation and integration are inverse operations.
The derivative allowed us to go from x(t) to vx(t) while
the integral took us from vx(t) to x(t). This inverse is
vx (t) not as simple as pushing a button on a calculator to go
vx (T) from ln to e x . Here we have to deal with limits on the
integration and a shift of variables from t to T. But
vx (T)∆t these two processes do allow us to go back and forth.

t
Figure 19 T T+∆t
The distance x (T+ ∆ t ) – x(T)
traveled during the time ∆ t .
Cal 1-20 Calculus 2000 - Chapter 1 Introduction to Calculus

A Fast Way to go Back and Forth however, we say that we started our trip at x(t i) = 0 ,
We introduced our discussion of integration by then we get the result
pointing out that equations T
x(T) = vx(t)dt (67)
dx(t) dvx(t) ti
vx(t) = ; a x(t) = (62a,b)
dt dt
representing the distance traveled since the start of
went the wrong way in that we were more likely to the trip.
know the acceleration a x(t) and from that want to
calculate the velocity vx(t) and distance traveled
Constant Acceleration Formulas
x(t). After many steps, we found that integration
was what we needed. The constant acceleration formulas, so well known
from high school physics courses, are an excellent
We do not want to repeat all those steps. Instead we application of the procedures we have just described.
would like a quick and simple way to go the other
way around. Here is how you do it. Think of the dt We will begin with motion in one dimension. Sup-
in (62a) as a small but finite time interval. That pose a car is traveling due east, in the x direction, and
means you can treat it like any other number and for a while has a constant acceleration a x . The car
multiply both sides of Equation (62a) through by it. passes us at a time t i = 0 , traveling at a speed vx0 .
At some later time T, if the acceleration a x remains
dx(t) constant, how far away from us will the car be?
vx(t) =
dt
dx(t) = vx(t)dt (63) We start with the equation
Now integrate both sides of Equation (63) from some dvx (t)
a x (t) = (68)
initial time t i to a final time T. (If you do the same thing dt
to both sides of an equation, both sides should still be Multiplying through by dt to get
equal to each other.) dvx(t) = a x(t)dt
T T then integrating from time t i = 0 to time tf = T, we
dx(t) = vx(t) dt (64) get
ti ti
T T
If dt is to be thought of as a small but finite time step, dvx (t) = a x (t)dt (69)
then dx(t) is the small but finite distance we moved in 0 0

the time dt. The integral on the left side of Equation Since the integral dvx (t) = vx (t) , we have
(64) is just the sum of all these short distances moved, T T
which is just the total distance moved during the time dvx (t) = vx (t) = vx (T) – vx (0) (70)
from t i to T. 0 0

T T
where vx (0) is the velocity vx0 of the car when it
dx(t) = x(t) = x(T) – x (ti) (65) passed us at time t = 0.
ti ti
While we can always do the left hand integral in
Thus we end up with the result Equation (69), we cannot do the right hand integral
T T until we know a x (t) . For the constant acceleration
x(t) = vx(t)dt (66) problem, however, we know that a x (t) = a x is
ti ti
constant, and we have
Equation (66) is a little more general than (62b) for T T
it allows for the fact that x(t i) might not be zero. If, a x (t)dt = a x dt (71)
0 0
 Calculus 2000 - Chapter 1 Introduction to Calculus Cal 1-21

Since constants can come outside the integral sign, One of the results of integration that you should
we get prove for yourself (just sketch the areas) is the rule
T T T f f f
a x dt = a x dt = a x t = axT (72) a(x) + b(x) dx = a(x)dx + b(x)dx (78)
0 0 0 i i i
where we used dt = t . Substituting Equations (70) thus we get
and (72) in (69) gives
T T T
vx T – vx0 = a x T (73) (vx0 + a xt)dt = vx0 dt + a xt dt (79)
0 0 0

Since Equation (73) applies for any time T, we can

Since constants can come outside the integrals, this
replace T by t to get the well known result
is equal to
vx(t) = vx0 + a xt (a x constant) (74) T T T
(vx0 + a xt)dt = vx0 dt + a x t dt (80)
0 0 0
Equation (74) tells us the speed of the car at any time Earlier we saw that
t after it passed us, as long as the acceleration
T T
remains constant. dt = t = T–0 = T (23)
0 0
To find out how far away the car is, we start with the
equation T
T t2 T2 T2
dx(t) tdt = = –0 = (30)
vx(t) = (62a) 0 2 2 2
dt 0
Multiplying through by dt to get Thus we get
dx(t) = vx(t) dt T 1
(vx0 + a xt)dt = vx0T + a xT2 (81)
then integrating from time t = 0 to time t = T gives 0 2
(as we saw earlier)
Using Equations (76) and (81) in (75) gives
T T
dx(t) = vx(t) dt (75) 1
0 0 x(T) – x 0 = vx0T + a xT2
2
The left hand side is Taking x 0 = 0 and replacing T by t gives the other
T T constant acceleration formula
dx(t) = x(t) = x(T) – x(0) (76)
0 0 1
x(t) = vx0t + a xt2 (a x constant) (82)
If we measure along the x axis, starting from where 2
we are (where the car was at t = 0) then x(0) = 0.
You can now see that the factor of t 2/2 in the constant
In order to do the right hand integral in Equation acceleration formulas comes from the integral tdt .
(75), we have to know what the function vx(t) is. But
for constant acceleration, we have from Equation
(74) vx(t) = vx0 + a xt , thus
T T
vx(t) dt = (vx 0 + a xt) dt (77)
0 0
Cal 1-22 Calculus 2000 - Chapter 1 Introduction to Calculus

Exercise 4 Then repeat, for each pair of equations, the steps that
Find the formula for the velocity v(t) and position x(t)
led to the constant acceleration formulas for motion
for a car moving with constant acceleration ax , that in the x direction. The results will be
was located at position xi at some initial time ti .
x(t) = vx0t + 1 a xt 2 vx(t) = vx0 + a xt
Start your calculation from the equations
2

dx(t)
y(t) = vy0t + 1 a yt 2 vy(t) = vy0 + a yt (84)
vx(t) = 2
dt
dvx(t) z(t) = vz0t + 1 a zt 2 vz(t) = vz0 + a zt
ax(t) = 2
dt
The final step is to combine these six equations into
and go through all the steps that we did to get the two vector equations
Equations (74) and (82). See if you can do this without
looking at the text.
x(t) = v0t + 1 at 2 ; v(t) = v0 + at (85)
2
If you have to look back to see what some steps are,
then finish the derivation looking at the text. Then a These are the equations we analyzed graphically in
day or so later, clean off your desk, get out a blank Chapter 3 of the Physics text, in Figure (3-34) and
sheet of paper, write down this problem, put the book Exercise (3-9). (There we wrote s instead of x(t) ,
away and do the derivation. Keep doing this until you and vi rather than v0 .)
can do the derivation of the constant acceleration
formulas without looking at the text. In many introductory physics courses, considerable
emphasis is placed on solving constant acceleration
problems. You can spend weeks practicing on
Constant Acceleration Formulas
solving these problems, and become very good at it.
in Three Dimensions
However, when you have done this, you have not
To handle the case of motion with constant accelera- learned very much physics because most forms of
tion in three dimensions, you start with the separate motion are not with constant acceleration, and thus
equations the formulas do not apply. The formulas were
important historically, for they were the first to
dx(t) dvx(t)
vx(t) = a x(t) = allow the accurate prediction of motion (of cannon-
dt dt balls). But if too much emphasis is placed on these
dy(t) dvy(t) problems, students tend to use them where they do
vy(t) = a y(t) = (83) not apply. For this reason we have placed the
dt dt
exercises using the constant acceleration equations
dz(t) dvz(t)
vz(t) = a z(t) = in an appendix at the end of Chapter 4 of the Physics
dt dt
text. There are plenty of problems there for all the
practice you will need with these equations. Doing
these exercises requires only algebra, there is no
practice with calculus. To get some experience with
calculus, be sure that you can confidently do Exer-
cise 4.
 Calculus 2000 - Chapter 1 Introduction to Calculus Cal 1-23

MORE ON DIFFERENTIATION When α is a number much smaller than 1 (α < < 1) ,

In our discussion of integration, we saw that the we can neglect α 2 compared to α (if α = .01,
basic idea was that the integral of some curve or α 2 = .0001 ), with the result that we can accurately
function f(t) was equal to the area under that curve. approximate (x + α) n by
That is an easy enough concept. The problems arose
(x + α) n ≈ x n + nαx n–1 α << 1 (87)
when we actually tried to find the formulas for the
areas under various curves. The only areas we Equation (87) gives us all the approximation formu-
actually calculated were the rectangular area under las found in Equations (1-20) through (1-25) on page
f(t) = constant and the triangular area under f(t) = at. 1-28 of the Physics text.
It was perhaps a surprise that the area under the
simple curve 1/t should turn out to be a logarithm. As an example of Equation (87), just to see that it
works, let us take x = 5, n = 7 and α = .01 to calculate
For differentiation, the basic idea of the process is (5.01) 7 . From the calculator we get
given by the formula
(5.01) 7 = 79225.3344 (88)
df(t) f(t + ∆t) – f(t)
= limit (54) repeated
dt ∆t→0 ∆t (To do this enter 5.01, press the y x button, then enter
7 and press the = button.) Let us now see how this
Equation (54) is short hand notation for a whole
result compares with
series of steps which we introduced through the use
of strobe photographs. The basic idea of differentia- (x + α) n ≈ x n + nαx n – 1
tion is more complex than integration, but, as we will (89)
now see, it is often a lot easier to find the derivative (5 + .01) 7 ≈ 5 7 + 7(.01)5 6
of a curve than its integral. We have
5 7 = 78125 (90)
Series Expansions
An easy way to find the formula for the derivative of 7 × .01 × 5 6 = 7 × .01 × 15625 = 1093.75 (91)
a curve is to use a series expansion. We will Adding the numbers in (90) and (91) together gives
illustrate the process by using the binomial expan-
sion to calculate the derivative of the function x n 5 7 + 7(.01)5 6 = 79218.75 (92)
where n is any constant.
Thus we end up with 79218 instead of 79225, which
We used the binomial expansion, or at least the first is not too bad a result. The smaller α is compared
two terms, in Chapter 1 of the Physics text. That was to one, the better the approximation.
during our discussion of the approximation formu-
las that are useful in relativistic calculations. As we
mentioned in Exercise (1-5), the binomial expan-
sion is
n(n – 1) 2 n – 2
(x + α) n = x n + nαx n – 1 + α x ⋅⋅⋅
2!
(86)
Cal 1-24 Calculus 2000 - Chapter 1 Introduction to Calculus

Derivative of the Function x n In our discussion of integration, we saw that a

constant could come outside the integral. The same
We are now ready to use our approximation formula
thing happens with a derivative. Consider, for
(87) to calculate the derivative of the function x n .
example,
From the definition of the derivative we have
d a f(x + ∆x) – a f(x)
limit (x + ∆x) – x
n n
d(x n) a f(x) = limit
= (93) dx ∆x→0 ∆x
dx ∆x→0 ∆x
Since ∆x is to become infinitesimally small, we can Since the constant a has nothing to do with the
use our approximation formula for ( x + α )n . We limiting process, this can be written
get
d f(x + ∆x) – f(x)
( x + α )n ≈ x n + n(α)x n–1 af(x) = a limit
∆x→0
(α << 1) dx ∆x
( x + ∆x) n ≈ x n + n(∆x)x n–1 (∆x << 1) (94) (99)
df(x)
= a
Using this in Equation (93) gives dx
Exercise 5
d(x n) [x n + n(∆x)x n–1 ] – x n
= limit (95) Calculate the derivative with respect to x (i.e., d/dx)
dx ∆x→0 ∆x of the following functions. (When negative powers of
We used an equal sign rather than an approximately x are involved, assume x is not equal to zero.)
equal sign in Equation (95) because our approxima- (a) x
tion formula (94) becomes exact when ∆x becomes
infinitesimally small. (b) x2

In Equation (95), the terms x n cancel and we are left (c) x3

with
(d) 5x2 – 3x
n n–1
d(x ) limit n(∆x)x (Before you do part (d), use the definition of the
= (96)
dx ∆x→0 ∆x derivative to prove that d f(x) + g(x) =
df(x) dg(x)
+ )
dx dx dx
At this point, the factors ∆x cancel and we have (e) x– 1
d(x n )
= limit
∆x→0
nx n–1 (97) (f) x– 2
dx
Since no ∆x's remain in our formula, we end up with (g) x
the exact result
(h) 1/ x
d(x n)
= nx n–1 (98)
dx (i) 3x.73

Equation (98) is the general formula for the deriva- (j) 7x – .2

tive of the function x n .
(k) 1

(In part (k) first show that this should be zero from the
definition of the derivative. Then write 1 = x0 and
show that Equation (98) also works, as long as x is not
zero.)

(l) 5
 Calculus 2000 - Chapter 1 Introduction to Calculus Cal 1-25

The Chain Rule Using (104) and (105) in the chain rule (100) gives
There is a simple trick called the chain rule that df(y) dy
makes it easy to differentiate a wide variety of = df × = (ny n–1 ) × ( 2x)
dx dy dx
functions. The rule is
= 2ny n–1x
df y(x) df(y) dy
= chain rule (100) = 2n(x 2 ) n–1x
dx dy dx

To see how this rule works, consider the function = 2n(x 2[n–1])x
(107)
f(x) = ( x 2 ) n (101) = 2n(x [2n–2])x
We know that this is just f(x) = x 2n , and the deriva- = 2n(x [2n– 2] + 1 )
tive is
df(x) = 2nx 2n– 1
= d (x 2n ) = 2nx 2n– 1 (102) which is the answer we expect.
dx dx
But suppose that we did not know this trick, and
In our example, using the chain rule was more
therefore did not know how to differentiate (x 2) n .
difficult than differentiating directly because we
We do, however, know how to differentiate powers
already knew how to differentiate x 2n . But we will
like x 2 and y n . The chain rule allows us to use this
shortly encounter examples of new functions that
knowledge in order to figure out how to differentiate
we do not know how to differentiate directly, but
the more complex function (x 2) n .
which can be written in the form f[y(x)], where we
We begin by defining y(x) as know df/dy and dy/dx. We can then use the chain
rule to evaluate the derivative df/dx. We will give
y(x) = x 2 (103) you practice with the chain rule when we encounter
these functions.
Then our function f(x) = (x 2) n can be written in
terms of y as follows
f(x) = (x 2) n = [ y(x) ] n = (y) n = f(y) Remembering the Chain Rule
The chain rule can be remembered by thinking of the
f(y) = (y)n (104) dy's as cancelling as shown.
Differentiating (103) and (104) gives
df(y) dy df(y) remembering
dy(x) = (108)
= d (x 2 ) = 2x (105) dy dx dx the chain rule
dx dx
df(y)
= d (y n ) = ny n–1 (106)
dy dy
Cal 1-26 Calculus 2000 - Chapter 1 Introduction to Calculus

Partial Proof of the Chain Rule (optional) (We call this a partial proof for the following reason.
The proof of the chain rule is closely related to For some functions y(x), the quantity
cancellation we showed in Equation (108). A ∆y = y(x + ∆x) – y(x) may be identically zero for
partial proof of the rule proceeds as follows. a small range of ∆x . In that case we would be
dividing by zero (the 1/∆y ) even before we took the
Suppose we have some function f(y) where y is a limit as ∆x goes to zero. A more complete proof
function of the variable x. As a result f[y(x)] is itself handles the special cases separately. The resulting
a function of x and can be differentiated with respect chain rule still works however, even for these special
to x. cases.)
d f y(x) = limit f y(x + ∆x) – f y(x) (123) Since ∆y = y(x + ∆x) – y(x) goes to zero as ∆x
dx ∆x→0 ∆x goes to zero, we can write Equation (127) as
Now define the quantity ∆y by d f y(x)
dx
∆y ≡ y(x + ∆x) – y(x) (124)
so that limit f(y + ∆y) – f(y)
= ∆y→0
∆y
y(x + ∆x) = y(x) + ∆y
f[ y(x + ∆x)] = f(y + ∆y) limit y(x + ∆x) – y(x)
× ∆x→0
∆x
and Equation (123) becomes
df(y) dy
d f y(x) = limit f(y + ∆y) – f(y) (125)
=
dy dx
(100) repeated
dx ∆x→0 ∆x
Now multiply (125) through by This rule works as long as the derivatives df/dy and
∆y y(x + ∆x) – y(x) dy/dx are meaningful, i.e., we stay away from kinks
1 = = (126)
∆y ∆y or discontinuities in f and y.
to get

d
f y(x)
dx

limit f(y + ∆y) – f(y) y(x + ∆x) – y(x)

= ×
∆x→0 ∆x ∆y

limit f(y + ∆y) – f(y) y(x + ∆x) – y(x)

= ×
∆x→0 ∆y ∆x

(127)
where we interchanged ∆x and ∆y in the denomina-
tor.
 Calculus 2000 - Chapter 1 Introduction to Calculus Cal 1-27

INTEGRATION FORMULAS Dividing through by (n+1) gives

Knowing the formula for the derivative of the func- T T
tion x n , and knowing that integration undoes differ- t ndt = 1 t n+1 (133)
entiation, we can now use Equation (98) ti n+1 ti

dx n = nx n – 1 (98) repeated
dx If we choose t i = 0 , we get the simpler result
to find the integral of the function x n . We will see T n+1
that this trick works for all cases except the special t n dt = T (134)
0 n+1
case where n = –1, i.e., the special case where the
integral is a natural logarithm. and the indefinite integral can be written
To integrate x n, let us go back to our calculation of n+1
the distance s x or x(t) traveled by an object moving t n dt = t (135) (also 34)
n+1
in the x direction at a velocity vx . This was given by
Equations (19) or (56) as This is the general rule we stated without proof back
T T in Equation (34). Note that this formula says noth-
x(t) = vx(t) dt (128) ing about the case n = –1, i.e., when we integrate
ti ti
t – 1 = 1/t , because n +1 = –1 +1 = 0 and we end up
where the instantaneous velocity vx(t) is defined as with division by zero. But for all other values of n,
dx(t) we now have derived a general formula for finding
vx(t) = (129)
dt the area under any curve of the form x n (or t n ). This
Suppose x(t) had the special form is a rather powerful result considering the problems
one encounters actually finding areas under curves.
x(t) = t n + 1 (a special case) (130) (If you did not do Exercise 1, the integration exer-
then we know from our derivative formulas that cises on page 14, or had difficulty with them, go back
dx(t) (n+1) and do them now.)
v(t) = = dt = (n+1)t n (131)
dt dt
Substituting x(t) = t n + 1 and v(t) = (n+1)t n into
Equation (128) gives
T T
x(t) = vx(t) dt (128)
ti ti

T T
tn + 1 = (n +1)t ndt
ti ti
T (132)
= (n +1) t ndt
ti
Cal 1-28 Calculus 2000 - Chapter 1 Introduction to Calculus

Derivative of the Exponential Function Let us now see how to use the series (136) for
The previous work shows us that if we have a series calculating the derivative of e x . We have, from the
expansion for a function, it is easy to obtain a definition of a derivative,
formula for the derivative of the function. We will d f(x) ≡ limit f(x + ∆x) – f(x) (56) repeat
now apply this technique to calculate the derivative dx ∆x→ 0 ∆x
and integral of the exponential function e x .
If f(x) = e x , we get
There is a series expansion for the function e x that
works for any value of x is but is most useful for
d(e x)
= limit e x + ∆x – e x (138)
dx ∆x→0 ∆x
small values of x = α << 1, is
To do this calculation, we have to evaluate the
eα ≈ 1 + α + α + α + ⋅ ⋅ ⋅
2 3
2! 3!
(136) quantity e x + ∆x . First, we use the fact that for
where 2! = 2 ×1 , 3! = 3 × 2 × 1 = 6 , etc. (The quan- exponentials
tities 2!, 3! are called factorials. For example 3! is ea + b = ea eb
called three factorial.)
(Remember that 10 2 + 3 = 10 2 × 10 3 = 10 5 .) Thus
To see how well the series (136) works, consider the
case α = .01 . From the series we have, up to the α 3 e x +∆x = e x e ∆x (139)
term
Now use the approximation formula (136), setting
α = .01 α = ∆x and throwing out the α 2 and α 3 and higher
α 2 = .0001 ; α 2/2 = .00005 terms because we are going to let ∆x go to zero
α 3 = .000001 ; α 3/ 6 = .000000167 e ∆x ≈ 1 + ∆x (140)
Giving us the approximate value Substituting (140) in (139) gives
1 + α + α + α = 1.010050167
2 3
(137) e x+∆x ≈ e x (1 + ∆x)
2! 3!
When we enter .01 into a scientific calculator and = e x + e x∆x (141)
press the e x button, we get exactly the same result. Next use (141) in (138) to get
Thus the calculator is no more accurate than includ-
ing the α 3 term in the series, for values of α equal d(e x ) limit (e + e ∆x) – e
x x x
to .01 or less. = ∆x→0 (142)
dx ∆x
The e x terms cancel and we are left with
d(e x ) limit e x∆x = limit e x
= ∆x→0 (143)
dx ∆x ∆x→0
Since the ∆x′s cancelled, we are left with the exact
result

d(e x )
= ex (144)
dx

We see that the exponential function e x has the special

property that it is its own derivative.
 Calculus 2000 - Chapter 1 Introduction to Calculus Cal 1-29

We will often want to know the derivative, not just Integral of the Exponential Function
of the function e x but of the slightly more general To calculate the integral of e ax , we will use the same
result e ax where a is a constant. That is, we want to trick as we used for the integral of x n , but we will be
find a bit more formal this time. Let us start with
d e ax (a = constant) (145) Equation (128) relating position x(t) and velocity
dx v(t) = dx(t)/dt go get
Solving this problem provides us with our first
meaningful application of the chain rule tf tf tf dx(t)
x(t) = vx(t) dt = dt (128)
ti ti ti dt
df(y) df(y) dy
= (100) repeated
dx dy dx Since Equation (128) holds for any function x(t) [we
If we set did not put any restrictions on x(t)], we can write
Equation (128) in a more abstract way relating any
y = ax (146) function f(x) to its derivative df(x)/dx
then we have
xf xf df(x)
de ax = de y dy (147) f(x) = dx (151)
dx dy dx xi xi dx
Now
de y = e y To calculate the integral of e ax , we set f(x) = e ax
(148) and df(x)/dx = ae ax to get
dy
xf xf
dy d dx e ax = a e ax dx (152)
= (ax) = a = a×1 = a (149) xi xi
dx dx dx
Using (148) and (149) in (147) gives Dividing (157) through by (a) gives us the definite
integral
de ax = (e y )(a) = (e ax )(a) = ae ax
dx xf xf
Thus we have e ax dx = 1a e ax (a = constant) (153)
xi xi

d ax
e = ae ax (150) The corresponding indefinite integral is
dx
e ax
This result will be used so often it is worth memoriz- e axdx = a (a = constant) (154)
ing.
Exercise 7
Exercise 6 The natural logarithm is defined by the equation
For further practice with the chain rule, show that
ln (x) = 1 dx (see Equations 33-40)
2 x
deax 2
= 2axeax
dx
Use Equation (151) to show that
2
Do this by choosing y = ax , and then do it again by
choosing y = x2 . d (ln x) = 1
dx x (155)

(Hint—integrate both sides of Equation (155) with

respect to x.)
Cal 1-30 Calculus 2000 - Chapter 1 Introduction to Calculus

DERIVATIVE AS THE SLOPE OF A CURVE Figure (20a) is a blowup of the curve in the region
Up to now, we have emphasized the idea that the between x and x + ∆x . If the distance ∆x is suffi-
derivative of a function f(x) is given by the limiting ciently small, the curve between x and x + ∆x
process should be approximately a straight line and that part
of the curve should be approximately the hypot-
df(x) limit f(x + ∆x) – f(x)
= ∆x→0 (55) repeat enuse of the right triangle abc seen in Figure (20a).
dx ∆x Since the side opposite to the angle θ * is
f(x + ∆x) – f(x) , and the adjacent side is ∆x , we
We saw that this form was convenient when we had have the result that the tangent of the angle θ * is
an explicit way of calculating f(x + ∆x) , as we did
f(x + ∆x) – f(x)
by using a series expansion. However, a lot of words tan(θ * ) = (156)
are required to explain the steps involved in doing ∆x
the limiting process indicated in Equation (55). In When we make ∆x smaller and smaller, take the
contrast, the idea of an integral as being the area limit as ∆x → 0 , we see that the angle θ * becomes
under a curve is much easier to state and visualize. more nearly equal to the angle θ shown in Figure
Now we will provide an easy way to state and (21), the angle of the curve when it passes through
interpret the derivative of a curve. the point x. Thus
f(x + ∆x) – f(x)
Consider the function f(x) graphed in Figure (20). tanθ = limit (157)
∆x→0 ∆x
At a distance x down the x axis, the curve had a
height f(x) as shown. Slightly farther down the x The tangent of the angle at which the curve passes
axis, at x + ∆x , the curve has risen to a height through the point x is called the slope of the curve at
f(x + ∆x) . the point x. Thus from Equation (157) we see that
the slope of the curve is equal to the derivative of the
f(x)
f(x+∆x) curve at that point. We now have the interpretation
that the derivative of a curve at some point is equal
f(x) to the slope of the curve at that point, while the
integral of a curve is equal to the area under the curve
up to that point.
f(x)
∆x θ
x
x x+∆x
Figure 20
Two points on a curve, a distance ∆ x apart.

c x
f(x+∆x)

f(x)
a θ*
}
b
f(x+∆x)
– f(x)
Figure 21
The tangent of the angle θ at which the curve
passes through the point x is called the slope
∆x
of the curve at that point.
Figure 20a
At this point, the curve is tilted
by approximately an angle θ *.
 Calculus 2000 - Chapter 1 Introduction to Calculus Cal 1-31

Negative Slope Exercise 8

In Figure (22) we compare the slopes of a rising and Estimate the numerical value of the slope of the curve
a falling curve. In (22a), where the curve is rising, shown in Figure (23) at points (a), (b), (c), (d) and (e).
the quantity f(x + ∆x) is greater than f(x) and the In each case do a sketch of f(x + ∆x) – f(x) for a small
derivative or slope ∆x , and let the slope be the ratio of f(x + ∆x) – f(x) to
∆x . Your answers should be roughly 1, 0, –1, + ∞ ,
df(x) limit f(x + ∆x) – f(x)
= ∆x→0 – ∞.
dx ∆x
is a positive number. f(x)
b
In contrast, for the downward curve of Figure (22b), e
c
f(x + ∆x) is less than f(x) and the slope is negative. a
d
For a curve headed downward, we have
x
df(x) downward heading
= – tan(θ) curve (158) Figure 23
dx
Estimate the slope at the
(For this case you can think of θ as a negative angle, various points indicated.
so that tan(θ) would automatically come out nega-
tive. However it is easier simply to remember that
the slope of an upward directed curve is positive and
that of a downward directed curve is negative.)

pe
e slo f(x+∆x)
itiv θ
os f(x)
p

f(x+∆x) – f(x) is positive

∆x
x x+∆x

f(x+∆x) – f(x) is negative

∆x
ne

ga f(x)
tiv θ
es
lo p f(x+∆x)
e

x x+∆x
Figure 22 a,b
Going uphill is a positive slope,
downhill is a negative slope.
Cal 1-32 Calculus 2000 - Chapter 1 Introduction to Calculus

THE EXPONENTIAL DECAY seconds, only 27 remained. The decay of these

A curve that we will encounter several times during muons is an example of an exponential decay of the
the course is the function e – ax shown in Figure (24), form
which we call an exponential decay. Since expo- number of number of
nents always have to be dimensionless numbers, we surviving = muons at × e – t/t0 (161)
muons time t = 0
are writing the constant (a) in the form 1/x 0 so that
the exponent x/x 0 is more obviously dimensionless. where t 0 is the time it takes for the number of muons
remaining to drop by a factor of 1/e = 1/2.7. That
The function e – x/x0 has several very special proper- time is called the muon lifetime.
ties. At x = 0, it has the numerical value 1 (e 0 = 1) .
When we get up to x = x 0 , the curve has dropped to We can use Equation (161) to estimate the muon
a value lifetime t 0 . In the movie, the number of muons at the
top of the graph, reproduced in Figure (25), is 648. That
e – x/x 0 = e – 1 = 1e (at x = x 0) is at time t = 0. Down at time t = 6 microseconds, the
(159) number surviving is 27. Putting these numbers into
≈ 1
2.7 Equation (161) gives
When we go out to x = 2x 0 , the curve has dropped surviving
muons × e
= 648 initial
27 muons – 6/t0
to
e – 2x 0 /x 0 = e – 2 = 12 (160) e – 6/t0 = 27 = .042 (162)
e 648
Out at x = 3x 0 , the curve has dropped by another Take the natural logarithm ln of both sides of Equa-
factor of e to (1/e)(1/e)(1/e). This decrease contin- tion (162), [remembering that ln e x = x ] gives
ues indefinitely. It is the characteristic feature of an
exponential decay. ln e – 6/t0 = –t 6 = ln .042 = – 3.17
0

Muon Lifetime where we entered .042 on a scientific calculator and

In the muon lifetime experiment, we saw that the pressed the ln key. Solving for t 0 we get
number of muons surviving decreased with time. At t 0 = 6 microsec = 1.9 microseconds (163)
3.17
the end of two microseconds, more than half of the
original 648 muons were still present. By 6 micro- This is close to the accepted value of t 0 = 2.2 0
microseconds which has been determined from the
study of many thousands of muon decays.
1

e–x/x0

1/e

1/e 2
x
0 x0 2x 0 3x 0
Figure 24
Figure 25
As we go out an additional distance The lifetime of each detected muon is represented
x 0 , the exponential curve drops by by the length of a vertical line. We can see that
another factor of 1/e. many muons live as long as 2 microseconds (2µs),
but few live as long as 6 microseconds.
 Calculus 2000 - Chapter 1 Introduction to Calculus Cal 1-33

Half Life To help illustrate the nature of exponential decays,

The exponential decay curve e – t/t0 decays to suppose that you started with a million muons. How
1/e = 1/2.7 of its value at time t 0 . While 1/e is a very long would you expect to wait before there was, on
convenient number from a mathematical point of the average, only one left?
view, it is easier to think of the time t 1/2 it takes for
half of the muons to decay. This time t 1/2 is called To solve this problem, you would want the number
the half life of the particle. e – t/t0 to be down by a factor of 1 million
e – t/t0 = 1 × 10 – 6
From Figure (26) we can see that the half life t 1/2 is
slightly shorter than the lifetime t 0 . To calculate the Taking the natural logarithm of both sides gives
half life from t 0 , we have
ln e – t/t0 = –t
t 0 = ln 1×10
–6
= –13.8 (166)
e – t/t0 = e – t1/2/t0 = 1 (164)
t = t 1/2 2
(To calculate ln 1×10 – 6 , enter 1, then press the
Again taking the natural logarithm of both sides of exp key and enter 6, then press the +/– key to change
Equation (164) gives it to –6. Finally press = to get the answer –13.8.)
–t
ln e – t1/2/t0 = t 1/2 = ln 1 = – .693 Solving Equation (166) for t gives
0 2
t = 13.8 t 0 = 13.8 × 2.2 µsec
t 1/2 = .693 t 0 (165)
t = 30 microseconds (167)
From Equation (165) you can see that a half life t 1/2
is about .7 of the lifetime t 0 . If the muon lifetime is That is the nature of an exponential decay. While
2.2 µsec (we will abbreviate microseconds as µsec ), you have nearly half a million left after around 2
and you start with a large number of muons, you microseconds, they are essentially all gone by 30
would expect about half to decay in a time of microseconds.

(t 1/2) muon = .693 × 2.2µsec = 1.5 µsec Exercise 9

The basic feature of the exponential decay curve How many factors of 1/2 do you have to multiply
e – t/t0 is that for every time t 0 that passes, the curve together to get approximately 1/1,000,000? Multiply
decreases by another factor of 1/e. The same applies this number by the muon half-life to see if you get
about 30 microseconds.
to the half life t 1/2 . After one half life, e – t/t0 has
decreased to half its value. After a second half life,
the curve is down to 1/4 = 1/2 × 1/2 . After 3 half 1
lives it is down to 1/8 = 1/2 × 1/2 × 1/2 as shown in
Figure (27). e–t /t 0
1 1/2

e–t /t 0 1/4
1/8
1/2
1/e 0 t 1/2 2t1/2 3t 1/2
Figure 27
After each half-life, the curve
decreases by another factor of 1/2.
t
0 t0
t 1/2
Figure 26
Comparison of the lifetime t0 and the half-life t1/2 .
Cal 1-34 Calculus 2000 - Chapter 1 Introduction to Calculus

Measuring the Time The height (y) of the point where we drew the
Constant from a Graph tangent curve is just the value of the function e – T/t0 .
The idea that the derivative of a curve is the slope of The tangent of the angle θ is the opposite side (y)
the curve, leads to an easy way to estimate a lifetime divided by the adjacent side (x)
t 0 from an exponential decay curve e – t/t0 . y e – T/t0
tanθ = x = x (169)
The formula for the derivative of an exponential
Equating the two magnitudes of tanθ in Equations
curve is
(169) in (168a) gives us
de at = ae at (150) repeated
dt 1 e – T/t0 = 1 e – T/t0
t0 x
Setting a = – 1/t 0 gives
d e – t/t0 = – 1 e – t/t0 which requires that
(168)
dt t0 x = t0 (170)
Since the derivative of a curve is the slope of the Equation (170) tells us that the distance (x), the
curve, we set the derivative equal to the tangent of distance down the axis where the tangent lines
the angle the curve makes with the horizontal axis. intersect the axis, is simply the time constant t 0 .
d e – t/t0 = – 1 e – t/t0 = tanθ (168a) The result gives us a very quick way of determining
dt t0 the time constant t 0 of an exponential decay curve.
The minus sign tells us that the curve is headed As illustrated in Figure (29), choose any point on the
down. curve, draw a tangent to the curve at that point and
measure the distance down the axis where the tan-
In Figure (28), we have drawn a line tangent to the gent line intersects the axis. That distance will be the
curve at the point t = T. This line intersects the (t) time constant t 0 . We will use this technique in
axis (the axis where e – t/t0 goes to zero) at a distance several laboratory exercises later in the course.
(x) down the t axis.
e–t/t 0
e–t /t 0

x
e –T/t 0 θ
y

t
x t t0
Figure 29
Figure 28
T
A quick way to estimate the time constant t0
A line, drawn tangent to the exponential decay for an exponential decay curve is to draw the
curve at some point T, intersects the axis a tangent line as shown.
distance x down the axis. We show that this
distance x is equal to the time constant t0 . This
is true no matter what point T we start with.
 Calculus 2000 - Chapter 1 Introduction to Calculus Cal 1-35

THE SINE AND COSINE FUNCTIONS Radian Measure

The final topic in our introduction to calculus will be We are brought up to measure angles in degrees, but
the functions sinθ and cosθ and their derivatives physicists and mathematicians usually measure
and integrals. We will need these functions when we angles in radians. The angle θ measured in radians
come to rotational motion and wave motion. is defined as the arc length subtended by the angle
θ on a circle of unit radius, as shown in Figure (32).
The definition of sinθ and cosθ , which should be
arc length subtended
familiar from trigonometry, are θ radians = by θ on a unit circle (173)
opposite
sinθ = ac hypotenuse (171a) (If we had a circle of radius c, then we would define
θ radians = /c , a dimensionless ratio. In the special
adjacent
case c = 1, this reduces to θ radians = .)
cosθ = bc hypotenuse (171b)
Since the circumference of a unit circle is 2π , we see
that θ for a complete circle is 2π radians, which is
c the same as 360 degrees. This tells us how to convert
a
from degrees to radians. We have the conversion
θ Figure 30
b factor
where θ is an angle of a right triangle as shown in 360 degrees
= 57.3
degrees
(174)
Figure (30), (a) is the length of the side opposite to 2π radians radian
θ , (b) the side adjacent to θ and (c) the hypotenuse. As an example of using this conversion factor,
The formulas are simplified if we consider a right suppose we want to convert 30 degrees to radians.
triangle whose hypotenuse is of length c = 1 as in We would have
Figure (31). Then we have 30 degrees
= .52 radians (175)
57.3 degrees/radian
sinθ = a (172a)
To decide whether to divide by or multiply by a
cosθ = b (172b) conversion factor, use the dimensions of the conver-
1 sion factor. For example, if we had multiplied 30
a degrees by our conversion factor, we would have
θ Figure 31 gotten
b
We can then fit our right triangle inside a circle of degrees degrees 2
30 degrees × 57.3 = 1719
radius 1 as shown in Figure (32). radian radian
This answer may be correct, but it is useless.

The numbers to remember in using radians are the

following:
90° = π/2 radians
1 180° = π radians
a (176)
270° = 3π/2 radians
θ
b 360° = 2π radians
The other values you can work out as you need them.
Figure 32
Fitting our right triangle inside a unit radius circle.
Cal 1-36 Calculus 2000 - Chapter 1 Introduction to Calculus

The Sine Function Our next step is to construct a graph in which θ is

In Figure (33) we have started with a circle of radius shown along the horizontal axis, and we plot the
1 and, in a somewhat random way, labeled 10 points value of sinθ = (a) on the vertical axis. The result
around the circle. The arc length up to each of these is shown in Figure (34). The eleven points, repre-
points is equal to the angle, in radian measure, senting the heights a 0 to a 10 at θ 0 to θ 10 are shown
subtended by that point. The special values are: as large dots in Figure (34). We have also sketched
in a smooth curve through these points, it is the curve
θ 0 = 0 radians
we would get if we had plotted the value of (a) for
θ 4 = π/2 radians (90°) every value of θ from θ = 0 to θ = 2π . The smooth
θ 6 = π radians (180°) curve is a graph of the function sin θ .
θ 8 = 3π/2 radians (270°)
θ 10 = 2π radians (360°) Exercise 10
Using the fact that the cosine function is defined as
In each case the sinθ is equal to the height (a) at that
point. For example cos θ = b (b is defined in Figures 31, 32)

sinθ 1 = a 1 plot the values of b0, b1, ⋅ ⋅ ⋅ , b10 on a graph similar to

sinθ 2 = a 2 Figure (34), and show that the cosine function cos θ
looks like the curve shown in Figure (35).
⋅⋅⋅⋅⋅⋅
sinθ 10 = a 10

We see that the height (a) starts out at a 0 = 0 for θ 0 ,

increases up to a 4 = 1 at the top of the circle, drops
back down to a 6 = 0 at θ 6 = π , goes negative, down
to a 8 = – 1 at θ 8 = 3π/2 , and returns to a 10 = 0 at
θ 10 = 2π .
θ4 θ3 1
θ5
θ2
a4 a2
a3 θ2
a5 a2 θ1
a1
a1 θ1
π 2π
θ6 θ0
θ 10 0 θ
π 3π
a7 a9 2 2

a8
a7 θ7
θ9
θ7

θ8 –1
Figure 33 Figure 34
The heights a i at various points around a unit circle. Graph of the function sin ( θ ) .
 Calculus 2000 - Chapter 1 Introduction to Calculus Cal 1-37

There is nothing that says we have to stop measuring Amplitude of a Sine Wave
the angle θ after we have gone around once. On the A graph of the function y(θ) = c sinθ looks just like
second trip around, θ increases from 2π up to 4π , the curve in Figure (36), except the curve goes up to
and the curve sinθ repeats itself. If we go around a height c and down to –c as shown in Figure (38).
several times, we get a result like that shown in We would get the curve of Figure (38) by plotting
Figure (36). We often call that a sine wave. points around a circle as in Figure (33), but using a
circle of radius c. We call this factor c the amplitude
Several cycles of the curve cosθ are shown in of the sine wave. The function sinθ has an ampli-
Figure (37). You can see that the only difference tude 1, while the sine wave in Figure (38) has an
between a sine and a cosine curve is where you set amplitude c (its values range from +c to –c).
θ = 0 . If you move the origin of the cosine axis back
(to the left) 90° (π/2) , you get a sine wave. c sin θ

1 c

0 θ 0
π π 3π 2π
2 2

–1 –c
Figure 35 Figure 38
The cosine function. A sine wave of amplitude c.

0 θ
π 2π 3π 4π 5π 6π

–1
Figure 36
Several cycles of the curve sin ( θ ) .

0 θ
π 2π 3π 4π 5π 6π

–1
Figure 37
Several cycles of the curve cos ( θ ) .
Cal 1-38 Calculus 2000 - Chapter 1 Introduction to Calculus

Derivative of the Sine Function Now draw a line vertically down from point (c) and
Since the sine and cosine functions are smooth horizontally over from point (b) to form the triangle
curves, we should be able to calculate the derivatives bcd shown in Figure (40). The important point is
and integrals of them. We will do this by first that the angle at point (c) in this tiny triangle is the
calculating the derivative, and then turning the pro- same as the angle θ at point (a). To prove this,
cess around to find the integral, just as we did for the consider the sketch in Figure (41). A line bf is drawn
functions x n and e x . tangent to the circle at point (b), so that the angle abf
is a right angle. That means the other two angles in
The derivative of the function sinθ is defined as the triangle add up to 90°, the total angle in any
usual by triangle being 180°
d(sinθ) sin(θ +∆θ) – sinθ θ + ϕ = 90° (178)
= limit (177)
dθ ∆θ→0 ∆θ Since the angle at (e) in triangle bef is also a right
angle, the other two angles in the triangle bef, must
where ∆θ is a small change in the angle θ .
also add up to 90°.
The easiest way to evaluate this limit is to go back to α + ϕ = 90° (179)
the unit circle of Figure (25) and construct both sinθ
and sin (θ +∆θ) as shown in Figure (39). We see For both Equations (178) and (179) to be true, we must
that sinθ is the height of the triangle with an angle have α = θ .
c
θ, while sin (θ +∆θ) is the height of the triangle
whose center angle is (θ +∆θ) . What we have to do
θ

∆θ
is calculate the difference in heights of these two
triangles.
d b c
In Figure (40) we start by focusing our attention on
the slender triangle abc with an angle ∆θ at (a) and d

∆θ
b
long sides of length 1 (since we have a unit circle). ∆θ θ
Since the angle ∆θ is small, the short side of this
triangle is essentially equal to the arc length along
the circle from point (b) to point (c). And since we a
θ
are using radian measure, this arc length is equal to Figure 40 r=1
the angle ∆θ . The difference between sinθ and
sin ( θ + ∆θ ) is equal to the height
of the side cd of the triangle cdb.

sin(θ+∆θ) b
sin(θ)
∆θ α
θ
θ φ
a
e f
θ + φ = 90°
r=1
α + φ = 90°
∴α = θ
Figure 41
Figure 39 Demonstration that the angle α equals the angle θ .
Triangles for the sinθ and the sin ( θ + ∆θ ) .
 Calculus 2000 - Chapter 1 Introduction to Calculus Cal 1-39

The final step is to note that when ∆θ in Figure (40) Exercise 11

is very small, the side cb of the very small triangle is Using a similar derivation, show that
essentially tangent to the circle, and thus parallel to
the side bf in Figure (41). As a result the angle d (cosθ) = – sinθ
between cb and the vertical is also the same angle θ . dθ
(182)

Because the tiny triangle, shown again in Figure (42)

has a hypotenuse ∆θ and a top angle θ , the vertical Exercise 12
side, which is equal to the difference between sinθ Using the chain rule for differentiation, show that
and sin (θ +∆θ) has a height (cos θ)∆θ . Thus we
have d (sinaθ) = a cosaθ
dθ
sin(θ +∆θ) – sinθ = (cosθ)∆θ (180) (a = constant) (183)
d (cosaθ) = – a sinaθ
Equation (180) becomes exact when ∆θ becomes dθ
an infinitesimal angle.
(Hint—if you need to, look at Equation (145) through
We can now evaluate the derivative (150).

d(sinθ) limit sin(θ +∆θ) – sinθ Exercise 13

= ∆θ→0
dθ ∆θ Using the fact that integration reverses differentia-
tion, as we did in integrating the function e x (Equa-
(cosθ)∆θ tions (151) through (154), show that
= limit
∆θ→0 ∆θ
θf
1 sinaθ θf
(cosaθ)dθ = a
θi θi
(184a)
= limit
∆θ→0
cosθ
(a = constant)
θf
Thus we get the exact result 1 cosaθ θf
(sinaθ)dθ = – a
θi θi
(184b)

d (sinθ) = cosθ (181)

dθ
Use sketches of the integrals from θi = 0 to θf = π /2 to
show that Equations (184a) and (184b) have the
∆θ cosθ

correct numerical sign. (Explicitly explain the minus

θ
∆θ

sign in (184b).

sin(θ+∆θ)
∆θ

sin(θ)
∆θ

a
θ
r=1
Figure 42
The difference between sinθ and sin ( θ + ∆θ ) is
equal to ∆θ cos θ . Check that this result is
reasonable by considering the special cases
θ = 0 and θ = 90° ( π /2) .
 Calculus 2000 - Chapter 2 Second Derivatives and 1D Wave Eq. Cal 2-1

Calculus 2000-Chapter 2
Second Derivatives and the One
Dimensional Wave Equation

CHAPTER 2 SECOND DERIVATIVES

AND THE ONE DIMENSIONAL WAVE
EQUATION

In our discussion of a wave pulse on a rope, in

Chapter 15 of Physics 2000, we used a combination
of physical observation and a somewhat tricky argu-
ment to show that the speed of the wave pulse was
given by the formula v = T/ρ . The physical obser-
vation was noting that a pulse travels down the rope
at an apparently uniform speed. The trick was to
analyze the behavior of the rope from the point of
view of someone moving along with the pulse (as on
pages 15-4, 5).

Another way to handle the problem is to directly

apply Newton's second law to a section of the rope.
When we use this direct approach, we end up with an
equation that involves second derivatives not only
with respect in time, but also with respect to space.
The resulting equation with its second derivatives is
what is known as the wave equation. The aim of this
chapter is to learn how to handle the wave equation,
at least for waves moving in one dimension. (Han-
dling three dimensional wave equations comes later.)

To use the wave equation with any real understand-

ing, not just manipulating formulas, requires more
familiarity with the properties of a second derivative
than we have needed so far. Thus we will begin this
chapter with a discussion of the second derivative,
and then go on to the one dimensional wave equa-
tion.
Cal 2-2 Calculus 2000 - Chapter 2 Second Derivatives and 1D Wave Eq.

THE SECOND DERIVATIVE Solution

Begin by taking the first derivative
We have already encountered the idea of a second
derivative in our discussion of velocity and accel- d sin(aθ)
eration. Consider a particle moving down the x axis, = a cos(aθ) (7)
dθ
whose position is described by the function x(t). The
Now differentiate again
particle's velocity vx(t) is given by

vx(t) =
dx(t) first
(1) d d sin(aθ) = d a cos(aθ)
dt derivative dθ dθ dθ
which is the first derivative, with respect to time, of
= a d cos(aθ)
x(t). The particle's acceleration a(t) is given by dθ
dvx(t) = a – a sin(aθ) (8)
a x(t) = (2)
dt
When we use (1) for v(t) in Equation (2) we get Thus we get

dx(t) d 2sin(aθ)
a x(t) = d second (3) = – a 2 sin(aθ) (9)
dt dt derivative dθ 2
In Equation (3), we see that a x(t) is obtained from We see that the second derivative of a sine curve is
x(t) by differentiating twice with respect to time. itself a sine curve, with a minus sign.
We say that a x(t) is the second derivative of x(t) and
use the simplified notation Exercise 1
2 simplified Calculate the following second derivatives
d dx(t) ≡ d x(t) notation for (4)
dt dt dt 2 second derivative d2
(a) cos (aθ)
dθ2
With this notation, the position x(t), velocity vx(t) ,
and acceleration a x(t) are related by d2
(b) e– ax
x(t) dx2

dx(t) d2
vx(t) = (c) ln(x)
dt dx2

d 2x(t) d2
a x(t) = (5) (d) ln (ax)
dt 2 dx2
There is nothing particularly difficult about carrying d2
out a second derivative, just do the derivative opera- (e) xn
dx2
tion twice as illustrated in the following example.
d2
(f) (ax)n
dx2
Example 1
d2
Calculate the second derivative, with respect to θ , (g) 1
of sin(aθ) dx2 (x)n

d2 1
d 2sin(aθ) (h)
= ? (6) dx2 (ax)n
dθ 2
 Calculus 2000 - Chapter 2 Second Derivatives and 1D Wave Eq. Cal 2-3

Geometrical Interpretation We will evaluate the second derivative at the bottom

of the Second Derivative of the circle, where the curve is horizontal and the
We have seen that the various calculus operations slope is zero.
have a geometrical interpretation. Integration was curve horizontal
equivalent to finding the area under a curve, while y′[x = 0] = 0 at x = 0 (12)
the first derivative represented the slope of a curve.
Now move down the x axis a distance ∆x as shown
We now want to obtain the geometrical interpreta-
in Figure (1). If ∆x is small, then ∆x is essentially
tion of the second derivative. We will see that the
equal to the arc length ∆ along the circle, and the
second derivative is equal to what we will call the
angle ∆θ in radian measure is the arc length divided
curvature of the curve. To see exactly what that is,
by the radius R of the circle
consider the following derivation.

Let y(x) be the section of a circle as shown in Figure (1). ∆θ = ∆ ≈ ∆x (13)

R R
Let us use notation found in a number of calculus If we draw a line tangent to the circle at position
texts, and denote the derivative of y(x) by y′(x) x = ∆x , this tangent line will make an angle ∆θ to the
dy(x) simplified horizontal as shown in Figure (1). (The two angles
y′(x) ≡ (10)
dx notation labeled ∆θ in Figure 1 are equal no matter how big ∆θ
In terms of y′(x) , the second derivative is is.) Thus the slope of the tangent line at x = ∆x is

d 2y(x) y′(x + ∆x) – y′(x) slope of circle = tan (∆θ) (14)

= limit (11) at x = ∆x
dx 2 ∆x → 0 ∆x
Remember that y′(x) = dy/dx is the slope of the Now if ∆θ is a small angle, which it will be if we take
curve at position x as shown in Figure (2) (For the limit as ∆x → 0 , we can use the approximation
example, see Figure 21 of Chapter 1). Thus Equa-
tion (11) tells us that to find the second derivative of tan (∆θ) ≈ ∆θ (15)
y(x) we have to find the change in slope as we move You can see why this approximation is good for
from x to x + ∆x . small angles from Figure (2a).
y
Thus the slope of the tangent line at x = ∆x is given by
slope of
∆θ tangent line = y′[x = ∆x] = ∆θ = ∆x (16)
at x = ∆x R

t where we used Equation (13) for ∆θ .

R gen
tan
line Now we have values of y′ at x = 0 (Equation 12) and
y(x)
∆θ at x = ∆x (Equation 16), we can use these values in
∆
x Equation (10) to get the value of d 2y/dx 2 at x = 0,
∆x
i.e., at the bottom of the circle.
Figure 1
Calculating the change in the slope of the
circle, as we go from x = 0 to x = ∆x . Figure 2a
R
y
For small angles, ∆θ
the angle and the x
y
tangent of the angle y
Figure 2 tan (∆θ) = x
are essentially
The slope is the y
equal. ∆θ = ≈
tangent of the angle. R x
θ
slope dy(x)
at x ≡ dx = tan θ x
Cal 2-4 Calculus 2000 - Chapter 2 Second Derivatives and 1D Wave Eq.

Introducing the notation CURVATURE

Consider the curve shown in Figure (3) representing
2
d y(x) means d 2y(x)/dx 2 some function y(x). At point x 0 we have drawn a
evaluated at the
dx 2 point x = 0 circle that just fits against the curve. The radius of
x=0
the circle is adjusted to give the closest match
We have from Equation (10) possible between the curve y(x 0) and the circle.
When we get this closest fit, both the first and the
d 2y(x) limit y′[x = ∆x] – y′[x = 0]
= ∆x second derivatives of the circle and y(x) are equal at
dx 2
x=0
→0 ∆x x = x 0 . In other words
(17)
With y′[x = ∆x] = ∆x/R (Equation 16) and d 2y(x)
y′[x = 0] = 0, we get = 1 (20)
dx 2 x = x0
R
d 2y(x) limit ∆x/R – 0
= ∆x
dx 2
x=0
→0 ∆x In Figure (3) the quantity R is called the radius of
curvature of the curve y(x) at the point x 0 , and 1/R
limit 1
= ∆x is called the curvature
→0 R
1 ≡ curvature of the curve (21)
Since the ∆x's canceled, we see that 1/R is the R
limiting value and we get
You can see intuitively why 1/R is called curvature.
If R is very large, the curve is almost flat and we
d 2y(x)
= 1 (18) would say it has little curvature. As R becomes
dx 2 x = 0 R smaller, the curve bends in a tighter circle, and the
curvature 1/R becomes greater.
With a slightly messier derivation we could calcu-
late d 2y/dx 2 anywhere around the circle, not just at This is the geometrical picture of the second deriva-
the bottom, and we get the same answer 1/R. Thus tive. While the first derivative was equal to the slope
we have the more general result of the curve at some point, the second derivative is
equal to the curvature of the curve at that point. The
curvature is explicitly the reciprocal of the radius of
d 2y(x) anywhere around curvature of the curve where the radius of curvature
= 1 (19)
dx 2 R the circle is found by fitting a circle to the curve as in Figure
(3). [Exercise: under what circumstances would the
second derivative or curvature be negative?]
y

x
x0
Figure 3
At any point along a curve, the curvature is
1/R or –1/R, where R is the radius of the
circle that just fits the curve as shown.
 Calculus 2000 - Chapter 2 Second Derivatives and 1D Wave Eq. Cal 2-5

Curve Fitting and Boat Lofting To avoid these bumps, the plans were taken up into
The problem of working with curves has a number the loft of the boat shed (hence the name lofting), and
of practical applications, one of the more interesting drawn full scale. Wooden splines, typically thin
of which, at least to a sailor, is the lofting of boats. strips of spruce, were bent along the points of the
curve. Since the splines bent along smooth curves,
It turns out that the eye is extremely good at judging any points that were out of place would not be fitted
the smoothness of a curve. We can, for example, by the spline and the points would be moved to fit the
easily spot the slightest wrinkle in what is supposed smooth curve. This process is called spline fitting.
to be the smooth side of a boat. (It is an interesting Once all the full scale curves were smoothed by
question as to how the eye and brain can do this so spline fitting, then the boat hull was constructed
well.) using these smoothed plans and the result, if done
correctly, was a smooth, good looking hull.
Through the 16th century, boats were rather crude
looking. Starting in the 17th century, better looking In the early 1970's, shortly after we had started using
boats were built using the following steps. The first the computer in teaching introductory physics, we
was to carve a model of the hull that was to be built. had lunch with a boat builder who described the
Then conceptually slice the model as you would rather tedious process of lofting a boat. He won-
slice a loaf of bread. Each of these cuts was called dered if lofting could be done more easily on the
a station. Typically one used about 15 stations, each computer. This was before the availability of inex-
representing a cross section of the hull at different pensive line plotters, so that the work would all have
distances along the length of the boat. Then points to be done numerically. We agreed to try, the
were taken from the model to represent the shape of incentive being a reduced price on a diesel engine for
the hull at each station. Figure (4) is a typical our boat if we successfully lofted the boat builder's
example of a hull cross section at a middle station. new lobster boat design.
Since the points showing the shape of the hull were The most successful part of the project was finding
taken from a small model, any errors in measure- an easy and very effective way to spot a smooth
ment would be greatly magnified when the hull was curve. Just print out a list of the third derivatives of
laid out full scale. An error of a fraction of a the curve. Since the second derivative is the curva-
millimeter in measuring the model would lead to a ture of the curve, the third derivative is the rate at
very obvious bump in the final hull shape. which the curvature is changing as you go along the
curve. If the curvature changes slowly, then the
curve looks smooth. A bump represents a sudden
change in curvature and therefore has a large third
derivative. What a spruce spline essentially does is
to minimize the third derivative.

About the same time that we wrote the lofting

program, a physicist, Peter Karos in Germany, also
wrote a boat lofting program. As one does not make
points, taken from much of a living from a lofting program, Karos
model, used to
draw plans
turned to the problem of using the computer to create
letter forms. The letters of the alphabet are con-
structed from different curves that depend upon
which font you are using. And just as in boat design,
Figure 4 the eye is very sensitive to the smoothness of the
Typical cross section. (Since boats are supposed curves, even for relatively small letters.
to be symmetric, only one side is usually drawn.)
Cal 2-6 Calculus 2000 - Chapter 2 Second Derivatives and 1D Wave Eq.

Karos based his boat lofting and letter design pro- THE BINOMIAL EXPANSION
grams on what are called Besier curves. To construct We have seen, starting in Chapter 1 of the Physics
a Besier curve through a series of points, at each point text, the usefulness of the binomial expansion
you specify the location y(x) of the point, the first
n(n – 1) 2
derivative y′(x) = dy/dx, and the second derivative (1 + α) n = 1 + nα + α + ⋅ ⋅ ⋅ (22)
y′′(x) = d 2y/dx 2 = 1/R . The section of curve be- 2!
tween two adjacent points is then constructed to which is valid for any value of α less than one, but
match the first and second derivative at the end points which gets better as α becomes smaller. For very
and minimize the third derivative in the region in small α , we could neglect all terms involving α 2 or
between. This uniquely determines the line. higher powers of α , giving us the approximation
Karos's techniques using Besier curves was built into formula
the Postscript™ language used for letter design. A (1 + α) n ≈ 1 + nα (α < < 1) (23)
way of graphically handling the construction of Besier
curves was developed and became the basis of the which is good for any value of n.
Adobe Illustrator™ program.
With calculus, we can easily derive the formula for
Those of you who have used Adobe Illustrator, or any the various terms in the binomial expansion. We
of the similar drawing programs, will be familiar begin with the assumption that the quantity (1 + α) n
with the constructing of Besier curves. You place the can be expanded in some kind of a series involving
pen tool at a point and press the mouse button. That powers of α . We will write the series in the form
establishes the point y(x). Then you drag the pen tool
in some direction. That direction establishes the (1 + α) n = A 0α 0 + A 1α 1 +A 2α 2 +A 3α 3 + ⋅ ⋅ ⋅
slope of the curve y′(x) at that point. How far out you (24)
drag the pen tool before you let up on the mouse where the A 0 , A 1 , A 2 , etc. are unknown coeffi-
button determines the radius of curvature R at that cients that we have to determine.
point, and thus establishes the second derivative
y′′(x) = d 2y/dx 2 = 1/R there (see Figure 5). When Equation (24) is supposed to be correct for small
you move the mouse to another point, press the values of α including α = 0 . Setting α = 0 gives
mouse button and drag, you determine y(x), y′(x)
and y′′(x) at the new point, and then the computer (1 + 0) n = A 00 0 + A 10 1 +A 20 2 +A 30 3 + ⋅ ⋅ ⋅
draws the smooth Besier curve between the two (25)
points. Here is a peculiar convention we use. We assume
that any number x 0 = 1 no matter what x is, includ-
When you are using Adobe Illustrator, or other draw-
ing 0 0 . Thus A 00 0 = A 0 , all the other terms on the
ing programs, think of the fact that you are control-
ling the position, the first derivative, and the second right side of Equation (25) are zero, and we get
derivative every time you place and drag the mouse. 1n = 1 = A 0 (26)
which determines A 0 .
Figure 5
po i n t slope
Constructing Besier
y(x) y′(x) (Writing A 0α 0 instead of just A 0 for the first term
curves with Adobe in the series is formalism that makes the series look
Illustrator®. In that cursor
more consistent, but is unnecessary if you do not like
program, the radius
of curvature is set to the idea of 0 0 = 1 .)
about 60% of the Distance from point to
distance that the cursor is proportional to
cursor is pulled out the radius of curvature.
from the point.
 Calculus 2000 - Chapter 2 Second Derivatives and 1D Wave Eq. Cal 2-7

To determine the value of A 1 , differentiate Equa- From Equation (34) you can see the general formula
tion (24) with respect to α . We get, using the chain emerging
rule n(n – 1)(n – 2)(n – 3)⋅ ⋅ ⋅
(1 + α) n d(1 + α)
d (1 + α) n = d An = (35)
n!
dα d(1 + α) dα
Thus by successive differentiation we can rather
= n(1 + α) n – 1 1 easily determine all the terms in the binomial expan-
sion.
= n(1 + α) n – 1 (27)
(One thing we have not worried about, but which is
Differentiating the right hand side of Equation (24) of major concern in calculus texts, is the range of
gives values of α for which the series is valid. Such
d A α0 + A α1 + A α2 + A α3 + ⋅ ⋅ ⋅ questions are important from a purely mathematical
dα 0 1 2 3 point of view, but are seldom of practical impor-
tance. From a practical point of view, you can
= 0 + A 1 + 2A 2α + 3A 3α 2 + ⋅ ⋅ ⋅ (28)
usually evaluate a few terms, and if the last ones are
negligibly small, the series is probably good enough.)
Thus the first derivative of Equation (24), with
respect to α , is
n(1 + α) n – 1 = A 1 + 2A 2α + 3A 3α 2 + ⋅ ⋅ ⋅ (29) The Taylor Series Expansion
The binomial expansion we have just discussed is a
Now set α = 0 and we get special case of the more general expansion called the
n(1+0) n – 1 = A 1 + 2A 2 ×0 + 3A 3 ×0 2 +⋅ ⋅ ⋅ (30) Taylor series expansion. In Figure (6) we have
sketched a curve representing some function
which gives us
n = A1 (31) y = f(x) (36)

and determines the coefficient A 1 . Suppose we know everything about the function at
the point x 0 and would like to figure out where the
To determine A 2 , differentiate Equation (29) with curve is going as we move away from that point. By
respect to α . With knowing everything about f(x) at the point x 0 , we
mean that we know f(x 0) as well as all the deriva-
d (1 + α) n – 1 = (n – 1)(1 + α) n – 2 tives of f(x) evaluated at x = x 0 .
dα
y
we get

n (n – 1)(1 + α) n – 2 = 2A 2 + 3(2α) + ⋅ ⋅ ⋅ (32)

Setting α = 0 gives =
y

f( x
) (x - x0 )
n(n – 1)
A2 = (33)
2
x
x0 x
Exercise 2
Figure 6
Differentiate Equation (32) with respect to α , set
If we know everything about the curve y = f (x )
α = 0 , and show that A3 is given by
at the point x0 , can we predict where the curve
n(n – 1)(n – 2) will be a short distance farther down the x axis?
A3 = (34)
3×2×1
Cal 2-8 Calculus 2000 - Chapter 2 Second Derivatives and 1D Wave Eq.

The derivation of the Taylor series expansion begins Exercise 3

with the assumption that the function f(x) can be Show that
expanded, in the vicinity of the point x 0 by the so
called power series d2 f(x) f″(x0 )
A2 = 1 = (42)
2 dx 2 2
0 1 x = x0
f(x) = A 0(x – x 0 ) + A 1(x – x 0 )
(37)
+ A 2(x – x 0 ) 2 + ⋅ ⋅ ⋅ 1 d3f(x) f′′′(x0 )
A3 = = (43)
3 × 2 × 1 dx3 3!
If you think of (x – x 0 ) as being some small distance x = x0
α , then the expansion in Equation (37) is the same
form as the expansion of the function (1 + α) n back From Exercise 3 you can see that the general form of
in Equation (24). The difference is that for different the Taylor series expansion is
functions f(x) we get different coefficients A n.
f(x – x 0 ) = f(x 0 ) + f′ x 0 (x – x 0 ) 1
To calculate the A n, we do the same thing that we
did in deriving the binomial expansion. We dif- + 1 f″(x 0 )(x – x 0 ) 2
2!
ferentiate both sides of the equation and then set
x = x 0 (which corresponds to setting α = (x – x 0 ) + 1 f′′′(x 0 )(x – x 0 ) 3+ ⋅ ⋅ ⋅
equal to zero). 3!
This can be written in the compact form
First we set x = x 0 in Equation (37) to get
∞ f n(x ) Taylor
f(x 0) = A 0(x – x 0 ) 0 + A 1(x – x 0 ) 1 + ⋅ ⋅ ⋅ f(x – x 0 ) = Σ
n=0
0
n!
(x – x 0 ) n series
expansion
= A 0(0) 0 + A 1(0) 1 + ⋅ ⋅ ⋅ (38)
(44)
= A0 where we used the notation
which determines the first coefficient A 0 . d n f(x)
f n(x 0 ) ≡ (45)
dx n x = x
Differentiating both sides of Equation (37) with 0

respect to x and then setting x = x 0 gives The tricky part of the mathematics of the Taylor
series expansion is how far you can go, how far x can
df(x) be away from x 0 , and still have a valid expansion.
f′(x) ≡ = A 1 + 2A 2(x – x 0 )
dx Perhaps more important to the physicist is how far
(39)
+ 3A 3(x – x 0 ) 3 + ⋅ ⋅ ⋅ you can go before you have to include too many
terms and the expansion is not useful.
where we used the chain rule to show that

d (x – x ) n = n(x – x ) n – 1 Exercise 4
0 0 (40) Apply the Taylor series expansion, Equation (44) to
dx
the function
Setting x = x 0 in Equation (39) gives
n
f(x) = (x – x0 )
df(x)
f′(x 0 ) ≡ = A1 (41) evaluated at x0 = 1 , and show that you get the bino-
dx x = x 0
mial expansion. (Hint—set α = x – x0 , i.e., substitute
all the other terms being zero. x = x0 + α at the end.)
 Calculus 2000 - Chapter 2 Second Derivatives and 1D Wave Eq. Cal 2-9

The Constant Acceleration Formulas If the particle is moving with constant acceleration,
While the Taylor series expansion, Equation (44), then all higher derivatives are zero. For example
looks like a very new topic, we have been using a
d 3y(t) da y(t) for
Taylor series expansion since the very beginning of y′′′(t) ≡ 3 = = 0 constant
our discussion of calculus. The constant accelera- dt dt acceleration
tion formulas are a simple example of this expan- (50)
sion. The Taylor series expansion for y directed motion
y(t) is
Figure (7) is a reproduction of our instantaneous ∞ (t – t 0 ) n
velocity drawing, Figure (3-32d) from Chapter 3 of y(t – t 0) = Σ y n(t 0)
n=0 n!
the Physics text and Figure (Cal 1-1d) of the Calcu-
lus text. At some instant of time, the ball is located dy(t) (t – t 0 )
= y 0(t 0) +
at some position (x 0, y 0) at time t 0 , and we wish to dt t 1!
0
predict the position of the ball at some later time t.
d 2y(t) (t – t 0 ) 2
The location of the ball is described by two functions + + ⋅⋅⋅
dt 2 t 2!
x(t) and y(t). We know x(t 0) , y(t 0) and all the 0
derivatives of these functions at time t 0 , they are With dy/dt = vy and d 2y/dt 2 = a y , we get
simply the velocity and acceleration
dx(t) y(t – t 0) = y 0(t 0) + vy(t 0)(t – t 0 )
x′(t) = = vx(t) (46) (51)
dt
+ 1 a y(t 0)(t – t 0 ) 2
d 2x(t) 2
x″(t) = = a x(t) (47)
dt 2 with all higher powers of (t – t 0 ) having zero coef-
dy(t) ficients.
y′(t) = = vy(t) (48)
dt
If we set t 0 = 0 Equation (51), we get the very
d 2y(t) familiar result
y′′(t) = = a y(t) (49)
dt 2 y(t) = y 0 + vy0t + 1 a yt 2 (52)
2
V
~i Here is an example of a Taylor series expansion that
is valid for any range of values (t – t 0 ) . It is good for
all times t because all derivatives of y(t) above the
second derivative are zero.)

Exercise 5
Suppose a particle is moving in the y direction with a
constantly increasing acceleration. I.e., assume that

day (t)
a′y (t) ≡ = constant
dt

Figure 7 Find the formula for y(t) for all future times t. (This is
Instantaneous velocity at time (t). one step above the constant acceleration formulas.)
Cal 2-10 Calculus 2000 - Chapter 2 Second Derivatives and 1D Wave Eq.

THE WAVE EQUATION Waves on a Rope

In the Physics text, we calculated the speed of a wave Our analysis of a wave pulse on a rope begins much
pulse on a rope in Chapter 15, pages 15-4 and 5 . As as it did in Chapter 15. Figures (8) and (9) are similar
we mentioned in the introduction, the calculation to Figures (15-3c) and (15-3d), except that we are
was relatively simple because of two tricks we were now standing still relative to the rope, and we are
able to pull. One was to walk along with the pulse, assuming the pulse is passing by us.
so that it looked as if the pulse were standing still and
In our current analysis of the wave pulse, we will be
the rope were passing through it. The second was to
somewhat more formal than we were in Chapter 15.
picture the top of the pulse as an arc of a circle, so that
We will say that the rope, at the present time, lies
we would know the acceleration of the rope as it
along a curve y(x) as shown in Figure (8). The
went around the arc. We got the right answer, but the
quantity x is the distance down the rope (say from
process did not generate much confidence that we
one end) and y(x) is the height of the pulse there, i.e.,
could handle more general cases, like calculating the
the distance the rope is displaced from its equilib-
speed of a sound wave pulse, or even of a compres-
rium position. From our various discussions of
sional pulse on a Slinky. (Remember that we used
derivatives, we know that dy(x)/dx is the slope of the
dimensional analysis, an important but approximate
rope at position x, and d 2y(x)/dx 2 = 1/R(x) is the
tool, to estimate the wave speeds in these cases.)
curvature, which is equal to the reciprocal of the
What we will do now is the more direct approach of radius of curvature R(x) at that point. In Figure (8)
applying Newton's laws to a section of the wave we have sketched in circles to show the radius of
pulse, get a differential equation, which happens to curvature at the two points x 1 and x 2 along the
involve second derivatives in both space and time, curve. The curvature is positive at x 1 and negative
and then solve the differential equation in the usual at x 2 .
way. That is, we guess a solution, plug it into the
Let us consider a short section of rope of length ∆
equation, and see if we made the correct guess. We
located at position x as shown in Figure (9). For now
will use as much physical insight as we can to guide
assume that this section begins at the top of the pulse
us in making the guess. The differential equation we
where the rope is horizontal. Shortly we will see that
will be working with is called the wave equation.
our results apply at any position along the rope.
Here we will be working with the wave equation for
The two ends of the section of rope are being pulled
waves moving in one dimension. The three dimen-
along the rope by the tension T. If the rope were
sional wave equation will be discussed later.
straight, if there were no curvature at this point, the
y tension forces would cancel each other and there
would be no net force on ∆ . Only because there is
curvature is there a net force which we have labeled
R(x1) Ty in Figure (9).
y(x) ∆
T

x x ∆θ Ty = Tsin ∆θ
x1 x2
2
∆θ T∆θ
1 = d y(x) R
T
R(x2) R(x) dx2

Figure 8 Figure 9
Wave pulse on a rope. The curvature is positive Due to the tension pulling on both
(points up) at x1 , and negative at x 2 . sides, this section of rope feels a
net downward force Ty ≈ T∆θ .
 Calculus 2000 - Chapter 2 Second Derivatives and 1D Wave Eq. Cal 2-11

As long as ∆ is short enough, this section of rope sides of the pulse do not become steep, the angle ϕ
will lie along the circle we have drawn to show the will be small, there will be very little rotation of
radius of curvature, and the two tension forces T Figure (9), and the net force Ty′ will point nearly
will be tangent to the circle at the two ends. The straight down and have a magnitude close to that
result, from geometry we have now seen several given by Equation (56).
times, is that the two angles labeled ∆θ will be equal
and the right hand tension force will have a down- On the other hand, if the pulse becomes steep, the net
ward pointing component Ty given by force is no longer y directed and our current analysis
will no longer apply. Whoever has watched ocean
Ty = T sin (∆θ) ≈ T∆θ (53) waves break as they approach the beach and become
where for small angles we can replace the sine of an steeper and steeper, will recognize that steep waves
angle by the angle itself. behave very differently from shallow ones. Here we
are working only with the theory of shallow waves.
From Figure (9) we see that the angle ∆θ is given by
Returning to Equation (56), which we have written
∆θ = ∆ (54) here again
R
so that d 2y(x)
Ty(x) = T∆ (56) repeated
dx 2
Ty = T∆θ = T∆ 1 (55)
R we want to point out that this equation gives us not
only the magnitude but also the direction of the net
Since 1/R is the curvature d 2y(x)/dx 2 at ∆ , we get
force Ty. Where the curvature d 2y(x)/dx 2 is posi-
d 2y(x) tive, as it is at point x 1 in Figure (8), the net force Ty
Ty(x) = T∆ (56) is directed upwards. Where the curvature is negative
dx 2
While Equation (56) was derived starting from the as at point x 2 , the net force Ty points down. Thus
Equation (56) for Ty(x) correctly changes sign
top of the pulse, we can see that as long as the sides
when the direction of the net force changes.
of the pulse are not steep, as long as we are dealing
with a shallow wave pulse, Equation (56) should Now that we have a reasonably general formula for
apply all along the wave. the net force Ty on a section ∆ of the rope (the only
approximation being the shallow wave approxima-
To see this, we have in Figure (10) analyzed the net
tion), we are ready to apply Newton's second law,
force Ty acting slightly to the left side of the top of
relating this net force to the mass ∆m and the
the pulse (at point x 2 in Figure (8)). Actually Figure
acceleration a(t) of this section.
(10) is the same as Figure (9), rotated by an angle
ϕ = dy(x)/dx which is the slope of the rope at point If the rope has a mass density µ kg/meter, then the
x 2 . Here is where the shallow wave approximation mass of a section of length ∆ is simply
comes in. As long as the wave is shallow and the
mass of
Ty' ∆m = µ∆ (57)
section ∆
∆θ T
∆
We need to think a bit more about the situation to
ϕ describe the acceleration of ∆m . So far we have
∆θ
R
described the rope by the curve y(x), which is
T essentially a single snapshot of the rope at some
special time t.
Figure 10
If the section of rope slopes at an angle ϕ , then the net
force Ty ′ slopes at the same angle. That has little effect
as long as the waves are shallow and ϕ remains small.
Cal 2-12 Calculus 2000 - Chapter 2 Second Derivatives and 1D Wave Eq.

Another way to look at a wave pulse is to look at one Partial Derivatives

point on the rope, and watch the point move up and When dealing with a function of two or more variables,
down as the pulse comes by. We can describe this like y(x,t), we have to be somewhat careful when we
changing height by the function y(t). The accelera- talk about derivatives. For now, we will always assume
tion a y(t) is then given by that if we are differentiating with respect to space, we
will hold the time variable constant, i.e., consider the
d 2y(t) curve at one instant of time. Conversely, if we are
a y(t) = (58)
dt 2 differentiating with respect to time, we will consider
Equation (58) is limited in that it describes the only one point in space, i.e., hold x constant. There is
motion of only one point of the rope. We can a special notation for these so called partial derivatives,
describe the motion of the whole rope for all times where we differentiate with respect to one variable
with a function y(x,t) that is a function of both space holding the other constant. In this notation we replace
and time. If we look at the rope at some instant of the d's, as in dx or dt by the symbol ∂ . Thus
time t 0 , then the shape of the rope is given by
dy(x,t) ∂y(x,t)
≡ (65)
y(x) = y(x,t) dx holding t constant ∂x
t = t0 (59)
while if we stand at one point x 0 , the motion of the dy(x,t) ∂y(x,t)
≡ (66)
rope is given by dx holding x constant ∂t
y(t) = y(x,t) With this notation we get, for
x = x0 (60)
An explicit example of such a function y(x,t) was y(x,t) = sin (kx – ωt) (67a)
our traveling wave formula of Equation (15-26) of
∂y(x,t)
the Physics text = k cos (kx – ωt) (67b)
∂x
y(x,t) = A sin (kx – ωt) (61) (also 15-26)
∂y(x,t)
= – ω cos kx – ωt (67c)
which as we saw represented a sinusoidal wave ∂t
traveling to the right at a speed
Using this new notation for partial derivatives, our
vwave = ω (62) (also 15-30) Equations (56) for the net force Ty on ∆ , and (58)
k for the acceleration a y of ∆ becomes
where the special frequency k is related to the
wavelength λ by k = 2π/λ , and the angular fre- ∂ 2y(x,t)
Ty(x,t) = T∆ (56a)
quency ω is related to period T by ω = 2π/ T . (As ∂x 2
a quick exercise, show that ω/k has the dimensions
of a velocity). ∂ 2 y(x,t)
a y (x,t) = (58a)
∂t 2
With Equation (61) for y(x,t), you can easily see that
if you look at the wave at one time, say t = 0, then To apply Newton's second law, we equate the net
force Ty(x,t) to the mass ∆m = µ∆ times the accel-
y(x,t) = y(x) = sin (kx) (63) eration a y (x,t) to get
t=0
is a pure spacial sine wave. If you look at one Ty(x,t) = ∆m a y(x,t)
particular point, for instance, x = 0, you get

y(x,t) = y(t) = sin (– ωt) (64) ∂ 2 y(x,t) ∂ 2 y(x,t)

T∆ = (µ∆ ) (68)
x=0 ∂x 2 ∂t 2
which is a pure sinusoidal oscillation.
 Calculus 2000 - Chapter 2 Second Derivatives and 1D Wave Eq. Cal 2-13

The factors of ∆ cancel, and after dividing through Using Equations (70b) and (70d) in Equation (69)
by µ we get gives
T (– Ak 2 ) sin(kx – ωt) =? – Aω 2sin(kx – ωt)
T ∂ y(x,t) = ∂ y(x,t)
2 2 µ
µ ∂x 2 ∂t 2
(69) (71)
The question mark in Equation (71) means that this
as our final differential equation for the motion of is a guess, and we still have to see if the guess works.
the wave pulse on the rope.
First we notice that the functions sin(kx – ωt) can-
How do you solve such a differential equation? As cel. We had to have this cancellation or there was no
we have mentioned several times, we guess an chance of making the two sides equal for all times t
answer for y(x,t), and plug the guess into the differ- and all positions x. We also note that the amplitudes
ential equation to see if we have made the correct A cancel, which means that the solution does not
guess. Also, we use whatever physics we have depend upon the amplitude A. After these cancella-
available to help us make a good guess. tions we get

Right now we do not have a formula for a single T (– k 2 ) = – ω 2

µ
pulse that we can use as a guess for a solution to
Equation (69). However we do have the formula in T = ω2 = v 2 (72)
Equation (61) for a sine wave traveling to the right µ k2
wave

at a speed v = ω/k where we noted that vwave = ω/k . Taking the square
root of Equation (72) gives
y(x,t) = A sin(kx – ωt) (61) repeated
T
To see if this traveling wave is a solution to our vwave = µ (73)
differential Equation (69), we have to take a number
of partial derivatives. They are which is the answer we got in the Physics text,
Equation (15-5), for the speed of a pulse on a rope.
∂y(x,t)
= ∂ A sin(kx – ωt)
∂x ∂x
= A k cos(kx – ωt) (70a)

∂ 2y(x,t)
= ∂ Ak cos(kx – ωt)
∂x 2 ∂x
= – A k 2 sin(kx – ωt) (70b)

∂y(x,t)
= ∂ Asin(kx – ωt)
∂t ∂t
= – ω Acos(kx – ωt) (70c)

∂ 2y(x,t)
= ∂ (– ω A)cos(kx – ωt)
∂t 2 ∂t
= – (– ω A)(– ω) sin(kx – ωt) (70d)
Cal 2-14 Calculus 2000 - Chapter 2 Second Derivatives and 1D Wave Eq.

The One Dimensional Wave Equation Exercise 7 gives us an important result. For our
If we go back to Equation (69), and replace T/µ by wave equation, which we got by considering wave
vwave2 , we get pulses that were not too steep, the sum of two or
more waves, each of which is a solution of the wave
∂ 2y(x,t) ∂ 2y(x,t) one equation, is itself a solution.
vwave2 = dimensional
∂x 2 ∂t 2 wave equation
In our discussion of Fourier analysis, introduced on
(74) page 16-6 of the Physics text, we saw that any
This is a general form of what is called the one continuous curve can be constructed from a sum of
dimensional wave equation. As we have just seen, a sine wave shapes. This suggests that we could
traveling sine wave, moving to the right at a speed construct a single wave pulse, moving to the left at
vwave is a solution to this equation. The following a speed vwave , by adding up a bunch of traveling sine
exercises demonstrate that waves traveling to the left, waves of different wavelengths λ i = 2π/k i, but all
and standing waves, are also solutions to this equation. with the same speed vwave = ω i /k i . The construc-
tion in Figure (11) suggests how we could add the
sine (actually cosine) waves to get a pulse. Since
Exercise 6
each wave is a solution to the same wave Equation
(a) The formula for a sine wave moving to the left at a
speed vwave = ω/k was given in Equation (15-33) of the
(73), the sum, i.e., the single pulse, is also a solution.
Physics text as From Figure (11), it should be clear that we can
y(x,t)wave moving left = A sin (kx + ωt) (15-33) construct a solution to the wave equation represent-
ing a pulse with very steep sides. However, in our
Show that this wave also obeys the wave Equation (73). analysis of the motion of the rope, we had to restrict
ourselves to shallow waves in order to derive the
(b) Later in Chapter 15 we saw that a standing wave,
wave equation for pulses on the rope. What this
which is the sum of a left moving and a right moving
means is that the wave equation has solutions that we
traveling wave, was given by the formula
will not see on the rope. The shallow pulses on the
y = A sin kx cos ωt (15-35) rope will obey the wave equation, but we should
expect that a steep pulse on the rope will behave
Show that this wave is also a solution to the wave differently. Not as differently as a breaking ocean
Equation. wave, but differently.
Exercise 7 Figure 11
Suppose you have two solutions y1(x,t) and y2(x,t) , both How to add cosine waves to get a pulse. At x = 0, all
of which are a solution to the wave equation with the the waves add to give a big amplitude y. As we go out
from x = 0, there is more and more cancellation until
same speed vwave . Show that the sum wave
the sum wave adds to zero. If all these are traveling
y(x,t) = y1(x,t) + y2(x,t) (75) waves moving to the right at the same speed
vwave = ω i k i , then the whole pulse must move at the
is also a solution of the same wave equation. same speed, maintaining its shape.
 Calculus 2000 - Chapter 2 Second Derivatives and 1D Wave Eq. Cal 2-15

Compressional Waves on a Spring There are two reasons why we have written ∆T as
When we came to the discussion of compressional kL(∆L /L) rather than just k∆L . The first is that
waves on a spring, in particular the compressional ∆L /L is the amount of stretch per unit length, a
Slinky wave we saw in Figure (Phys1-6) reproduced quantity engineers call strain. It is a more inherent
here, we resorted to dimensional analysis in Chapter property of the spring than the total stretch ∆L .
15 of the Physics text because there are no obvious
tricks to calculate the speed of the pulse. Now we are The second reason is that the product kL is also an
in a position to set up a differential equation describ- inherent property of the spring. In Chapter 15, page
ing the motion of a short segment ∆ of the spring. 15-7 of the Physics text, we saw that if you had two
We will get the wave equation, and from that we can identical springs of spring constant k, and attached
immediately tell the speed of the pulse. them together, you got a spring twice as long but
with half the spring constant. It is the product kL that
Suppose we have a stretched spring of length L as does not change when you connect identical springs
shown in Figure (12). The force required to stretch or cut a spring in half. Engineers would call this
the spring, which is equal to the tension T in the inherent property kL of the spring a spring modulus.
spring, is given by Hook's law as
To describe the stretched spring, we will introduce
T = k(L – L 0 ) (76) a function y(x) that represents the displacement of a
where L 0 is the unstretched length of the spring. point on the spring from its equilibrium (or initial)
position. When we stretch a spring from a length L
Now suppose that we stretch the spring an additional to a length L + ∆L , as shown in Figure (13), every
amount ∆L. The tension will increase by an amount point on the spring moves to the right a distance y(x)
∆T given by given by the formula
T + ∆T = k(L + ∆L – L 0) displacement
y(x) of a point = x ∆L (78)
= k(L – L 0) + k∆L on the spring L

Using Equation (76) to cancel the T and k(L – L 0 ) where x is the distance down the spring, starting at
terms, we are left with the left end. You can see where we got Equation
(78). If we are at the left end where x = 0, y(x) = 0
∆L
∆T = k∆L = kL (77) and there is no displacement. At the right end, where
L x = L, we get the full displacement
y(L) = (L/L)∆L = ∆L . In Equation (78) we are
assuming that the displacement increases uniformly
as we go down the spring.
L0 ∆L displacement
y(x)
Figure 1-6 (Physics 2000)
Compressional wave on a Slinky.
T
L
Figure 13
Figure 12 The displacement y(x) increases as we
A tension T L0 go down the spring. With the formula
stretches the y (x) = (x/L) ∆ L , we are assuming that
spring from T the displacement is increasing
a length L 0 L uniformly.
to a length L.
Cal 2-16 Calculus 2000 - Chapter 2 Second Derivatives and 1D Wave Eq.

If we differentiate y(x) with respect to x we get This gives us as the formula for the tension T(x) at
point x
= d x ∆L = ∆L
dy(x)
(79) T(x) = T0 + ∆T(x)
dx dx L L
Thus for a uniformly stretched spring, T(x) = T0 + kLy′(x) (82a)
y′(x) = dy(x)/dx is the amount of stretch per unit
where T0 is the equilibrium tension, and kLy′(x) is
length, which we have called the strain of the spring.
the change in tension caused by the displacement of
If the strain is not uniform, if for example, we have parts of the spring from their equilibrium position.
a compressional wave on the spring, the strain is still
Let us now apply Equation (82a) to a short section of
given by
spring of length ∆x , as shown in Figure (14). If the
local stretch dy(x) tension were uniform, the tension forces would
strain = per unit =
(80)
length dx cancel and there would be no net force on this section
of the spring. A net force arises only if there is a
To help see that y′(x) = dy(x)/dx is the local amount change in tension as we go from x to x + ∆x . This
of stretching per unit length, note that when we net force will be
integrate the local stretching per unit length over the net
total length of the spring, we get the total stretch ∆L . force = T(x +∆x) – T(x)
on ∆x
L L L
dy(x)
dx = dy(x) = y(x) = T0 +kLy′(x +∆x) – T0 +kLy′(x)
0
dx 0 0

= y(L) – y(0) (81) = kL y′(x +∆x) – y′(x)

= ∆L – 0 = ∆L
y′(x +∆x) – y′(x)
= kL∆x (83)
where y(L) = ∆L the total displacement at the end. ∆x

Now go back to Equation (77) We immediately see that the last quantity in the square
brackets is going to become, in the limit as ∆x → 0 ,
∆T = kL ∆L (77) repeated the second derivative of y(x) with respect to x. Thus our
L formula for the net force on a section of length ∆x is
which said that the change in tension in the spring is
proportional to the strain ∆L/L . We proved this was net d 2y(x)
force = kL∆x (84)
true for a uniform strain ∆L/L . The obvious gener- on ∆x dx 2
alization when the strain is not uniform is to replace
the average strain ∆L/L by the local strain If the spring has a mass per unit length of µ kg/meter ,
y′(x) = dy(x)/dx to get the mass ∆m of a length ∆x is

dy(x) ∆m = µ∆x (85)

∆T(x) = kL = kLy′(x) (82)
dx
∆x
where ∆T(x) is the increase in the tension in a point x T(x) T(x+∆x)
due to the local strain y′(x) .
x x+∆x
Figure 14
There will be a net force on this short section of spring
if the tension changes as we go from x to x+∆x.
 Calculus 2000 - Chapter 2 Second Derivatives and 1D Wave Eq. Cal 2-17

If we allow waves on the spring, the displacement The Speed of Sound

y(x) from equilibrium depends not only on the The analysis of compressional sound waves in air
position x down the spring, but also on the time (t) . can be carried out along lines very similar to our
Thus the displacement is described by the function analysis of a compressional wave on a spring. How-
y(x,t). The acceleration a y(x,t) at position x on the ever to do this, we need to build on our discussion of
spring is the behavior of an ideal gas in Chapters 17 and 18 of
∂ 2y(x,t) the Physics text. Thus we will assume that the reader
a y(x,t) = (86) is familiar with this material, including the discus-
∂t 2
sion on adiabatic expansion in the Chapter 18 appen-
We are using the partial derivative symbol ∂ be-
dix.
cause we want to measure the change in y(x,t) with
time at a fixed position x. Consider a column of gas with a cross sectional area
In terms of partial derivatives, Equation (84) for the A and length L as shown in Figure (15). We can
net force on ∆x is think of the gas as being in a cylinder with friction-
less walls, but it could be a hypothetical column in
net ∂ 2y(x,t) a large volume of gas. Let the variable x measure the
force = kL∆x (84a)
on ∆x ∂x 2 distance down the column, starting at the left end,
and imagine that we have a frictionless piston at the
With Equations (84a), (85) and (86), Newton's sec- right end.
ond law applied to ∆m gives
If we pull the piston out a small distance ∆L , we
net change the volume of the gas by an amount
force = (∆m) a y(x,t) Newton's law
F = ma
on ∆m ∆V = A∆L (90)
∂ 2y(x,t) ∂ 2y(x,t) and in so doing, decrease the pressure p.
kL∆x = (µ∆x) (87)
∂x 2 ∂t 2 How much the pressure changes depends upon the
The factors of ∆x cancel and we are left with way the gas is expanded. If we expand it very slowly
so that heat has time to flow into the gas and the
kL ∂ y(x,t) = ∂ y(x,t)
2 2
temperature remains constant (this is called an isother-
µ ∂x 2
(88)
∂t 2 mal expansion) then we have, from the ideal gas law
pV = NRT = constant isothermal (91)
We recognize Equation (88) as the wave equation expansion

2 ∂ 2y(x,t) ∂ 2y(x,t) where N is the number of moles of gas in the cylinder,

vwave = R is the gas constant, and T the temperature in kelvins.
∂x 2 ∂t 2
where we can identify the wave speed as However in a sound wave, expansions and compres-
sions happen so rapidly that there is not enough time
compressional kL for heat to flow in or out, and the temperature changes.
v Slinky wave = µ (89)

We got this same answer on page (15-8) of the area x ∆x ∆L

Physics text using dimensional analysis. However, A
with dimensional analysis we were not sure whether
a factor of 2 or π might be missing. Having derived
the wave equation, we know that kL /µ is the L
correct answer with no missing constant factors.
Figure 15
Column of gas of cross-sectional area A and length L.
Cal 2-18 Calculus 2000 - Chapter 2 Second Derivatives and 1D Wave Eq.

When we expand a gas with no heat flow, we call this We can use the fact that ∆V is very small compared
an adiabatic expansion. As we saw in the appendix to V0 to get
to Chapter 18, in an adiabatic expansion the gas γ γ
obeys the equation (V0 + ∆V) γ = V0 1 + ∆V = V0 γ 1 + ∆V
V0 V0
pV γ = constant adiabatic
expansion (92)
Using the approximation (1 + α) γ ≈ 1 + γα for a
where
small α , we have
cp
γ = c (93) γ
v 1 + ∆V ≈ 1 + γ ∆V (96)
is the ratio of the specific heat c p at constant V0 V0
pressure to the specific heat c v at constant volume. Using (96) in (95b), with p 0 + ∆p = p 0 (1 + ∆p/p 0) ,
It is Equation (92) for an adiabatic expansion rather gives
than Equation (91) for an isothermal expansion that
we need to use to describe the relationship between p 0V0 γ = (p 0 + ∆p)(V0 + ∆V) γ
(97)
pressure and volume for a sound wave. ∆p
= p 0 1 + p V0 γ 1 +γ ∆V
The quantity γ = c p/c v depends, as we saw at the 0 V0
beginning of Chapter 18, on the number of effective
degrees of freedom of the gas molecules. As you Multiplying this out gives
found if you did Exercise 2 of Chapter 18, for a
∆p γ∆V γ
monatomic gas like helium or argon with no rota- p 0V0 γ = p 0V0 γ 1 + p + + ∆p∆V
tional degrees of freedom, γ = 1.66 (5 3) . For di- 0 V0 p 0V0
atomic gases like oxygen, nitrogen, and of course
(98)
air, that have two rotational degrees of freedom,
The factors p 0V0 γ cancel, and we can neglect the
γ = 1.40 . When we get to more complex structures
second order term ∆p∆V , giving
like CO 2 and NH 4 , then γ drops to 1.28.
∆p γ∆V
1 = 1+ p +
We will now use Equation (92) for an adiabatic 0 V0
expansion to calculate the change ∆p in pressure
when we change the volume of the gas in the After canceling the 1's and multiplying through by
cylinder by an amount ∆V . Before we compress we p 0 we get for the pressure change ∆p
have ∆p = – γp 0 ∆V (99)
γ γ
V0
pV = p 0V0 (94)
If you look at the appendix to Chapter 18 in our
where p 0 and V0 are our original pressure and
discussion of the adiabatic expansion, you see that
volume. After the expansion, V goes to V0 + ∆V
we started with the equation
and p goes to p 0 + ∆p , where we know that ∆p is
negative for an expansion. Thus after the expansion γ p ∆V + ∆pV = 0 (18-A8)
0 0
we have
[which is Equation (99) if we solve for ∆p] and went
γ γ
pV = (p 0 + ∆p)(V0 + ∆V) (95a) through a number of calculus steps to derive
pV γ = constant . What we have done in going from
With pV γ = p 0V0 γ = constant, we get pV γ = constant to Equation (99) is to undo the
calculus steps in that appendix. However one typi-
p 0V0 γ = (p 0 + ∆p)(V0 + ∆V) γ (95b) cally remembers the equation pV γ = constant for
adiabatic expansions rather than Equation (18-A8),
and it seemed worthwhile to show how to get from
pV γ = constant to our formula for ∆p .
 Calculus 2000 - Chapter 2 Second Derivatives and 1D Wave Eq. Cal 2-19

Now that we have Equation (99) for ∆p , we can Now consider a section of the cylinder of length ∆x
follow essentially the same steps that we did earlier located at x as shown in Figure (16). The gas external
to calculate the speed of a compressional wave pulse to ∆x on the left, where the pressure is p(x), exerts
on a spring. a right directed force of magnitude

If the cylinder in Figure (15) has a cross sectional F(x) = Ap(x) (106)
area A, length L, and we move the piston out a while the gas on the right exerts a left directed force
distance ∆L , we have of magnitude
V0 = AL F(x + ∆x) = Ap(x + ∆x) (107)
∆V = A∆L (100) where we have used the fact that the force is the
thus from Equation (99) we have pressure times the area. The net force on ∆x is thus
Fnet on ∆x = F(x) – F(x + ∆x)
∆p = – γp 0 ∆V = – γp 0 A∆L (108)
V AL
= A p(x) – p(x + ∆x)
= – γp 0 ∆L (101)
L
Using Equation (105) for p(x) we get
In moving the piston out, the average displacement
of a molecule y(x) at position x will be
Fnet = A p 0 – γp 0 y′(x) – p 0 – γp 0 y′(x + ∆x)
y(x) = x ∆L (102)
L The p 0 terms cancel and we are left with
which is the same as our Equation (78) for the
average displacement of a piece of spring at position Fnet = Aγp 0 y′(x + ∆x) – y′(x)
x. Differentiating Equation (102) with respect to x
gives We can multiply by ∆x/∆x to get

= ∆L
dy(x)
y′(x) = (103)
dx L y′(x + ∆x) – y′(x)
Fnet = Aγp 0∆x (109)
Thus we see that for a uniform displacement of the gas ∆x
molecules, the strain, the displacement per unit length,
is y′(x) = dy(x)/dx. We will now assume that even for As in the case of the spring, we will end up taking the
non uniform displacements such as the kind we would limit as ∆x goes to zero, so that the term in the
have in a pressure pulse, y′(x) represents the local square brackets in Equation (109) becomes the sec-
strain or displacement per unit length. In terms of this ond derivative d 2y(x)/dx 2 .
local strain, our formula (101) for ∆p(x) becomes
local ∆x
∆p(x) = – γp 0y′(x) pressure (104)
change area p(x) p(x +∆x)
A
As in our discussion of springs, we can write this x
equation in the form Figure 16
Pressure forces acting on a small section
p(x) = p 0 + ∆p(x) of gas in our hypothetical cylinder.
(105)
p(x) = p 0 – γp 0 y′( x)
where we see that variations from the static pressure
p 0 are caused by local strains y′(x) .
Cal 2-20 Calculus 2000 - Chapter 2 Second Derivatives and 1D Wave Eq.

We will also let y(x) become a function of time was called the bulk modulus of the gas. Going back
y(x,t), so that second derivative becomes a partial to Equation (101), we have
derivative with respect to x only, and we get
∆p = = – γp 0 ∆L (101)
d 2y(x,t) L
Fnet = γp 0(A∆x) (110)
dx 2 = – γp 0 ∆V (117)
V
as our final formula for the net force on the gas in ∆x .
for an adiabatic expansion, and the same with a +
The next step is to calculate the mass ∆m of the gas in sign for compression. Thus
the region ∆x . If the density of the gas is ρ kg/meter 3
and the volume inside ∆x is (A∆x) meters 3, we have ∆p for adiabatic
= γp 0 = B (118)
∆V/V compression
∆m = ρA∆x (111)
The acceleration of the gas in ∆x is and our old formula for the speed of sound can be
written as
∂ 2y(x,t)
a x(t) = (112) γp 0
∂t 2 B =
vsound = ρ ρ (119)
Using Equations (110), (111), and (112) in Newton's which is the same result we got from the wave
second law gives equation.
Fnet on ∆x = ∆m a x(t)
Using the ideal gas law, we can re-express the
∂ y(x,t)
2
∂ y(x,t)2 quantity p 0/ρ in our formula for the speed of sound
γp 0(A∆x) = ρ(A∆x) (113) in terms of the temperature T of the gas and some
∂x 2 ∂t 2
other constants. First we will write the density ρ as
The factor A∆x cancels and we are left with
kg M kg/mole × N moles
γp 0 ∂ 2y(x,t) ∂ 2y(x,t) ρ 3
= (120)
ρ = (114) meter V meters 3
∂x 2 ∂t 2
and we get the wave equation where M is the mass of one mole of the gas (an
Avogadro's number of the gas molecules), N is the
2 ∂ 2y(x,t) ∂ 2y(x,t) number of moles in our cylinder, and V the volume
vwave = (74) repeated
∂x 2 ∂t 2 of the cylinder.
where we immediately see that the speed of the Next write the ideal gas law pV = NRT as
sound wave is given by
N = p (121)
γp 0 V RT
vsound = ρ (115) where R is the gas constant and T the temperature in
kelvins. Combining Equations (120) and (121) gives
In our discussion of sound waves in Chapter 15 of
Mp
the Physics text, where we used dimensional analy- ρ = MN =
sis to predict the speed of sound, we came up with the V RT
formula or we have
B p RT
vsound = ρ (116)
ρ = M (122)
where and our formula for the speed of sound becomes
∆p
B ≡ γp γRT
∆V/V vsound = ρ = M
(123)
 Calculus 2000 - Chapter 2 Second Derivatives and 1D Wave Eq. Cal 2-21

To interpret the physics of Equation (123), it is Aside from its dependence on the mass of the gas
perhaps clearer to express the answer in terms of molecules, the other important feature is that the
mass of the gas molecules involved. We have speed of sound is proportional to the square root of
kilograms/mole temperature. Thus the warmer the gas the greater the
m molecule = M speed. This dependence of the speed of sound on the
N A molecules/mole
square root of temperature leads to a close connec-
kilograms tion between the speed of sound and the average
= M
N A molecule speed of the air molecules due to their thermal
where N A is Avogadro's number, and motion.

joules /mole kelvin joules In our discussion of the ideal gas law, we used the
k = R = R fact that the temperature was a measure of the
N A molecules/mole N A kelvin
average thermal kinetic energy of the gas, the pre-
is Boltzman's constant. Thus cise relationship being
R = N Ak
= m k (124) 1m v 2 = 3 kT (126)
M N Am molecule molecule 2 molecule 2
and in terms of the molecular mass m molecule we get where v 2 is the average of the square of the speed of
the gas molecules (v 2 = vx 2 + vy 2 + vz2 ) . Writing
Equation (126) in the form
γkT
vsound = m molecule v2
(125) kT = (127)
m molecule 3
From Equation (125), we immediately see that for a
and using this in Equation (125) gives
gas like hydrogen consisting of light molecules, the
speed of sound is considerably greater than in a gas
with heavy molecules. γkT γv 2
vsound = =
m molecule 3
Exercise 8
Calculate the speed of sound at a temperature of 300
γ
kelvin, in hydrogen, helium, nitrogen and CO2 . Use vsound = v
the fact that a hydrogen molecule has the mass of 2 3
(128)
protons, a helium atom the mass of 4 protons (with a
nucleus of 2 protons and 2 neutrons), a nitrogen Several times we mentioned that the speed of sound
molecule the mass of 28 protons (each nucleus has
is closely related to the speed of the air molecules
7 protons and usually 7 neutrons) and a CO2 mol-
ecule has a mass of around 44 protons (carbon
due to their thermal motion. Equation (128) gives us
nucleus has 6 protons and 6 or 7 neutrons, oxygen the precise relationship. For air, for example, where
has 8 protons and 8 neutrons, for a total of 12 + 16 + γ = 1.28 we get
16 = 44 nuclear particles).
1.28
vsound = v = .65 v (129)
3

Sound travels over half as fast as the average speed

v of the air molecules.
 Calculus 2000 - Chapter 3 The Gradient Cal 3-1

Calculus 2000-Chapter 3
The Gradient
CHAPTER 3 THE GRADIENT

The gradient operation represents the fundamental

way that we go from a scalar field like the electric
voltage V to a vector field like the electric field E .

In this chapter, we present two distinct ways to

introduce the gradient operation. One is to use the
fact that electric fields are related to electric voltage
the same way that forces are related to potential
energy. The second, more geometrical way, is to
picture the electric voltage as being described by a
contour map, and that the electric field is described
by the lines of steepest decent in the map. We present
these two points of view as separate sections, View
1 and View 2, that can be read in either order.

We end the chapter with View 3, an application to

fluids, where we see that the pressure force f p acting
on fluid particles is the gradient of the pressure field
p. This represents a straightforward example of
obtaining a vector field f p from a scalar field p.
Cal 3-2 Calculus 2000 - Chapter 3 The Gradient

TWO VIEWS OF THE GRADIENT While mapping the voltage V(x,y) at various points
In the Physics text, our first laboratory exercise on in the water was a straightforward process, our
electric phenomena was the potential plotting ex- construction of the electric field lines E(x,y) was not
periment illustrated in Figure (25-10) reproduced so obvious. Our procedure was to map E(x,y) by
here. Two small brass cylinders connected to a drawing a set of lines perpendicular to the equipoten-
battery were placed in a shallow tray of slightly tial lines as shown in Figure (25-12). With this
conducting water. In order to measure the distribu- technique we were just barely able to tell whether the
tion of voltages V(x,y) at various points (x,y) in the resulting field E(x,y) more closely resembled the
water, we had two probes of bent, stiff, wire attached field of line charges or point charges. Our technique
to blocks of wood, adjusted so that the tips of the was conceptually correct, but a very crude way to
wire stuck down in the water. The other end of the determine the electric field E(x,y) from a map of the
wire probes were attached to a voltmeter as shown. voltage V(x,y).

By leaving one probe fixed, and moving the other in

a way that the reading on the voltmeter remained
constant, we could map out lines of constant voltage
in the water. The results from a student lab notebook
are shown slightly cleaned up in Figure (25-11).
These lines of constant voltage are also known by
the name equipotential lines or lines of equal elec-
tric potential. We also pointed out that these lines
were analogous to lines of equal height, the contour
lines in a contour map of the countryside.
battery
probes
Figure 25-11
Plot of the equipotential lines from a student project by
B. J. Grattan. Instead of a tray of water, Grattan used a
A volt sheet of conductive paper, painting two circles with
V meter aluminum paint to replace the brass cylinders. We used
brass B the Adobe Illustrator® program to draw the lines
cylinders through Grattan's data points.

tap water pyrex dish

Figure 25-10 (from Physics text) A
Simple setup for plotting fields. You plot equipotentials
by placing one probe (A) at a given position and moving
the other (B) around. Whenever the voltage V on the
voltmeter reads zero, the probes are at points of equal
potential.

Figure 25-12
To sketch the field lines, draw smooth lines, always
perpendicular to the equipotential lines, and maintain
any symmetry that should be there.
 Calculus 2000 - Chapter 3 The Gradient Cal 3-3

After this initial experiment, we resorted to com- build our discussion of the gradient operation either
puter plots, like the one shown in Figure (25-15), to upon our knowledge of the mathematics of the
see the relationship between the electric field and a electric field, or by developing the ideas from a
voltage map. discussion of the nature of a hilly terrain. While both
approaches are equivalent, we see the subject from
The computer plots, and the models we constructed two rather different points of view. The electric field
from them, nicely illustrate the geometrical relation- approach is more efficient, while the hilly terrain
ship between a voltage map and the electric field approach develops some concepts that we will need
lines, but did not provide a convenient technique for later on.
actually calculating the field. The missing tech-
nique, which is the subject of this chapter, is the As we mentioned in the introduction, we will begin
mathematical procedure called the gradient, a pro- this chapter with the electric field approach, and
cedure involving the partial derivatives of the volt- later discuss the hilly terrain viewpoint separately in
age function V(x,y). View 2. You should study both approaches to see
this important topic from two points of view. It does
As Figure (25-15) illustrates, there is a complete not really matter which one you study first.
analogy between the contour map of a hilly terrain
and electric field plots from a voltage map. We can

V = .1

V = .2

V = .3

V = .4

V = .5

.1
V=
0 1
.
V=
–.
V=

–1 +3
Figure 25-15
Computer plot of the field lines and equipotentials for a charge distribution
consisting of a positive charge + 3 and a negative charge – 1. These lines
were then used to construct the plywood model.
Cal 3-4 Calculus 2000 - Chapter 3 The Gradient

View 1
The Gradient from a
Force – Energy Perspective
CALCULATING THE ELECTRIC FIELD Now imagine that I move the test particle along the
Figure (1) shows a small section of the voltage map dashed line from point A at .3 volts over to point B
of Figure (25-15) on the previous page. The solid at .4 volts. The potential energy of the particle has
lines are equipotential lines, lines of constant volt- increased from .3 joules to .4 joules. Thus to move
age spaced .1 volts apart. We want to imagine that the particle, I must supply (.1) joules of energy to the
we actually have a detailed map of the voltage particle.
V(x,y) at every point (x,y), and want to mathemati- Imagine that I move the test particle slowly, so that
cally determine, from that map, the electric field the force Fme(x,y) that I exert on the particle is just
E(x,y) at every point. enough to oppose the force E(x,y) that the electric
In the Physics text, we emphasized the idea that the field is exerting on the particle. Thus for the entire
electric voltage V(x,y) was the electric potential trip from A to B we have
energy of a unit test charge, while the electric field Fme(x,y) = – E(x,y) (1)
E(x,y) was the electric force on a unit test charge.
Thus the connection between E and V is the rela- The amount of work I do in moving the particle is
tionship between force and potential energy. given by the formula first discussed in Chapter 10 of
the Physics text (see page 10-15, Equation (10-25)).
To review this relationship, imagine that I place a
work I do in B
unit test particle at point A in Figure (1), where the moving the = F me⋅ d (2)
voltage is VA = .3 volts. Since the voltage is the test particle A
potential energy, in joules, of a unit test charge, our test
Because I am moving the particle slowly so that all
particle at point A has a potential energy of .3 joules.
the work I do is stored as electric potential energy,
and because the increase of potential energy of the
V = .2 unit test charge is VB – VA , we have
B
F me⋅ d = VB – VA (3)
A
We can get me out of the equation by using Equation
V = .3
(1) to give

V = .4 B
B – E ⋅d = V B – VA (4)
A
A V = .5
Equation (4) is the integral equation that relates the
voltage V(x,y) to the electric field E(x,y) . It is a
relationship we used extensively in the Physics text.
Figure 1
In the Calculus text, we will often translate from
A small section of the voltage map, showing
equipotential lines spaced .1 volts apart. We will integral to differential equations, and this chapter on
calculate the amount of work required to move a the gradient will be our first example of how this is
unit test charge from point A to point B. done.
 Calculus 2000 - Chapter 3 The Gradient Cal 3-5

The first step in going to a differential equation is to In going along the x axis, we have
focus in on a very small region of Figure (1), a region
V3 – V1 = – E(x i,y i ) ⋅ ∆ x
shown in Figure (2), centered at the point (x i,y i ) on (7)
the path from A to B. We have zoomed in so closely = – E x(x i,y i )∆x
to the point (x i,y i ) in Figure (2), we have so greatly
where the dot product of E with the x directed
magnified the plot, that the equipotential lines and
displacement ∆x leaves us with the x component
the field lines in this region are simply straight lines
E x . Writing out V1 and V3 in the form
at right angles to each other.
V1 = V(x i,y i )
Now suppose we move our test particle from point
V3 = V(x i +∆x, y i )
(1) at (x i,y i) over a distance ∆ along the path to
point (2) as shown. Equation (4) applied to this short Equation (7) becomes
displacement is V(x i +∆x, y i ) – V(x i,y i ) = – E x(x i,y i )∆x (8)
2 Dividing through by –∆x gives
V2 – V1 = – E(x,y) ⋅ d (5)
1 V(x i +∆x, y i ) – V(x i,y i )
For this short path, we can assume that E(x,y) is E x(x i,y i ) = –
∆x
essentially constant and replace the integral by the
product E(x i,y i) ⋅ ∆ , giving us When we take the limit that ∆ goes to zero, both ∆x
and ∆y will go to zero, giving
V2 – V1 = – E(x i,y i) ⋅ ∆ (6)
V(x i +∆x, y i ) – V(x i,y i )
[You can see that in going from Equation (5) to (6) E x(x i,y i ) = – limit
∆x→ 0 ∆x
we are essentially undoing the step we took in
Chapter (10) to derive the integral Equation (4).] (9)
By now you should recognize that the limit in
We are discussing the electric field of point charges.
Equation (9) is the partial x derivative of the function
This is a conservative field, which is a fancy way of
V(x,y) evaluated at the point (x i,y i ) . Since this is
saying that the change in potential energy when we
true for any point (x,y), we get
move a particle between two points does not depend
upon the path we take. Thus if we first go a distance ∂V(x,y)
∆x along the x axis to point (3), then up the y axis E x(x,y) = –
∂x (10)
a distance ∆y to point (2), we should get the same
change in voltage V2 – V1 that we got by going where the symbol ∂ is used for partial derivatives.
directly from point (1) to point (2) along ∆ .
Exercise 1
y Use the above line of reasoning to show that
to B ∂V(x,y) (11)
E y(x,y) = –
∂y
(2)
∆ i Introducing the unit vectors x and y , we can combine
∆y Equations (10) and (11) into the single vector equation
(1)
y
i ∆x (3) E(x,y) = x E x(x,y) + y E y(x,y)
E(x i,yi )
∂V(x,y) ∂V(x,y)
E(x,y) = – x +y
to A x ∂x ∂y (12)
xi
Figure 2
If we zoom in far enough, we reach a point where the
Equation (12) is the differential equation we can use
equipotential lines and contour lines are straight lines
perpendicular to each other. to calculate the electric field E(x,y) at every point
from a knowledge of the voltage V(x,y).
Cal 3-6 Calculus 2000 - Chapter 3 The Gradient

Interpretation To interpret Equation (14), imagine that we smooth out our

plywood model of the voltage surface, then saw the model
∂V(x,y) ∂V(x,y) in two, cutting through the point A with the saw blade
E(x,y) = – x +y (12)
∂x ∂y oriented along the x axis, along the dotted line in Figure (3).
A side view of the upper piece is shown at the bottom of
To help interpret Equation (12) repeated above, let us go Figure (4). You can see that the voltage at the beginning of
back to Figure (25-15) where we started with a plot of the the cut, point C, is somewhat greater than .1 volts, and rises
equipotential lines of the voltage V(x,y) and constructed to just over .4 volts at the end, point D. The mathematical
a three dimensional plywood model of the voltage. The formula for the curve we see in Figure (4) is V(x,y A) , and
equipotential lines became the contour lines of this the partial derivative with respect to x at point A is the slope
model, and the perpendicular electric field lines are the of the curve V(x,y A) at x = x A . This is just the tangent
lines of steepest slope. If you were standing on terrain of the angle θ in Figure (4).
represented by this model, and the slope became slip-
pery, the field line is the direction you would start to
slope at point ∂V(x,y A)
A going in = = tanθ (15)
slide. Ski instructors call this direction of steepest slope x direction ∂x x = xA
the fall line.
This is the maximum slope at point A. If we sawed through
To simplify the job of interpreting Equation (12), imagine point A, orienting the saw blade in any other direction, the
slop at point A would be less. In particular the slope would
that we are standing at the point A = (x A,y A ) shown in
be zero if we oriented the saw in the y direction.
Figure (3), where the contour line happens to be running in
the y direction. If we move along a contour line there is no From this discussion we see that the vector E(x,y) points
change in height, thus the partial derivative of V(x,y) with in the direction of the maximum slope and has a magnitude
respect to y—the rate of change of V(x,y) in the y direc- equal to that slope. The minus sign results from the fact that
tion—is zero at point A. the force E is in the downward direction toward lower
energy, while the positive slope, or gradient as we will call
∂V(x,y) it, is in the upward direction.
= 0
∂y x = xA
(13)
y = yA

The formula for E(x,y) at point A becomes .2

=
∂V(x,y)
V
E(x,y) = –x
x = xA
y = yA
∂x x = xA
y = yA
(14)
=.
3
V
y path through A
C A D
.4
5
=.
=
V

.1
V=
0 1
.
V=
–.
V=

C D top view
V(x,yA)
.4 D
.3
A θ
.2 oug h A
height of path thr
C .1

xC side view xA xD
x
Figure 3 Figure 4
The V = .2 volt contour line passes straight up through The top view shows the point A and the horizontal path
the point labeled A. Imagine that the surface is through that point. The side view shows the path we
smoothed out and you walk along the dotted line. would have to climb if the surface were smooth. The
steepest slope at the point A is in the +x direction and is
the tangent of the angle labeled θ.
 Calculus 2000 - Chapter 3 The Gradient Cal 3-7

THE GRADIENT OPERATOR Another mathematical concept, which we did not

The extension of Equation (12) to the case where the bother naming in the Physics text, is the scalar field.
voltage varies in three dimensions, where V = V(x,y,z) It is a quantity that has a scalar or numerical value at
is fairly obvious. It is every point in space. An example of a scalar field is
voltage, the potential energy of a unit test charge. At
∂V(x,y,z) ∂V(x,y,z) ∂V(x,y,z) every point in space that we place the unit test
E(x,y,z) = – x +y +z
∂x ∂y ∂z charge, we get a voltage reading. Since energy has
a magnitude but does not point anywhere, this read-
(16) ing has a scalar or numerical value only.
Until the beginning of the 20th century, research
papers and textbooks dealing with partial deriva- From Equation (19), we see that the gradient opera-
tives used notation similar to Equation (16), and the tor ∇ , operating on a scalar field V creates the vector
formulas could become cumbersome and difficult to field E = – ∇V . The vector ∇V has a numerical
read. It was Willard Gibbs who introduced the value equal to the maximum slope of V(x,y,z), and
gradient operation ∇ defined by the equation points opposite to the direction where the slope is
greatest.
∂ ∂ ∂
∇ ≡ x +y +z
∂x ∂y ∂z In the remainder of this part of the chapter, we will
give examples of using the gradient operation to
≡ x ∇x + y ∇y + z ∇z (17)
calculate the electric field from the voltage. In only
where ∇ x = ∂/∂x , etc. a few cases, like the example of the parallel plate
capacitor, is a Cartesian coordinate system (x,y,z)
We call ∇ an operator because it does not have an the most convenient coordinate system to use. In our
explicit meaning until it operates on something like study of electric and magnetic phenomena, we often
the voltage function V(x,y,z). dealt with point charges where there is spherical
symmetry or line charges with cylindrical symme-
∂V ∂V ∂V
∇V(x,y,z) = x +y +z try. We will see that to handle problems with
∂x ∂y ∂z spherical or cylindrical symmetry, it is much easier
(18)
= x ∇ xV + y ∇ yV + z ∇ zV to work with the gradient ∇V expressed in spherical
or cylindrical coordinate systems. Much of the
With this notation, the formula for the electric field detailed work for the remainder of the chapter will
E(x,y,z) in terms of the voltage V(x,y,z) is be to work out the formulas for the gradient in these
coordinate systems. (You do these derivations once,
E(x,y,z) = – ∇V(x,y,z) and then use the results for the remainder of your
(19) scientific career.)
We say that the electric field E is minus the As we mentioned, we have View 2 later in the
gradient of the voltage V. chapter, where we look at the gradient from a more
geometrical and mathematical point of view. We
In the Physics text, we defined a vector field as a
end up with Equation (16) as the formula for the
quantity with a vector value at every point in space.
gradient, but explicitly demonstrate that the compo-
We began our discussion of vector fields in Chapter
nents ∇ xV and ∇ yV of the gradient transform
23 with the velocity field rather than the electric field
(change) the same way the components of a dis-
because the velocity field is easier to visualize. At
placement vector change when we go to a rotated
any point in space the vector is simply the velocity
coordinate system. Such discussions will become
vector of the fluid particle located there. For the
very useful later on.
electric field we first have to invent the concept of a
tiny unit test charge before we can visualize the force
vector at each point in space.
Cal 3-8 Calculus 2000 - Chapter 3 The Gradient

THE PARALLEL PLATE CAPACITOR The other partial derivatives are

We introduced the parallel plate capacitor in Chap-
∂V(x,y,z)
= ∂ 0
ter 26, page 26-14 of the Physics text. We dealt with V y V
= 0 (24)
an idealized situation where we assumed that the ∂y ∂y d d
plate diameters were much greater than the separa-
tion. Then we could neglect edge effects and assume ∂V(x,y,z)
= ∂ 0
V y
= 0 (25)
that the electric field was uniform between the ∂z ∂z d
plates, as shown in Figure (26-27) reproduced here.
Using Equations (23), (24), and (25) in (22) gives us
Since E is the force on a unit test charge, and the
voltage V is its potential energy, we can calculate the V0
voltage V between the plates by calculating the amount E = –y (26)
d
of work required to lift the unit charge a distance y
above the bottom plate. Since the force E we have to which says that E points down in the –y direction,
work against is constant, the work we do is simply the and has a magnitude V0/d which we already know
force of magnitude E times the height y. If we say that from Equation (21). We see that the calculation of
the bottom plate is grounded, i.e., define the potential E from V using E = – ∇V is a fairly straightfor-
energy or voltage as being zero at the bottom plate, then ward process.
the formula for the voltage between the plates is simply

V(x,y,z) = E y (20) battery

To evaluate E, we note that when we get to the top

+ A
plate where y = d, the voltage is up to V0 , the voltage d
to which we charged the capacitor –
V0 = Ed (21) capacitor plates
Figure 26-25
Thus E = V0/d , and the voltage between the plates is The parallel plate capacitor. The capacitor is charged
given by up by connecting a battery across the plates as shown.

V0
V(x,y,z) = y (22)
d
Let us now turn the problem around and use the + + + + + + + + + + + + +
gradient formula E = – ∇V to calculate the electric
field E from our voltage formula Equation (22). – – – – – – – – – – – – –
Writing out all the components of – ∇V as partial
derivatives, we have from Equation (16)
Figure 26-26

E(x,y,z) = – x ∂V + y ∂V + z ∂V
The electric field between and around
(16)
∂x ∂y ∂z the edge of the capacitor plates.
plate of area A
The x partial derivative is + + + + + + + + + + + + +

∂V(x,y,z) d
= ∂ 0
V y
= 0 (23)
∂x ∂x d – – – – – – – – – – – – –

This is zero because there is no x dependence in our Figure 26-27

In our idealized parallel plate capacitor the field
formula for V. When we take the partial derivative
lines go straight from the positive to the negative
with respect to x, we hold y and z constant. Thus plate, and the field is uniform between the plates.
nothing in the formula V0y/d changes when we
change x, and this partial derivative is zero.
 Calculus 2000 - Chapter 3 The Gradient Cal 3-9

Voltage Inside a Conductor V0 d 0

The main idea of Chapter 26 of the Physics text was
that you cannot have a static electric field inside a
conductor if there is no flow of charge. The equiva-
lent statement in terms of electric voltage is that the
voltage is constant inside a conductor
E
V(x,y,z) inside a conductor = constant (27)

To see that this gives a zero electric field, we have

E = –∇V inside a conductor = 0 (28)

All the components are zero because the partial V0

derivative of a constant is zero.

To provide an explicit example, suppose we turn our

parallel plate capacitor on its side and assume that it 0
is constructed from thick metal plates as shown in Figure 5
Figure (5). The voltage as a function of distance is Voltage in a parallel plate capacitor. The voltage is
shown below the drawing of the plates. Inside the constant inside the plates and, for the assumed uniform
left plate the voltage has the constant value V0 , field structure, drops uniformly between the plates.
which gives zero field inside. Between the plates the
voltage drops uniformly. It has a constant gradient,
which gives us a constant electric field E = – ∇V
pointing in the direction of the downward slope. The
voltage is again constant (V = 0) in the left hand
plate.
Cal 3-10 Calculus 2000 - Chapter 3 The Gradient

ELECTRIC FIELD OF A POINT CHARGE Our first step will be to write out the vector equation
Our first example of an electric field in the Physics E = – ∇V as three component equations
text was the field of a point charge. If we have a E x = – ∂V ; E y = – ∂V ; E z = – ∂V
charge Q located at the origin of our coordinate ∂x ∂y ∂z
system, then the electric field at a position r = (x,y,z) (31)
as shown in Figure (6) is given by Focusing on the x component equation we have

E x = – ∂V = – ∂ r
kQ
Q kQ
E(r) = r 2
= r 2 (29) ∂x ∂x
4πε 0r r
Taking the constant kQ outside the derivative we have

E x = – kQ ∂ 1r
where r is a unit vector in the r direction and
(32)
k = 1/4πε0 . ∂x
To go any farther, we have to express the distance r
In the Physics text, we mentioned, but never accu-
as a function of the coordinate x. This is done by the
rately derived, that the voltage V(r) of a point charge
three dimensional Pythagorean theorem
was
Q kQ r = x2 + y2 + z2
V(r) = = r (30) To calculate the derivative of (1/r) with respect to x
4πε 0r
now becomes an exercise in the use of the chain rule
when we chose the zero of potential energy at for differentiation. Let us start with
r = infinity. What we want to do now is to show that
the formula for E(r) follows directly from Equation r 2 = x2 + y2 + z2
(30) for V(r) when we use the relationship which is easy to differentiate. We get
E = – ∇V (14) repeated ∂r 2 = ∂ x 2 + y 2 + z 2 = 2x (33)
∂x ∂x
The work is a bit messy, because we will be using a Next look at
Cartesian coordinate system to solve a problem with
∂r = ∂ r 2 = ∂ r ∂r 2
2
spherical symmetry. Later we will find the formula (34)
for the gradient in spherical coordinates, and then ∂x ∂x ∂r 2 ∂x
see that it is very easy to evaluate E = – ∇V for a To evaluate ∂ r 2 /∂r 2 , set y = r 2 so that we have,
point charge. using ∂y n/∂y = ny n – 1
z
∂ r2 ∂ y
= = ∂ y .5 = 1 y – .5 = 1 (35)
r ∂r 2 ∂y ∂y 2 2r
r Thus using Equation (33) and (35) in (34) gives

∂r = ∂ r ∂r 2 = 1 2x
Q y 2
∂x ∂r 2 ∂x 2 2r

x ∂r = x
∂x r
(36)
Figure 6
Out at a point given by the coordinate which is a fairly simple result considering what we
vector r , we have the unit vector r . went through.
 Calculus 2000 - Chapter 3 The Gradient Cal 3-11

Finally we have The messiness we encountered calculating the field

of a point charge from V = kQ/r resulted from our
∂ 1 = ∂r –1 = ∂r –1 ∂r
∂x r ∂x ∂r ∂x calculating x, y, and z components of E when we
knew that E pointed in the radial direction. If we use
= –r – 2 ∂r = – 12 xr what is called a spherical coordinate system, we
∂x r
will find that the formula for the radial component of
∂ 1 = – x the electric field is simply
∂x r r3 (37)
– ∂V(r)
Er = (41)
and our formula for E x becomes ∂r
With V(r) = kQ/r we get
E x = – kQ ∂ 1r = – kQ – x3
∂x r
E r = – kQ ∂ 1r = – kQ – 12 = 2 (42)
kQ
E x = kQ x3 ∂r r r
r (38a)
and we get the final answer in a one line calculation.
Clearly the y and z components are
To get this simple result requires, however, a fair
y amount of work deriving the formula for the gradi-
E y = kQ 3
r (38b) ent in spherical coordinates. First we have to define
precisely what a spherical coordinate system is,
E z = kQ z3 show what the unit vectors are, and then calculate
r the components of the gradient when we move in the
(38c)
directions defined by the unit vectors. When this is
To check that we got the right answer, we can go all done, when we have the formula for the gradient
back to Equation (29) in spherical coordinates, we can use the formula
kQ without ever going through the derivation again.
E(r) = r (29) repeated
r2 In the Physics text we encountered problems with
and replace the unit vector r with its definition r /r plane symmetry, like the parallel plate capacitor,
giving cylindrical symmetry, like the field of a line charge,
r = (r x,r y,r z ) = (x,y,z) and spherical symmetry like the field of a point
charge we have just discussed. The plane symmetry
problems are most easily handled in a Cartesian
r = rr = 1r (x,y,z) coordinate system, the cylindrical problems in what
y is called a cylindrical coordinate system, and spheri-
r x = xr ; r y = r ; r z = zr (39) cal problems in a spherical coordinate system. We
will now discuss these three coordinate systems and
Equation (39) says, for example, that the x compo- develop the formulas for the components of the
nent of the unit vector r has a length x/r. Thus the x gradient vector in each coordinate system. Since we
component of E in Equation (29) is have already done this for the Cartesian coordinate
kQ kQ system, that discussion will serve as a review of the
E x = r x 2 = xr 2 = kQ x3 (40) procedure we will use.
r r r
with similar equations for E y and E z. Since Equations
(38) and (40) are the same, we have verified that
E = – ∇V gives the correct result for V = kQ/r.
Cal 3-12 Calculus 2000 - Chapter 3 The Gradient

GRADIENT IN THE CARTESIAN Using Equation (14) in (45), we can eliminate E and
COORDINATE SYSTEM get the relationship between the small change in
voltage ∆V and the voltage gradient ∇V
An example of a right handed Cartesian coordinate
system is shown in Figure (7). Out at some
point r = (x,y,z) the unit vectors x , y , and z are ∆V = (∇V) ⋅ ∆ (46)
parallel to the x, y, and z axis as shown. It is called
a right handed coordinate system because the unit
vectors obey the relationship Equation (46) will allow us to find the formula for
the gradient in the various coordinate systems.
x×y = z (43) To see how we are going to use Equation (46), we
when we use the right hand rule for the cross prod- will start with the Cartesian coordinate system and
uct. (If we used a left hand rule, the z axis would choose ∆ to be a short step ∆x in the x direction.
point the other way.) Explicitly we will start at a point (x,y,z) and move to
the point (x + ∆x, y, z) so that ∆V , ∆ and
(∇V) ⋅ ∆ become
Exercise 1
Show that y × z = x and z × x = y . ∆V = V(x + ∆x, y, z) – V(x, y, z) (47)

We will use the force/potential energy relationship ∆ = x∆x (48)

to define the gradient vector. If I move a unit test
charge a short distance ∆ , exerting a force Fme = – E (∇V) ⋅ ∆ = (∇V) x ∆x (49)
to just overcome the electric field E , the work ∆W
I do is Using (47) and (49) in (46) gives
∆W = Fme ⋅ ∆ = – E ⋅ ∆ (44) V(x + ∆x, y, z) – V(x, y, z) = (∇V) x ∆x (50)
Since this work is the change ∆V in the potential Dividing through by ∆x and taking the limit as ∆x
energy of the unit test charge, we have goes to zero gives
∆V = – E ⋅ ∆ (45) V(x + ∆x, y, z) – V(x, y, z)
(∇V) x = limit
But the voltage V is related to the field E by the ∆x → 0 ∆x
gradient
(51)
E = – ∇V (14) repeated which is the definition of the partial derivative. Thus
z ∂V(x, y, z)
z (∇V) x = (52)
∂x
p which is our earlier result. This procedure does not
y
r give us anything new for a Cartesian coordinate
x system, but will give us new results for other coor-
y
dinate systems.

(On the next page you will find two pictures of our
x model of the electric field of two point charges. We
Figure 7 put the pictures there so that the discussion of the
The unit vectors x , y , z out at the point r. gradient in cylindrical and spherical coordinates
would each be completed on facing pages.)
 Calculus 2000 - Chapter 3 The Gradient Cal 3-13

Figure 25-14 (from Physics text)

Different views of the model of the electric field of two
point charges Q+ = +3 and Q – = –1.
Cal 3-14 Calculus 2000 - Chapter 3 The Gradient

GRADIENT IN CYLINDRICAL COORDINATES To calculate the first component (∇V) r , we will start
In a cylindrical coordinate system, we define the at the point p at (r, θ, z) and move a short distance
location of a point p by giving the distance r out ∆r in the r direction, to the point (r +∆r, θ, z) . Our
from the z axis, the angle θ over from the x axis, and change in voltage ∆V , displacement ∆ and the dot
the height z above the xy plane as shown in Figure product (∇V) ⋅ ∆ are for this move
(8). The unit vectors are r which points radially out ∆V = V(r + ∆r, θ, z) – V(r, θ, z) (55)
from the z axis, z which points in the z direction, and
θ which is perpendicular to the r z plane. The ∆ = r ∆r (56)
direction of θ is the direction we move when in-
creasing the angle θ . This gives us a right handed
coordinate system where the unit vectors are related (∇V)⋅∆ = r (∇V) r + θ (∇V) θ + z (∇V) z ⋅ r ∆r
by
(57)
r×θ =z (53) Since the unit vectors are all at right angles to each
other, r ⋅ r = 1 , θ ⋅ r = 0 and z ⋅ r = 0 , giving us
You should check for yourself that Equation (53)
works for the unit vectors shown in Figure (8), and ∆V = (∇V) ⋅ ∆ = (∇V) r ∆r (58)
that θ × z = r and z × r = θ .
Dividing (58) through by ∆r , using (55) for ∆V and
We will assume that in cylindrical coordinates, the taking the limit as ∆r goes to zero gives
gradient vector at point p is given by the equation
V(r + ∆r, θ, z) – V(r, θ, z)
∇V = r (∇V) r + θ (∇V) θ + z (∇V) z (54) (∇V) r = limit
∆r → 0 ∆r
where (∇V) r , (∇V) θ and (∇V) z are the compo- (59)
nents of the gradient vector that we want to deter- The right side of Equation (59) is what we will define
mine. to be the partial derivative of V(r, θ, z) with respect
z to r in cylindrical coordinates
r z
θ ∂V(r, θ, z) limit V(r + ∆r, θ, z) – V(r, θ, z)
≡ ∆r
∂r →0 ∆r
p r (60)
This is the rate of change of the function V(r, θ, z)
y as we change the r coordinate. With this definition,
we get
θ
∂V(r, θ, z)
x (∇V) r = (61)
∂r
y

θ r θ

x
top view r
looking
down
Figure 8
The unit vectors r, θ , z in cylindrical coordinates.
 Calculus 2000 - Chapter 3 The Gradient Cal 3-15

So far, our results look very much like what we had Dividing Equation (65) through by r∆θ and then
for Cartesian coordinates. However, we get some- taking the limit as ∆θ goes to zero gives
thing new when our step ∆ is in the θ direction.
Suppose we are at the position (r, θ, z) , and move to V(r, θ+∆θ, z) – V(r, θ, z)
the new point (r, θ+∆θ, z) where we increased the (∇V) θ = 1r limit
∆θ → 0 ∆θ
θ coordinate angle by ∆θ as shown in Figure (9).
Since the angle θ +∆θ is measured in radians, the (66)
arc length ∆ that we move when going from θ to We define the quantity in curly brackets to be the
θ +∆θ is partial derivative of V(r, θ, z) with respect to the
∆ = r∆θ variable θ

You will notice that the vector displacement ∆ is in ∂V(r, θ, z) limit V(r, θ+∆θ, z) – V(r, θ, z)
≡ ∆θ
the same direction as the θ unit vector, thus ∂θ →0 ∆θ
(67)
∆ = θr∆θ (62) Thus we end up with the equation
The change in voltage ∆V and the dot product
(∇V) ⋅ ∆ are thus ∂V(r, θ, z)
(∇V) θ = 1r
∂θ
∆V = V(r, θ+∆θ, z) – V(r, θ, z) (63) (68)
and we get a factor of 1/r in our formula for the
θ component of the gradient in cylindrical coordi-
∇V⋅∆ = r(∇V) r + θ (∇V) θ + z (∇V) z ⋅θ r∆θ
nates. The factor of 1/r appears because the partial
derivative with respect to θ measures the rate of
= (∇V) θ r∆θ (64) change of V for a given change ∆θ in angle, while
the gradient measures the rate of change of V with
where we used θ ⋅ θ = 1 , r ⋅ θ = z ⋅ θ = 0 . respect to a given step in distance. When we make
a change ∆θ in angle, the distance we move is r∆θ
Using (63) and (64) in our equation ∆V = ∇V ⋅ ∆ , which increases with r. The factor of r has to be
we get divided out to get the rate of change of V with
V(r, θ+∆θ, z) – V(r, θ, z) = (∇V) θ r∆θ (65) distance.

Exercise 2
Following the above steps, show that
z
∂V(r, θ, z)
(∇V)z =
∆θ ∂z
(69)

p ∆ This should look the same as our derivation for the

Cartesian coordinate system.
y

∆θ
r
x ∆

Figure 9
The displacement ∆ when we
increase the angle θ by ∇ θ .
Cal 3-16 Calculus 2000 - Chapter 3 The Gradient

GRADIENT IN SPHERICAL COORDINATES Exercise 3

While the steps are fresh, let us derive the formulas Start at the point (r, θ, φ) and move a distance ∆ to the
for the components of the gradient vector in spheri- point V(r + ∆r, θ, φ) and show that the r component of the
cal coordinates. We will then return to various gradient in spherical coordinates is
applications of the new gradient formulas.
∂ V(r, θ, φ)
In the spherical coordinate system shown in Figure ( ∇V) r = (71)
∂r
(10), a point p is located by the displacement r from
the origin, the angle θ that the coordinate vector r where
makes with the z axis, and the angle φ that the
projection of r on the x,y plane makes with the x ∂V(r, θ, φ) V(r + ∆r, θ, φ) – V(r, θ, φ)
∂r = limit
∆r → 0
(72)
axis. The unit vectors are r pointing out in the r ∆r
direction, θ which lies in the r z plane pointing in
It was Equation (71) that we used to show in one line that
the direction of increasing θ , and φ which is per-
the voltage V = kQ/r leads to the field E = r kQ/r2 .
pendicular to the r z plane, in the direction of
increasing φ . This gives us a right handed coordi-
In spherical coordinates, the radial component of the
nate system where
gradient is simply the partial derivative, as we asked
r×θ = φ (70) you to show in Exercise 3. We get new results when
we look at the θ and φ components, where the
(Again, show for yourself that θ × φ = r and φ × r = θ.) change in distance ∆ is not equal to ∆θ or ∆φ alone.
z
First let ∆ be in the θ direction, so that we go from
φ the point (r, θ, φ) to (r, θ +∆θ, φ) . The distance ∆
p r is shown in Figure (11) where we are looking squarely
at the rz plane. You can see that ∆ is in the θ
θ r direction and has a magnitude ∆ = r∆θ so that
θ
y ∆ = θ r∆θ (73)
r si
φ nθ
φ The change in voltage ∆V and the dot product
∇V ⋅ ∆ are
x
Figure 10 ∆V = V(r, θ +∆θ, φ) – V(r, θ, φ) (74)
The unit vectors r, θ , φ for a
spherical coordinate system.
∇V⋅∆ = r (∇V) r + θ (∇V) θ +φ(∇V) φ ⋅ θ r∆θ
z
= (∇V)θ r ∆θ (75)
p
∆
where θ ⋅ θ = 1 , r ⋅ θ = φ ⋅ θ = 0 .
r
θ ∆θ Equating ∆V from (74) with ∇V ⋅ ∆ in (75), then
dividing through by r∆θ and taking the limit as ∆θ
goes to zero, gives
φ
V(r, θ +∆θ, φ) – V(r, θ, φ)
(∇V)θ = 1r limit
∆θ → 0 ∆θ
x
Figure 11 (76)
The step ∆ when we increase θ by ∆θ .
We are directly facing the rz plane.
 Calculus 2000 - Chapter 3 The Gradient Cal 3-17

We define the partial derivative of V r, θ, φ with The quantity ∇V ⋅ ∆ is

respect to θ in spherical coordinates as
∇V ⋅ ∆ = r (∇V) r + θ (∇V) θ +φ (∇V) φ ⋅ φ (r sin θ)∆φ
∂V(r, θ, φ) limit V(r, θ +∆θ, φ) – V(r, θ, φ)
≡ ∆θ
∂θ →0 ∆θ
= (∇V) φ (r sin θ)∆φ
(77) (81)
so that we get because φ ⋅ φ = 1 and r ⋅ φ = θ ⋅ φ = 0 .

∂V(r, θ, φ) Equating ∇V ⋅ ∆ in Equation (81) to ∆V in (80)

(∇V)θ = 1r (78)
∂θ gives

V(r, θ, φ +∆φ) – V(r, θ, φ) = (∇V) φ (r sin θ)∆φ

as the formula for the θ component of the gradient
vector in spherical coordinates. (82)
Dividing (82) through by (r sin θ)∆φ and taking the
Finally we will derive the φ component of ∇V by limit at ∆φ goes to zero gives
taking a step ∆ in the φ direction. The geometry
is shown in Figure (12). The first thing to note is that 1 limit V(r,θ,φ+∆φ) –V(r,θ,φ)
( ∇V) φ =
the projection of the coordinate vector r down on r sinθ ∆φ→0 ∆φ
the xy plane has a length ( r sin θ ). This is the
distance the point p is out from the z axis. When we (83)
rotate an angle ∆φ about the z axis, the arc length ∆ We define the partial derivative with respect to φ
out a distance ( r sin θ ) is (r sin θ)∆φ . This distance in spherical coordinates as
is in the direction of the unit vector φ , thus
∂V(r, θ, φ) V(r, θ, φ +∆φ) – V(r, θ, φ)
= limit
∆ = φ (r sin θ)∆φ (79) ∂φ ∆φ → 0 ∆φ

The change in voltage, going from (r, θ, φ) to (84)

(r, θ, φ +∆φ) is to get the result

∆V = V(r, θ, φ +∆φ) – V(r, θ, φ) (80)

( ∇V) φ = 1 ∂V(r, θ, φ) (85)
z r sin θ ∂φ

r si
nθ

∆
θ r p

y
r sin θ
φ
∆φ
x ∆
Figure 12
The step ∆ when we increase φ by ∆φ .
Note that we are out a distance r sin θ
from the z axis.
Cal 3-18 Calculus 2000 - Chapter 3 The Gradient

SUMMARY OF GRADIENT FORMULAS EXAMPLES

We collect in one place the formulas for the gradient
Electric Field of a Point Charge
in Cartesian, cylindrical and spherical coordinates.
Let us now see explicitly how the formula for the
Cartesian Coordinates gradient in spherical coordinates, Equation (88),
makes it easy to calculate the electric field of a point
∇V(x,y,z) = x ∂V + y ∂V + z ∂V (86) charge, starting from the voltage formula
∂x ∂y ∂z
kQ
z V(r) = r (27) repeated
z
The formula for the gradient in spherical coordi-
y nates is
φ ∂V
x ∇V = r ∂V + θr ∂V + (88) repeat
y ∂r ∂θ r sin θ ∂φ

While Equation (88) looks somewhat messy, the

x thing to note is that V(r) has no dependence on the
variables θ and φ , thus the partial derivatives with
Cylindrical Coordinates respect to these variables are zero
(87) ∂V(r) ∂V(r)
∇V(r,φ,z) = r ∂V + θr ∂V + z ∂V
=0 ; = 0 (89)
∂θ ∂φ
∂r ∂θ ∂z
and all we are left with is
z
∂V(r)
∇V = r (90)
r z ∂r
θ
We have for ∂V(r)/∂r
r
∂ kQ = kQ ∂ (r – 1) = – 1 2
kQ
(91)
y ∂r r ∂r r
θ thus we get
kQ kQ
x E = – ∇V = – – r = r (92)
r2 r2
Spherical Coordinates
which is the correct answer.
φ ∂V
∇V(r,θ,φ) = r ∂V + θr ∂V + The advantage of using spherical coordinates to
∂r ∂θ r sin θ ∂φ calculate the field of a point charge was that, two out
of three of the components of the gradient were zero,
z (88) and we had only a simple derivative for the remain-
ing component. This is the kind of simplification
r you get when you use a coordinate system that
θ r matches the symmetry of the problem at hand. Our
θ next example will be the calculation of the electric
y field of a line charge. That problem has cylindrical
φ symmetry, and is most easily handled using a cylin-
φ drical coordinate system.
x
 Calculus 2000 - Chapter 3 The Gradient Cal 3-19

Electric Field of a Line Charge We can assume, because of the cylindrical symme-
In the Physics text, our first calculation of the elec- try of the problem, that the voltage V depends only
tric field of an extended object was to show that the on the radial distance r out from the wire. That is,
radially directed electric field of a charged wire, that V = V(r). Thus the partial derivatives with
shown in Figure (24-27) repeated here, had a mag- respect to the variables θ and z (using cylindrical
nitude coordinates) should be zero and we should be left
with
E(r) = λ (24-43) repeated
2π ε0r
where λ is the amount of charge per meter on the ∂V(r) θ ∂V(r) ∂V(r)
E = – ∇V = – r + r +z
wire and r is the radial distance out from the wire. To ∂r ∂θ ∂z
simplify the constants, we will set k = 1/2πε 0 so ∂V(r)
= –r (94)
that the vector formula for E is ∂r

E(r) = r kλ 1 where we used Equation (87) for the gradient in

r ; k =
2π ε0
(93)
cylindrical coordinates.
In the Physics text we never did say what the voltage Comparing Equations (93) and (94) for E we get
was in the vicinity of a charged wire. You will see why
shortly. ∂V(r)
E = r kλ
r = r – ∂r (95)

L As a result, the voltage V(r) should obey the equa-

tion
λ coulombs
r per meter ∂V(r)
= – kλ 1r (96)
+ + + + + + + ∂r
The question we have now is, what function of r,
when differentiated with respect to r, gives 1/r? The
side
view
answer, you may recall from Chapter 1 of the Calcu-
A = 2πrL lus text, is the natural logarithm. Explicitly
d (ln r) = 1 (97)
dr r
E
Thus the appropriate voltage V(r) is

V(r) = – k λln r (98)

r
Going back from this V(r) to E we have
end
view ∇V(r) = r ∂ (– k λ ln r)
∂r
(99)
= r – k λ ∂ ln r = r – kr λ
Figure 24-27 (repeated)
Using Gauss' law to calculate the electric field of a line ∂r
charge. Draw the Gaussian surface around a section and
of the rod. The flux all flows out through the
cylindrical surface. E(r) = –∇V(r) = r krλ (100)

This explicitly checks that the voltage –(k λ ln r) leads

to the electric field of a line charge.
Cal 3-20 Calculus 2000 - Chapter 3 The Gradient

The logarithm ln(r) that appears in Equation (100) is At large distances, there is no problem with the
an interesting function in that it is zero at r = 1, goes formula for the voltage of a point charge. At r = ∞ ,
to – ∞ at r = 0 and + ∞ at r = ∞ as shown in Figure the voltage V = kQ/r goes to zero, which is what we
(13). Thus, for example, at r = 0 we get wanted for the potential energy of a test charge
infinitely far away. But for a line charge, Equation
V(r) = – kr ln(0) = – kr – ∞ = + ∞ (101) (94) gives
r=0
V(r) = – k λ ln(+ ∞) (102)
and the voltage becomes infinite. This tells us that r=+∞
it is not physically reasonable to put a finite charge
This predicts a voltage or potential energy of minus
density λ on an infinitely thin wire. We had the
infinity when we are infinitely far away from a line
same problem with a point charge. The formula V
charge! How did this happen?
= kQ/r also goes to infinity at r = 0 which tells us we
have a problem with the potential energy of a point Either the mathematics is wrong, or our physical
charge of zero radius. (The modern theory of quan- interpretation is wrong. The answer lies with the
tum electrodynamics treats the electron as a point physical interpretation. What is wrong is that you
charge of zero radius. The tricky part of the theory cannot get infinitely far away from a line charge.
is to get around the infinities that result from this.) Any real physical piece of wire must have a finite
ln(x) length. The wire may look infinitely long when you
are close to it, but as you move away, you will
eventually be able to see both ends. The farther
1 away you move, the shorter the wire looks. Move
infinitely far from the wire and the wire looks like a
point charge and the voltage it produces goes to zero.
0 x Thus physically we will not encounter the infinity
that appears at large distances in the formula for the
voltage of a line charge.
–1
As we have often mentioned, in any formula for
–2 potential energy, we can arbitrarily choose the zero
of potential energy (the floor) wherever we want.
For point charges, we usually choose the zero of
–3 potential energy out at r = ∞ . We have seen that we
cannot make the same choice for a line charge. What
0 1 2 3 4 5 we have to do is write the formula for the potential
Figure 13 energy in the more general form
The function ln(x) starts out at minus infinity at x = 0,
goes through zero at x = 1, and slowly goes to plus V(r) = – k λ ln(r) + constant (98a)
infinity at x = infinity.
and adjust the constant so that V(r) is zero at some
convenient place. We can see how this works in the
following discussion of a coaxial cable.
 Calculus 2000 - Chapter 3 The Gradient Cal 3-21

The Coaxial Cable Thus V(r) in Equation (99) can be more compactly
A physical example where our voltage formula written
(98a) makes sense is the coaxial cable. Suppose we r
have a cable whose inner conductor has a radius r i V(r) = k λ ln r0 (107)
and the outer shield has an inside radius r 0 as shown
With the constant k written out as 1/2π ε 0 (see
in Figure (14). Assume that the inner conductor has
Equation 93), we get
a charge density λ coulombs per meter, and the
outer conductor is grounded (i.e., we say that the
V(r) = λ ln r 0
voltage V(r) is zero at r = r 0 .) What is the voltage
2π ε 0 r
throughout the cable? (108)
First of all, we know that the voltage inside a At the outer shield, at r = r 0, we have
conductor must be constant so that the field ln(r 0/ r) = ln(1) = 0
E = – ∇V inside is zero. Since the outer conductor and the voltage goes to zero. This is what we wanted
is grounded, the voltage throughout the shield (for for a grounded shield.
r > r 0 ) will be zero as shown in Figure (15). The
voltage on the inner conductor will have some As demonstrated in Exercise 4 below, Equation
constant value Vi (for r < r 0 ). (108) allows us to calculate the charge density λ on
the inner conductor of a coaxial cable when the outer
Between the conductors, in the region between r i conductor is grounded and the inner conductor is
and r 0 , the voltage must have the logarithmic de- raised to some voltage Vi.
pendence given by Equation (98a)
V(r) = – k λ lnr + constant (103) Exercise 4
We can evaluate the constant by setting the voltage (a) For the coaxial cable of Figure (14), find the formula
equal to zero out at the grounded shield, at r = r 0 . for the charge density λ when the inner conductor is at
This gives a voltage Vi volts.
V(r 0) = – k λ lnr 0 + constant = 0
(b) Suppose Vi = 100 volts, ri = .5 mm, r0 = 2 mm and
constant = k λ lnr 0 (104) recall that ε0 ≈ 9 × 10– 12 . Then what is λ in coulombs
and V(r) becomes per meter?
V(r) = – k λ lnr + k λ lnr 0 (105)
(c) What is the general formula for the capacitance per
Logarithms have the peculiar property meter of the coaxial cable in Figure (14)?
lna – lnb = ln a (106) V(r)
b
voltage constant
inside conductor
Vi

ri

r0

0 r
Figure 14 ri r0
A coaxial cable, where the inner wire has a radius r i Figure 15
and the outer grounded shield an inner radius r 0 . Voltage in the coaxial cable.
Cal 3-22 Calculus 2000 - Chapter 3 The Gradient

View 2
The Gradient from a
Geometrical Perspective
CH 3 VIEW 2 THE GRADIENT FROM A GEOMETRICAL PERSPECTIVE

In the first part of this chapter, we used the relation-

ship between force and potential energy to define
what we meant by the gradient vector. We then used V = .1

that relationship to derive the formulas for the

gradient in cylindrical and spherical coordinates.
V = .2

What we want to do now is to approach the gradient

from a geometrical point of view. This is the point of V = .3
view we began to develop when we constructed the
V = .4
physical models of electric voltage like the one V = .5
shown in Figure (25-15) reproduced again here.
Once we have developed a geometrical definition of V=
.1

the gradient we will check that the gradient behaves 0 1

.
V=
–.
V=

like a vector. To do that, we show that the compo- –1 +3

nents of the gradient change or transform the same
way that the components of a displacement vector
when we rotate the coordinate system. This idea of
testing the vector nature of a new quantity will
become particularly important when we get to a
mathematically advanced discussion of special rela-
tivity.

This discussion of the gradient is designed to be

independent of the first part of the chapter, so that
you can start from either approach. This leads to
some repetition of definitions, but the points of view
are sufficiently different that some duplication should
not be a problem. We, of course, end up with the
same definition of the gradient vector from the two
points of view. Figure 25-15 (repeated)
Computer plot of the field lines and equipotentials for a
charge distribution consisting of a positive charge + 3
and a negative charge – 1. These lines were then used
to construct the plywood model.
 Calculus 2000 - Chapter 3 The Gradient Cal 3-23

SLOPE IN TWO DIMENSIONS However the interesting part about going out in the
Imagine that you are planning a trip in a desert with desert is that you do not have to follow any particular
hills and valleys. One possibility is to follow a path path. If you do not want to climb very much, you can
that heads due east through the desert. If you draw walk along a contour line. If you are anxious to get
the path on a contour map, and note where the path to the top of a hill and want the steepest climb
crosses different contours, you can create a plot of possible, you walk at right angles to a contour line,
the height (h) of the path as a function of the distance along what we have called a field line, or what ski
(x) of the path. The result might look like a plot of instructors call the fall line. At any point you can
h(x) shown in Figure (1). This should at least choose a path whose slope ranges from zero along a
represent a smoothed version of the terrain you will contour line to the maximum along the field line. To
encounter. define the slope at some point, you have to state the
direction you are traveling.
Your curve h(x) tells you roughly how steep the
path should be at any point x 0 . Mathematically, you To handle this new feature mathematically, we first
can define the steepness as the slope of the tangent introduce a coordinate system (x,y), where the x
line at the point x 0 , which is equal to the first direction, for example, could be east-west and the y
derivative of h(x). direction north-south. The terrain is then described
by a function h(x,y) giving the height of the land at
slope at x 0 = h′(x 0) any point (x,y).
dh(x) To describe the slope of a one dimensional curve
=
dx x = x 0 h(x) at some point x 0 , we drew a tangent line at x 0
as shown in Figure (1). To describe slopes for a two
= tan θ (1) dimensional function h(x,y), at some point (x 0, y 0)
we look at the tangent plane at that point. This
As long as you stay on the path, the slope at any point
assumes that the function h(x,y) is smooth enough
is uniquely determined by Equation (1).
that, as we get closer and closer to the point (x 0, y 0)
h(x) the landscape looks smoother and smoother. It
assumes that when we get very close, the landscape
looks flat and we are looking at the tangent plane.
θ Not all functions h(x,y) are necessarily that smooth.
Curves describing real landscapes, like the shape of
x a coastline, look just as rough no matter how close
x0 we look. Such curves are described by what is called
fractal geometry. What we will be discussing are
Figure 1
curves, or surfaces that become smooth when we
Imagine that you are walking due east (x direction) in
the desert. We will call h(x) the height of your path. At look close enough. A sufficient mathematical crite-
some point x0 , the slope of your path is dh(x)/dx ria for such smoothness is that all derivatives with
evaluated at x0 , which is the tangent of the angle θ . respect to any variable are finite.

If the terrain h(x,y) is smooth enough to have a

unique tangent plane at every point, then our discus-
sion of the nature of slopes on a curved surface can
begin with a study of how slopes behave in a tangent
plane. What we learn from the study of one tangent
plane can then be applied to all tangent planes in the
terrain.
Cal 3-24 Calculus 2000 - Chapter 3 The Gradient

To visualize a tangent plane at some point (x 0, y 0) , plane. Unless this tangent plane happens to be
start by imagining that the point is on the surface of horizontal, there will be a unique horizontal line in
a table, and construct a coordinate axis (x,y,z) whose the plane that passes through the point (x 0, y 0) . This
origin is at (x 0, y 0) as shown in Figure (2). The xy horizontal line corresponds to the x axis in Figure
plane is the table surface and the z axis points (3). In a region very close to the point (x 0, y 0) this
straight up. Let us assume that the x axis faces east horizontal line will coincide with the contour line of
and the y axis north. h(x,y) that passes through that point.

To represent a tangent plane, take a thin flat object Perpendicular to the x axis in the tangent plane will
like a piece of cardboard, and place it on the table be a line of maximum slope heading in the y direc-
surface, tilted at an angle θ as shown in Figure (3). tion of Figure (3). The field line of our curved
Orient the cardboard so that the line of contact with surface h(x,y) that passes through the point (x 0, y 0)
the table is the x axis. will be y oriented for a small region around (x 0, y 0) .
As a result, in this small region the contour lines and
It is easy to see that in our flat tilted surface, all lines the field lines of the curved surface have the same
parallel to the x axis are contour lines, and that all properties as the contour and field lines in the
lines parallel to the y axis headed north are field lines tangent plane. In particular, even for curved sur-
with a maximum slope. It is also clear that the field faces, contour lines and field lines will always be
lines are perpendicular to the contour lines. perpendicular to each other where the contour lines
are in the direction of zero slope and the field lines
These features carry over to a smooth curved surface
in the direction of maximum slope.
h(x,y). At any point (x 0, y 0) construct a tangent

tangent plane
z up through (x 0,y0)

north x
y east
x
east
y
north θ (x0 , y0)

Figure 2 Figure 3
Our coordinate system. The tangent plane. All lines in the tangent plane that
are parallel to the x axis are lines of equal height, or
contour lines. Lines in the perpendicular y direction
are lines of maximum slope, or field lines.
 Calculus 2000 - Chapter 3 The Gradient Cal 3-25

THE GRADIENT To see the physical significance of the gradient, we

When you have a mathematical function h(x,y) that will evaluate ∇h(x,y) at some point (x 0, y 0) , using
describes a surface, the slope of that surface in some a coordinate system where the x axis is parallel to the
direction is given by the partial derivative in that contour line passing through that point. That is the
direction. Explicitly the slope in the x direction at same coordinate system we used in our discussion of
the point (x 0, y 0) is given by the tangent plane in Figure (3). Since the x axis lies
along a contour line at the point of interest, there is
slope in x ∂h(x,y) no change in height as we move a short distance in
direction at = (2a)
∂x x = x 0 , y = y0
the x direction, and thus the partial derivative in the
(x 0 , y0) x direction is zero.
and the slope in the y direction is
∂h(x,y) for an x axis
= 0 lying along (4)
∂x x = x0
slope in y ∂h(x,y) y = y0 a contour line
direction at = (2b)
∂y x = x 0 , y = y0
(x 0 , y0)
What remains of the gradient is
What we will do now is to define a quantity we will
call the gradient, and represent it by the symbol ∂h(x,y) for an x axis
∇h(x,y) = y lying along
∇h(x,y) . Explicitly we define ∇h(x,y) by the x = x0 ∂y x = x0
a contour line
y = y0 y = y0
equation
(5)
∂h(x,y) ∂h(x,y) For this coordinate system, the gradient is purely y
∇h(x,y) = x +y (3)
∂x ∂y oriented, which is the direction of the field line
through (x 0, y 0) . Also the magnitude of the gradient
where x and y are unit vectors in the x and y is equal to the magnitude of the steepest slope at
directions respectively. (x 0, y 0) . As a result, physical significance of the
The gradient ∇h(x,y) looks like a vector with x and gradient, at least in this special coordinate system, is
y components equal to the slope of h(x,y) in the x and that it describes both the direction and magnitude
y directions. However a vector is more than a of the steepest slope.
quantity with some components. We saw in Chapter Thus the gradient has both a magnitude and a direc-
2 of the Physics text that a vector has a basic physical tion like the displacement vectors we discussed in
significance that does not depend upon the coordi- Chapter 2 of the Physics text. If the components of
nate system used to define the vector. What we need the gradient change (transform) in the same way as
to do for our gradient is to find the basic significance the components of a displacement vector, then the
of the quantity ∇h(x,y) and then show that the magnitude and direction will be preserved when we
physical picture does not change when the gradient go to a new (rotated) coordinate system. The
is evaluated in a different coordinate system. components will look different, but the magnitude
and direction will be unchanged.
Cal 3-26 Calculus 2000 - Chapter 3 The Gradient

To see whether the components of the gradient When we go from the coordinate system (x,y) to the
transform (change) like the components of a dis- rotated coordinate system (x′,y′) , the gradient
placement vector, let us first review what happens to
a purely y oriented displacement vector B when we ∂h(x,y) ∂h(x,y)
∇h(x,y) = x +y (3) repeat
go to a new coordinate system (x′,y′) that is rotated ∂x ∂y
by an angle φ about the z axis as shown in Figure (4). becomes
You can easily see that in the x′,y′ coordinate
∂h(x′,y′) ∂h(x′,y′)
system, the components of B are ∇h(x′,y′) = x′ + y′ (9)
∂x′ ∂y′
B x′ = B sin φ To calculate the new components ∂h(x′,y′)/∂x′ and
(6)
B y′ = B cos φ ∂h(x′,y′)/∂y′ at some arbitrary point (x,y) we will
use our familiar tangent plane of Figure (3) repro-
duced here as Figure (5). We have also drawn in the
Exercise 1
rotated coordinate system (x′,y′) seen in the top
(a) Show that for a purely x oriented vector A the
view of Figure (5). The coordinate axes x, y and
components of A in the rotated (x′, y′) coordinate
x′,y′ all lie in the table top surface, what we can call
system are
the z = 0 plane.
Ax′ = A cos φ
Ay′ = –A sin φ
(7) The partial derivative, for example ∂h(x,y)/∂y , rep-
resents the rate of change of the height h as we go out
(b) Now show that if you start with a vector along the y axis. For the rotated coordinate system,
C = xA + yB ≡ xCx + yCy the partial derivative ∂h(x′,y′)/∂x′ represents the
rate of change of the height h as we go out along the
which has components Cx = A in the x direction and x′ axis. We will use these ideas to calculate the
Cy = B in the y direction, then in the rotated coordinate height ∆h of the point A shown in Figure (5), a point
system, the components of C are
that is a distance ∆x′ down the x′ axis.
z directed straight up
Cx′ = + Cx cos φ + Cy sin φ
(8) (no x' t)
Cy′ = – Cx sin φ + Cy cos φ r th x (eas
)y A φ
y'
φ ∆x' ∆y'
(Equations (8) are the general formula for the trans- ∆x
θ
formation of the x and y components of a vector when
we rotate the coordinate system by an angle φ about
the z axis.)
y' y

B
y' y
B co

φ φ top view
B sin x'
φ
sφ

φ x A x'
∆x' ∆y'
z axis up out of paper φ
x
Figure 4 ∆x
When we rotate the coordinate system about the
z axis, the y directed vector B gets components Figure 5
in both the x′′ and y ′ directions. Our tangent plane of Figure (3) showing the
rotated coordinate system x',y', and the point
A, a distance ∆x′ down the x' axis.
 Calculus 2000 - Chapter 3 The Gradient Cal 3-27

There are two distinct ways to get to the point A. One (In our drawing of Figure (5), we have shown the x
is to go down the x′ axis directly, a distance ∆x′ . For axis as being horizontal, so that the slope
this route we get as the formula for ∆h ∂h(x,y)/∂x would be zero. This makes the drawing
easier to interpret, but we do not need to assume the
slope in distance we x slope is zero for the current discussion.)
∆h = the x′ × go in the
direction x′ direction The final step in calculating the height ∆h of point
A from the second route is to relate ∆x and ∆y to the
∂h(x′,y′)
∆h = × ∆x′ (10) distance ∆x′ traveled along the x′ axis. From the top
∂x′ view of Figure (5) it is clear that
The other way to get to point A is to go down the x
axis a distance ∆x , gaining a height ∆h x given by ∆x = ∆x′cos φ
∆y = ∆x′sin φ (14)
∂h(x,y)
∆h x = ∆x (11)
∂x Using these values in Equation (13) give us
and then go out a distance ∆y in the old y direction,
∂h(x,y) ∂h(x,y)
giving us an additional height ∆h y given by ∆h = ∆x′cos φ + ∆x′sin φ
∂x ∂y
∂h(x,y) (15)
∆h y = ∆y (12)
∂y We can now equate our two formulas, Equation (10)
and Equation (15) for the height ∆h at point A. The
The height ∆h at point A will be the sum of these two
factors of ∆x′ cancel and we are left with
heights
∆h = ∆h x + ∆h y ∂h(x′,y′) ∂h(x,y) ∂h(x,y)
= (cos φ) + (sin φ)
∂x′ ∂x ∂y
∂h(x,y) ∂h(x,y) (16)
= ∆x + ∆y (13)
∂x ∂y Comparing Equation (15) with Equation (8) for the
transformation of the x component of the displace-
z directed straight up
ment vector C
(no x' t)
r th
)y A φ
x (eas Cx′ = Cx cos φ + Cy sin φ (8a) repeated
y'
φ ∆x' ∆y'
∆x
we see that the x component of the gradient trans-
θ forms (changes) in the same way as a displacement
vector when we rotate the coordinate system by an
angle φ .

Exercise 2
y' y Using similar arguments, show that the y′ slope
∂h(x′, y′)/ ∂ y′ is given by
top view
∂h(x′, y′) ∂h(x, y) ∂h(x, y)
φ = (– sin φ) + (cos φ) (17)
x' ∂ y′ ∂x ∂y
A
∆x' ∆y' which is the same as the transformation of the y
φ component of a displacement vector.
x
∆x
Figure 5 repeated
Our tangent plane of Figure (3) showing the rotated
coordinate system x', y', and the point A, a distance
∆x down the x' axis.
Cal 3-28 Calculus 2000 - Chapter 3 The Gradient

Gradient as a Vector Field As we saw in the first part of this chapter, the
What is the significance of our demonstration that extension of Equation (3) to the gradient of a three
the quantity ∇h(x,y) , defined by dimensional function is

∂h(x,y) ∂h(x,y) ∇h(x,y,z) = x ∇x h + y ∇y h + z ∇z h (18)

∇h(x,y) = x +y (3) Repeat
∂x ∂y where ∇x , ∇y and ∇z are the partial derivatives
transforms like a vector at each point (x,y) in space? ∂/∂x , ∂/∂y , and ∂/∂z . Equation (18) here is equiva-
As we pointed out in Chapter 29 of the Physics text, lent to Equation (16) in the first part of the chapter
a vector field, which is a vector at every point in relating E to ∇V(x,y,z) .
space, is uniquely determined if we have general
formulas for the surface integral and the line integral This completes our discussion of the gradient vector
of the field. There were four Maxwell's equations ∇h(x,y,z) from a geometrical point of view. If you
because we needed formulas for the surface and the have not done so already, now is the time to look at
line integrals of both the electric and magnetic applications of the gradient vector to electric field
fields. problems, starting with the discussion of the gradi-
ent vector just before Equation (16) of the first part
In the Physics text and the first part of this chapter, of the chapter.
we knew that the electric field was a vector field
because of its definition as the force vector acting on
a unit test charge. The knowledge that forces trans-
form as vectors was sufficient to tell us that any
correct formula for E gave us a vector field. In this
section with the definition of Equation (3), the
gradient is given a geometrical definition, which at
first sight might or might not make ∇h(x,y) behave
as a vector field. The demonstration that ∇h(x,y)
transforms as a vector means that concepts like line
and surface integrals can be applied to any gradient
fields.
 Calculus 2000 - Chapter 3 The Gradient Cal 3-29

View 3
Pressure Force as a Gradient
CH 3 VIEW 3 PRESSURE FORCE
AS A GRADIENT

We end the chapter with View 3, an application to PRESSURE FORCE AS A GRADIENT

fluids, where we see that the pressure force f p acting In the Physics text, there were two main places
on the fluid particles is the gradient of the pressure where we dealt with the concept of pressure. The
field p. This represents a straightforward example of first was in Chapters 17 and 18 on the ideal gas law,
obtaining a vector field f p from a scalar field p. and the second was in Chapter 23 during our discus-
sion of Bernoulli's equation. In both cases we
mentioned that pressure had the dimensions of a
force per unit area, but was itself a scalar field
p(x,y, z) that did not point anywhere. We pointed
out that the pressure force acting on an area ∆A was
directed perpendicular to the area and had a magni-
tude
∆F = p∆A (1)
We will now use the concept of a gradient to show
that the pressure force per unit volume f p , acting on
the fluid particles, is equal to minus the gradient of
the pressure p(x,y,z)

f p = –∇p(x,y,z) (2)
This is analogous to the electric field being equal to
minus the gradient of the electric voltage

E = – ∇V(x,y,z) (3-19)
Cal 3-30 Calculus 2000 - Chapter 3 The Gradient

which we saw back in Equation (3-19) of this chap-

ter. p(x,y,z +∆z) – p(x,y,z) ∂p(x,y,z)
limit =
∆z → 0 ∆z ∂z
To calculate the pressure force, we start with a small (6)
volume ∆V = ∆x∆y∆z shown in Figure (1). This Thus Equation (5) can be written in the somewhat
volume element has a left face located at z and a right mixed form
face at z + ∆z . The center of the faces are located at ∂p(x,y,z)
∆F 1 + ∆F 2 = – z ∆x∆y∆z (7)
(x, y) where the pressures are p(x,y,z) and ∂z
p(x,y,z + ∆z ) respectively. where we will shortly think in terms of the limit that
The pressure force ∆F 1 exerted on the left face of ∆V = ∆x∆y∆z goes to zero.
∆V is equal to the force per unit area p 1(x,y,z) times Before we do, let us add in the pressure forces F 3
the area ∆A 1 = ∆x∆y of that face. The pressure and F 4 acting on the bottom and top faces respec-
force is directed into the volume, toward the right in tively, and F 5 and F 6 acting on the back and front
the z direction, as shown faces to get the total pressure force ∆F p acting on
∆F 1 = zp(x,y,z)∆x∆y (3) ∆V . Following the same steps used to derive
Equation (7), we get
On the right side, the force is directed back into ∆V ,
∆Fp
in the –z direction, and has a value

∆F 2 = – zp(x,y,z +∆z)∆x∆y (4) = ∆F1 +∆F2 +∆F3 +∆F4 +∆F5 +∆F6

The net force on these two sides is

∂p(x,y,z) ∂p(x,y,z) ∂p(x,y,z)
=– z +y +x ∆x∆y∆z
∂z ∂y ∂x
∆F 1 + ∆F 2
(8)
= – z p(x,y,z +∆z) – p(x,y,z) ∆x∆y The quantity in the square brackets in Equation (8)
is the gradient ∇p of the pressure field. Thus we
p(x,y,z +∆z) – p(x,y,z) have, after dividing both sides by ∆V = ∆x∆y∆z
= –z ∆x∆y∆z
∆z
∆Fp
(5) = –∇p(x,y,z) (9)
∆V
You can immediately see that when we take the limit
that ∆V is an infinitesimal volume and ∆z goes to We recognize the left side of Equation (9) as the total
zero, the quantity in the square brackets in Equation pressure force acting on ∆V divided by the volume
(5) becomes the partial derivative of p(x,y,z) with ∆V . It is therefore the pressure force per unit
respect to z. volume f p(x,y,z) acting in that region of the fluid,
and we get our advertized result
pressure
∆z force
f p(x,y,z) = – ∇p(x,y,z)
y ∆x per unit
volume
∆F1 (2) ∆F2
(1) (2) repeated
∆y With Equation (2) we have a powerful way of
calculating pressure forces, since we can evaluate
x the gradient in any of the coordinate systems we
have been discussing, such as cylindrical or spheri-
z cal polar coordinates.
z z+∆z
Figure 1
The volume element ∆x ∆y∆z .
 Calculus 2000 - Chapter 4 Del Squared Cal 4-1

Calculus 2000-Chapter 4
The Operator ∇ 2 ≡ ∇ ⋅ ∇
(The Laplacian)

In our earliest discussion of vectors in Chapter 2 of

second derivative we encountered in Calculus Chap-
the Physics text, we were introduced to the vector
ter 2, during our discussion of the one dimensional
dot product wave equation. Thus we should expect ∇ 2 to appear
when we begin to discuss three dimensional wave
CHAPTER 4 THE OPERATOR ∇ 2≡∇⋅∇
A ⋅ B ≡ A xB x + A yB y + A z B z (1) equations in the next few chapters.
as having the special property of being a scalar
quantity. That is, the quantity A ⋅ B had the same Fluid theory
value no matter what coordinate system we used to Another area of physics where the operator ∇ 2
evaluate it. Having just seen that the gradient plays a prominent role is in fluid dynamics. For
operator ∇ operating on a scalar field h(x,y,z) common fluids like water and air, the viscous force
produces a vector field, one might wonder what we acting on the fluid particles turns out to be propor-
get when we take the dot product of two gradient tional to the Laplacian of the velocity field, namely
operations acting on a scalar field. The answer is ∇ 2 v . We will derive that result starting from an
that we get another scalar field. assumption that Issac Newton made about the na-
ture of viscous forces.
The standard name for this dot product of two
gradient operators is del squared, written as As an application of the theory of viscous forces, we
will look at the steady flow of a viscous fluid in a
∇2 ≡ ∇ ⋅∇ (2) pipe. This example provides a way to measure the so
called coefficient of viscosity that appears in
It is often called the Laplacian operator after the Newton's theory. It also provides an example of the
French mathematician Laplace. This operator is use of the operator ∇ 2 acting on a vector field.
essentially an extension to three dimensions of the
Cal 4 -2 Calculus 2000 - Chapter 4 Del Squared

Schrödinger's Equation The Formulary

One of the glaring omissions in the Physics text In the main part of this chapter we will simply state
resulted from our inability to calculate the electron the formula, in spherical polar coordinates, for
wave patterns in the hydrogen atom. All we were ∇ 2 acting on a scalar field ψ . This is the formula
able to do is show drawings of a few of the lowest we will use in Chapter 6 in our discussion of the
energy wave patterns, describe the electron's en- hydrogen atom. In the appendix, however, we will
ergy and angular momentum in these wave patterns, derive the formula, showing you exactly how the
and then state that these patterns came from a wave changing unit vectors affect the results. We have
equation called Schrödinger's equation. We were placed this derivation in an appendix because it is
neither able to write down or solve the equation the kind of derivation you probably want to observe
itself. only once in your life, to find out where the rather
messy results come from.
To handle Schrödinger's equation as applied to the
hydrogen atom, we needed two mathematical con- When you are actually working problems involving
cepts we did not then have. One is the operator quantities like ∇ 2 in cylindrical or spherical coor-
∇ 2 which we are introducing in this chapter, the dinates, you do not want to derive the formulas
other is the concept of a complex variable which we yourself because the chances of your getting the
will introduce in the next chapter, Chapter 5. Once right answer are too small. You are not likely to
we develop these two mathematical tools, we will be memorize them correctly either, unless you use a
ready to approach Schrödinger's equation in Chap- particular formula often. Instead, the best proce-
ter 6. dure is to look up the result in a table of formulas,
sometimes called a formulary. We provide a for-
When we apply Schrödinger's equation to the hydro- mulary at the end of this text, one adapted from a
gen atom, we are dealing with a system that has formulary developed by David Book of the Naval
spherical symmetry. As a result it is much easier to Research Laboratory.
deal with the theory using a coordinate system that
has the same symmetry. The problem is that the In our discussion of viscous forces in this chapter,
operator ∇ 2 , which in Cartesian coordinates is a we use the formulary to find the formula in cylindri-
straightforward extension of the second derivative, cal coordinates for ∇ 2 acting on the vector field v .
becomes quite complex when we work in other
coordinate systems like spherical polar coordinates.
The reason for the complexity is that in any coordi-
nate system except Cartesian coordinates, the unit
vectors may change direction as we move from one
point in space to another. This change in the
direction of the unit vectors complicates the formu-
las for ∇ 2 .
 Calculus 2000 - Chapter 4 Del Squared Cal 4-3

∇ 2 IN CARTESIAN COORDINATES As a result, all we are left with, when we evaluate

∇ 2f in Cartesian coordinates is
We will first take a careful look at ∇ 2 = ∇⋅∇ in
∇ 2f(x,y,z) = ∂ f2 + ∂ f2 + ∂ f2
Cartesian coordinates before we approach the spheri- 2 2 2
cal case. Using the unit vector notation for ∇ we (9)
∂x ∂y ∂z
have
∇ = x ∂ +y ∂ +z ∂
which is an obvious extension to three dimensions of
(3)
∂x ∂y ∂z the second derivative ∂ 2f/∂x 2 that appeared in our
where x, y and z are unit vectors pointing in the x, one dimensional wave equation in Chapter 2 of the
y, and z directions respectively. The dot prod- Calculus text.
uct ∇ ⋅ ∇ acting on some function f(x,y,z) should be
given by
∇ 2 in Spherical Polar Coordinates
∇⋅∇f(x,y,z) As we mentioned, the results are not so simple when
we are working in other coordinate systems. In
= x ∂ + y ∂ + z ∂ ⋅ x ∂f + y ∂f + z ∂f spherical polar coordinates, when ∇ 2 is acting on a
∂x ∂y ∂z ∂x ∂y ∂z scalar function, we get the following result which is
derived in the appendix to this chapter.
= x ∂ ⋅ x ∂f + x ∂ ⋅ y ∂f + ⋅ ⋅ ⋅
∂x ∂x ∂x ∂y
∂2
∇ 2f = 1r 2 (rf)
+ ⋅ ⋅ ⋅ + z ∂ ⋅ z ∂f (4) ∂r
∂z ∂z

+ 12 1 ∂ sin θ ∂f + 12 ∂ f2
2
Being very careful with our differentiation, we have, ∂θ
r sinθ ∂θ sin θ ∂φ
for example,

x ∂ ⋅ x ∂f = x ⋅ ∂x ∂f + x ∂ f2
2 (10)
(5)
∂x ∂x ∂x ∂x ∂x where r, θ , and ϕ are the polar coordinates shown
in Figure (1). Much of this complexity comes from
We have been overly careful because the unit vec-
the fact that the unit vectors are not constant, and
tors x , y and z are constant in both magnitude and
have to be differentiated. You will see how this
direction, thus
works by going to the appendix.
∂x = 0 (6)
∂x (We should note that, in non Cartesian coordi-
and we are left with nates, ∇ 2 acting on a vector, e.g. ∇ 2E , has an even
more complex formula, which is given in the formu-
x ∂ ⋅ x ∂f = x ⋅ x ∂ f2 = ∂ f2
2 2
(7)
∂x ∂x ∂x ∂x lary at the end of the text.)
Similarly z

x ∂ ⋅ y ∂f = x ⋅ ∂y ∂f + y ∂ f
2
∂x ∂y ∂x ∂y ∂x∂y p
θ
= x⋅y ∂ f = 0
2
r
∂x∂y (8)

because ∂y/∂x = 0 and x ⋅ y = 0 .

φ

x
Figure 1
Spherical polar coordinates.
Cal 4 -4 Calculus 2000 - Chapter 4 Del Squared

NEWTONIAN FLUIDS This combination of oppositely directed forces on

We now move on to our example of the use of the opposite sides of the plane is called a stress, in this
Laplacian operator to describe viscosity in a case a stress generated by the action of viscosity. For
Newtonian fluid. a so called Newtonian fluid, the stress τ is assumed
to be directly proportional to the rate at which the
Newton proposed that viscous effects in a fluid velocity field is changing as we move up, which for
resulted from the shearing motion of one layer of our x directed flow is
fluid over another. This shearing force can be intro-
duced as follows. ∂vx (y)
τ = µ (11)
∂y
Suppose we have a simple flow where all the fluid is
moving in the x direction, and the velocity is increasing The quantity µ is called the coefficient of viscosity
in the y direction as shown in Figure (2).
µ = coefficient of viscosity (12)
To analyze the forces involved, consider a horizon-
tal plane indicated by the dashed line labeled by
A----B. The fluid above the plane, which is travel- For a Newtonian fluid, µ is assumed to be a constant
ling faster, drags the fluid below forward. The fluid throughout the fluid. In many situations, both water
below, which is going slower, drags the upper fluid and air behave as Newtonian fluids.
back. Let τ + be the force per unit area exerted by the
upper fluid on the lower fluid, and τ – , the force
exerted by the lower fluid on the upper. In Figure (2)
we have drawn the forces τ + and τ – inside the fluids
upon which they act.
y
v

τ–
A B
τ+

Figure 2
Diagram of a simple flow where the velocity field v is
x directed and increasing in the y direction.
 Calculus 2000 - Chapter 4 Del Squared Cal 4-5

VISCOUS FORCE ON A FLUID ELEMENT With Equations (14) and (15) we see that the total
Suppose again that we have a simple x directed viscous force on the fluid in our volume element can
velocity field whose velocity profile is shown in be written
Figure (3). Now consider a small volume element ∆Fx = ∆F–(y) +∆F+(y+∆y)
with sides ∆x , ∆y and ∆z , the bottom of which is
located at (y) and the top at (y + ∆y ) is shown. The ∂vx(y) ∂v (y+∆y)
fluid below the plane A----B at y is dragging the fluid = –µ +µ x ∆x∆z
∂y ∂y
above, back with a force per unit area τ–(y)
(16)
∂v (y) force per unit
τ–(y) = – µ x area at the bottom (13) Multiplying the right side by ∆y/∆y gives
∂y of volume element

The total force at the bottom is the force per unit area ∂vx(y+∆y) ∂vx(y)
τ–(y) times the area ∆x∆z upon which it is acting –
∂y ∂y
∆Fx = µ ∆x∆y∆z (17)
∆y
∆F–(y) = τ–(y)∆x∆z
(14)
∂v (y) The quantity in the square brackets should be recog-
= –µ x ∆x∆z
∂y nized as the second derivative of v x(y) with respect
to y. Dividing through both sides by the volume
Up at the top of the volume element, the faster fluid ∆x∆y∆z gives us the viscous force per unit volume
above the C----D plane at (y + ∆y ), is pulling for-
ward the slower fluid below with a total force ∆Fx viscous force per unit
= fνx volume acting on
∆F+(y+∆y) = τ+(y+∆y)∆x∆z ∆x∆y∆z the fluid element
(15)
∂v (y+∆y)
= +µ x ∆x ∆z ∂ 2vx(y)
∂y fνx = µ (18)
∂y 2
y
v This is the formula for the viscous force per unit
volume acting on the fluid particles when we have a
purely x directed flow of a Newtonian fluid whose
speed varies only in the y direction. In the next
y+∆y section we generalize the result to three dimensional
C τ+(y+∆y) D flows.
∆y
τ– (y)
y
A B

Figure 3
Calculating the viscous force on a fluid element.
Cal 4 -6 Calculus 2000 - Chapter 4 Del Squared

VISCOUS FORCE FOR THREE What we want is the most general combination we
DIMENSIONAL FLOWS can make out of constants, two derivatives ∇ , and a
velocity field v . Basically we have three vectors ∇ ,
At first sight, there seems to be a rather obvious
∇ , v , and we must multiply them together to get a
extension of Equation (18) to three dimensional
single vector. To do this, we have to take the dot
flows. In a chapter devoted to discussing the opera-
product of two of them. The possibilities are (∇ ⋅ ∇)v
tor ∇ 2 , we might expect that the generalization of
and ∇(∇ ⋅ v) .* As a result, our most general for-
our formula for the viscous force f ν per unit volume
mula for a Newtonian fluid with constant coeffi-
should be
cients is
f ν = µ∇ 2 v (19)
f ν = µ 1(∇ ⋅ ∇)v + µ 2∇(∇ ⋅ v) (21)
To check that Equation (19) reduces to our result in
where µ 1 and µ 2 are constants. There is no other
Equation (18), when v is the one dimensional flow
combination of constants and second derivatives of
v x(y) , we have
the velocity field that transforms as a vector when
fνx = µ∇ 2vx(y) = µ ∂ 2 + ∂ 2 + ∂ 2 vx(y)
2 2 2
we rotate the coordinate system.
∂x ∂y ∂z
If we are dealing with a constant density fluid,
∂ 2vx(y)
= µ (20) ∇ ⋅ v = 0 and we are left with
∂y 2
Thus we get the desired result for one dimensional f ν = µ 1(∇ ⋅ ∇)v = µ∇ 2v (19a)
flows.
which is the result we guessed back in Equation (19),
However, complications arise in three dimensional with µ 1 = µ .
flows that we did not consider in our analysis of the
simple one dimensional flow pattern. In three di- Equation (21) suggests that it is possible to have a
mensions, fluids flow around corners and x directed second kind of viscosity when the fluid is compress-
flows can become y or z directed. The definition of ible and ∇⋅ v is not zero. This has in fact been
viscous stress we gave in Equation (11) simply observed, and µ 2 is sometimes called the second
cannot handle changes in the direction of the flow. viscosity coefficient. (Some texts use a second
viscosity coefficient defined as λ = µ 2 – µ .) In this
An effective way to deal with viscous forces in three text we will only deal with incompressible fluids
dimensional flows is to note that the resulting force where there is no second viscosity, and f ν is simply
f ν per unit volume must be a vector field. That is, given by the Laplacian operator ∇ 2 acting on v ,
f ν must transform like a vector field when we rotate namely f ν = µ∇ 2v .
the coordinate system. (See the discussion of the
transformation of vector fields at the end of the
geometrical discussion of the gradient in Chapter 3.)

We will also require that f ν be made up of some * (You might also consider vector cross products
combination of constants and second derivatives of involving ∇ , ∇ , and v . The possibilities are
the velocity field. These requirements on f ν are ∇ × (∇ ⋅ v) , ∇ ⋅ (∇ × v) , and ∇ × (∇ × v) . At the be-
essentially what we mean by a Newtonian fluid with ginning of Chapter 9, we find that the first two of
constant coefficients. If the viscous forces are more these are identically zero, and the third turns out to be
complex, which they can be for something like a
liquid crystal, then we say that the fluid is ∇ × (∇ × v) = ∇ (∇ ⋅ v) – (∇ ⋅ ∇)v
non Newtonian. which involves only the two terms we got from dot
products. Thus we get nothing new by considering
cross products. )
 Calculus 2000 - Chapter 4 Del Squared Cal 4-7

Viscous Force in Looking farther down in the formulary we find for

Cylindrical Coordinates the components of ∇ 2A
Now that we have the formula for the viscous force ∂A θ A r
f ν = µ∇ 2v , which applies to any fluid that we will (∇ 2A) r = ∇ 2A r – 22 – 2 (23a)
r ∂θ r
consider in this text, we are free to use general
formulas we have in the formulary for ∇ 2 in various ∂A r A θ
(∇ 2A) θ = ∇ 2A θ + 22 – 2 (23b)
coordinate systems. We are about to study the flow r ∂θ r
of a viscous fluid in a pipe, a problem that obviously
has cylindrical symmetry. Thus to analyze the (∇ 2A) z = ∇ 2A z (23c)
viscous forces, we should work with ∇ 2v in cylin-
drical coordinates. where, for example, ∇ 2A z means apply Equation
(22) to A z
We mentioned earlier that ∇ 2 acting on a vector
field is more complex than ∇ 2 acting on a scalar ∂ ∂A z ∂ 2A z ∂ 2A z
∇ 2A z(r,θ,z) = 1r r + 12 +
field in anything except Cartesian coordinates. Thus ∂r ∂r r ∂θ 2 ∂z 2
evaluating ∇ 2v in cylindrical coordinates will give (24)
us some practice in correctly using the formulary. All this looks like a terrible mess. But suppose we
have a fluid flowing smoothly along a pipe as shown
From the formulary we find the following formula in Figure (5). Taking the z direction down the pipe
for ∇ 2 acting on a scalar field f and a vector field A . and r the distance out from the axis of the pipe, we
can assume, for cylindrical symmetry, that v(r,θ,z)
∇ 2f = 1r ∂ r ∂f + 12 ∂ f2 + ∂ f2
2 2
(22) is purely z directed and depends only on the radius r.
∂r ∂r r ∂θ ∂z
where the coordinates r, θ , z are the unit vectors r , v(r,θ,z) = zv z(r) (25)
θ , z shown in Figure (4).
Now let us work out ∇ 2v for this simple case using
z Equations (23) for ∇ 2 in cylindrical coordinates.
z
Because vr and vθ are zero, we do not worry about
r
θ Equations (23a) and (23b). From (23c) we have
p r (∇ 2v ) z = ∇ 2vz (26)

y
Thus for this case we do not have to worry about the
extra stuff that comes in when we take ∇ 2 of a
θ vector.
z
x
Figure 4 r
Cylindrical coordinates. vz(r)
R
z
Figure 5
Velocity profile for the
uniform flow in a pipe.
Cal 4 -8 Calculus 2000 - Chapter 4 Del Squared

Next we note that vz = vz(r) , thus we can ignore the Thus (∇ 2v ) z becomes
∂vz /∂θ and ∂vz /∂z terms in (13a) and we are left ∂ ∂vz(r)
(∇ 2v ) z = 1r r
with ∂r ∂r
∂ ∂vz(r)
(∇ 2v ) z = 1r r (27) ∂ 2V
∂r ∂r = 1r – 20 r 2
∂r R
which is not such a difficult thing to work with after
= 1r – 20 ∂r
all. 2V 2
R ∂r (32)
To get a feeling for what the viscous force looks like
for pipe flow, we look up in a fluids text what the so 2V
= 1r – 20 2r
called laminar (i.e., non turbulent) velocity profile R
is in a pipe. The result they give is
parabolic The r's cancel and we are left with
V
vz(r) = 02 (R 2 – r 2 ) velocity (28) 4V
R profile (∇ 2v ) z = – 20 (33)
R
where R is the radius of the pipe, V0 the flow speed
The viscous force f ν = µ∇ 2v becomes
at the center, and r the radial distance from the axis.
This is the parabolic profile shown in Figure (5).
4V0
You can see that at the edge of the pipe, where r = R, fνz = – µ (34)
the velocity goes to zero. At the center where r = 0, R2
vz = V0 is a maximum.
We end up with the result that f ν points in the – z
To calculate the viscous force per unit volume for direction (it has only a negative z component) and is
this parabolic profile, we have constant in magnitude throughout the pipe. This is
a wonderfully simple result considering the stagger-
f ν = µ∇ 2v (29)
ing mess of terms we faced in Equation (23).
∂ ∂vz(r)
(fν ) z = µ(∇ 2v) z = µ 1r r (30) We will see that the physics of the parabolic laminar
∂r ∂r
flow is that this uniform – z oriented viscous force
With Equation (28) written as is balanced by a uniform + z oriented pressure
gradient down the tube. Thus there is no net force on
V0 2 each fluid element and the fluid moves down the
vz(r) = – r + V0 (28a)
R2 pipe without acceleration, i.e., at constant velocity.
we easily get

∂vz(r) 2V
= – 20 r
∂r R

∂vz(r) 2V
r = – 20 r 2 (31)
∂r R
 Calculus 2000 - Chapter 4 Del Squared Cal 4-9

Measuring the Viscosity Coefficient The pressure force on the fluid at the front end of the
If we have an apparatus where we know the pressure pipe is p 2 A 2 = p 2 A where A is the cross sectional
gradient, we can use that to measure the viscosity area of the pipe. At the far end it is – p 3 A , the minus
coefficient µ of the fluid. Such an apparatus is sign is used because the pressure force is in the
sketched in Figure (6), a sketch taken from the –z direction. Thus the net pressure force Fp is
excellent fluid dynamics text by Tritton.
Fp = z (p 2 A – p 3 A)
Since there is essentially no viscosity acting in the
region between points (1) at the top of the fluid in the = z (p 2 – p 3 )A (38)
container, and point (2) near the entrance to the pipe, = zρghA
we can use Bernoulli's equation to get
If we divide Fp by the volume AL of the pipe, we get
ρv 2 ρv 2 the average pressure force per unit volume f p .
p 1 + 1 + ρgh 1 = p 2 + 2 + ρgh 2 (35)
2 2
Fp z
With v 1 = 0 and h 1 – h 2 = h , we get fp = = ρghA
AL AL
ρv 2
p 2 – p 1 = ρgh – 2 (36) ρgh average pressure
2
fp = z force per (39)
If we use a sufficiently long and small diameter pipe, L unit area
the pipe flow velocity will be sufficiently small that As we mentioned, for steady laminar flow, the
we can neglect v 2 2 compared to gh. Noting that viscous force should be exactly opposed by the
both p 3 and p 1 are both atmospheric pressure and pressure force so that there is no acceleration of the
thus equal, we get for the pressure difference (p 2 – p 3 ) fluid. Since the viscous force per unit volume is
at the ends of the pipe uniform throughout the fluid for parabolic pipe
pressure difference flow, the pressure force per unit volume should also
(p 2 – p 3 ) = ρgh between ends of (37) be uniform, with the result that Equation (39) for f P
the pipe should apply at all points in the fluid in the pipe.
(There will always be some disturbance at the begin-
P1 ning of the flow that we are neglecting.)
(1)
overflow

h
supply outlet to
atmosphere
(2) v
long, small diameter pipe (3)
L
Figure 6
Apparatus to measure the viscosity coefficient.
Cal 4 -10 Calculus 2000 - Chapter 4 Del Squared

Saying that the viscous and pressure forces oppose dΦ = 2πrv(r) dr (43)
each other throughout the pipe flow gives us from
Equations (34) for f ν and (39) for f p With v(r) given by the parabolic profile
( V0 /R 2 )( R 2 – r 2 ) , we get for the total flux
fp = – fν R R
V0 2 2
ρgh 4V Φ = dΦ = R – r 2πrdr
z = – – z 20 µ R2
L R 0 0
R R
ρgh 4V0 µ
= (40) = 2πV0 rdr – 12 r 3dr
L R2 R
0 0
We are left with an equation involving measurable
constants and the viscosity coefficient µ . R R
= 2πV0 r2 – 12 r
4

Later in the text, we will see that the ratio µ/ρ , which 2 R 4
0 0
is called the kinematic viscosity coefficient ν , is
more convenient for theoretical work. Equation 2 2 V
(40) gives us for this ratio Φ = 2πV0 R – R = 0 π R 2 (44)
2 4 2
kinematic viscosity Since V0( π R 2 ) is the flux we would get if the
µ gh R 2 determined from
ν ≡ ρ = × parabolic pipe (41) velocity were a uniform V0 across the pipe, we see
L 4V0
flow that the flow rate for a parabolic profile is half that
The only constant that may be a bit difficult to for a uniform flow.
measure directly is the stream velocity V 0 at the With Equations (41) and (44) we can now express
center. This can be accurately determined by mea- the kinematic viscosity ν in terms of the easily
suring the flow rate which we will call Φ (phi), and measured volume flux Φ . From Equation (44) we
then express V 0 in terms of Φ . We have called the get
flow rate Φ because it is simply the flux Φ of the
fluid through the pipe, given by our old flux formula

Φ = v ⋅ dA (42)
area of
tube
r
and is measured, in the MKS system, in cubic meters
per second. dr
To calculate Φ , we divide the cross sectional area
into circular bands of radius r, thickness dr, as shown R
in Figure (7). The area of a band is 2πrdr and the
flux dΦ through the band is Figure 7
The integration area is the
area 2 π R dr of the band.
 Calculus 2000 - Chapter 4 Del Squared Cal 4-11

The two fluids that we will most often use in any

V0 = 2Φ2 ; 1 = π R2
discussion of fluid dynamics are water and air. At
πR V0 2Φ
room temperature and pressure, the kinematic vis-
and from Equation (41) we get cosity ν of these two fluids are approximately
µ gh R 2 1
ν = ρ = ν water = 1.0 × 10 – 6meter 2/second
L 4 V0
gh R 2 π R 2 ν air = 1.5 × 10 – 5meter 2/second (46)
=
L 4 2Φ
Intuitively you would think that air would be much
formula for less viscous than water, but the two coefficients νair
π ghR 4 measuring and νwater are quite close, with air having the greater
ν = (45)
8L Φ kinematic
viscosity value. What has happened is that we have divided by
the density, which brings the viscosity coefficients
Although rather a mess of constants appears in our
much closer together.
formula for the kinematic viscosity ν , all are quite
easily measured. Note that by going to the kinematic
viscosity, the result is independent of the density of
the fluid.

Exercise 1
Show that the kinematic viscosity ν has the dimensions
of meters 2 /second .
Cal 4 -12 Calculus 2000 - Chapter 4 Del Squared

Appendix: The Operator ∇ 2

in Spherical Polar Coordinates
APPENDIX THE OPERATOR ∇ 2 IN SPHERICAL POLAR COORDINATES
SPHERICAL POLAR COORDINATES This product involves terms like
We will begin with a review of spherical coordinates
θ ∂ ⋅ r ∂f = θr ⋅ ∂r ∂f + r ∂ f
2
discussed in Chapter 3. In spherical polar coordi-
nates, the three unit vectors are r , θ , φ are shown in r ∂θ ∂r ∂θ ∂r ∂θ∂r
Figure (A1) which is Figure (3-10) repeated. We
= 1r θ ⋅ ∂r ∂f + (θ ⋅ r ) 1r ∂ f
2
have a complication in evaluating ∇ 2f in spherical (A2)
polar coordinates because these unit vectors change ∂θ ∂r ∂θ∂r
direction as we move about, and we can no longer set Because the unit vectors always remain perpendicu-
the derivatives of the unit vectors to zero. Thus we lar to each other as we move around in space,
have to evaluate derivatives of the unit vectors as θ⋅ r = 0 and the second term in Equation (A2) is
well as use the rather messy formula for ∇f we zero. However, when we change the angle θ , the
derived in Equation (3-88) unit vector r changes direction. For example, at
φ ∂f θ = 0 , r points straight up, but at θ = 90° , r is
∇f(r,θ,φ) = r ∂f + θr ∂f + (3-88) horizontal. Thus ∂r /∂θ is not zero and has to be
∂r ∂θ r sinθ ∂φ
evaluated.
What we have to evaluate is the complete expression
In order to evaluate Equation (A1) for ∇ 2f , we will
∇ 2f(r,θ,φ) ≡ ∇ ⋅ ∇f(r,θ,φ) first calculate all the derivatives of all the unit
vectors, and then plug the whole mess together. We
φ find derivatives like ∂r /∂θ by evaluating the change
= r ∂ + θr ∂ + ∂
∆ r as we make a change ∆θ and then taking the
∂r ∂θ r sinθ ∂φ
limit as ∆θ goes to zero. The nine derivatives are as
(A1) follows.
φ ∂f
⋅ r ∂f + θr ∂f +
∂r ∂θ r sinθ ∂φ

φ
p r

θ r
θ
y
r si
φ nθ
φ
x
Figure A1 (3-10 repeated)
Unit vectors in spherical polar coordinates.
 Calculus 2000 - Chapter 4 Del Squared Cal 4-13

Derivatives of r 3) Change of r with φ

1) Change of r with r In Figure (A3), when we go from φ to φ + ∆φ , the

unit vector r goes to the unit vector r ′ . The projec-
because r does not
∂r = 0 change direction as we tions of r and r ′ in the horizontal plane have a length
∂r go out along a radius r sinφ = 1 sinφ , and differ in direction by an angle
∆φ . The change ∆r = r ′ – r points in the φ direc-
2) Change of r with θ
tion, and has the same length as the change in the
Figure (A2) shows ∆r that we get when θ increases horizontal projections of r and r ′ , which from the
by ∆θ . We see that ∆r points in the θ direction and small triangle is seen to be (sinφ)(∆φ) . Thus
has a length ∆r = r ∆θ = 1 × ∆θ . Thus we get
∆r = θ(∆θ) ∆r = φ(sinφ)∆φ

∆r = θ ⇒ ∂r = θ ∆r
= φ(sinφ) = ∂r (A4)
∆θ ∂θ
(A3)
∆φ ∂φ

z
z

r r'
r
θ r r+∆r
∆θ θ r'
y r
y
φ
∆φ |r'| si
nθ
x
x
∆r is the change in the unit vector r |r| s
in θ
when we increase θ by ∆θ.
∆r is the change in the unit vector r
r ∆r = θ ∆θ
when we increase φ by ∆φ.
∆θ
r+∆r |r'| si
nθ =
sin θ
in θ ∆φ
Unit vectors enlarged. |r| = |r+∆r| = 1 |r| s
=s
in θ
z ∆r = φ (sin θ)∆φ
Unit vectors enlarged.
r
Figure A3
θ r Evaluation of ∂r/ ∂φ.
θ
y

x
∆r points in the θ direction

Figure A2
Evaluation of ∂r/ ∂θ .
Cal 4 -14 Calculus 2000 - Chapter 4 Del Squared

Derivatives of θ 6) Change of θ with φ

4) Change of θ with r From Figure (A5), we see that θ changes to θ′ as φ

goes to φ + ∆φ . The change ∆θ points in the φ
None of the unit vectors change direction as we go
direction. To determine the magnitude of ∆θ , note
out along the radius, thus
that ∆θ and its projection in the horizontal plane are
the same. Since the projections of θ and θ′ have a
∂θ = 0
∂r (A5) length of cosθ , and an angle ∆φ between them, the
length of ∆θ is cosθ∆φ as seen in the small horizon-
5) Change of θ with θ tal triangle. Thus
From Figure (A4) we see that as we increase θ to ∆θ = φ(cosφ)∆φ
θ + ∆θ , the unit vector θ goes to θ ′ = θ + ∆θ .
From the small triangle, we see that the change ∆θ
∆θ = φcosθ = ∂θ
points in the – r direction, and has a magnitude ∆θ . ∆φ
(A7)
∂φ
Thus we have

∆θ = (–r )∆θ
Derivatives of φ
7) Change of φ with r
∆θ = – r = ∂θ (A6)
∆θ ∂θ As we noted earlier, the unit vectors do not change
with r, thus
z
∂φ
= 0 (A8)
∂r
z
θ r θ
∆θ
θ' θ'
y
θ
θ

x
∆θ is the change in the unit vector θ y
when we increase θ by ∆θ. φ |θ'| c
∆φ os θ

|θ|
x co
Unit vectors sθ
enlarged.
∆θ is the change in the unit vector θ
θ' θ
when we increase φ by ∆φ.
∆θ
|θ'| c |θ| = |θ'| = 1
os θ
|θ| ∆φ
∆θ = −r ∆θ co
sθ
∆θ = φ(cos θ)∆φ
∆θ points in the −r direction
∆θ points in the φ direction
Figure A4
Evaluation of ∂θ / ∂θ . Figure A5
Evaluation of ∂θ / ∂φ.
 Calculus 2000 - Chapter 4 Del Squared Cal 4-15

8) Change of φ with θ z axis straight up

y
As we can see from Figure (A6), the unit vector φ does
r'
not change direction when we change the angle θ . φ'
φ ∆φ
r
For example, when r is in the xz plane, φ points in φ
the +y direction for all angles θ . Thus
x
∂φ
= 0 (A9) ∆φ is the change in the unit vector φ
∂θ
when we increase φ by ∆φ.
9) Change of φ with φ ∆φ = (–u)∆φ
Finally, we have to figure out how the unit vector φ φ'
changes with the angle φ . This time we will take a ∆φ
φ
top down view as shown in Figure (A7). When we
change φ to φ + ∆φ , the unit vector φ goes to φ ′ .
From the small triangle we see that the change ∆ φ
points toward the z axis and has a magnitude ∆φ .
Unit vectors enlarged. The unit
In Figure (A8), we see that a unit vector u pointing vector –u points toward the z axis.
toward the z axis is given by Figure A7
Evaluation of ∂θ / ∂φ .
unit vector
pointing
= – r sinθ – θ cosθ
toward z
z axis r
u
Thus ∆φ = ∆φ(– r sinθ – θ cosθ) and we get θ
r
u = unit vector
∆φ ∂φ
= – r sinθ – θ cosθ = ∂φ (A10) θ
pointing toward
∆φ z axis

z
y
φ
φ
φ
θ r
∆θ
x
y z
θr
u
x
θ
The unit vector φ does not change r
when we increase θ by ∆θ.
θ
Figure A6 r-θ plane
Evaluation of ∂φ / ∂θ .

u = (–r ) sin θ + (–θ) cos θ

Figure A8
The unit vector we call u that points toward the z axis.
Cal 4 -16 Calculus 2000 - Chapter 4 Del Squared

Summary of Derivatives of Unit Vectors The terms in Equation (58) with a single line through
In summary, we get them are zero because the unit vectors are orthogo-
nal: i.e., r ⋅ θ = 0 , θ ⋅ φ = 0 , etc. Next we use our
∂r ∂r ∂r
= 0; = θ; = φsinθ summary, Equation (57) to evaluate the following
∂r ∂θ ∂φ
terms.
∂θ = 0 ; ∂θ = – r ; ∂θ = θcosθ r ⋅ ∂r = 0 because ∂r = 0 (A13a)
∂r ∂θ ∂φ ∂r ∂r

∂φ ∂φ ∂φ
= – r sinθ – θ cosθ r ⋅ ∂θ = 0 because ∂θ = 0 (A13b)
∂r
= 0;
∂θ
=0;
∂φ ∂r ∂r
(A11) ∂φ ∂φ
r⋅ = 0 because = 0 (A13c)
∂r ∂r
Calculation of ∇ 2 f
We are now ready to calculate ∇ 2f given again by θ ⋅ ∂r = θ ⋅ θ = 1 (A13d)
Equation (A1) ∂θ

∇ 2f(r,θ,φ) ≡ ∇ ⋅ ∇f(r,θ,φ) θ ⋅ ∂θ = θ ⋅ (– r ) = 0 (A13e)

∂θ
φ
= r ∂ + θr ∂ + ∂
θ⋅
∂φ ∂φ
∂r ∂θ r sinθ ∂φ ∂θ
= 0 because
∂θ
= 0 (A13f)

φ ∂f φ ⋅ ∂r = φ ⋅ (φ sinθ) = sinθ
⋅ r ∂f + θr ∂f +
(A13g)
∂φ
∂r ∂θ r sinθ ∂φ
φ ⋅ ∂θ = φ ⋅ (φ cosθ) = cosθ (A13h)
∂φ
= r ⋅ ∂r ∂f + r ∂ 2f + ∂θ 1r ∂f + θ ∂ 1r ∂f
∂r ∂r ∂r 2 ∂r ∂θ ∂r ∂θ ∂φ
φ⋅ = φ⋅ (– r sinθ – θ cosθ) = 0 (A13i)
∂φ
∂φ 1 ∂f
+ +φ∂ 1 ∂f
The terms in Equation (A12) for ∇ 2f , that are zero
∂r r sinθ ∂φ ∂r r sinθ ∂φ
because of Equations (A13), have a double line
through them. We are left with
+ θr ⋅ ∂r ∂f + r ∂ f + ∂θ 1r ∂f + θr ∂ f2
2 2
∇ 2f = r ⋅ r ∂ 2f + θr ⋅ ∂r ∂f
2
∂θ ∂r ∂θ∂r ∂θ ∂θ ∂θ ∂θ ∂r
∂r

+ ∂φ 1 ∂f + φ ∂ 1 ∂f
+ θr ⋅ θr ∂ f2 +
2 φ
⋅ ∂r ∂f
∂θ r sinθ ∂φ ∂θ r sinθ ∂φ ∂θ r sinθ ∂φ ∂r

φ φ φ ∂ 2f
⋅ ∂θ 1r ∂f + ⋅
+ φ ⋅ ∂r ∂f + r ∂ f + ∂θ 1r ∂f
2 +
r sinθ ∂φ ∂θ r sinθ r sinθ ∂φ 2
r sinθ ∂φ ∂r ∂φ∂r ∂φ ∂θ
(A14)
+ θr ∂ f + ∂φ 1 ∂f
2
∂φ∂θ ∂φ r sinθ ∂φ

+φ 1 ∂f 2
r sinθ ∂φ 2 (A12)
 Calculus 2000 - Chapter 4 Del Squared Cal 4-17

Using Equations (A13), Equation (A14) becomes

∂f 1 ∂ 2f
∇ 2f = ∂ 2f + 1r
2
+
∂r ∂r r 2 ∂θ 2

+ 1 sinθ ∂f + 1 cosθ 1 ∂f
r sinθ ∂r r sinθ r ∂θ

+ 1 ∂ 2f (A15)
r 2 sin 2 θ ∂φ 2
This becomes

∂f
∇ 2f = ∂ 2f + 2r
2

∂r ∂r

∂f ∂ 2f
+ 12 cosθ +
r sinθ ∂θ ∂θ 2
(A16)
+ 2 1 2 ∂ f2
2

r sin θ ∂φ

In most textbooks, you will find the equivalent

formula

∂2
∇ 2f = 1r 2 (rf)
∂r

+ 12 1 ∂ sin θ ∂f + 12 ∂ f2
2

r sinθ ∂θ ∂θ sin θ ∂φ

(10)
which is the result we stated earlier in the chapter.

Exercise 2
Show that Equation (A15) follows from Equations (A13)
and (A14).

Exercise 3
Show that Equations (A16) and (10) are equivalent.
 Calculus 2000 - Chapter 5 Complex Variables Cal 5-1

Calculus 2000-Chapter 5
Introduction to Complex Variables
CHAPTER 5 INTRODUCTION TO
COMPLEX VARIABLES

A ROAD MAP INTRODUCTION

In this chapter you will see that the use of complex After introducing the concepts of imaginary and
variables greatly simplifies the analysis of RLC complex numbers, we find that an important feature
circuits and other forms of sinusoidal behavior. of a complex number is that it can be expressed as a
This chapter does not depend on previous chapters complex exponential. We then go on to two major
of the Calculus text and may be studied directly in applications of complex variables that we just men-
connection with the related material in Chapters 27 tioned. One is the analysis of RLC circuits, which
and 31 of the Physics text. can be handled using real variables only, but where
there is an enormous simplification if we use com-
This chapter is also background material for the plex variables. Then in the next chapter, we discuss
next chapter, Chapter 6, on the Schrödinger's equa- the Schrödinger's equation where the equation itself
tion. The wave equations we have discussed so far involves complex variables.
can be solved using either real variables or complex
variables. Schrödinger's wave equation is different There are other topics involving the theory of com-
in that the equation itself involves complex numbers plex variables that we will not discuss in these
and cannot be handled by real variables alone. That introductory chapters. It is possible to construct
is why this chapter is a prerequisite. Also, to solve fascinating maps of complex functions and distort
Schrödinger's equation for the hydrogen atom re- these maps in intriguing ways (not completely unlike
quires the use of ∇ 2 in spherical polar coordinates, the distortion of images one can create on the
which we discussed in the last chapter. computer). Complex variables are also useful in
finding the formulas for various integrals. These
Once we finish Chapters 5 and 6 on complex vari- advanced topics are usually covered in a graduate
ables and Schrödinger's equation, we return to level mathematical physics course.
basic calculus operations, discussing divergence in
Chapter 7 and curl in Chapter 8. We then apply
divergence and curl to electromagnetism in Chap-
ters 9, 10, and 11, and to fluid dynamics in Chapters
12 and 13.
Cal 5-2 Calculus 2000 - Chapter 5 Complex Variables

IMAGINARY NUMBERS photographs showing the motion of a ball, we noted

What number, when multiplied by itself gives (-1)? that the position of the ball could be described by the
The answer is none of the ordinary numbers. This coordinate vector r , as shown in Figure (1). For the
number, – 1 is not one of the real numbers like 5, strobe photographs, which only show two dimen-
–2, 3, etc. It belongs to a completely different sions, the coordinate vector r was completely speci-
system of numbers which we call imaginary num- fied by its (x) and (y) components. Thus two
bers. The number – 1 is denoted by the letter i, and dimensional coordinate vectors and complex num-
the square root of any negative number can be bers are similar in that they both consist of two
written as a real number times i. For example independent components.

example This similarity suggests that we can treat a complex

–7 = 7 × (– 1) = 7 × –1 of an
(1) number in the same way we handle a two dimen-
imaginary
= ( 7 )i number
sional coordinate vector, plotting the real and imagi-
nary parts along different axes. It is traditional to
All numbers with one factor of i are imaginary.
plot the real part along the x axis and the imaginary
part along the y axis. Thus, for example, the com-
COMPLEX NUMBERS plex number (4 + 3i) can be represented by a point
We can make things a bit more complicated by whose coordinate vector has an x component of 4
adding together a real number and an imaginary and a y component of 3 as shown in Figure (2).
number, such as (4 + 3i). Such a mixture with both
a real part (4) and an imaginary part (3i) is called In this chapter you will see that in some cases there
a complex number. These two parts are distinct; is considerable simplification of the mathematics
there is no way we can confuse the real and imagi- and much greater insight when we use complex
nary parts because imaginary numbers are not part numbers. This is illustrated in our analysis of the
of the real number system. RLC circuit where we will see that a sinusoidal
oscillation and an exponential decay can both be
This is not the first time we have encountered a handled by one simple complex function.
quantity that has two distinct parts. In our strobe
imaginary

y
3i

r r

Figure 1 x Figure 2 4 real

The coordinate vector for a two Plot of the complex number (4 + 3i), where the real
dimensional strobe photograph. part is plotted along the x axis and the imaginary part
long the y axis.
 Calculus 2000 - Chapter 5 Complex Variables Cal 5-3

EXPONENTIAL FORM Now let us express z in terms of the variables r and θ

OF THE COMPLEX NUMBER rather than x and y. Since from Figure (2a) we see that
Once we start plotting complex numbers on x and y x = r cos θ
axes, we will find that any complex number can be y = r sin θ (5)
expressed in the exponential form re i θ. How we get
we can write (z) as
to this rather remarkable result can be seen in the
following way. z = x + iy = rcos θ + i rsin θ
Let us go back to Figure (2), showing the plot of the z = r(cos θ + i sin θ) (6)
complex number (4 + 3i). One way to describe that
point is to give its x and y coordinates (x = 4, y = 3i). It is the function (cos θ + i sin θ) that we wish to
An equally good description, shown in Figure (2a), study in detail.
is to give the distance r from the origin to the point,
Let us first look at the derivative of (cos θ + i sin θ) .
and the angle θ that r makes with the x or real axis.
Since
From the Pythagorean theorem we have
d cos θ = – sin θ; d sin θ = cos θ (7)
r 2 = x 2 + y 2 = 4 2 + 3 2 = 16 + 9 = 25 dθ dθ
r=5 (2) we get
The tangent of the angle θ is the opposite side y d (cos θ + i sin θ) = – sin θ + i cos θ
divided by the adjacent side x dθ
y Since (– 1) = i 2 , this can be written
tan θ = x = 3 = .75 (3)
4
d (cos θ + i sin θ) = i 2 sin θ + i cos θ
Entering .75 in our calculator and pressing the tan – 1 dθ
(8)
button gives 36.9°. Thus the point is located at a = i (cos θ + i sin θ)
distance r = 5 from the origin at an angle θ = 36.9° .
To express this result more formally, let us write
It is traditional to use the letter (z) to describe a
complex number. Thus if a complex number (z) has f(θ) = (cos θ + i sin θ) (9)
a real part (x) and an imaginary part (iy), we can write Then Equation (8) becomes
z = x + iy (4) d f(θ) = i f(θ) (8a)
dθ
To within a constant (i), the function f(θ) is equal to
imaginary
its own derivative. What function that you are
already familiar with, behaves this way? The expo-
nential function! Recall that
y d e ax = ae ax (10)
dx
r Thus if we replace (x) by θ and (a) by (i) , we get
d e iθ = i e iθ (11)
dθ
θ
x real
Comparing Equations (8) and (11), we see that the
Figure 2a
function (cos θ + i sin θ) and the function e iθ obey
Plot of the complex number (4 + 3i),
showing the angle θ . the same rule for differentiation.
Cal 5-4 Calculus 2000 - Chapter 5 Complex Variables

When two functions (cos θ + i sin θ) and e iθ have Let us check the accuracy of these expansions for
the same derivatives, does that mean that they are the θ = .1 radians. We have, keeping three terms,
same functions? It will if we show that both func-
(.1) 2 (.1) 4
tions start off with the same value for small values of cos (.1) = 1 – +
2! 4!
θ . Then as we increase θ , if both functions have the
same derivative or slope, they must continue to be = 1 – .01 + .0001
2 4×3×2
the same function for all values of θ . (15)
= 1 – .005 + .0000004166
Small Angle Approximations = .995004166
We can show that (cos θ + i sin θ) and e iθ have the
Changing our calculator from degrees to radians and
same values for small θ by using the small angle
taking the cos(.1) gives
approximations for sin θ , cos θ and e iθ. In our
discussion of the exponential in Chapter 1 of the cos (.1) = .995004165 (16)
Calculus text, (Cal 1, Eq. 136), we had
We see that we get almost a nine place accuracy by
x2 x3 keeping the first three terms of the expansion.
ex = 1 + x + + + ⋅ ⋅ ⋅ (1-136)
2! 3!
For sin(.1), keeping the first three terms, we have
While this expansion is true for any value of x, it is
most useful for small values of x where we do not (.1) 3 (.1) 5
have to keep many terms to get an accurate answer. sin (.1) = .1 – +
3! 5!
Setting x = iθ gives
= .1 – .001 + .0001
3×2 5×4×3×2
e iθ = 1 + iθ + i θ + i θ + ⋅ ⋅ ⋅
2 2 3 3
(12) = .1 – .000166666 + .000000083
2! 3!
(Since our previous discussion of exponents only = .009833417 (17)
dealt with real numbers, we can consider Equation
(12) as the definition of what we mean when the The calculator gives
exponent is a complex number).
sin (.1) = .099833416 (18)
What we did not discuss earlier were the expansions
for cosθ and sinθ . Let us state them and check their which is again accurate to almost nine places.
accuracy now. They are
If you can't figure out how to get your calculator to
work in radians, you can convert .1 radians to
cos θ = 1 – θ + θ + ⋅ ⋅ ⋅
2 4
(13) degrees by using the conversion factor
2! 4!

sin θ = θ – θ + θ + ⋅ ⋅ ⋅
3 5 360 degrees/cycle degrees
(14) = 57.29577951 (19)
3! 5! 2 π radians/cycle radian

where θ is in radians. Again these expansions are

valid for any value of θ , but most useful for small
values where we do not have to keep many terms.
 Calculus 2000 - Chapter 5 Complex Variables Cal 5-5

Now that we have checked the expansions (13) and If we add Equations (22) and (23), the sin θ terms
(14) let us see what the expansion for (cos θ + i sin θ) cancel, and we are have, after dividing through by 2,
is. We get, replacing – signs by i 2 , and using
i 4 = + 1, e iθ + e –iθ
cos θ = (24)
cos θ = 1 – θ + θ = 1 + i θ + i θ
2 4 2 2 4 4 2
2! 4! 2! 4!
Subtracting Equation (23) from (22) cancels the
i sin θ = i θ – θ + θ = iθ + i θ + i θ
3 5 3 3 5 5
cos θ terms, leaving, after dividing through by 2i,
3! 5! 3! 5!
(20) e iθ – e –iθ
sin θ = (25)
2i
Adding these gives

(cos θ + i sin θ) = 1 + iθ + i θ + i θ
2 2 3 3 If we note that
2! 3! 1 = 1 × i = i = i = –i
i i i i2 –1
+i θ +i θ + ⋅ ⋅ ⋅
4 4 5 5
4! 5! we can write Equation (25) as
(21) e iθ – e –iθ
which is just the expansion we had for e iθ in Equa- sin θ = (– i)
2
tion (12).
e –iθ – e iθ
Comparing Equations (20) and (21), you can see that sin θ = i (25a)
2
the expansions for cos θ and i sin θ fit together to
produce the much more regular expansion of e iθ. Equations (22) through (25) give a complete pre-
We will also see that it is often much easier to work scription of how to go back and forth between cos θ,
with the complete function e iθ than with cos θ and sin θ , e iθ and e – iθ.
sin θ separately.
Finally returning to our complex function
In summary, if we define a complex exponential by
the series expansion of Equation (12), then we have z = x + iy imaginary
shown that
= r cos θ + i r sin θ
e iθ = cos θ + i sin θ (22) y
= r (cos θ + i sin θ)
r
Even though we checked the sin and cosine expan- we now have
sions for a small value of θ , the fact that e iθ and θ
(cos θ + i sin θ) have the same derivative properties z = r e iθ (26) x real

means that Equation (22) holds for all values of θ .

If we replace θ by – θ in Equation (22) we get as the other way of expressing a complex number,
where r is the distance from the origin and θ the angle
e –iθ = cos(–θ) + i sin(–θ) the coordinate vector r makes with the x or real axis.
Since cos(–θ) = cos θ , sin(– θ) = – sin θ , this gives
Exercise 1
(a) Construct a series expansion for e– iθ.
e –iθ = cos θ – i sin θ (23)
(b) Using the series expansions for eiθ and e– iθ in
Equation (25a), show that you end up with the series
expansion for sin(θ).
Cal 5-6 Calculus 2000 - Chapter 5 Complex Variables

The Complex Conjugate Z* DIFFERENTIAL EQUATIONS

The complex conjugate of a complex number is FOR R, L, C CIRCUITS
defined as the number we get by replacing all factors One of the most convenient uses of complex vari-
of (i) by (–i) in the formula for the number. We ables is in the analysis of electric circuits involving
generally denote the complex conjugate by placing resistors, capacitors and inductors. We will see that
an asterisk after the number. For example, if using complex variables unifies the analysis and
z = x + iy greatly simplifies the work involved.
then
z* = x – iy (26a) The RC Circuit
Let us begin with the RC circuit shown in Figure (3).
If we start with
If we charge up the capacitor to a voltage V0, and
z = r e iθ close the switch, a current flows out of the capacitor
then through the resistor, and the voltage VC on the
z* = r e –iθ (26b) capacitor decays exponentially.
The main reason for defining a complex conjugate is The formulas for the capacitor voltage VC and resis-
that the product of a complex number z with its tor voltage VR are
complex conjugate z* is always a real positive
Q
number, equal to the square of the distance r that the VC = ; VR = iR (27)
complex point is from the origin. For our two C
examples above, we have where Q is the charge on the capacitor, C the
capacitor's capacitance, (i) the current through the
z*z = (x – iy)(x + iy) y resistor and R the resistor's resistance in ohms. It is
= x 2 – ixy + iyx – i 2y 2 2 +
2
y assumed that C and R are constants and that (i) is the
x
r2
= rate at which charge Q is leaving the capacitor. That is,
= x2 + y2 = r2
x dQ
and i = – (28a)
dt
z*z = (r e –iθ )(r e iθ ) = r 2 (26c)
Setting the sum of the voltage rises around the circuit
equal to zero (see Equation 27-41 in the Physics text)
and using (28a) for i, gives us
VC – VR = 0
switch i Q Q dQ
– iR = + R = 0 (28b)
C C dt
Dividing through by R, we get
Q R VR = iR
VC = C C
dQ Q
+ = 0 (29)
dt RC

Figure 3 as the differential equation for the amount of charge

The RC circuit. When we walk around in the Q remaining in the capacitor.
direction shown by the circular arrow, we go
with VC but against VR , giving VC – VR as the
sum of the voltage rises.
 Calculus 2000 - Chapter 5 Complex Variables Cal 5-7

An Aside on Labeling Voltages With i = + dQ/dt , this gives

To avoid worrying about minus signs like the
i = – dQ/dt for the discharging capacitor, we will dQ Q dQ Q
R + = 0 ; + = 0 (29a)
obtain the differential equations for our L, R, and C dt C dt RC
circuits by sketching the voltages when the rate of
This is just the same as Equation (29) for the dis-
change of charge in the capacitor and change of the
charge of the capacitor.
current in an inductor are all positive. If we had an
increasing current running down through three cir- Figure (5) appears less intuitive than Figure (3)
cuit elements R, L, and C, all because we have drawn a current flowing into the
i dQ
the voltages would point up as i=
dt capacitor, while we know that the current actually
shown in Figure (4). The re- flows out. But the fact that we analyzed the circuit
sistance voltage VR is always in Figure (5), assuming the wrong direction for the
R VR = i R
directed opposite to the cur- current, does not affect the resulting differential
rent. If the downward current equation for the circuit. When using Kirchoff's laws
is increasing, then the induc- to analyze a circuit, you do not have to know the
tor opposes the increase and correct direction for the currents ahead of time. If
points upward. With a posi- L VL = L di you make the wrong guess, the resulting equations
dt
tive current flowing into the will fix things up by giving you a minus sign.
capacitor, the current is equal
to + dQ/dt . If the capacitor While Figure (5) is less intuitive than Figure (3), it
C VC = Q
started off with zero charge, C is much more straightforward to stick with all posi-
then the upper plate is becom- tive quantities and always label our circuit element
ing positively charged by the Figure 4 voltages and currents as shown in Figure (4). With
positive current flowing into it. Direction of the more complex circuits it is the only way to maintain
voltages for an sanity and get the right differential equation.
Using these conventions for increasing
the current and voltages, we downward
current.
can construct an RC circuit
from Figure (4) by pulling out the inductor and i i=
dQ
connecting the back side of the circuit as shown in dt
Figure 5
Figure (5). Setting the sum of the voltages to zero The RC circuit for a
around Figure (5) gives (walking around the circuit positive increasing R VR = i R
counterclockwise as shown by the circular arrow) current i.

VR + VC = 0
C VC = Q
Q C
iR + = 0 (30)
C
Cal 5-8 Calculus 2000 - Chapter 5 Complex Variables

Solving the RC Circuit Equation The LC Circuit

Solving the differential equation We will construct an LC circuit from Figure (4) by
dQ Q taking out the resistor and connecting the back side
+ = 0 (29) repeated as shown in Figure (6). Setting the sum of the
dt RC
voltage rises around this circuit equal to zero gives
for the capacitor discharge was quite straightfor-
ward. We first looked at the circuit experimentally VL + VC = 0 (35)
and saw that the voltage Q/C appeared to decay
exponentially as shown in Figure (27-44c) from the Q dQ
L di + = 0 ; i = + (36)
Physics text, reproduced here. This suggested that dt C dt
we try, as a guess, a solution of the form Writing di/dt = d 2 Q/dt 2 , and dividing through by L
Q = Q 0 e – αt (31) gives
dQ
= – αQ 0 e – αt (32) d 2Q
+
Q
= 0
dt 2 (37)
dt LC
Plugging our guess into Equation (30) gives
Q 0 e – αt Now suppose we naïvely try the same exponential
– αQ 0 e – αt + ? 0
= decay solution we had for the RC circuit
RC
The constants Q 0 and the functions e – αt cancel,
dQ
and we are left with Q = Q 0e – αt ; = – αQ 0e – αt (guess)
dt
–α + 1 = ? 0
RC d 2Q
We can satisfy the differential equation if α has the = – αQ 0 – αe – αt = α 2Q 0e – αt (38)
dt 2
value
Plugging our guess (38) into the differential Equa-
1
α = (33)
tion (37) gives
RC
Q 0 e – αt
The formula for Q becomes α 2Q 0 e – αt + =? 0
LC
Again the Q 0′s and e – αt ′s cancel and we are left with
Q = Q 0 e – t /RC
(34)
We see the time constant for the decay of the charge α 2 + 1 =? 0 (39)
LC
Q is T = RC. I.e., when t gets up to T = RC, the value
of the charge has decreased to e – 1 = 1/e of its The differential equation will be solved if we can set
original value. α2 = – 1 (40)
LC
V

α = – 1 –1 (41)
LC

Figure 6 L VL = L di
T dt
R = 10K, d = 2mm
The LC circuit.
Figure 27-44c
Discharge of our aluminum plate capacitor
(separation 2mm) through a 10KΩ resistor. C VC = Q
C
The inset is the experimental data and the
solid curve is drawn from that data.
 Calculus 2000 - Chapter 5 Complex Variables Cal 5-9

When we tried this in the Physics text, we noted However in Equations (24) and (25) we saw that we
that α comes out imaginary. We also noted that the could construct the real functions cosθ and sinθ
LC circuit oscillated rather than decayed. Thus we from the complex functions e iθ and e –iθ . Replacing
concluded that we had guessed the wrong function, θ by ω0t , we have
and tried a sine wave
cosω0t = 1 e i ω 0t + e – i ω 0t (47a)
Q = Q 0 sinω0t (42) 2

instead. When you plug the guess (42) into the differ- sinω0t = i e – i ω 0t – e i ω 0t (47b)
ential Equation (37) you end up with 2
One of the features of differential equations like the
ω0 2 = 1 ; ω0 = 1 (43) one for the LC circuit* is that if the equation has
LC LC more than one solution, any combination of the
solutions is also a solution of the equation. In our
which avoids imaginary numbers and gives a result in
case the two solutions are Q 1 = Q 0e iω 0t and
agreement with experiment. The quantity ω0 = 1/ LC
Q 2 = Q 0e –iω 0t . Thus the combination
is the resonant frequency of the LC circuit. (If you do
not remember plugging the guess Q = Q 0 sin(ω0t) Q = aQ 1 + bQ 2 a,b,constants (48)
into the LC differential equation, do so now.)
must also be a solution, as you will check for yourself
Knowing more about handling imaginary numbers, let in Exercise 2. Choosing the constants a = 1/2,
us see what happens if we take our guess Q = e – αt b = +1/2 gives
seriously for the LC circuit. We still have to satisfy
Q = Q 0 cosω0t (49a)
Equation (40),
and choosing a = – i/2 , b = i/2 gives
α 2 = – 1/LC
Q = Q 0 sinω0t (49b)
Writing 1/LC = ω0 2 , we get
These are both real functions which can describe the
α 2 = – ω0 2 (44) electric charge in the capacitor.
which has two solutions, namely
Thus we see that for both the RC and the LC circuit,
α = iω 0 (45a) we can use the same trial function Q = Q 0e – αt . For
α = –iω 0 (45b) the RC circuit, α was a real number, which gave us
2
You can see this by noting that both i = –1 and the exponential decay Q = Q 0 e – t /RC. For the LC
(–i) 2 = – 1. Thus the possible solutions for Q are circuit, α turned out to be imaginary which gave us
real oscillating solutions like Q = Q 0 cos ω0t . By
Q 1 = Q 0e i ω 0 t (46a) using complex numbers, we are able to handle both
Q 2 = Q 0e – i ω 0t (46b) the RC and the LC circuits with the same trial
function. Whether α turns out to be real or imagi-
While Equations (46a, b) are both mathematical nary tells us whether the circuit decays or oscillates.
solutions to the differential equation for the LC
circuit, both are complex functions. But the amount
of charge Q in the capacitor must be described by a
real number. No imaginary charge resides there. *This is an example of what is called a homogenous differential
equation. We will have more to say about them shortly.
Cal 5-10 Calculus 2000 - Chapter 5 Complex Variables

In the next section we will consider the RLC circuit, A FASTER WAY TO FIND REAL SOLUTIONS
which is an LC circuit with resistance included. When we got the complex solutions e +i ω 0t and
Experimentally we saw that such a circuit could e – i ω 0t for the LC circuit differential equation, we
have a decaying oscillation. When we plug the were careful to construct real combinations of these
guess Q = Q 0e – αt into the equation for the RLC complex solutions. You might think that it was
circuit, α will turn out in some cases to be complex, lucky that we just happened to know that the combi-
i.e., have both a real and an imaginary part. The 2 e i ω 0t – e – i ω 0t ) was the real function
nation 1 2(
imaginary part will describe the oscillation of the cosω0t . You might be concerned that for some other
circuit while the real part will tell us how the oscil- differential equations you would not be so lucky.
lation decays. But before we get to the RLC circuit,
we need to discuss a simpler way to get real solutions Don't worry. If you find a complex solution for a
from complex solutions of differential equations. homogeneous differential equation, you can simply
Before that, do Exercise 2 to see that Q = aQ 1 + bQ 2 take the real part of the complex solution and throw
is a solution of our LC equation. away the imaginary part. This works because both
the real part and the imaginary part must separately
Exercise 2 be solutions of the differential equation. (You could
The differential equation for an LC circuit is also keep the imaginary part without the (i) and
throw away the real part.)
d2Q + Q = 0 (37) repeated
dt2 LC To see why both the real and imaginary parts are
This is called a homogenous differential equation be- solutions, let us write the complex solution for Q in
cause it contains only terms involving Q or its deriva- the form
tives. An example of a non homogeneous differential Q = Q real + iQ imaginary (52)
equation will be
where both Q real and Q imaginary are real functions.
d2Q + Q = a sin ω t (50) Plugging Equation (52) into the LC differential
1
dt2 LC equation gives
This will represent an LC circuit that is being forced to d 2Q Q d 2Q real Q real
oscillate at some frequency ω1 . The appearance of the + = +
term ( a sin ω1t ) with no factor of Q makes this a non
dt 2 LC dt 2 LC
homogeneous equation. We will discuss this equation
shortly, to show what effect the non homogeneous term d 2Q imaginary Q imaginary
has. For now we will limit our discussion to homoge- +i +
dt 2 LC
neous equations.
= 0 (53)
You have seen that Q1 = ae i ω 0t and Q2 = be– i ω 0t are
both solutions to Equation (37) when a and b are Since both Q real and Q imaginary are real functions,
constants and ω0 = 1/ LC . Now explicitly plug in their derivatives must also be real functions, and the
Q = ae i ω 0t + be– i ω 0t (51) quantities inside both square brackets in Equation
(53) must be real. As a result the first square bracket
into Equation (37), and show that this is a solution for any is purely real, and the second square bracket with its
constant values of a and b. This demonstrates that any factor of (i) must be purely imaginary. The only way
linear combination of e i ω 0t and e– i ω 0t is also a solution. you can add purely real and purely imaginary func-
tions together to get zero is for both functions to be
separately equal to zero.
 Calculus 2000 - Chapter 5 Complex Variables Cal 5-11

That is, we must have THE RLC CIRCUIT

d 2Q real Q real Adding a resistor to an LC circuit gives us the RLC
+ = 0 (54a) circuit shown in Figure (7). If the resistance R is not
dt 2 LC
too large, we get a decaying oscillation like that
d 2Q imaginary Q imaginary shown in Figure (31-A9) taken from the Physics text.
+ = 0 (54b)
dt 2 LC
The equation for the RLC circuit is obtained by
Equations (54) tell us that both functions Q real and setting to zero the sum of the voltage rises around the
Q imaginary must be solutions of the LC differential circuit, giving
equation. If we want a real solution, we can use
either Q real , Q imaginary or any linear combination of VR + VL + VC = 0 (56)
the two. A similar argument applies to the solution Q
of any homogeneous differential equation. iR + L di + = 0 (57)
dt C
Setting
As an explicit example for our LC equation, suppose
2
we had come up with the solution dQ di = d Q
i = ; (58)
dt dt dt 2
Q = Q 0e i ω 0 t (55)
and dividing through by L gives
and had not noticed that Q = Q 0e – i ω 0t was also a
solution. Instead of hunting for another complex d 2Q + R dQ + Q = 0 the LRC (59)
equation
solution and then trying to find real combinations, dt 2 L dt LC
we could just break e i ω 0t into its real and imaginary As a trial function, suggested by the decaying oscil-
parts using e iθ = cos θ + i sin θ to get lation of Figure (31-A9), we could try the solution
Q = Q 0 e – αt cosωt guess (60)
Q = Q 0e i ω 0t = Q 0 cosω0t + iQ 0 sinω0t (55a)
Then we immediately know that the real functions
Q 0 cosω0t and Q 0 sinω0t are solutions of the LC
differential equation. We can use either one or some
linear combinations of the two. (Using a linear
combination is equivalent to using an arbitrary phase
angle, like Q = Q 0 sin(ω0t + φ) . See the Physics experimental data
text, pages 15-17 or 16-31.)
i

R VR = i R

Figure 31-A9 – Ringing like a bell

Figure 7 We hit the RLC circuit with a square wave and the
The RLC circuit. circuit responded like a bell struck by a hammer.
L VL = L di We are looking at the voltage across the capacitor.
dt

C VC = Q
C
Cal 5-12 Calculus 2000 - Chapter 5 Complex Variables

If you plug the guess (60) into Equation (59), you get The Easy Way
many terms involving both sin ωt and cos ωt . Working with separate sines and cosines is the
Q = Q 0 e – αt cosωt guess (60) repeated difficult way to handle the RLC circuit. Using
complex variables which provide a unified treat-
d 2Q + R dQ + Q = 0 (59) repeated ment of both decay and oscillation is the easy way.
dt 2 L dt LC
For a trial solution, let us use
To see where the terms come from, consider
dQ
dQ Q = Q 0 e – at ; = – aQ 0 e – at
= Q 0(– α) e – αt cosωt + Q 0 e – αt (– sinωt) dt
dt
(61) d 2Q
– αt = a 2Q 0 e – at (62)
where we had to differentiate the two terms e and dt 2
cosωt separately. Differentiating again we get four It looks much easier already. Substituting this trial
terms for d 2Q/dt 2 , two with a cosωt and two with a solution into the LCR differential equation gives
sinωt . When we plug this all back into Equation (59),
we end up with seven terms, four with cos ωt and three d 2Q R dQ Q
+ + = 0 (59) repeated
with sin ωt . In order for all this to be equal to zero, you dt 2 L dt LC
have to separately set the sinωt and the cosωt terms
Q e – at
equal to zero. This leads to two equations, from which a 2Q 0 e – at – aR Q 0 e – at + 0 =? 0 (63)
you can determine both the constants α and ω . If you L LC
are careful, your chances of getting the answer without The function e – at and constant Q 0 cancel and we
making a mistake may be as high as 50%. In other are left with
words this is the hard way to solve the problem.
a 2 – aR + 1 =? 0 (64)
L LC
Exercise 3 This is a standard quadratic equation of the form
Try finding the coefficients α and ω by using Equation x 2 + bx + c = 0 (65)
(60) as a trial solution for Equation (59). Then check whose solution is
your answer with the one we get in the next section.
x = –b± b 2 – 4c (66)
2
For our case, –b = R/L, c = 1/LC, thus (a) is given by

a = 1 R± R2 – 4
2 L L 2 LC

= R ± R2 – 1 (67a)
2L 4L 2 LC

= R ± (– 1) 1 – R2 (67b)
2L LC 4L 2
Setting 1/LC = ω0 2 , where ω0 is the resonant fre-
quency of the undamped (R = 0) circuit, and taking
– 1 outside the square root as a factor of (i) gives

Figure 31-A9 (repeated) 2

We are looking at the voltage across a = R ±i ω0 2 – R 2 (68)
2L 4L
the capacitor in an RLC circuit.
 Calculus 2000 - Chapter 5 Complex Variables Cal 5-13

We now introduce the notation As R and α = R/2L are increased, the oscillation
frequency ω = ω0 2 – α 2 decreases until we reach
α = R α = ω0 . At that point, ω = ω0 2 – α 2 = 0 , oscilla-
2L
(69) tion ceases, and we have what is called critical
2 damping. The time constant for decay at critical
ω = ω0 2 – R 2 = ω0 2 – α 2
4L damping is just the length of time it takes the
So that undamped circuit to go through one radian of oscil-
lation, or 1/2π of a complete cycle. You can see that
a = α ± iω (70) result from dimensions. We have
and our trial solution Q = Q 0e –at becomes 1 = ω1 seconds (76)
Q = Q 0e – (α ± iω) t ω0 radians 0 radian
second
Q = Q 0e –αt e ±i ωt
solution for the
RLC circuit (71) and at critical damping, where α = ω0 ,

where 1 = 1 seconds
T = α (77)
ω0 radian
α = R ; ω = ω0 2 – α 2 (69) repeated At critical damping, there is only one unique solu-
2L
tion for the RLC circuit. As we increase the resis-
For the case where ω0 is bigger than α , tance beyond critical damping, when α = R/2L be-
ω = ω0 2 – α 2 is a real number, the real part of comes larger than ω0 , the solution becomes
e iωt is cosωt and we get the real solution overdamped. For α > ω0 , it is easiest to go back to
writing the solution in the form
Q 1 = Q 0e – αt cosωt (72)
Q = Q 0e – at (from Eq.62)
The imaginary part of e iωt is proportional to sinωt ,
which gives us the other real solution a = R ± R2 – 1 = α ± α 2 – ω 2
0
2L 4L 2 LC
Q 2 = Q 0e – αt sinωt (73) (from Eq.67a)
As in the case of the LC circuit, the sine and cosine and we see that we now have two exponential decay
waves can be combined as a sine wave with an solutions
arbitrary phase angle φ to give the general solution Q 1 = Q 0 e – α + α2 – ω02 t (78a)

Q = Q 0e – αt sin(ωt + φ) damped oscillation

of an RLC circuit Q 2 = Q 0 e – α – α 2 – ω02 t (78b)
(74) If we increase the resistance so much that ω02 is
Equation (74) represents a damped oscillation of completely negligible compared to α 2 , then the two
frequency ω = ω0 2 – α 2 and a damping time con- solutions become
stant T given by
Q 1 → Q 0 e – 2αt α 2 > > ω0 2 (79a)
1 = 2L
T = α
damping time
Q2 → Q0 e0 = Q0 (79b)
R constant (75)
In this limit we easily see that the solution Q 1 damps
Imagine that we start with an RLC circuit that more rapidly than Q 2 . For the Q 2 solution, we have
initially has negligible resistance, and that we gradu- increased the resistance so much that no charge
ally increase the resistance. When R = 0, then α = 0 leaves the capacitor and the charge remains at Q 0 .
and the oscillation frequency is ω = ω0 2 = ω0 ,
where ω0 is the undamped frequency.
Cal 5-14 Calculus 2000 - Chapter 5 Complex Variables

We can get a better insight into the solution Q 2 by Exercise 4

assuming that ω0 2 is small but not quite zero. In this To make our study of the RLC circuit more concrete,
case we can write Q 2 as suppose that in the circuit you use a 0.10 microfarad
capacitor and one millihenry inductor, so that
Q 2 = Q 0 e–α 1 – 1 – ω 20 α 2 t (80)
L = 10– 3 hy
Since ω02
α 2 << 1 , we can use the approximation
formula C = 10– 5 farads

1–x ≈ 1– x x << 1 (a) What is the resonant frequency ω0 radians/second

2 and f0 cycles/second, when R = 0?
We get, for x = ω02 α 2 ,
(b) What is the length of time it takes the R = 0 circuit to
go through one radian of its oscillation?
ω02 ω02
α 1– 1– ≈ α 1– 1– (c) What value of resistance R C should you use for
α2 2α 2
critical damping?

ω02 (d) What is the time constant for the decay at critical
= α = 1 ω02 1 damping?
2α 2 2 α
(e) Suppose you raise R from its critical value R C up to
With ω02 = 1/LC and α = R/2L, we get 2 R C. What are the time constants T1 and T2 for the
decay of the solutions Q1 and Q2 respectively? (Partial
= 1 × 1 × 2L
2 LC R answer: Q2 takes twice as long as Q1 to decay when
R = 2 R C.)
= 1 (81)
RC
Thus for α >> ω0 we have
2 2

Q 2 = Q 0 e – t /RC α 2 >> ω02 (82)

This is just the solution for the decay of an RC circuit

with a time constant T = RC.
The condition α 2 >> ω02 can be written as

R 2 >> 1 or R 2C >> 1 (83)

4L 2 LC 4L
We can meet this condition for finite values of R and
C by making L small enough.
 Calculus 2000 - Chapter 5 Complex Variables Cal 5-15

IMPEDANCE The individual voltages were calculated noting that

Circuits commonly encountered are AC circuits d sin ωt = ω cos ωt
where the current has a sinusoidal form dt
i = i 0 sin ωt (84) 1 cos ωt .
sin ωt dt = – ω
For standard American households, the household
current has a frequency of 60 cycles/second, or The voltage V across all three elements is just the
ω = 2π × 60 radians/second. In much of the rest of sum of the individual voltages
the world the standard household frequency is 50
cycles per second. World War II aircraft used a V = i 0 R sin ωt + Lω cos ωt – 1 cos ωt
Cω
standard frequency of 400 cycles per second which
resulted in smaller and lighter transformers.
= i 0 R sin ωt + Lω – 1 cos ωt
The concept of impedance, which involves complex ωC
variables, provides an easy way to handle the volt-
ages across R, L, and C circuit elements in an AC V = i 0 [Asin ωt + Bcos ωt] (86)
circuit. To demonstrate the advantage of the com-
plex variable approach, we will first analyze these where
voltages using our standard real variables, and then
see how much the calculations are simplified by A = R; B = Lω – 1 (87)
ωC
complex variables.
We want to express the term [Asin ωt + Bcos ωt] as
Suppose we have three circuit elements, an R, L, and C, a single sine wave with an amplitude which we will
connected in series as shown in Figure (8), and run call Z 0 , and a phase angle φ
an AC current through them. In the diagram we
show the formula for the voltage across each circuit [Asin ωt + Bcos ωt] = Z 0 sin(ωt + φ) (88)
element. What we wish to calculate is the total To do this we use the trigonometric identity
voltage V across all three elements.
i = i 0 sinωt sin (a + b) = cos b sin a + sin b cos a
to write
sin (ωt + φ) = cos φ sin ωt + sin φ cosωt (89)
R VR = iR = i 0 R sinωt (85a)
Multiplying through by Z 0 gives
Z 0 sin (ωt + φ) = (Z 0 cos φ) sin ωt + (Z 0 sin φ) cosωt

= Asin ωt + Bcos ωt (90)

V= ?
L VL = L di = i 0 Lω cosωt where
dt
(85b) A = Z 0 cos φ ; B = Z 0 sinφ (91)

B = sin φ = tan φ
VC = Q = 1 idt (92)
C C C A cos φ

=
– i0
cosωt (85c) A 2 + B 2 = Z 0 2(cos 2 φ + sin 2 φ) = Z 0 2 (93)
Cω
Figure 8
AC voltages in the R, L, and C circuit elements.
Cal 5-16 Calculus 2000 - Chapter 5 Complex Variables

Applying Equations (91), (92), and (93) to our formula Now let us see how much more quickly we can arrive
at the amplitude Z 0 and phase shift φ using the
V = i 0 [Asin ωt + Bcos ωt] (86) repeated complex variables shown in Figure (9).
gives
In Figure (9) we have a current i given by the formula
V = i 0 Z 0 sin (ωt + φ) (94) i = i 0 e i ωt (97)
and the resulting voltage across the three circuit
where from Equations (92) and (87) elements is the sum of the individual voltages which
can easily be written in the form
tan φ = B = Lω – 1/ωC (95)
A R
V = i 0 R + i Lω – 1 e i ωt (98)
ωC
and from Equation (93)
The quantity in square brackets is the complex
2 number R + i( Lω – 1/ωC ) graphed in Figure (10).
Z 0 2 = R 2 + Lω – 1 (96) It can be represented by an arrow whose length is Z 0
ωC
given by the Pythagorean theorem as
After a fair amount of calculation, we see that the 2
voltage across all three circuit elements is still propor- Z 0 2 = R 2 + Lω – 1 (99)
ωC
tional to sin ωt. Its amplitude Z 0 is given by
Equation (96) and there is a phase shift by an angle and is oriented at an angle φ whose tangent is
φ that is given by Equation (95).
tan φ = Lω – 1/ωC (100)
i = i 0 e i ωt R
Notice that the formulas for Z 0 and tan φ are the
same as in Equations (96) and (95), which we got
after so much more work.
VR = iR = i 0Re i ωt
R
imaginary
Lω

V= ?
L VL = L di = L i 0(iω) e i ωt Lω – 1
ωC
dt C0
φ
R real
VC = Q = 1 idt
C C C
– 1
ωC
i i
= 0 e i ωt = – i 0 e i ωt
iωC ωC
Figure 9 Figure 10
AC voltages in the R, L, and C circuit Graph of the complex number R + i ( L ω – 1/ω C) .
elements, using complex notation.
 Calculus 2000 - Chapter 5 Complex Variables Cal 5-17

From our earliest work with complex variables we Equation (102) is our complex formula for the voltage
saw that the complex number across the three circuit elements.
z = x + iy (4) repeated To find the real voltage, we simply take the real (or
could be written as the exponential imaginary) part of the complex voltage. Choosing
the imaginary part (without the i) to get a sine wave,
z = re iθ (26) repeated
we get
where z is graphed in Figure (2a) repeated here.
V = i 0 Z 0 sin (ωt + φ)
Thus the complex number R + i( Lω – 1/ωC ) , (103)
graphed in Figure (10) can also be written in the
which is the same answer, Equation (94), that we got
exponential form
from the real analysis.
R + i Lω – 1 = Z 0 e iθ (101)
ωC The main advantage of the complex analysis is that
where Z 0 is the distance from the origin and φ the all the voltages had the same factor e i ωt, so that we
angle above the real axis. could simply add the voltages without using the
fairly messy trigonometric identities. Also note that
Using Equation (100) for the square brackets in the main result of all the work of the real analysis
Equation (98) for the voltage V gives was to calculate the amplitude Z 0 and the phase
angle φ . We got Z 0 and φ immediately in the
V = i 0 R + i Lω – 1 e i ωt
ωC complex analysis, as soon as we graphed the com-
plex coefficient of e i ωt in Figure (10).
= i 0 Z 0 e i φ e i ωt

V = i 0 Z 0 e i (ωt + φ)
(102)
2
Z 0 2 = R 2 + Lω – 1 (99) repeated
ωC

tan φ = Lω – 1/ωC (100) repeated

R
imaginary

φ
x real
Figure 2a (repeated)
Plot of the complex number (4 + 3i),
showing the angle φ .
Cal 5-18 Calculus 2000 - Chapter 5 Complex Variables

Impedance Formulas If we define the total impedance Z of the three circuit

The concept of a complex impedance which we will elements connected in series by the equation
now introduce, allows you to determine the ampli-
tude Z 0 and phase angle φ by inspection, without Z = ZR + ZL + ZC
(107)
doing hardly any calculation at all.
then our formula for the complex voltage is
In Figure (11), we have redrawn our three circuit
elements, introduced a complex current i = i 0 e i ωt , V = iZ (108)
and expressed voltage in terms of i and the complex
impedances Z R, Z L , Z C defined by Comparing this with Ohm's law for a single resistor
ZR ≡ R (104a) VR = iR Ohm's law (Physics 27-1)
Z L ≡ iωL (104b) we see that we can think of Equation (108) as simply
a complex form of Ohm's law.
ZC ≡ – i
ωC (104c)
When we graph the complex impedance Z we can
In terms of these Z′s, the voltages are immediately read off the amplitude Z 0 and phase
VR = iZ R angle φ , as shown in Figure (12). We have

VL = iZ L complex
Z = R + i Lω – 1 = Z 0 e iφ impedance
(105) ωC
VC = iZ C
(109)
The sum of the three voltages V becomes where
2
magnitude of
V = VR + VL + VC Z 0 2 = R 2 + Lω – 1 impedance (110a)
ωC
= i(Z R + Z L + Z C ) (106)
phase of
tan φ = Lω – 1/ωC impedance
(110b)
i = i i0 e i ωt R
In Equation (109), we introduced the exponential
form Z 0 e iφ for the complex variable Z.
R VR = (i 0 e i ωt )R = (i)Z R
Z imag = Lω – 1
ωC

V Z0
L VL = (i 0 e i ωt )Liω = (i)Z L φ
Z real = R

C VC = (i 0 e i ωt ) – i = (i)Z C Figure 12
ωC The complex impedance can be pictured as an arrow of
imag oriented at an angle φ .
2 + Z2
length Z0 = Zreal
Figure 11
The voltages VR , VL ,and VC expressed
in terms of impedances Z.
 Calculus 2000 - Chapter 5 Complex Variables Cal 5-19

The Driven RLC Circuit Setting the sum of the voltage rises to zero around
Our first demonstration in the physics course was the the circuit in Figure (13) gives, (walking counter
driven RLC circuit, which could be used to measure the clockwise),
speed of light without looking at light. (This was a crucial
point in our discussion of special relativity.) In Chapter 31 VC + VL + VR – Vind = 0 (112a)
we calculated the resonant frequency of an LC circuit and
i Z R + i Z L + i Z C = V0e i ωt (112b)
wrote down some formulas for the driven RLC circuit.
But we did not derive the formulas because the work is Solving for the current i in the circuit gives
messy when we have to use real functions. However with
the complex analysis we have developed in this chapter, V0e i ωt
we get, almost by inspection, not only the formulas but i =
Z (113)
considerable insight into the behavior of the circuit.
where Z = Z R + Z L + Z C = Z 0 e iφ is the total im-
In the lecture demonstration, we drove the LRC circuit
pedance of the circuit.
by wrapping a couple of turns of wire around the outside
of the inductor and attaching the wire to an oscillator. Using the exponential form for Z in Equation (113)
The oscillating magnetic flux produced by these few for the current i gives
turns induces a voltage Vind in the coil and drives the
circuit to oscillate. V0e i ωt V0 i (ωt – φ)
i = ; i = e (114)
The important thing is that we did not put the oscillator Z 0 e iφ Z0
directly in the circuit, for the oscillator has its own
internal resistance, capacitance and inductance that could Equation (114) tells us that if we drive an RLC
completely alter the behavior of the circuit. The idea is circuit with an induced voltage Vind = V0 e i ωt the
to give the circuit a gentle voltage shove of the form circuit will respond with a current i that has an
Vind = V0e i ωt (111) amplitude (V0/Z 0 ) and a phase (– φ) relative to the
driving voltage. We get this result almost without
as indicated in Figure (13), and see how the circuit doing any calculation. To get the same result using
responds. real functions sin ωt and cos ωt would have taken
i
several pages of algebra and trigonometric identities.
imaginary

R VR = iZ R Lω

Vind = V0e i ωt
Lω – 1
ωC
L VL = iZ L Z0
φ
R real
Z = Z 0 e iφ
C VC = iZC

– 1
ωC
Figure 13
The driven RLC Figure 14
circuit. Photo is Complex impedance for an RLC circuit.
Figure (1-10) from
the Physics text.
Cal 5-20 Calculus 2000 - Chapter 5 Complex Variables

Let us look at the physics contained in Equation (114). From Equations (117), we see that at low frequen-
cies, the phase of the current is π 2 ahead of the
V0 i (ωt – φ)
i = e (114) repeated induced voltage, and the amplitude goes to zero as ω
Z0 goes to zero.
For very low frequencies, for sufficiently small ω ,
the quantity 1/ωC is much larger than either Lω or The other extreme, at high frequencies where ωL is
R, the impedance is essentially all capacitive as much bigger than R or 1/ωC , we have
indicated in Figure (15). For this case, Z 0 ≈ Lω (118)
Z 0 ≈ 1 ; φ ≈ – 90° = – π (115) φ ≈ +90° (π /2) (119)
ωC 2
and the formula for the current in the circuit caused And we get
by the induced voltage Vind is V0 i (ωt – π/2) current
i = e at high (120)
current Lω frequencies
i = V0ωCe i (ωt + π/2) at low (116a)
frequencies Taking the real part gives us the real current
complex V0
Vind = V0e i ωt induced (116b) i = cos (ωt – π /2)
voltage Lω
large ω
Taking the real part of Equations (116) gives us the Vind = V0 cos ωt (121)
real current for a real induced voltage

i = V0ωC cos (ωt + π /2) We see that at high frequencies the phase of the current
small ω is π 2 behind of the induced voltage, and the amplitude
Vind = V0 cos ωt (117) goes to zero as ω goes to infinity.

imaginary imaginary

Lω
Lω R Lω – 1
ωC
φ real

Z0
Figure 15 Figure 16
Z for small ω . Z for large ω .

Z0

φ
Lω – 1
ωC
– 1 – 1 R real
ωC ωC
 Calculus 2000 - Chapter 5 Complex Variables Cal 5-21

There is a special frequency, call it ω0 , where the Taking the real part of Equation 24 gives
capacitive impedance Z C = – 1/ω0C just cancels the
inductive impedance Z L = Lω0 , leaving us with a V0
i = cos ω0t
pure resistive impedance Z R = R , as shown in Fig- R
at
ure (17). resonance (125)
Vind = V0 cos ω0t
This happens when
We see that, at resonance, the current and the in-
Z L = –Z C duced voltage are in phase with each other, and the
only thing that limits the current is the actual resis-
ω 0L = + 1 (122)
ω0C tance R in the circuit.

Comparing Equations (117, 121, and 125), we see

ω0 2 = 1 that the phase of the current shifts by 180 degrees ( π )
LC (123)
as we go from well below to well above the reso-
This special frequency is the resonant frequency nance. The smaller the value of R, the sharper the
ω0 = 1/ LC of the RLC circuit. We now see that resonance, and the faster this phase shift occurs. The
the resonance occurs when the capacitive and induc- shape of the resonance curves, for three different
tive impedances cancel, leaving only the resistance values of R were shown in the Physics text, Figure
to dampen the current in the circuit. Also note that (14-31) repeated here.
V0
at this frequency the phase angle φ is zero, and the V =
Z0
current i is given by

V0 i (ω t) current
i = e 0 at (124)
R resonance

imaginary

Lω

Z0
R real

ω/ω 0 0.6 0.8 1.0 1.2 1.4

frequency
– 1 Figure 14-31
ωC Amplitude of the oscillation for various values of the
resistance R. The peak occurs at ω = ω0 because the
Figure 17 inductive and capacitive impedances cancel at the
At resonance, the capacitive and inductive resonant frequency ω0 .
impedances cancel, and we are left with only the
resistive impedance.
Cal 5-22 Calculus 2000 - Chapter 5 Complex Variables

TRANSIENTS Equation (127) is an example of a non-homoge-

While the above discussion of the driven RLC neous differential equation. It is non-homoge-
circuit describes what you most likely will see when neous because of the driving term
you study the circuit in the lab, it is not the whole – V 0ωd/L sin ωdt which does not have a factor of
story. There are other solutions for the circuit, the variable (i) or a derivative of (i). This is called
solutions which die out as time goes on, and thus are the inhomogeneous term.
called transient solutions. To see where the tran- In the previous section, we found that Equation
sients come from, we need to go back to the differ- (127) has the solution
ential equation for the driven circuit. We get the
equation from Figure (18) which is simply Figure V0 i (ωt – φ) particular
(13) with some labels changed. To make the circuit ip = e solution (114) repeat
Z0
more nearly what we deal with in the lab, we are
writing the induced voltage as a real function where
V0 cos ωdt , where we are now calling the driving 2
frequency ωd . Z 0 2 = R 2 + ( Lω – 1 ) (99) repeated
ωC

Particular Solution tan φ = Lω – 1/ωC (100) repeated

R
Setting the sum of the voltages around the circuit
equal to zero gives The value of i p from Equation (113) is called the
particular solution of the differential equation (127).
VR + VL + VC = Vind (110) repeated
Transient Solutions
Q To see what the other solutions are, let us look at the
iR + L di + = V0 cos ωdt (126)
dt C homogeneous differential equation
This time, let us express everything in terms of the d 2i + R di + i = 0
current i rather than the charge Q, by differentiating (128)
dt 2 L dt LC
Equation (126) with respect to time and using
i = dQ/dt. We get, after dividing through by L which represents an RLC circuit with no driving
term. I.e., it is Equation (127) without the inhomoge-
d 2i + R di + i = –V 0ωd sin ω t neous term.
dt 2 L dt LC L d (127)
As a review, let us see how quickly we can solve
Equation (128). Using the trial solution
where we used d(cos ωdt)/dt = – ω d sin ωdt .
i = i 0 e – at ; di = – ae – at ; d 2i = a 2 e – at
i dQ
i=
dt
dt dt 2
gives
R VR = i R
a2 – R a + 1 = 0
L LC
Vind = V0cos ωdt This is a quadratic equation in a, of the form
L VL = L di a 2 +ba +c = 0 which has the solution
dt

a = –b ± b 2 – 4c = –b ± b2 – c
C VC = Q
2 2 4
C

Figure 18
The driven RLC circuit again.
 Calculus 2000 - Chapter 5 Complex Variables Cal 5-23

With b = R/L and c = 1/LC we get Adding Equations (131) and (132) together gives

a = R ± R2 – 1 d 2(i p+ai T) d(i p+ai T) (i p+ai T)

2L 4L 2 LC +R +
dt 2 L dt LC
= R ±i 1 – R2 = i d sin ωdt (133)
2L LC 4L 2
and we see that i new = (i p+ai T) obeys the same
Thus the solution to Equation (128) is
equation as i p alone. Thus i new is a solution of the
so called equation of the driven RLC circuit, for any value of
i T = i 0 e – αt e ± i ωt transient (129)
solution the constant (a).
Where This result tells us that to the driven or particular
2 solution i p , we can add any amount of the homoge-
α = R ; ω = 1 – R
2L LC 4L 2 neous solution i T, and we still have a solution for the
We can write ω in the form driven RLC circuit.

ω 2 = ω0 2 – α 2 The solutions i T for the homogeneous equation are

where fundamentally different from the particular solution
i p . The driven solution
ω0 = 1
LC V0 i (ωt – φ)
ip = e (113) repeated
is the resonant frequency. Equation (129) is just Z0
Equation (71) expressed in terms of the current i goes on at a constant amplitude V0/Z 0 for as long as
rather than the charge Q. We are calling this a the driving voltage is attached. The transient solution
transient solution i T. The reason for the name will
become apparent shortly. i T = i 0 e – αt e i ωt (129) repeated

Combined Solutions dies out exponentially with a time constant T = 1/α .

Let us now go back to Equation (127) for the driven Because such solutions do not last, they are called
circuit, and write i d for the constant ( – V 0ωd/L ) in transient solutions.
order to simplify the equation's appearence What you will observe in the lab is the following.
2 When you first turn on or suddenly change the
d i + R di + i = i sin ω t (127a)
dt 2 L dt LC d d driving voltage V0cos ωdt , you will see not only the
Now try the solution particular solution i p , but also some transients mixed
in. If you wait for several time constants T = 1/α ,
i new = i p + ai T (130) and keep the driving voltage amplitude V0 con-
where i p is the particular solution (113), i T is the stant, the transients will die out and the pure driven
transient solution of Equation (129), and (a) is an solution will appear on your oscilloscope. If you
arbitrary constant. We know that want to see the transient solutions, you have to look
within a time constant 1/α of the time you changed
d 2(i p ) d(i p ) (i p ) the driving voltage.
+R + = i d sin ωdt (131)
dt 2 L dt LC
non-homogeneous equation This finishes our discussion of the application of
complex variables to the analysis of circuits. We
d 2(ai T) R d(ai T) (ai T) now move on to the use of complex variables to
+ + = 0 (132)
dt 2 L dt LC describe wave motion.
homogeneous equation
Cal 5-24 Calculus 2000 - Chapter 5 Complex Variables

SOLUTIONS OF THE ONE As we saw in Chapter 15 of the Physics text and

DIMENSIONAL WAVE EQUATION Chapter 2 of the Calculus text,
In Chapter 2 of the Calculus text we discussed the sine wave moving
one dimensional wave equation applied to both y 1 = A sin(kx – ωt) to the right at a (139)
waves on a rope and sound waves. Applied to waves speed vwave = ω /k
on a rope, the equation was
sine wave moving
∂ 2 y(x,t) ∂2 y y 2 = A sin(kx + ωt) to the left at a (140)
= v (134)
∂t 2
wave
∂x 2 speed vwave = ω /k
(Calculus 2-73)
If we add y 1 and y 2 we get the standing wave
where y(x,t) represented the height of the rope
above its equilibrium position at some point x along standing
the rope at some time t. (For a sound wave, replace y 1 + y 2 = 2A sinkx cos ωt wave (141)
y(x,t) by p(x,t) where p(x,t) is the change in pressure
due to the sound wave at some point x and time t.) You can use the trigonometric identity
sin(a + b) = sina cosb + cosa sinb , noting that
(Recall that when we are working with more than sin(– b) = – sinb , and cos(– b) = cosb to check
one variable, like x and t, we use the notation Equation (141).
∂f(x,t)/∂t to mean the derivative of f(x,t) with re-
spect to t, holding x constant. This is called a partial Rather than use the real function sin(kx – ωt) , we
derivative with respect to time). can, as a trial solution to the wave equation, use the
complex function
We solved Equation (134) with a trial function of the
y = Ae i (kx – ωt) (142)
form
y(x,t) = A sin(kx – ωt) (135) ∂y ∂y
= ikAe i (kx – ωt) ; = (– iω)Ae i (kx – ωt)
∂x ∂t
∂2 y ∂2y
= – k 2y ; = – ω 2y (136) ∂2 y
∂x 2 ∂t 2 = (ik) 2Ae i (kx – ωt) = – k 2 y
to get ∂x 2

– ω 2y = – vwave
2 k 2y ∂2 y
= (– iω) 2Ae i (kx – ωt) = – ω 2y (143)
∂t 2
= ω2
2
2
vwave where (– i) 2 = –1 .
k
We are now right back to Equation (136) and get the
v wave = ω (137) same solution vwave2 = ω 2/k 2 . In this case it is
k
actually easier to work with the real function
In the solution sin(kx – ωt) , ω is, as we have noted sin(kx – ωt) rather than the complex function
many times, the angular frequency, of the number e i (kx – ωt) because you do not have to take the real
of radians per second. The quantity k, which is part of the complex function at the end. Working
called by the rather bland name wave number is with the real variables was not difficult in this case
actually the spacial frequency or the number of because the wave equation did not mix up sine and
radians per centimeter. When we take the ratio ω/k cosine functions as the RLC equation did.
we get
ω radians/second = ω centimeters (138)
k radians/centimeter k second
which is clearly a velocity.
 Calculus 2000 - Chapter 5 Complex Variables Cal 5-25

For completeness we have

complex sine wave
y 1 = A e i kx – ωt = moving to the right (142)
at a speed ω/k

complex sine wave

y 2 = A e i kx + ωt = moving to the left (143)
at a speed ω/k

The standing wave solution is

y standing = y 1 + y 2 = A e i (kx – ωt) + e i (kx + ωt)

= A e i kx e – i ωt + e i kx e i ωt

e –i ωt + e i ωt
= 2Ae i kx
2

= 2Ae i kx cos ωt

= 2A(cos kx + i sin kx)cos ωt

= 2Acos kx cos ωt + i 2A sin kx cos ωt

(144)
The imaginary part of y standing is

(y standing) imag = 2A sin kx cos ωt (145)

which is the standing wave solution we got using

real variables. Using complex variables to get the
standing wave solution was not easier than using
real variables.
 Calculus 2000 - Chapter 6 Introduction to Schrödinger's Equation Cal 6-1

Calculus 2000-Chapter 6
Introduction to the
Schrödinger Wave Equation
CHAPTER 6 INTRODUCTION TO

SCHRODINGER'S EQUATION

In the introduction to Chapter 37 of the Physics text, As we mentioned, Schrödinger took Debye’s advice,
we quoted the following story from an address by and in the following months devised a wave equation
Felix Block to the American Physical Society in for the electron wave, an equation from which one
1976. could calculate the electron energy levels. That
wave equation is now the foundation of chemistry.
“Once at the end of a colloquium I heard Debye
saying something like: ‘Schrödinger, you are not In this chapter we sketch the ideas that led
working right now on very important problems...why Schrödinger to formulate an equation involving
don’t you tell us some time about that thesis of de complex variables to describe the electron. We then
Broglie, which seems to have attracted some atten- go on to solve that equation for the lowest energy
tion?’ So in one of the next colloquia, Schrödinger spherically symmetric wave functions for the elec-
gave a beautifully clear account of how de Broglie tron in a hydrogen atom. This is enough to show that
associated a wave with a particle, and how he could the Schrödinger equation, without any extra as-
obtain the quantization rules ... by demanding that sumptions, is enough to explain the quantized en-
an integer number of waves should be fitted along a ergy levels of hydrogen.
stationary orbit. When he had finished, Debye casu-
ally remarked that he thought this way of talking
was rather childish ... To deal properly with waves,
one had to have a wave equation.”
Cal 6-2 Calculus 2000 - Chapter 6 Introduction to Schrödinger's Equation

SCHRÖDINGER'S WAVE EQUATION Using the standard notation

Schrödinger's approach to finding a wave equation h ≡ h h "bar" (4)
for the electron was roughly as follows. 2π
we get
De Broglie, suspecting that the electron, like the pho-
ton, had a wave nature as well as a particle nature, went E = hf = h ω = hω
2π
back to Einstein's formula for the energy of a photon
p = h = h k = hk
E = hf (1) λ 2π
Thus we get the very simple formulas
where h is Planck's constant, ( f = c/λ ) the frequency
of the photon and λ its wavelength. Setting E = mc 2
E = hω ; p = hk (5)
where m is the mass of the photon gives

mc 2 = hf = h c ; m = h as the relationship between a particle's energy E and

λ λc momentum p, and its wave's frequency ω and wave
Since photons travel at the speed c, the photon's number k.
momentum p should be its mass m times its speed c, or
Schrödinger's first attempt at finding a wave equa-
p = mc = h c tion was to start with the relativistic relationship
λc
between the energy and momentum of a particle.
That relationship, as we saw in the section on par-
p = h (2)
λ ticle accelerators, page 28-24 of the Physics text, is
relativistic
Equation (2) is the famous de Broglie formula for the E 2 = p 2c 2 + m 0 2c 4 relationship (6)
relationship between the wavelength and momen- between E and p
tum of any particle. De Broglie explained the
where m 0 is the rest mass of the particle.
quantization of angular momentum in the Bohr
theory by assuming that the allowed Bohr orbits To see how to construct a wave equation, let us start
were those in which exactly an integral number of with the simple case of a zero rest mass particle,
wavelengths fit around the orbit. namely the photon. For the photon, we have simply
Schrödinger's job was to find a wave equation based zero rest mass
E 2 = p 2c 2 particle (7)
on the two fundamental relationships E = hf for the
particle energy and p = h/λ for the particle wave- We will see that the one dimensional wave equation
length. Because we have been writing wave that leads to Equation (7) is
equations in terms of the angular frequency
ω radians/second rather than the regular frequency ∂ 2ψ 2∂ ψ
2
= c (8)
f cycles/second, and the wave number (spacial fre- ∂t 2 ∂x 2
quency) k radians/cm rather than the wavelength where ψ (psi) is a Greek letter to represent the wave
λ cm/cycle, let us first re-express E and p in terms amplitude. (For rubber rope waves ψ = y , the wave
of ω and k rather than f and λ . Using dimensions height. For sound waves ψ = p , the excess pres-
we have sure.) To check that Equation (8) is the correct
ω cycles equation, use the trial function
= ω radians/sec =
cycles
f
second 2π radians/cycle 2π second
ψ = ψ 0 e i (kx – ωt) (9)
1 cycles which, as we saw at the end of the last chapter (see
= 1 = k radians/cm
λ cm/cycle λ cm 2π radians/cycle Equation 5-142), represents a wave travelling to the
right at a speed ω /k.
cycles
= k cm (3)
2π
 Calculus 2000 - Chapter 6 Introduction to Schrödinger's Equation Cal 6-3

We have Now that we have some experience constructing

ψ = ψ 0 e i (kx – ωt) wave equations, let us go for the equation for a particle
with rest mass. This time let us first convert the
∂ψ ∂ 2ψ relationship between the particle energy E and momen-
= – iωψ ; = (– iω) 2ψ = – ω 2ψ
∂t ∂t 2 tum p into a relationship between ω and k. We have
∂ψ ∂ 2ψ
= – ikψ ; = (– ik) 2ψ = – k 2ψ E 2 = p 2c 2 + m 0 2c 4
∂x ∂x 2
Plugging these values into Equation (8) gives Setting E = hω and p = hk gives

∂ 2ψ 2∂ ψ
2
h 2ω 2 = h 2k 2 c 2 + m 0 2 c 4
= c (8) repeated
∂t 2 ∂x 2
Dividing through by h 2 gives
– ω 2ψ = c 2(– k 2 )ψ
m 0 2c 4
ω 2 = c2 k 2 + (13)
The factor – ψ cancels and we get h2
ω 2 = c 2 k2 (10) Using a ∂/∂t for each ω and a ∂/∂x for each k
Multiply through by h 2 and noting that E = hω suggests the wave equation
and p = hk we get
∂ 2ψ 2 ∂ ψ – m0 c ψ
2 2 4
h 2ω 2 = c 2 (h 2k 2 ) = c
∂t 2 ∂x 2 h2
(14)
E2 = c 2 p2 (11)
Plugging in the trial solution
which is the result we wanted.
ψ = ψ 0 e i (kx – ωt)
Exercise 1
For a traveling wave, use the trial function ∂ 2ψ ∂ 2ψ
= – ω 2ψ ; = – k 2ψ
ψ = ψ0 sin (kx – ωt) ∂t 2 ∂x 2

and show that you get the same result. gives

m 0 2c 4
You can see that the process is quite straightforward. – ω 2ψ = – c 2k 2ψ – ψ (15)
For each factor of ω you want from your differential h2
equation, you put a ∂/∂t into the equation. For each cancelling the factor of – ψ gives
factor of (k), you include a ∂/∂x .
m 0 2c 4
If we set ψ = E or B in Equation (8) we get the wave ω 2 = c 2k 2 + (16)
equations h2
∂2 E = c2 ∂2 E (12a)
which is the result we wanted.
∂t 2 ∂x 2
Equation (14) is the one dimensional form of
∂ B = c2 ∂ B
2 2
(12b)
Schrödinger's relativistic wave equation. This is
∂t 2 ∂x 2 the first wave equation Schrödinger found, but he
ran into trouble with it.
These turn out to be the differential form (in one
dimension) of the electromagnetic wave we discussed
in Chapter 32 in the Physics text. (These are Equations
(24a) and (24b) of Chapter 9 of the Calculus text, if we
set c 2 = 1/µ 0ε 0 .) This should not be surprising, be-
cause an electromagnetic wave just represents the
wave nature for the zero rest mass photon.
Cal 6-4 Calculus 2000 - Chapter 6 Introduction to Schrödinger's Equation

Consider the case of a particle at rest, or nearly at That relationship is

rest, so that we can neglect p 2c 2 compared to
m 0 2c 4 . Then the square of the energy E is approxi- kinetic E = 1 mv 2 = 1 (m 2 v 2) (24)
energy
2 2m
mately equal to the square of the rest energy m 0c 2
where v is the speed of the particle, m the rest mass,
for
E 2 ≈ m 0 2c 4 small p
(17) and mv = p is the momentum. Thus E and p are
related nonrelativistically by
This equation has two solutions
(mv) 2 p2
E = = (25)
E 1 = m 0c 2 2m 2m
(18)
E 2 = – m 0c 2 Writing E = hω , p = hk , the nonrelativistic rela-
tionship between ω and k is
Solution (2) appears to represent a particle with a
negative rest energy, a very un-physical thing. The nonrelativistic
2 2
corresponding wave solutions are hω = h k relationship
(26)
2m between ω and k
ψ 1 = ψ 0 e i (kx – ω1 t) ; hω1 = E 1 (19)
Schrödinger went to the nonrelativistic form be-
cause the relationship E = p 2/2m does not involve
ψ 2 = ψ 0 e i (kx – ω2 t) ; hω2 = E 2 (20)
negative rest masses.
When you encounter two solutions to a physical
To construct a wave equation that gives this nonrela-
problem, and one is nonsense, you usually throw the
tivistic relationship between ω and k, we need one
bad solution out. For example, the hypotenuse of a
time derivative to give the one factor of ω , and two
right triangle is given by the equation
x derivatives to give the factor of k 2 . What works,
c2 = a2 + b2 (21) as we will check, is
which has two solutions one dimensional
∂ψ 2 ∂ 2ψ
= – h
Schrödinger's
ih
c1 = + a 2 + b2 (22) ∂t 2m ∂x 2 equation for (27)
a free electron
c2 = – a 2 + b2 (23) With the trial solution
Since you know that you cannot have a negative ψ = ψ 0 e i (kx – ωt)
hypotenuse, you just throw out the un-physical solu-
tion c 2 . ∂ψ ∂ 2ψ
= – iωψ ; = – k 2ψ (28)
∂t ∂x 2
Schrödinger tried to throw out the un-physical solu- we get
tion ψ 2 of his relativistic wave equation, but ran into
2
the following problem. If he started with pure ψ 1 ih(–iωψ) = h k 2 ψ
waves for the electrons, and let the electrons inter- 2m
act, ψ 2 waves were generated. In other words, if he 2
threw out the un-physical ψ 2 waves, the equations – i 2ω hψ = h k 2 ψ (29)
2m
put them back in. We did not have this problem with The ψ′s cancel, and with – i 2 = 1 , we are left with
the Pythagorean theorem. the desired result
Schrödinger gave up on the relativistic wave equation 2 2
hω = h k (26) repeated
and decided to use the nonrelativistic relationship 2m
between the kinetic energy E and momentum p of a Equation (27) is the one dimensional form of
slowly moving particle. Schrödinger's equation for a free particle.
 Calculus 2000 - Chapter 6 Introduction to Schrödinger's Equation Cal 6-5

In Chapter 2 of the Calculus text, we saw that the Exercise 2

equations for rope waves, sound waves, and electro- In three dimensions, the momentum vector
magnetic waves all had second derivatives of both p = (px, py, pz ) has a magnitude p given by the
space and time. That is how we got the oscillating Pythagorean theorem as
solutions. In our study of the RLC circuit, we saw
that the presence of a first derivative, the R term in p2 = (px2+ py2+ pz2 ) (30)

d 2Q R dQ Q With p = hk , we have
2
+ + = 0 (5-59) repeated
dt L dt LC
p2 = h2(kx2+ ky2+ kz2 ) (31)
led to an exponential decay.
We got the one dimensional wave equation by replac-
One might wonder, since there is only a first deriva- ing k2x by ∂2/∂x2 . This suggests that the extension of
tive with respect to time in Schrödinger's equation, Equation (27) to describe three dimensional plane
shouldn't that lead to an exponential decay with waves should be
time, of the wave amplitude ψ ? It did not do so
∂ψ 2 ∂2ψ ∂2ψ ∂2ψ
because of the explicit factor of (i) in Schrödinger's ih =– h + + (32)
∂t 2m ∂ x2 ∂ y2 ∂ z2
equation. With the trial solution ψ = ψ 0 e i (kx – ωt)
the (–i) from the first derivative with respect to time As a trial solution, try the guess
was turned into a 1 by the i in the ∂/∂t term. Thus
i (k x + kyy + kzz – ωt)
by having an (i) in Schrödinger's equation itself, we ψ = ei (k⋅x – ωt) = e x (33)
can get an oscillating solution with a first time
derivative. and show that the guess implies
h2 2
The reason we have introduced Schrödinger's equa- hω = (k + k2y +k2z )
2m x
(34)
tion after a chapter on complex variables is that
and
factor of (i) in the equation itself. With the other
p2
differential equations we have discussed so far, we E =
2m
had the choice of using real or complex variables.
But we cannot write, let alone solve, Schrödinger's
equation without the use of complex variables.
Cal 6-6 Calculus 2000 - Chapter 6 Introduction to Schrödinger's Equation

POTENTIAL ENERGY & THE HYDROGEN ATOM

SCHRÖDINGER'S EQUATION The reason Schrödinger developed his wave equation
The relationship E = p 2/2m = mv 2/2 is for a free was to handle the electron waves in hydrogen in a
particle traveling at a constant speed v. If the particle mathematically rigorous way. To apply Schrödinger's
has a potential energy V(x) , like spring potential equation of the hydrogen atom, you use the fact that the
energy electron is bound to the proton nucleus by a Coulomb
spring force of magnitude e 2/r 2 whose potential V(r) is
V(x) = – 1 Kx 2 potential (35) Coulomb
2 energy 2
V(r) = – re potential (42)
where K is the spring constant, then the formula for energy
the total nonrelativistic energy E is
With this potential energy, Schrödinger's equation (41)
p2 for the hydrogen atom becomes
E = 1 mv 2 + V(x) = + V(x) (36)
2 2m
In terms of ω and k we have Schrödinger's
h ∂ψ = h 2 ∇ 2ψ – e 2 ψ equation for
2 2 i ∂t r
hω = h k + V(x) (37) 2m hydrogen atom
2m (43)
and the corresponding one dimensional wave equa- Solving Equation (43) is not easy. The first problem
tion should be we encounter is the fact that we have been writing
∇ 2 = ∂ 2/∂x 2 + ∂ 2/∂y 2 +∂ 2/∂z 2 using Cartesian co-
one
∂ψ 2 ∂ 2ψ dimensional ordinates x, y, z, while the Coulomb potential – e 2/r
ih = – h + V(x)ψ Schrödinger has spherical symmetry. The best way to handle the
∂t 2m ∂x 2 equation situation is to use a coordinate system that has the
(38) same symmetry as the potential energy.
If you did Exercise (2), it is clear that the three dimen-
sional form of Schrödinger's equation is expected to be The coordinate system of choice is the
spherical polar coordinate system that has an inherent
∂ψ 2 ∂ 2ψ ∂ 2ψ ∂ 2ψ spherical symmetry. This coordinate system is de-
ih =– h + + + V(x,y,z)ψ
∂t 2m ∂x 2 ∂y 2 ∂z 2 scribed in Chapter 4 of the Calculus text and indicated
in Figure (1). Instead of locating a point by giving its
(39)
x, y, and z coordinates, we locate it by the r, θ and φ
In Chapter 4 of the Calculus text, we discussed the
coordinates. The quantity r is the distance from the
combination of derivatives ∂ 2/∂x 2 + ∂ 2/∂y 2 + ∂ 2/∂z 2
origin, θ the angle down from the z axis, and φ the
and gave them the special name
angle over from the x axis, as shown.
∇2 ≡ ∂ + ∂ + ∂
2 2 2 definition z
(40)
∂x 2 ∂y 2 ∂z 2 of ∇2

With this notation, the three dimensional form of

Schrödinger's equation can be written in the more θ r
compact and familiar form
y
∂ψ 2 full φ
ih = – h ∇ 2ψ + V(x,y,z)ψ Schrödinger
∂t 2m equation
x
(41)
We can immediately get back to the one dimensional Figure 1
Spherical polar coordinates.
Schrödinger's equation by replacing ∇ 2 by ∂ 2/∂x 2 .
 Calculus 2000 - Chapter 6 Introduction to Schrödinger's Equation Cal 6-7

In the appendix to Chapter 4 of the Calculus text, we A special feature we discover when we solve
calculated ∇ 2 in spherical polar coordinates. The Schrödinger's equation in Appendix II, is that in
result was order for ψ 1 and ψ 2 to be solutions of Schrödinger's
∂ψ
equation (43), the frequencies ω1 and ω2 have to
∇ 2ψ = 1r ∂ 2 (rψ) + 2 1 ∂ sinθ
2
have the following values
∂r r sinθ ∂θ ∂θ
4
hω1 = – e 2m = – 13.6 eV (48)
∂ ψ 2
+ 21 (44) 2h
r sinθ ∂φ 2 4
hω 2 = – e 2m = – 3.60 eV (49)
This surely does not look simpler than 8h
∇ 2 = ∂ 2/∂x 2 + ∂ 2/∂y 2 +∂ 2/∂z 2 , but it does allow You can immediately see that hω1 is the energy of the
you to find solutions to Schrödinger's equation for electron in the lowest hydrogen energy level, and hω2
the hydrogen atom. is the electron energy in the second energy level. Just
looking at the spherically symmetric solutions begins
In the appendix to this chapter, we calculate some to tell us that Schrödinger's equation is going to explain,
spherically symmetric solutions to Schrödinger's in a natural way, the hydrogen energy levels.
equations. These are solutions that depend only on
r, namely ψ = ψ(r) , so that ∂ψ/∂θ = 0 and As we mentioned in our discussion of the hydrogen
∂ψ/∂φ = 0, which eliminates the second and third atom in Chapter 38 of the Physics text, there are many
terms in Equation (44). The solutions we get, (we allowed standing wave patterns for the electron in
solve one and leave the second as a homework hydrogen. In Figure (38-1), reproduced on the next
exercise) are page, we show sketches of the six lowest energy
patterns ψ n, ,m labeled by their energy quantum num-
ψ 1 = e – r /a 0 e – i ω 1t (45) ber (n), angular momentum quantum number ( ) and
z projection of angular momentum quantum number
(m). We noted that all the zero angular momentum
ψ 2 = 1 – r e – r /2a0 e – i ω 2t patterns ( = 0) are spherically symmetric. By solving
2a 0 (46) Schrödinger's equation for spherically symmetric stand-
ing waves, we began to generate the = 0 patterns.
where a 0 has the value
Explicitly, the waves we got are
2
a0 = h 2 Bohr radius (47) ψ 1,0,0 = ψ 1 (of Equation 45)
me
This quantity a 0 is the Bohr radius, the radius of the ψ 2,0,0 = ψ 2 (of Equation 46)
smallest orbit in the Bohr theory of hydrogen. (See
Exercise 7 in Chapter 35 of the Physics text.) To solve for the non symmetric patterns like ψ 2,1,1 that
have angular momentum, you have to be able to handle
Exercise 3 angular terms involving θ and φ in the formula (44)
Go to Appendix II of this chapter (page 6-14) and study for ∇ 2 . Differential equations involving ∇ 2 have
the steps that led to the solution ψ1 . Then work Exercise been studied for well over a century, and the angular
5 to find the solution ψ2 . After that return here and terms, which are common to many of these equations,
continue reading. have been carefully worked out with standardized
notation. The angular dependence of the non spherical
standing waves involve what are called spherical
harmonics which are briefly discussed in Appendix II
of this chapter.
 Cal 6-8 Calculus 2000 - Chapter 6 Introduction to Schrödinger's Equation

There are 8 more n = 3 patterns

E = –1.51eV

in addition to the one shown.

The and m quantum numbers
are
= 1; m = 1, 0, –1
= 2; m = 2, 1, 0, –1, –2.

(i)
n = 3, = 0, m = 0
E = –3.40eV

top view (c) top view (e) top view (g)

(b)
n = 2, = 0, m = 0
r
ψ 2(r) = 1 – e – r /2a0
2a 0
side view (d) side view (f ) side view (h)
n = 2, = 1, m = 1 n = 2, = 1, m = 0 n = 2, = 1, m = –1
E = –13.6eV

ψ 1(r) = e – r /a0

(a)
n = 1, = 0, m = 0

Figure 38-1 (page 38-3 of the Physics text)

The lowest energy standing wave patterns in hydrogen. The intensity is what you
would see looking through the wave. We have labeled ψ 1 and ψ 2 on the diagram.
 Calculus 2000 - Chapter 6 Introduction to Schrödinger's Equation Cal 6-9

INTERPRETATION OF SOLUTIONS If we can get a positive number for a vector field by

TO SCHRÖDINGER'S EQUATION taking the dot product of the vector with itself, what
do we do to get a positive number from a complex
Bohr's theory of the hydrogen atom, although quite
ψ ? The answer, as we mentioned at the beginning
successful, was based on Newtonian mechanics with
of Chapter 5 (see Equation 5-26), is that we get a real
the ad hoc assumption that angular momentum was
positive number from a complex number by multi-
quantized in units of h . De Broglie's theory suggested
plying by the complex conjugate. To remind you
that the reason for the quantization of angular momen-
how this works, suppose that we have separated ψ
tum was due to the wave nature of the electron, but he
into its real and imaginary parts
also treated the electron wave in a rather ad hoc manner.
If one assumes that Schrödinger's equation rather than ψ = ψ real + iψ imag (50)
Newtonian mechanics provides the basic theory for the where both ψ real and ψ imag are real numbers. Then
electron in hydrogen, then all the quantized energy the complex conjugate, which we designate by ψ* ,
levels follow a direct consequence of the theory. No is defined by changing (i) to (–i)
extra assumptions have to be fed in. Schrödinger had
found the theory to replace Newtonian mechanics in ψ* = ψ real – iψ imag (51)
describing atoms. To calculate the complex conjugate ψ* you do not
But questions remained. The electron's wave nature have to separate the function into real and imaginary
was well established, but what was the meaning of the parts ahead of time. You get the same result by
electron wave? The answer to that was provided a replacing all (i) by (–i) in the complex formula.
couple of years later by Max Born, who was calculating When you multiply a complex number ψ by its
how electron waves would be scattered by atoms. The complex conjugate ψ* , the result is a real positive
calculations suggested to him that the electron wave number, as you can see below
should be interpreted as a probability wave, as we
discussed in Chapter 40 of the Physics text. ψ* ψ = (ψ real – iψ imag) (ψ real + iψ imag)
One of the main features of a probability wave is that = ψ realψ real + iψ realψ imag
it has to be represented by a real, positive number. You
cannot have negative probabilities or imaginary prob- – iψ imagψ real – i 2ψ imagψ imag
abilities. But so far, our electron waves are described
by a complex variable ψ , obtained from an equation The iψ realψ imag terms cancel, and with – i 2 = 1 we get
that was itself complex. How do we get real positive
numbers from the complex ψ ? ψ* ψ = ψ 2real + ψ 2imag
(52)
We ran into a somewhat similar problem in our discus- and thus ψ* ψ is a real, positive number.
sion of electromagnetic radiation. Maxwell's equa-
tions predict that light waves consist of electric and For electron waves, the positive number ψ* ψ repre-
magnetic fields E and B . Yet most of the time we are sents the intensity of the wave in much the same way
concerned with the intensity or energy density of a light that (E ⋅ E + B ⋅ B) represented the intensity of the
wave. To predict the intensity from Maxwell's theory, electromagnetic wave.
we have to know how to calculate the intensity from the
vectors E and B . The answer is that the intensity is
proportional to the square of E and B . If we use the
correct units, the intensity is proportional to
(E ⋅ E + B ⋅ B) . These dot products E ⋅ E and B ⋅ B
are always positive numbers and therefore can repre-
sent an energy density or intensity.
Cal 6-10 Calculus 2000 - Chapter 6 Introduction to Schrödinger's Equation

Normalization Let us see what the integral of ψ *1ψ 1 over all space
In describing probabilities, one usually represents a is. We have
probability of 1 as being certainty, and that the
ψ 1 = e – r /a 0 e – i ω 1t (56a)
probability of an event as being allowed to range
from zero to one. If the wave function ψ is to ψ*1 = e – r /a 0 e + i ω 1t (change – i to i) (56b)
represent a probability wave for an electron, we have
so that
to include the idea that the probability of something
ranges from zero to one. ψ*1 ψ 1 = e – r /a 0 e + i ω 1t e – r /a 0 e – i ω 1t

The intensity ψ* ψ is a density that varies over space. ψ*1 ψ 1 = e – 2r/a0

If you have an energy density, call it E , then the total (57)
energy E is the integral over all of space of the The e i ω 1t ′s cancelled and we end up with a real
energy density E . We can write this symbolically as positive density.
E = E (x,y,z)d 3V (53) To integrate ψ* ψ over all space, we notice that since
all space ψ* ψ is spherically symmetric, we can take d 3V as the
where, if we are using Cartesian coordinates, the volume of the spherical shell shown in Figure (2), a
volume element d 3V would be (dx×dy×dz) . shell of radius r and thickness dr. That volume is
d 3V = (4πr 2)dr (58)
If we are to interpret ψ* ψ as a probability density, then
the total probability should be the integral of the because 4πr 2is the area of a sphere of radius r.
probability density over all space. We can write this as Throughout the shell, ψ* ψ has the same value
e – 2r/a0 , thus our volume integral is simply
total = ψ* ψ d 3V (54) ∞
probability
all space ψ* ψ 3
dV = e – 2r /a0 (4πr 2)dr (59)
all space r =0
The question is, this is the total probability of what? If
we are talking about the electron wave in hydrogen, and Being somewhat lazy, we look up in our short table
we think of ψ *ψ d 3V as the probability of finding the of integrals, the integral of r 2 e – αr . After some
electron in some small volume element d 3V , then if manipulation shown in Appendix 1, we get
we sum these probabilities over all space, we should ∞
end up with the total probability of finding the electron 4π r 2 e – 2r /a0dr = π (a 0) 3 (60)
somewhere in space. If the hydrogen atom has one 0
electron, and you look everywhere, you should eventu- The result is that the integral of ψ* ψ over all space
ally find the electron with a probability (1). Thus the is π (a 0) 3 instead of the desired value of 1.
total probability should be given by the formula z
spherical shell
of thickness dr
1 = ψ* ψ d 3V (55)
all space

The wave functions ψ 1 and ψ 2 that we presented r

you in Equations (45) and (46) do not have this y
property.

x
Figure 2
We can use as the volume element d 3V the spherical
shell of radius r and thickness dr.
 Calculus 2000 - Chapter 6 Introduction to Schrödinger's Equation Cal 6-11

To fix this problem, we use a so-called normalized When you look at tables of wave functions, you will
wave function (ψ 1) normalized , which is simply ψ 1 see factors like 1/ π (a 0) 3 or 3/8π . They are
multiplied by an appropriate normalization constant merely the normalization constants. In one sense,
C. To find out what C should be, write the normalization constants just make the formulas
look complicated. Most of the physics in our equa-
(ψ 1) normalized = Cψ 1 (61a) tion for ψ 1 is contained in the factor e – r/a0 . It tells
us that the electron wave decays exponentially as we
(ψ *1) normalized = C *ψ *1 (61b) go out from the proton, decaying by a factor of 1/e
when we go out one Bohr radius a 0 . The intensity,
where, if we want, the normalization constant can be or probability ψ * ψ is proportional to e – 2r/a0 and
complex. Then we have thus drops off by a factor 1/e 2 when we are a Bohr
radius from the proton. We also calculated the
(ψ*1) normalized(ψ 1) normalized = (C * C) ψ*1 ψ 1 energy levels E 1 and E 2 without worrying about
the normalization constants. It is nice to have a table
1 = 3
(ψ*1) normalized (ψ 1) normalized d V that gives you the normalization constants, but you
all space
get a better insight into the shape of the standing
wave patterns if you have another table without
= C *C ψ* ψ d 3V (62) them.
all space
Exercise 4
= C *C π (a 0) 3 At what finite radius is there zero probability of finding
an electron when the electron is in the n = 2, = 0,
Thus m = 0 standing wave pattern? Explain why and sketch
C* C = 1 (63) the intensity ψ*2,0,0ψ2,0,0
π (a 0) 3
The simplest choice is to take C real, giving

C = 1 normalization
(64)
constant for ψ1
π (a 0) 3

As a result our normalized wave function becomes

(ψ 1) normalized = 1 e – 2r/a0 e – iωt

π (a 0) 3

(65)
Cal 6-12 Calculus 2000 - Chapter 6 Introduction to Schrödinger's Equation

THE DIRAC EQUATION Dirac found that by using a certain combination of

Our story is incomplete if we stop our discussion of spinors, he could write a relativistic wave equation
particle wave equations with Schrödinger's equa- for the electron that had only a first order time
tion. As successful as that equation is, it still does derivative ∂ψ/∂t . He hoped that this equation would
not handle relativistic effects. As we saw, avoid the un-physical negative mass solutions.
Schrödinger could avoid the negative rest mass Dirac's equation was successful in that it not only
solutions by starting with the nonrelativistic formula included all the results of Schrödinger's and Pauli's
E = p 2/2m rather than the relativistic one equations, but it also correctly predicted tiny relativ-
E 2 = p 2c 2 + m 0 2c 4 . istic effects that could be detected in the spectra of
It appeared to Dirac that the reason Schrödinger hydrogen. However, Dirac soon found that his
could avoid the nonphysical solutions is because the equation also led to the apparently negative mass
nonrelativistic equation involves only the first de- solutions.
rivative with respect to time ∂ψ/∂t , rather than the Dirac could not throw his equation away because it
second derivative ∂ 2ψ/∂t 2 that appeared in the successfully predicted relativistic effects that were
relativistic equation (see Equation (14). Dirac thought observed by experiment. Instead he found a new
that if he could develop a relativistic wave equation interpretation of the previously undesirable solu-
that avoided second time derivatives, then perhaps tions. He found that these solutions could be reinter-
he could avoid the un-physical negative mass solu- preted as the wave for a particle whose mass was
tions. positive but whose electric charge was of the oppo-
By 1929, when Dirac was working on the problem, site sign. The equation led to the prediction that
it was known that the electron had two spin states, there should exist a particle with the same rest mass
spin up and spin down. It was these two spin states, as the electron but with a positive electric charge.
along with the Pauli exclusion principle, that led to That particle was observed four years later in Carl
an understanding of the structure of the periodic Anderson's cloud chamber in the basement of the
table. These spin states are not included in or physics building at Caltech. It became known as the
explained by Schrödinger's equation. positron.

Slightly earlier, Wolfgang Pauli had introduced a We now know that any relativistic wave equation for
new mathematical quantity called a spinor to de- a particle has two kinds of waves for a solution. One
scribe the spin state of the electron. Spinors are represents matter particles, and the other, like the
quantities, involving complex numbers, that are in a wave for the positron represents antimatter. If you
sense half way between a scalar number and a have a relativistic wave equation, even if you start
vector. The existence of such a mathematical quan- only with matter particles, the equation contains the
tity was unknown until its invention was required to mechanism for particle-antiparticle pair creation.
explain the electron. Pauli was able to modify You let the matter particles interact, and antimatter
Schrödinger's equation with the use of spinors to has a finite probability of being created. That is why
include the effects of electron spin. Schrödinger and Dirac could not suppress the anti-
matter waves in the relativistic equations. However,
by going to a nonrelativistic equation, representing
situations where not enough energy is available to
create electron positron pairs, Schrödinger could
avoid the antimatter waves.
 Calculus 2000 - Chapter 6 Introduction to Schrödinger's Equation Cal 6-13

Appendix I – Evaluation of a Normalization Integral

Our normalization integral is The exponential decay is so powerful that in the limit
∞ of large R, a term of the form R n e – aR goes to zero
ψ* ψ d 3V = 4π r 2 e – 2r /a0 dr (59) repeat for any value of n for positive (a). Thus all terms with
all space r =0 a e – aR go to 0 as R goes to infinity. With e 0 = 1, we
are left with
Looking for the integral of r 2 e – αr in our short table ∞
of integrals in the formulary, we find instead r 2 e – ar dr = 23 (68)
a
0
x 2 e – ax dx = 13 (a 2x 2 + 2ax + 2)e – ax (66)
a Now set a = 1/ 2a 0 and we get
If we set x = r and integrate from 0 to infinity, we have ∞
4π r 2 e – 2r /a0 dr = 4π 2
∞ (2/a 0) 3
∞
r 2 e – ar dr = 13 (a 2r 2 + 2ar + 2)e – ar 0
(69)
a 0 = π (a 0) 3
0

= 13 (a 2R 2 + 2aR + 2)e – aR
a R=∞

– 13 (a 20 2 + 2a×0 + 2)e – a×0

a
(67)
Cal 6-14 Calculus 2000 - Chapter 6 Introduction to Schrödinger's Equation

An introduction to Schrödinger's
APPENDIX II -
Equation Applied to the Hydrogen Atom
The Hydrogen Atom The next step is to note that it is not convenient to
Schrödinger's first major success with his wave handle a spherically symmetric potential V(r) = – e 2/r
equation was to solve for the electron standing using Cartesian coordinates x, y, and z. In the Chapter
waves in hydrogen, and to determine the electron 4 of the Calculus text we derived the formula for ∇ 2 in
energies in each of the standing wave patterns. For spherical polar coordinates r, θ , φ which are shown
an electron in hydrogen, the potential energy is in Figure (1) reproduced here. In these spherical
given by Coulomb's law as coordinates we show, after considerable work, that
2 ∇ 2ψ is given by Equation (4-10) as
V(r) = – er (42) repeated
∇ 2ψ = 1r ∂ 2 (rψ)
2
where –e is the charge on the electron and r is the
separation of the electron and proton. Thus the ∂r
equation Schrödinger had to solve for hydrogen is
+ 1 ∂ sinθ ∂ψ
the three dimensional equation ∂θ
r 2sinθ ∂θ
∂ 2ψ (4-10)
∂ψ 2 2 Schrödinger′s 1
ih = – h ∇ 2ψ – re ψ equation for +
∂t 2m hydrogen atom r 2 sin 2 θ ∂φ 2

(43) repeated (Note: many texts write the first term as

Quite a few steps are required to obtain solutions to 1/r 2 ∂/∂r (r 2 ∂ψ/∂r) which is an equivalent but usu-
Equation (43). The first is to look for solutions of ally less convenient form.)
definite frequency ω or energy E = hω by using the z
trial function
ψ = ψ(x,y,z) e – iωt
(70) θ r
= ψ(x) e – iω t
where we will use the bold face x to stand for (x,y,z). y
Plugging this guess into Equation (43) gives φ

2
ih (– iω)ψ(x) e – iω t = – h ∇ 2ψ(x) e – iω t x
2m
Figure 1 (repeated)
2
– er ψ(x) e – iω t
Spherical polar coordinates.

The factor e – iωt cancels and we are left with

2 2
hω ψ(x) = – h ∇ 2ψ(x) – er ψ(x) (71)
2m
With hω = E , this becomes

2 2
E ψ(x) = – h ∇ 2ψ(x) – er ψ(x) (72)
2m
 Calculus 2000 - Chapter 6 Introduction to Schrödinger's Equation Cal 6-15

If we look at only the spherically symmetric solu- The common factor e – αr cancels and we are left with
tions where
– 2α + a + r α 2 + b = 0 (80)
spherically
ψ(x,y,z) = ψ(r) symmetric
The only way we can satisfy Equation (80) for
wave (73)
arbitrary values of r is to set both square brackets
then ∂ψ(r)/∂θ = 0 , ∂ψ/∂φ = 0 , and only the radial
separately equal to zero, giving
part of ∇ 2ψ(r) survives. Schrödinger's equation for
the spherically symmetric waves of energy E becomes 2α = a ; α = a/2 (81a)

E ψ = – h 1r ∂ 2 (rψ) – er ψ
2 2 2
(74) α2 = – b (81b)
2m ∂r
Squaring Equation (81a) gives
Multiplying through by 2mr/h 2 , Equation (74) can 2
be written in the form α2 = a (81c)
4
∂ 2 (rψ) + a + b rψ = 0 (75) For Equations (81b) and (81c) to be consistent, the
∂r 2 r
constants (a) and (b) must satisfy the relationship
where
2
a = 2me
2
; b = 2mE (76) –b = a (82)
h 2 h2 4
If we define the variable u(r) by To see what Equation (82) implies, let us put back in
the values of (a) and (b)
u = rψ ; ψ = ur (77)
2 a 2 = 1 × 4m 2e 4
our equation for u becomes a = 2me ; (83a)
h2 4 4 h4
∂ 2u + a + b u = 0 (78)
∂r 2 r
–b = – 2mE (83b)
h2
Exercise 5
Thus Equation (82) requires
Derive Equation (78) starting from Equation (74).
2 4
Equation (78) is a differential equation we have not – 2mE = m 4e
h2 h
encountered before. Neither of our familiar guesses or
for a solution, like u = e – αr or u = sinωr , will work,
as you can check for yourself. What does work is the 4
function we will call u 1 , which is E = – me2 = – 13.6 eV (84)
2h
u 1(r) = re – αr guess (79)
In our study of the Bohr theory, we found that the
Plugging our guess into Equation (79) gives lowest energy level of the hydrogen atom was
du 1 E 1 = – me 4/2h 2 which turns out to be –13.6 elec-
= e – αr – αre – αr tron volts. We now see that if the hydrogen wave
dr
amplitude is given by the solution u 1 , or ψ 1 = u 1r ,
d 2u 1 then the energy of the electron in this wave pattern
= – αe – αr – αe – αr + α 2 re – αr must be the same as the lowest energy level of the
dr 2 Bohr theory. This is a prediction of Schrödinger's
d 2u 1 wave equation without any arbitrary added assump-
Thus + ar + b u = 0 becomes
dr 2 tions like assuming angular momentum is quan-
– 2αe – αr + α 2re – αr + ar re – αr + bre – αr = 0 tized.
Cal 6-16 Calculus 2000 - Chapter 6 Introduction to Schrödinger's Equation

To see what the wave pattern is that corresponds to The Second Energy Level
the energy level E 1 , note that the Bohr radius a 0 , the In the following exercise you will find another
radius of the smallest Bohr orbit in the Bohr theory, spherically symmetric solution for the hydrogen
is given by atom.
2
a0 = h 2 Bohr radius (85)
me Exercise 6
Thus our constant (a) in Equation (77) can be written Try the guess
2 u2 (r) = (r + cr2 )e– αr , u2 = r ψ 2 (89)
a = 2me2
= a2 (86)
h 0 as a possible solution to Equation (78) where (c) is an
unknown constant. Show that for (89) to be a solution,
Thus Equation (81a) requires that you have to satisfy the conditions
– 2α + 2c + a = 0 (90a)
a = 2α = a2 ; α = a1 (87)
0 0 α 2 – 4cα + ac + b = 0 (90b)
and the wave function ψ 1(r) is given by α 2c + bc = 0 (90c)
Then show that this requires α = –b as before, and that
2

ψ 1(r) =
u 1(r) r e – αr = e – αr 2 4
r = r 2
–b = a ⇒
– 2mE2
= 1 × 4m e (91)
16 h2 16 h4

ψ 1(r) = e – r/a0 (88) or

me4 – 13.6 eV
E2 = – 1 = = – 3.60 eV (92)
The electron wave decays exponentially as we go 4 2h 2 4
out from the nucleus, decaying by a factor of 1/e
when we go out one Bohr radius. We have just used Then show that ψ 2 (r) is given by
Schrödinger's equation to solve for the ground state
wave function, the lowest energy level standing ψ 2 = (1 – 2ar )e– r/2a0
0
wave pattern in hydrogen.

E2 = – 3.6 eV (93)
 Calculus 2000 - Chapter 6 Introduction to Schrödinger's Equation Cal 6-17

In the Bohr theory, the energy levels E n are given by If you try a guess of the form

E1 u 3(r) = (1 + c 2r + c 3r 2)e – αr (96)

En = 2
= – 13.62eV (94)
n n you end up with a spherical wave pattern ψ 3(r) that
The second energy level E 2 is thus has two spherical nodes, and has an energy

E1 E1 E1
E2 = = = – 3.6 eV E3 = (97)
2 4 32
(2)
which is the third energy level.
Thus the wave pattern you solved for in Exercise (8)
is the spherically symmetric standing wave pattern You can now see the pattern. We can generate all the
in the second energy level. It is what we have called spherically symmetric = 0 wave patterns by add-
the n = 2, = 0 wave pattern. Note that in the ing terms like c 4r 3 , c 5r 4 , ⋅ ⋅ ⋅ c nr n–1 to our guess for
solution u n(r) . Solving for all the constants, we end up with
E1
ψ 2(r) = 1 – r e – r /2a0 (93) repeated En = (98)
2a 0 n2
when we are at a distance which is the energy level structure Bohr discovered.
r = 1 ; r = 2a (95)
0
2a 0
the wave pattern in Equation (93) goes to zero. This
means that the standing wave ψ 2(r) has a spherical
node out at a distance r = 2a 0 . This is the spherical
node we saw in the ψ(n = 2, = 0) pattern shown in
the Physics text, Figure (38-1) repeated here.

Figure 38-1a
Hydrogen atom
standing wave
pattern for
n = 2, = 0.

Figure 38-1i
Wave pattern for n = 3, = 0.
Figure 3
Tacoma Narrows
bridge in an n = 2
second harmonic
standing wave
pattern.

(Movie. Press esc to stop)

Cal 6-18 Calculus 2000 - Chapter 6 Introduction to Schrödinger's Equation

Non Spherically Symmetric Solutions The functions Y m(θ,φ) , which are called spherical
It was fairly easy to handle the spherically symmet- harmonics, start off quite simply for small , m, n,
ric solutions to Schrödinger's equation for hydro- but become more complex as and m increase. The
gen, because we did not have to deal with the angular simplest are
terms involving θ and φ in Equation (4-10) for ∇ 2 .
To find non spherically symmetric solutions, we Y0,0(θ,φ) = 1 no angular dependence
have to work with the complete equation
2 Y1,0 = cos θ
Eψ = – h ∇ 2ψ + V(r)ψ
2m
Y1,1 = – 1 sinθ e iφ
∇ 2ψ = 1r ∂ 2 (rψ)
2 2
∂r
Y1,– 1 = 1 sinθ e – iφ (102)
∂ψ
+ 2 1 ∂ sinθ
2
r sinθ ∂θ ∂θ Since Y0,0 has no angular dependence, all solutions
(99)
1 ∂ ψ
2 of the form
+
r2 sin θ ∂φ 2
2
ψ n,0,0 = ψ n(r)Y0,0 = ψ n(r) (103)
Differential equations involving ∇ 2 in spherical
coordinates have been studied for a long time and are the spherically symmetric solutions we have
standard procedures have been carefully worked out already been studying. We calculated ψ 1(r) and
to handle the angular dependence of the solutions of had you calculate ψ 2(r) , which corresponds to the
these equations. As long as the equation has no other values n = 1 and n = 2 respectively.
angular terms except those that appear in ∇ 2 , then
When we worked out the solution ψ 1(r) we found
the solutions are of the form
that it represented an electron in the lowest, n = 1,
f(r,θ,φ) = R n m(r)Y m(θ,φ) (100) energy level. You were to show that ψ 2(r) repre-
sented an electron in the second, n = 2, energy level.
where R n m(r) are functions that depend only on the We can see that for the symmetric solutions, the
variable (r), and the Y m(θ,φ) are functions only of integer subscript n is the energy quantum number for
the angles θ and φ . The subscripts n, and m can the electron.
take on only integer values.
It turns out that the integer subscripts and m define
When we are dealing with Schrödinger's equation, the amount of angular momentum the electron has in
the solutions are of the form a particular wave pattern. When = 0, m = 0, the
electron has no angular momentum. Thus the sym-
ψ(r,θ,φ) = ψ n m(r)Y m(θ,φ) (101) metric solutions represent an electron with no angu-
lar momentum.
where each different allowed integer value of the
subscripts n, , and m corresponds to a different The quantum number is related to the total orbital
allowed standing wave pattern for the electron. angular momentum of the electron, and m is propor-
tional to the z component L z of orbital angular
momentum. Explicitly
L z = mh (104)
 Calculus 2000 - Chapter 6 Introduction to Schrödinger's Equation Cal 6-19

The fact that the numbers , m and n have to have

integer values is simply a consequence that for any
confined wave, there is an explicit set of allowed
standing wave patterns. The electron in the hydro-
gen atom is confined by the Coulomb force of the
proton. When you work out the mathematics to
handle ∇ 2 in spherical coordinates, you find that the
allowed standing wave patterns can be identified by
the integers , m and n.

There are certain rules for the possible values of ,

m and n. When n = 1, there is only one solution
which we found. It corresponds to = m = 0. For
n = 2, the possible solutions are:

n m
possible values of
2 0 0
and m for n = 2
2 1 0

2 1 1

2 1 –1
In general, n ranges from 1 to infinity, can have
values from 0 up to n - 1, and m can range in integer
steps from + down to - . These are the rules that
define the possible standing wave patterns of the
electron in hydrogen.
 Calculus 2000 - Chapter 7 Divergence Cal 7-1

Calculus 2000-Chapter 7
Divergence

CHAPTER 7 DIVERGENCE

In the Physics text we pointed out that a vector field

was uniquely determined by formulas for the surface
integral and the line integral. As we have mentioned
several times, that is why there are four Maxwell
equations, since we need equations for the surface
and line integral of both the electric and magnetic
fields. The divergence and curl are the surface and
line integrals shrunk down to an infinitesimal or
differential scale. We will discuss divergence in this
chapter and curl in the next.
Cal 7-2 Calculus 2000 - Chapter 7 Divergence

THE DIVERGENCE Back to Gauss' law, Equation (29-5). Before we

As we mentioned, the divergence is a surface inte- shrink the law to an infinitesmal scale, we would like
gral shrunk down to an infinitesimal or differential to change the right hand side, expressng the total
scale. To see how this shrinking takes place, we will charge Q in in terms of the charge density ρ(x,y,z)
start with the concept of the surface integral as that is within the volume bounded by the closed
expressed by Gauss' law and see how we can apply surface.
it on a very small scale. We do this by considering a small volume element
We begin with Equation (29-5) of the Physics text ∆Vi = (∆x ∆y ∆z) i . If the charge density at point
Q (i) is ρ(x i,y i,z i ) then the amount of charge ∆Q i at
E⋅dA = ε in ∆Vi is
0
closed (29-5)
surface
∆z ∆V
Equation (29-4) says that for any closed surface, the ∆Q i = ρ(x i,y i,z i )∆Vi ∆y (2)
integral of E ⋅ dA over the surface is equal to 1/ε 0 ∆x
times the total charge Q in inside the volume bounded Adding up all the ∆Q i that reside inside the surface
by the surface. gives us
The interpretation we gave to this equation was to
call E ⋅ dA the flux of the field E out through the
Q in = Σi ∆Qi = Σi ρi∆Vi
(3)
area element dA . The integral over the closed = Σi ρ(xi,yi,zi ) ∆xi ∆yi ∆zi
surface is the total flux flowing out through the
surface. We said that this net flux out was created by Taking the limit as the ∆x , ∆y and ∆z go to zero
the electric charge inside. By calculating the flux of gives us the integral
E out through a spherical surface centered on a
point charge, we found that the amount of flux Q in = ρ(x,y,z) dxdydz
created by a charge Q was Q/ε 0 . volume
bounded by
(4)
The fact that Equation (29-4) applies to a surface of closed surface
arbitrary shape follows from the fact that the electric To shorten the notation, let V be the volume bounded
field of a point charge is mathematically similar to by the closed surface S, and introduce the notation
the velocity field of a point source in an incompress-
ible fluid like water. We described a point source of d 3V ≡ dxdydz (5)
a velocity field as some sort of "magic" device that
created water molecules. The physical content of Then Equation (4) can be written
Gauss' law applied to water was that the total flux of
water out through any closed surface had to be equal Q in = ρ(x,y,z) d 3V (6)
to the rate at which water molecules were being V
created inside.
Using Equation (6) in Gauss' law (29-5) gives us
Of course for a real situation there are no "magic"
sources creating water molecules, with the result
that there is no net flux of water out through any E⋅dA = ε1 ρ(x,y,z)d 3V (7)
0
closed surface, and the velocity field of water obeys S V
the equation Equation (7) is a more general integral form of
Gauss' law, relating the surface integral of E over a
closed
v ⋅ dA = 0 (1) closed surface S to the volume integral of ρ over the
surface volume bounded by S. It is Equation (7) that we
Equation (1) is the condition that the velocity field is would now like to shrink down to an infinitesmal
a purely solenoidal field like the magnetic field. scale.
 Calculus 2000 - Chapter 7 Divergence Cal 7-3

We know how to go to the small scale version of the pointing out of the surface. Thus ∆A 2 is x directed
volume integral of ρ , just undo the steps (2) through with a magnitude equal to the area ∆y∆z of that side,
(6) that we used to derive the volume integral. In while ∆A 1 points in the –x direction and has the
particular we will focus our attention on one small same magnitude.
volume element ∆Vi = ∆x i ∆y i ∆z i and apply Gauss'
law to this volume We can formally write
∆Q ∆A 1 = – x∆y∆z ; ∆A 2 = x∆y∆z
E⋅dA = ε i = ε1 ρ(x i,y i,z i)∆Vi (9)
0 0
surface (8) where x is the unit vector in the x direction. Similar
bounding
∆Vi formulas hold for the area vectors for the other four
It is clear how we got the total charge Q in when we faces of ∆V . For example, on the top face we have
added up all the ∆Q i inside the volume V. But how ∆A 3 = z∆x∆y .
do we handle the surface integral of E ? How do we
interpret adding a bunch of surface integrals over the To calculate the total flux of E out of ∆V , we have
small volume elements ∆Vi to get the surface inte- to calculate the flux out through each of the six faces.
gral over the entire surface S? For the two x oriented areas ∆A 1 , and ∆A 2 , only
the x component of E will contribute to the dot
The way to picture it is to remember that the surface products E⋅ ∆A . Let E x(x,y,z) be the average
integral over the surface of ∆Vi is equal to the flux value of E x at face 1, and E x(x +∆x,y,z) be the
of E created inside ∆Vi . From this point of view, average value of E x at face 2, which is a distance ∆x
the total flux flowing out through the surface of the down the x axis from face 1. The flux out of face 2
entire volume will be the sum of the fluxes created will be
within each volume element. To calculate this sum,
we first have to calculate the flux flowing out of the flux out = E (x +∆x,y,z)∆A
of face 2 x 2
volume element ∆Vi . (10)
= E x(x +∆x,y,z)∆y∆z
In Figure (1), we show the volume element ∆Vi
located at (x i,y i,z i) , with sides ∆x , ∆y and ∆z . At face 1, where ∆A 1 = – x∆y∆z , the dot product
Flowing through this volume element is the electric E⋅ ∆A can be written
field E(x,y,z) .
E⋅∆A 1 = (xE x + yE y + zE z )⋅(–x∆y∆z)
Also in Figure (1) we have drawn the surface area (11)
vectors ∆A 1 , and ∆A 2 for the left and right vertical = –E x(x,y,z)∆y∆z
faces. Recall that for a surface integral, the area
vector ∆A or dA is perpendicular to the surface, where x ⋅ x = 1 , y ⋅ x = z ⋅ x = 0 . We wrote the full
dot product in Equation (11) so that you could see
∆A 3 explicitly where the minus sign came from.
z Combining Equations (10) and (11) for the total flux
out of the two x directed faces of ∆V , we get
y
flux out
of x
directed = E x(x +∆x,y,z) – E x(xy,z) ∆y∆z
∆A 1 ∆A 2 faces
(12)
∆z
∆y
(x i, yi , z i ) x
∆x
Figure 1
The volume element ∆Vi .
Cal 7-4 Calculus 2000 - Chapter 7 Divergence

If we multiply Equation (12) by ∆x/∆x = 1 we get We put this discussion in the appendix because it
takes some effort which distracts from our goal of
flux out reducing Gauss' law to a differential equation. How-
E x(x+∆x,y,z) – E x(x,y,z)
of x = ∆x∆y∆z
directed ∆x ever it is important to know how to figure out when
faces of ∆V certain terms or dependencies can be neglected
(13)
when we take calculus limits. Thus the appendix
At this point, E x(x +∆x,y,z) and E x(x,y,z) are the
should not be skipped.
average values of E x , averaged over the x directed
faces at x + ∆x and x respectively, while the func- Assuming that we can replace E x by E x in Equation
tions without averaging, namely E x(x +∆x,y,z) and (13), noting that ∆x∆y∆z = ∆V , and taking the
E x(x,y,z) are just the values of E x at the lower front limit as ∆x goes to zero gives us
corners of the x oriented faces as shown in Figure
(2). Any difference between the average values of flux out of E x(x+∆x,y,z) – E x(x,y,z)
E x and the corner values E x will be due to y and z x directed = limit ∆V
variations of E x over the area ∆y∆z . faces of ∆V ∆x→0 ∆x
(14)
In Equation (13) we see that the change of E x , as we The limit is clearly the partial derivative
move in the x direction, is going to become very ∂E x(x,y,z)/∂x and we get
important. It should be clear that we are going to get
a partial derivative of E x with respect to x. What we flux out of ∂E x(x,y,z)
x directed = ∆V (15a)
are going to do now is say that variations of E x in the
faces of ∆V ∂x
x direction are important but variations of E x in the
y and z direction are not, and as a result we can Similar equations should apply to the y and z faces,
replace the average values of E x with the corner giving us
values E x .
flux out of ∂E y(x,y,z)
The above paragraph was intended to sound like a y directed = ∆V (15b)
∂y
questionable procedure. If we do it, Equation (13) faces of ∆V
immediately simplifies, as we will see shortly. But
how do we justify such a step? The answer, which flux out of ∂E z(x,y,z)
z directed = ∆V (15c)
we work out in detail in the appendix to this chapter,
faces of ∆V ∂z
is that when we take the limit as ∆V goes to zero,
contributions due to y and z variations of E x go to
Exercise 1
zero faster than the contribution from the x variation.
Neglecting the y and z variations turns out to be Draw the appropriate sketches and reproduce the
similar to neglecting α 2 terms compared to α terms arguments needed to derive Equation (15b) or (15c).
in an expansion of (1 + α) n when α is a small
number.
z

∆x ∆y
∆z
E x(x, y, z) E x(x+∆x, y, z)
x
(x, y, z) (x+∆x, y, z)
Figure 2
Electric field at the lower front corners.
 Calculus 2000 - Chapter 7 Divergence Cal 7-5

When we add up the flux out of all six faces, we get As we take the limit at ∆Vi goes to zero size, the sum
the total flux out of ∆V becomes an integral, and we end up with

total flux = ∂E x + ∂E y + ∂E z ∆V divergence

out of ∆V
(16) E⋅dA = ∇⋅E d 3V theorem (21)
∂x ∂y ∂z
closed surface V
bounding
You should spot immediately that the notation in volume V
Equation (16) can be simplified by introducing the
partial derivative operator where we are using the notation of Equation (5) that
d 3V ≡ dx dy dz .
∂ ∂ ∂
∇ = x +y +z
∂x ∂y ∂z Equation (21) is known as the divergence theorem,
(17) and the quantity ∇⋅E is known as the divergence of
≡ (x ∇ x + y ∇ y + z ∇ z) the vector field E . We saw the same operator ∇ in
the Chapter 3 when it acted on a scalar field f(x,y,z).
From the definition of the vector dot product we have
Then we had what was called a gradient
∇⋅ E gradient of
∇f a scalar field
∂ ∂ ∂
= x +y + z ⋅(x E x +y E y +z E z ) (22)
∂x ∂y ∂z ∇⋅ E
divergence of
a vector field
∂E x ∂E y ∂E z
= + + (18) You can see that ∇ operating on a scalar field
∂x ∂y ∂z
f(x,y,z) creates a vector field ∇f . In contrast, the dot
where we used x ⋅ x = 1 , x ⋅ y = 0 , etc., and noted product of ∇ with a vector field E creates a scalar
that the unit vectors are constants that can be taken field ∇ ⋅ E that has a value at every point in space
outside the derivative. For example, but does not point anywhere.
∂ ∂E Equation (21), the divergence theorem, is an ex-
(x E x ) = x x (18a)
∂x ∂x tremely useful result for it allows us to go back and
Using the notation of Equation (18), we get for the total forth between a surface integral and a volume inte-
flux out of ∆V gral. In Equation (7) reproduced here,

total flux = (∇⋅E)∆V (19) E⋅dA = ε1 ρ(x,y,z)d 3V (7) repeated

out of ∆V 0
S V
Equation (19) applies to each ∆Vi at each point we had a mixed bag with a surface integral over a
(x i, y i, z i ) within any volume V bounded by a closed surface on the left and a volume integral over
closed surface S. The total flux out through the the enclosed volume V on the right. Back then, there
surface S, which is the surface integral of E , will be was not much more we could do with that equation.
equal to the sum of all the flux created inside in all
the ∆Vi . Thus we get

E⋅dA = Σi (∇⋅E)∆Vi (20)

surface
bounding V
Cal 7-6 Calculus 2000 - Chapter 7 Divergence

But now we can replace the surface integral of E Electric Field of a Point Charge
with a volume integral of ∇⋅ E to get Until now, in both the Physics and Calculus texts,
when we obtained a new differential equation, we
E⋅dA = ε1 ρ(x,y,z)d 3V (7) repeated illustrated its use with explicit examples. This time
0
S V we do not yet have a good example for our new
Equation (25) ∇⋅ E = ρ/ε 0 . This is the differential
∇⋅E d 3V = ε1 ρ(x,y,z) d 3V (23) form of Gauss' law, and our best example for the use
0 of Gauss' law was in calculating the electric field of
V V
a point charge. The problem is that, at the point
Since we are integrating over the same volume V for
charge itself, the field E and its partial derivatives
both integrals, we can write (23) as
are infinite and the assumptions we made in deriving
ρ(x,y,z) 3 Equation (25) do not apply.
∇⋅ E (x,y,z) – ε0 d V = 0 (24)
When we are dealing with the electric field of a point
V charge, the field E is well behaved and all partial
The next argument is one often used in physics. derivatives are finite, except at the charge. The way
Since the integral in Equation (24) has to be zero we can handle point charges is to use Equation (25)
for any volume V we choose, the only way that can ∇⋅ E = ρ/ε 0 everywhere except in a small region
happen is if the integrand, the stuff in the square around the charge. In that region we revert to the
brackets, is zero. This gives us the differential integral form of Gauss' law which allows us to work
equation just outside the point charge and avoid the infinities.
Gauss'
ρ(x,y,z) Here is an outline of the way we handle the problem
∇⋅E(x,y,z) = ε0
law in
differential (25) of a point charge. We are working with Equation (25)
form ρ(x,y,z)
∇⋅E(x,y,z) = ε0 (25) repeated
Equation (25) is the differential equation represent-
ing Gauss' law. When Maxwell's equations are and everything is going well until we come up to a
written as differential equations, this will be one of point charge located at the point (x 0,y 0,z 0 ) . In a
the four. small region surrounding the point charge, we inte-
grate Equation (25) over the volume, getting
Exercise 2
ρ 3
Another of Maxwell's equations in integral form is ∇⋅E d 3V = ε0 d V (26)
volume volume
B ⋅ dA = 0 surrounding surrounding
closed charge charge
surface

What is the corresponding differential equation? The volume integral of the charge density ρ over the
region of the point charge is simply the charge Q
itself, thus we can immediately do that volume
integral, giving us
Q
∇⋅E d 3V = ε (27)
0
volume
surrounding
charge
 Calculus 2000 - Chapter 7 Divergence Cal 7-7

We still have the problem that ∇⋅ E is infinite at the The question that remains unanswered, is whether
charge itself. But we can avoid this problem by the electron is truly a point particle, or does it have
converting the volume integral of ∇⋅ E to a surface some size that is so small that we have not been able
integral of E using the divergence theorem, Equa- to see the structure yet? The important feature of
tion (21) quantum electrodynamics is that it makes testable
predictions without any reference to the electron's
∇⋅ E d 3V = E⋅ dA structure. We get the same predictions whether the
volume surface electron has no size, or is some structure that is too
surrounding enclosing (21) repeated
charge charge small to see. Our handling of the electric field of a
to get point charge is your first example of how such a
theory can be constructed. By converting to a
Q surface integral surrounding the charge, it makes no
E⋅ dA = ε
0 difference whether the charge is truly a point, or
surface (28)
surrounding confined to some region too small to see.
charge
By the way, in the current picture of elementary
In Equation (28), which we recognize as the form of particles, in what is often called the standard model,
Gauss' law we started with in the Physics text, the the true elementary particles are all point particles.
electric field is evaluated only at the surface sur- These elementary particles are the six electron type
rounding the point charge, and not at the charge particles called leptons (they are the electron, the
itself. Away from the charge, the field is finite and muon, the tau particle, and three kinds of neutrinos)
we have no problem with Equation (28). and six kinds of quarks. The standard model makes
There is a mathematical problem with the concept of many successful predictions but appears to have one
a point charge, where a finite amount of charge is critical flaw. The problem is that no one has yet
crammed into a region of zero volume, giving us succeeded in constructing a theory for the interac-
infinite charge densities and infinite fields there. We tion of point particles with gravity, the so called
have just shown how these infinities can be avoided quantum theory of gravity. Every attempt to do so
mathematically, at least for Gauss' law, by convert- has thus far led to infinities that could not be gotten
ing the volume integral of ∇⋅ E at the charge to a rid of by any known mathematical technique.
surface integral of E out from the charge. Was this
This failure to develop a quantum theory of gravity
just a mathematical exercise, or in physics do we
in which gravity interacts with point particles, has
really have to deal with point charges?
led to theories such as string theory where the
The theory of quantum electrodynamics, which de- elementary particles have a finite, but tiny size.
scribes the interaction of electrons with light (with String theory appears to avoid the infinities in the
photons), is the most precisely verified theory in sci- gravitational interaction, but the strings, from which
ence. It explains, for example, the very smallest particles are assumed to be made, are predicted to be
relativistic corrections observed in the spectrum of the so small that no way has been found to test whether
hydrogen atom. This theory treats the electron as an they actually exist or not. It is interesting that so far
actual point particle with a finite amount of mass and our only evidence that elementary particles actually
charge confined to a region of zero volume. The trick have structure is our failure to construct a theory of
we just pulled to handle the electric field of a point gravity.
charge was quite simple compared to the tricks that the
inventors of quantum electrodynamics, Feynman,
Schwinger, and Tomonaga, had to pull to handle the
infinite mass and energy densities they encountered.
The remarkable accomplishment was that they suc-
ceeded in constructing a theory of point particles, a
theory that gave finite and correct, answers.
Cal 7-8 Calculus 2000 - Chapter 7 Divergence

THE δ FUNCTION Now take the limit as ∆x → 0 , and we end up with

a curve, whose total area remains 1, but whose width
When we applied the differential form of Gauss' law
goes to zero and height goes to infinity. We will call
∇⋅ E = ρ/ε 0 to the field of a point charge, we avoided
this curve δ(x 0)
the problem of mathematical infinities by integrat-
ing the equation over a small volume surrounding of the curve of width
the charge. We never did say what the charge δ(x 0) ≡ lim ∆x and height 1/∆x, (30)
∆x → 0
centered at x0
density ρ(x,y,z) was for a point charge Q, because
we knew that if we integrated ρ(x,y,z) over the Even though δ(x 0) is infinitely high at the point x 0 ,
region of the charge, the answer would be simply Q its integral over any region that includes the point x 0
itself. is just the number 1
In physics we often run into quantities like the x greater than x 0
charge density of a point charge where the density at δ(x 0)dx = 1 (31)
the charge looks infinite, but when we integrate the
x less than x 0
density over the region of the charge, we get a finite,
reasonable answer. There is a convenient way to Actually the only important property of the δ func-
handle such problems by using what is called the tion is Equation (31). The curve does not have to be
delta ( δ ) function. a rectangle, it could be the limit of some smooth
curve like that shown in Figure (4). As long as, in the
The one dimensional δ function is a curve with a limit that ∆x → 0 , the curve becomes infinitely
unit area under it, but all the area is confined to a high, infinitely narrow, and has a unit area under it,
region of zero width. We obtain such a curve it is a δ function.
mathematically through the use of a limiting process.
In three dimensions, the δ function δ(x 0,y 0,z 0) is a
Consider the curve shown in Figure (3) that is zero quantity that is zero everywhere except at the point
everywhere except in the region around the point (x 0,y 0,z 0) , but whose integral over that region is 1
x 0 . In that region it is a rectangle of width ∆x and
height 1/∆x . The area under this curve is δ(x 0,y 0,z 0)dV = 1
any volume
area under = (∆x) 1 = 1 (29) including the (32)
rectangle ∆x point (x 0,y 0,z0)

An example of such a δ function is the function whose

∆x value is zero everywhere except within a distance ∆x
of x 0 , ∆y of y 0 , and ∆z of z 0 . In that region the value
is (1/∆x)(1/∆y)(1/∆z) , so that the total volume is 1.
1 Then take the limit as ∆x → 0 , ∆y → 0 , and ∆z → 0 .
∆x

x0
Figure 3
When we take the limit as ∆ x goes to zero, 1
we get a one dimensional delta function. ∆x
∆x
x0
Figure 4
We have a delta function as long as the area
remains 1, and the width goes to zero.
 Calculus 2000 - Chapter 7 Divergence Cal 7-9

We can now use the δ function to describe the From this example, you can see that the δ function
charge density of a point charge. If a point charge allows us to write an explicit formula for the charge
has a total charge Q and is located at the point density of a point charge, and you can see that the
(x 0,y 0,z 0) , then the charge density ρ(x,y,z) is only things we have to know about a δ function is
that δ(x 0,y 0,z 0 ) is zero except at (x 0,y 0,z 0 ) and
charge density
ρ(x,y,z) = Qδ(x 0,y 0,z 0 ) of point charge (33) that its volume integral around that point is 1. As you
at x0,y0,z0 go farther in physics, you will encounter the δ
function more and more often. It is rather nice in that
The differential form of Gauss' law applied to this there is no function easier to integrate.
charge density is

ρ(x,y,z) Exercise 3
∇⋅ E = ε0 Explain why the following mathematical relationship is
(34) true for any continuous function f(x,y,z)
Q
∇⋅ E = ε δ(x 0,y 0,z 0 ) f(x,y, z)δ(x0,y0,z0)d3V = f (x0,y0,z0)
0
any volume
To handle Equation (34), we use our old trick of (38)
includingthe
going back to the integral form by first integrating point (x0,y0,z0)
over a volume that includes the charge
Q
∇⋅ E dV = ε 0 δ(x 0,y 0,z 0 )dV (35)
volume volume
including including point
charge x 0, y 0, z0

Since Q/ε 0 is a constant, it can be taken outside the

integral on the right side of Equation (35), giving

Q Q
ε0 δ(x 0,y 0,z 0 )dV = ε ×1 (36)
0
volume
including point
x 0, y 0, z 0

where we used the fact that the integral of the δ

function was 1. Now convert the volume integral of
∇⋅ E to a surface integral

∇⋅ E dV = E⋅ dA (37)
volume surface
including point surrounding
x 0, y 0, z0 x 0, y 0, z 0

Using (36) and (37) gives

Q
E⋅ dA = ε
0
closed surface
including Q
which is our integral form of Gauss' law.
Cal 7-10 Calculus 2000 - Chapter 7 Divergence

DIVERGENCE FREE FIELDS What kind of solutions are possible for a divergence
It may seem a bit discouraging that we did all this free field? What are the solutions to the equation
work to derive the differential form of Gauss' law ∇ ⋅ v=0?
∇ ⋅ E = ρ/ε 0 , and then end up, when we want to The answer is at least as complex as the behavior of
actually solve a problem, going back to the integral water. You have seen water flow smoothly in a lazy
form of the equation. At this point, that is about all river. That is called laminar flow. Such laminar
we can do to solve for explicit field patterns E . flow is one solution to ∇ ⋅ v = 0 . But in a fast
However, the differential form begins to tell us flowing stream there can be complex eddies called
about some general features of a vector field as we turbulence. Turbulent flow is also a solution to the
shall now see. With a lot more practice with the equation ∇ ⋅ v = 0 .
differential form of the field equations, and perhaps
a computer thrown in, one can begin to solve for You can now see that the equation ∇ ⋅ v = 0 puts a
complex field shapes. In this text we will focus on restriction on the field v , but still allows an enor-
what we can learn about general features and leave mous range of solutions. Because of your familiar-
the solution of complex field shapes to a later course. ity with the flow of water you have some insight into
what these solutions can be.
To see what we can learn about general features of
a field, suppose that we have a velocity field v(x,y,z) ,
whose divergence is zero, i.e., it obeys the equation

∇⋅ v(x,y,z) = 0 (39)
We say that such a field is divergence free. What can
we say about the properties of such a field?

To answer that question, we will again go back to the

integral form, by integrating Equation (1) over some
volume V to get

∇⋅ v d 3V = 0 (40)
volume V
Now use the divergence theorem to convert this
volume integral to a surface integral, giving

v⋅ dA = 0 (41)
closed
surface
Equation (41) is our old equation for a vector field
that has no sources or sinks. It is the equation for an
incompressible, constant density fluid, a real one
like water where water molecules are not being
created or destroyed. Thus the condition that a
vector field be divergence free, i.e., ∇ ⋅ v = 0 or
∇ ⋅ E = 0 or ∇ ⋅ B = 0 , is that the field behaves like
the velocity field of an incompressible fluid.
 Calculus 2000 - Chapter 7 Divergence Cal 7-11

APPENDIX — DERIVATION To evaluate E x at (x, y, z + ∆z) , up at point (3), we

can use a Taylor series expansion. So far we have
OF FLUX EQUATION (14) discussed a Taylor series expansion only of a func-
Earlier in the chapter we had the following formula tion of a single variable f(x). The expansion was,
for the flux out of the x directed faces of the small from Equation (2-44 of Calculus Chapter 2)
cube ∆V = ∆x∆y∆z
f(x–x 0) = f(x 0) + ∂f (x–x 0) + 1 ∂ f2 (x–x 0) 2 +⋅⋅⋅
2

flux out
∂x 2! ∂x
E (x+∆x,y,z) – E x(x,y,z)
of x directed = x ∆x∆y∆z (2-44) repeated
faces of ∆V ∆x
which is good for small steps (x–x 0) .
(13) repeated
where E x(x+∆x,y,z) and E x(x,y,z) were the average What we are doing when we go from point (1) to
values of E x on the two x directed faces of the cube. point (3) in Figure (2), is keeping the values of x and
y constant, and looking at the change in E x as we
In Equation (14) we replaced the average values E x vary z. Thus in going up, we have a function E x(z)
by the values E x(x+∆x,y,z) and E x(x,y,z) at the that is only a function of z, and we can use our old
lower front corners as shown in Figure (2), repeated Taylor series expansion to get
here giving
flux out E x(x+∆x,y,z) – E x(x,y,z)
E x(x,y,z+∆z) = E x(x,y,z)
of x directed = limit ∆V
faces of ∆V ∆x→0 ∆x ∂E x(x,y,z)
+ (∆z)
(14) repeated ∂z
What we are doing is in going from Equation (13) to ∂ 2E x(x,y,z) (42)
(14) is to neglect the y and z dependence of E x while +1 (∆z) 2
2 ∂z 2
developing an equation for the x dependence. This
step needs justification. +⋅⋅⋅
where ∆z is analogous to the step (x–x 0) in the
To see what effect the y and z dependence has, let us Taylor series formula.
start by approximating the average value of E x over
the entire x faces by the average of the top and Because we are eventually going to take the limit as
bottom values of the front side of ∆A X , i.e., the ∆z goes to zero, we will be able to neglect terms of
average of E x at points (1) and (3) on the left and order (∆z) 2 compared to ∆z . Because of that, it is
points (2) and (4) on the right as shown in Figure (5). sufficient to write

This is a rather crude approximation for the average E x(x,y,z+∆z) = E x(x,y,z)

over the face, but begins to show us what the effect ∂E x
+ ∆z (42a)
of the y and z dependence of E x is. ∂z
z + terms of order ∆z 2
E x (x, y, z+∆z)
∆x ∆y E x (x+∆x, y, z+∆z)
(3)
(4)
∆z
E x(x,y,z) E x(x+∆x,y,z) (2)
x (1)
E x (x, y, z) E x (x+∆x, y, z)
(x, y, z) (x+∆x, y, z)
Figure 2 (repeated) Figure 5
Electric field at the lower front corners. Electric field at four positions.
Cal 7-12 Calculus 2000 - Chapter 7 Divergence

When we take the average of E x at points (1) and

(3), a result we will call E x(x) 1,3 , we get limit E x(x+∆x,y,z) – E x(x,y,z) = ∂E x(x,y,z)
∆x→0 ∆x ∂x
E x(x,y,z) + E x(x,y,z+∆z) (46)
E x(x) 1,3 =
2
∂E x(x+∆x,y,z) ∂E x(x,y,z)
∂E (x,y,z) –
= E x(x,y,z) + 1 x ∆z + O(∆z 2) limit ∂z ∂z ∂ 2E x(x,y,z)
2 ∂z =
(43) ∆x→0 ∆x ∂x∂z
2 2
where O(∆z ) means terms of order (∆z ) . (47)
A similar argument gives the average E x(x+∆x) 2,4 Thus Equation (45) is taking on the form
at points (2) and (4)
flux out of
∂E x ∂ 2E x
E (x+∆x,y,z) + E x(x+∆x,y,z +∆z) x faces of = ∆V + ∆z + O(∆z 2)
E x(x+∆x) 2,4 = x ∆V for 2 ∂x ∂x∂z
2 point average
∂E (x+∆x,y,z) (48)
= E x(x+∆x,y,z) + 1 x ∆z + O(∆z 2)
2 ∂z We see that corrections due to the z dependence of
(44) E x are of magnitude ∆z times the partial second
derivative ∂ 2E x/∂x∂z . As long as all derivatives of
Using our 2 point averages in Equation (13) for the
E x are bounded, stay finite as we take the limit as
flux out of ∆V gives us
∆x , ∆y , and ∆z go to zero, then the ∆z term in
flux out of Equation (48) becomes negligently small, which
E x(x+∆x) 2,4 – E x(x) 1,3
x directed face =
of ∆V for 2
∆V means that in the limit we can neglect the z depen-
point average ∆x dence of E x , at least in this two point approximation.

Our 2 point approximation to the average of E x can

∂E (x+∆x,y,z) be improved by using more points. If we included
= ∆V E x(x+∆x,y,z) + 1 x ∆z
∆x 2 ∂z the back points at (y+∆y) , we would add terms to
Equation (48) of the form
∂E (x,y,z) ∂ 2E x
– E x(x,y,z) – 1 x ∆z + O(∆z 2) ∆y + O(∆z 2)
2 ∂z ∂x∂y
(49)

terms which would go to zero in the limit ∆y → 0 .

E (x+∆x,y,z) – E x(x,y,z) All points we add in to the average will give terms
= x ∆V
∆x proportional to ∆x or ∆y or some combination, and
all these terms will go to zero when we take the limit
∂E x(x+∆x,y,z) ∂E x(x,y,z) as ∆x , ∆y , and ∆z goes to zero. Thus, it is an exact
–
1 ∂z ∂z result that, in the limit that ∆v → 0 , only the x
+ ∆z∆V (45)
2 ∆x dependence of E x has to be taken into account,
provided all derivatives of E x are finite.
+ O(∆z 2)∆V
When we go to the limit that ∆x goes to zero, we see
that we get the partial derivatives
 Calculus 2000 - Chapter 8 Curl Cal 8-1

Calculus 2000-Chapter 8
Curl

CHAPTER 8 CURL

ABOUT THE CURL

In the Physics text, we saw that a vector field was
uniquely determined by formulas for the surface
integral and the line integral. In the last chapter, we
saw that the divergence, such as ∇ ⋅ E , represented
the surface integral shrunk down to an infinitesimal
scale. In this chapter, we study the curl, which is the
line integral shrunk down to an infinitesimal scale.
Here our emphasis will be on the application of the
curl to electric and magnetic fields. In the final
chapters of this text, Chapters 12 and 13, we develop
an intuitive picture of the curl applied to the velocity
field of fluids such as water and superfluid helium.
The curl of the velocity field is called vorticity, a
concept that plays a fundamental role in under-
standing such phenomena as quantum vortices and
turbulence.
Cal 8-2 Calculus 2000 - Chapter 8 Curl

INTRODUCTION TO THE CURL motion. You will recall that the angular analogy to
The partial derivative operator Newton's second law was

τ = dL
∇ = x ∂ +y ∂ +z ∂
(7)
∂x ∂y ∂z dt
has now appeared in our formulas for the gradient of where the torque, τ = r × F , is what we called the
a scalar field f(x,y,z) angular force, and L = r × p is the angular momen-
tum. Despite the appearance of two cross products in
∇f(x,y,z) = x ∂f + y ∂f + z ∂f (1) Equation (7), the equation led to a very successful
∂x ∂y ∂z prediction of the motion of a gyroscope at the end of
in the divergence of a vector field E(x,y,z) Chapter 12 in the Physics text (see page12-18).
∂E x ∂E y ∂E z With this background, we see that there is one more
∇⋅E = + + (2)
∂x ∂y ∂z natural vector product involving the operator ∇ . It
and in the Laplacian is the cross product of ∇ with some vector field like
E , B , or v . The cross product, for example with B ,
∇⋅∇f = ∇ 2f = ∂ f2 + ∂ f2 + ∂ f2
2 2 2
(3) is called the curl of B .
∂x ∂y ∂z
∇ × B = x(∇ yB z – ∇ zB y )
While ∇ is an operator in the sense that it only has a
value when operating on some field, we see that it acts + y(∇ zB x – ∇ xB z ) curl (8)
very much like a vector. This suggests that we may
+ z(∇ xB y – ∇ yB x )
encounter other vector like operations involving ∇.
With all these derivatives in the formula for ∇ × B , the
In our discussion of vectors in Chapter 2 of the concept of the curl looks rather formidable. Later in
Physics text, we saw that there were two kinds of this chapter we will discuss the formula for the curl in
vector products, the scalar or dot product cylindrical coordinates. That formula looks even worse
than Equation (8). However when we apply the curl in
C = A⋅B = (A xB x + A yB y + A zB z ) scalar
product (4) cylindrical coordinates to a problem with cylindrical
and the vector cross product symmetry, we end up with a simple, easily applied
formula (which we will see in Equation 58).
C=A×B vector cross product (5)
C=A B C = AB(sinθ)
where the formulas for the components of C were
B
C x = A yB z – A zB y
C y = A zB x – A xB z
θ A
(6)
C z = A xB y – A yB x C=A B
We saw that the vector C = A × B was oriented
perpendicular to the plane of the vectors A and B ,
the choice of which direction being given by the B
right hand rule as shown in Figure (1). The magni-
tude was C = AB sin θ which is maximum when A A
and B are perpendicular and zero when parallel.
Figure 1
The vector cross product seems like a rather peculiar
Right hand rule for the cross product.
mathematical construct, but it plays an important (Discussed in Physics 2000, page 2-15.)
role in physics, particularly in describing rotational
 Calculus 2000 - Chapter 8 Curl Cal 8-3

As we have mentioned several times now, to deter- that theorem to the theory of electricity and magne-
mine a vector field we need formulas for the surface tism. This allows us to finish translating Maxwell's
integral and the line integral. In the last chapter we equations from the integral to the differential form.
saw that when we go to the small scale limit, the
volume integral becomes a divergence. An example In Chapter 9 we derive a set of equations called vector
was Gauss' law which in the integral form was identities that simplify working with formulas involv-
ing the curl. We will use the vector identities to show
Q
E ⋅ dA = ε in (9) that Maxwell's equations in empty space become the
0
wave equations for electromagnetic fields.
It became the differential equation
ρ In Chapter 11 we find that the wave equation for
∇⋅E = ε (10) electromagnetic fields in the presence of electric charge
0
In this chapter we will see that the differential limit of and current is considerably simplified by expressing
the line integral is the curl. We will see, for example, the magnetic field as the curl of a new kind of a vector
that the old form of Ampere's law (when ∂E/∂t = 0 ) field called the vector potential A . This is a rather
technical subject, the study of which can be put off for
B⋅d = µ 0i in (11) a while. We placed this material where we did so that
you could see what happens to the electromagnetic
becomes the differential equation wave equation when sources are present.
∇ × B = µ 0 i (x,y,z ) (12) In Chapter 12 we apply the curl to the velocity field v .
where i (x,y,z ) is the current density. It is in that chapter where you can develop the best
intuitive picture of the curl. If you want to put off for
In our discussion of divergence, one of the important a while studying the wave equation for electromagnetic
results was the divergence theorem fields, you can go directly from this chapter to Chapter
divergence 12 and build your intuition for curl.
E ⋅ dA = ∇ ⋅ E d 3V theorem (13)
S V In case you were wondering about Chapter 10, it deals
where V is the volume bounded by a closed surface with the extension of the continuity equation to handle
S and d 3V = dxdydz . The divergence theorem compressible conserved flows, like the flow of electric
allowed us to immediately go back and forth be- charge. We discover from this work a rather remark-
tween surface integrals and volume integrals. able result, namely that Maxwell's equations require
that electric charge be conserved. This is one of the
An important result of this chapter is what one could first completely new physical predictions we get by
call the curl theorem, but which is known as Stokes' going to the differential form of Maxwell's equations.
law. It is

B⋅ d = (∇ × B ) ⋅ dA Stokes' (14)
law
around area of
closed path closed path

which relates the line integral of B around a closed path

to an integral of the curl of B over any area bounded
by the closed path. An example of a closed path is the
wire loop shown in Figure (2). One of the areas
bounded by this closed path is that of the soap film.

Our discussion of the curl will proceed through the

remaining chapters of the text. In this chapter we Figure 2
will focus on deriving Stokes' theorem and applying Example of a surface bounded
by a closed path (wire loop).
Cal 8-4 Calculus 2000 - Chapter 8 Curl

STOKES' LAW The integral up the right hand side becomes

As we noted, Stokes' law, Equation (14) allows us to
3 3
convert from a line integral around a closed path to B⋅d = B yd y
a surface integral over the area bounded by the path. 2 2
(18)
Once we have derived Stokes' law, it will be quite = B y (x+∆x/2,y)∆y
easy to use it to convert to differential equations the
two Maxwell equations involving path integrals. where B y (x+∆x/2,y) is the average value of B y along
the right side, out at a distance ∆x/2 from the center.
To derive Stokes' law, we begin by calculating the
path integral of some vector field B around a small On the top side, we are integrating in the –x direc-
rectangular path of sides ∆x and ∆y shown in tion, the dot product B ⋅ d is negative, and we get
Figure (3). Our arguments will be somewhat similar
to those we used to derive the divergence theorem. 4 4
B⋅d = –B xd x
The line integral around the rectangle ∆x∆y can be 3 3
(19)
written as the four integrals = –B x (x,y+∆y/2)∆x
2 3
B⋅d = B⋅d + B⋅d where B x (x,y+∆y/2) is the average value of B x on
1 2 the top edge.
around
∆x∆y
4 1 (15) Going back down from point (4) to point (1) we are
+ B⋅d + B⋅d going in the –y direction, B ⋅ d = – B yd y and we
3 4
get
Along the path from point (1) to point (2), along the
bottom of the rectangle, we are integrating in the x 1 1
direction, thus B⋅d = –B yd y
4 4
2 2 (20)
B⋅d = B xd x (16) = –B y (x–∆x/2,y)∆y
1 1
The integral of B xd x over the bottom side can be Using Equations (17) through (20) in (15) gives,
written as after some rearranging
2
B xd x = B x (x,y–∆y/2)∆x (17) B y(x+∆x/2,y) – B y(x–∆x/2,y)
1 B⋅d = ∆x∆y
∆x
around
where B x (x,y–∆y/2) is the average value of B x ∆x∆y
along the lower edge, a distance ∆y/2 below the B x(x,y+∆y/2) – B x(x,y–∆y/2)
center (x,y) of the rectangle. – ∆x∆y
∆y
(4) (3) (21)
As a first approximation to Equation (21), we could
replace the average values of B x , B y on the four
B(x, y) ∆y sides by the actual values of B x , B y at the center of
each side. For example, since the center of the side
from (2) to (3) is at the point (x+∆x/2,y) , we would
∆x be making the substitution for that side of
B y(x+∆x/2,y) → B y(x+∆x/2,y) (22)
(1) (2)
Figure 3 I.e., we would be removing the bars over the values
Calculating the integral of B ⋅ dl around a small of B in Equation (21).
rectangular path centered at the point (x,y).
 Calcul