Elements of Newtonian Mechanics

Elements of Newtonian Mechanics
Advanced Texts in Physics

This program of advanced texts covers a broad spectrum of topics which are of
current and emerging interest in physics. Each book provides a comprehensive and
yet current and emerging interest in physics. Each book provides a comprehensive
and yet accessible introduction to a field at the forefront of modern research. As
such, these texts are intended for senior undergraduate and graduate students
at the MS and PhD level; however, research scientists seeking an introduction to
particular areas of physics will also benefit from the titles in this collection.
ONLINE L1BRARY
Physics and Astronomy
http://www.springer.de/phys/
Springer-Verlag Berlin Heidelberg GmbH
Jens M. Knudsen Paul G. Hjorth
Elements
of Newtonian Mechanics
Including Nonline ar Dynamics
Third Revised and Enlarged Edition
With 341 Figures and 131 Problems with Solutions

and 89 Worked Examples
i Springer
Dr. Jens M. Knudsen Dr. PouI G. Hjorth
0rsted Laboratory MathematicaI Institute, B303
Niels Bohr Institute Technical University of Denmark
Universitetsparken 5 2800 Lyngby
2100 Copenhagen Denmark
Denmark
Library of Congress Cataloging-in -Publication Data

Knudsen, J.M. (Jens M.), 1955-
Elements of Newtonian mechanics: including nonlinear dynamics / Jens M. Knudsen,
Poul G. Hjorth. - - 3rd rev. and enl. ed.
p. cm. - - (Advanced texts in physics, ISSN 1439-2674)
Includes bibliographical references and index.
ISBN 978-3-540-67652-2 ISBN 978-3-642-57234-0 (eBook)
DOI 10.1007/978-3-642-57234-0
1. Mechanics. 2. Mechanics-Problems, exercises, etc. 1. Hjorth, P.G. (Poul G.), 1930-
II. Title. III. Series.
QC125.2 .K48 2000 531 - - dC21 00-044012
Third Edition 2000. Corrected 2nd Print ing 2002
ISSN 1439-2674
ISBN 978-3-540-67652-2
This work is subject to copyright. AII rights are reserved, whether the whole or part of the material
is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broad-
casting, reproduction on microfilm or in any other way, and storage in data banks. Duplication of
this publication or parts thereof is permitted only under the provisions of the German Copyright Law
of September 9, 1965, in its current version, and permission for use must aIways be obtained from
Springer-Verlag. Violations are liable for prosecution under the German Copyright Law.
http://www.springer.de
© Springer-Verlag Berlin Heidelberg 1995, 1996, 2000
Originally published by Springer-Verlag Berlin Heidelberg New York in 2000
The use of general descriptive names, registered names, trademarks, etc. in this publication does not
imply, even in the absence of a specific statement, that such names are exempt from the relevant pro-
tective laws and regulations and therefore free for general use.
Typesetting: Data prepared by the author using a Springer TIYC macro package
Cover design: design & production GmbH, Heidelberg
Printed on acid -free paper SPIN 10914670 56/3111 5 4 3 2 1
There is no physical experiment
by which you can determine
and later recover
a single point in astronomical space
A Galaxy (M-74) in the constellation Pisces. This galaxy is about 20 000 000 light years
from Earth. This is how our own galaxy, the Milky Way, would appear if seen from a great
distance (Photograph © Hale Observatories, Pasadena)
A Note for the Reader
Nearly half of this book consists of examples. These examples form an ab-
solutely essential part of the text and should by no means be skipped. The
authors believe that Newton's ideas on the subtle problem of motion can
be assimilated only by working with Newton's equations applied to concrete
examples.
Listening to lectures is not enough. All processes of learning are somehow
connected to active participation, and the learning of physics is no exception.
To underline this viewpoint we have, at the beginning of the course, always
written on the blackboard, as a kind of motto:
At Home
by Your Desk.
Nearly all the chapters in the book are followed by a set of problems.
Very few of these problems are simple "plug-in" exercises. Most problems
will demand some independent thinking. If you cannot solve all the problems
at first try, do not despair. We have good advice which has worked for many
students: study the text, and in particular the examples, one, two, ... many
times over. In the end, you will succeed.
For several problems you will need some parameters, e.g., the mass of
the Earth, the distance of the Earth from the Sun, etc. The numbers you
need can all be found on the inside of the cover of the book. Answers to the
problems are found at the back of the book.
Throughout the book SI units have been used. Relevant SI units can be
found on the inside of the cover.
A Note for the Instructor

In Copenhagen, less than one quarter of the examples were presented in the
lectures. Most of the remaining examples were assigned for recitation sessions.
JMK & PGH

Preface to the Third Edition
In the third edition a number of minor misprints that appeared in the second
edition have have been corrected. Furthermore, 17 new problems have been
added, at the end of chapters 6, 8, 9, 11, 12, 13, and 14. The answers to
these 17 problems have not been listed in the 'Answers' section at the end
of the book. This will permit the problems to be used as hand-in problems
or perhaps in mid-term exams.
Copenhagen JMK €9 PGH

May 2000
Preface to the Second Edition
In the second edition, a number of misprints that appeared in the first edition
have been corrected. In addition to this, we have made improvements based
on the experience gathered in the use of the first English edition of the book
in the introductory course in physics at the University of Copenhagen.
A chapter introducing nonlinear dynamics has been added. The purpose
of this chapter is to provide supplementary reading for the students who
are interested in this area of active research, where Newtonian mechanics
plays an essential role. The students who wish to dig deeper, should consult
texts dedicated to the study of nonlinear dynamical systems and chaos. The
literature list at the end of this book contains several references for the topic.
The book still contains a one-semester (15 weeks) first university course
on Newtonian mechanics. This necessarily introduces some constraints on
the choice of topics and the level of mathematical sophistication expected
from the reader. If one looks for discussions of technical issues, such as the
physics behind various manifestations of friction, or the tensorial nature of
the rotation vector, one will look in vain. The book contains what we feel are
the essential aspects of Newtonian Mechanics.
It is a pleasure again to thank Springer-Verlag and in particular Dr. H. J.
Kolsch and the staff at the Heidelberg office for helpfulness and professional
collaboration.
Copenhagen JMK & PGH

June 1996
Preface to the First Edition
This book is intended as a textbook for an entry-level university course in

Newtonian mechanics for students of physics, astronomy, and the engineering
sciences. The material has been used as a first-semester text for first-year
undergraduates at the Niels Bohr Institute, which is part of the University
of Copenhagen.
Our way of presenting Newtonian mechanics is influenced by the writings
of the late Max Born. Also, the Feynman Lectures on Physics have been an
important source of inspiration. In fact, the idea for the book came when
we read Section 16.1 of Volume 1 of the Feynman Lectures. Ideas from the
well-known Berkeley Physics Course may also be traced in the text. All of
the books quoted in the literature list have, in one way or another, served as
a source for our lectures for undergraduates.
It is assumed that the students already have a rudimentary knowledge
of Newtonian mechanics, say at the high-school level. Some background in
vectors and elementary calculus is also required, i.e., the students should know
how to add vectors as well as how to differentiate and integrate elementary
functions. The Appendix contains the required background for the use of
vectors in Newtonian mechanics.
Careful study of the many worked examples will give the student the
ability to use the powerful tools of Newtonian mechanics. Furthermore, we
emphasize the fundamental problem of motion from the very beginning. This
prepares the students for an understanding of Einstein's special and general
theory of relativity. The text will demonstrate for the student that the answer
to the question "What does it mean that a body moves?" is far from simple.
The authors wish to express their gratitude to many colleagues for discus-
sions and for the encouragement given. Special thanks goes to J. Lyngesen
who has read the entire manuscript and provided many valuable suggestions.
It is a pleasure for the authors to thank cando polyt. E. B. Beran at the
Technical University of Denmark, B. Kaluza and Blue Sky Research (makers
of Textures) for assistance in 1l\TE]X typesetting, and Springer-Verlag Heidel-
berg, in particular Dr. H.J. Kolsch, for excellent cooperation in the prepara-
tion of the manuscript.
Copenhagen lens Martin Knudsen
January 1995 Poul Georg Hjorth
Contents
1. The Foundation of Classical Mechanics .. . . . . . . . . . . . . . . 1. . .

1.1 Principia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.2 Prerequisites for Newton . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.3 The Masterpiece ...................... . .......... . ... 5
The Acceleration of Gravity. . . . . . . . . . . . . . . . . . . . . . .... . .11
Circular Motion. . . . . . . . . . . . . . . . . . . . . . . . . . .... . .12 .. .. .
Communication Satellite .............................. 13
Horizontal Throw. . . . . . . . . . . . . . . . . . . . . . . . ... . 14 . .. . .. . . .
The Gravitational Constant. . . . . . . . . . . . . . . . . . ... . 17 .. . .. .
String Force. . . . . . . . ............ . . . . . . . . .... . . 18
. . .....
Forces and Tension. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... 19
Dimensional Analysis . . . . . . . . . . . . . . . . . . . . . . .. . 20 . .. . . .. .
1.4 Concluding Remarks. . . . . . . . . . . . . . . . . . . . . . .. . . . .21 . . . . . .
1.5 Problems. . . . . . . . . .......... . . . . . . . . . . . . . . ..
. . . .22 . ..
2. Newton's Five Laws ............. .. .... .. .... . ...... .. .. . . 27
2.1 Newton's Laws of Motion ................. . ............ 27
2.2 Integration of the Equation of Motion. . . . . . . . . . . . ... . .31. .
Constant Force. . . . . . . . . . . . ... . . ..... . . . . . .31 ..... . ...
The Harmonic Oscillator. . . . . . . . . . . . .. ........ 33. . . . . . ..
Mass on a Spring in the Gravitational Field of Earth. . . .. 36
Sphere Falling Through a Liquid. . . . . . . . . . . . . . . ... . . 38. . . .
Solid Against Solid . . . . . . . . . . . . . . . . . . . . . . . . . . 40 . . . . . . .
The Atwood Machine. . . . . . . . . . . . . . . . . . . . . . .... 41. . . . . . .
Force in Harmonic Motion .......................... . . 42
Charged Particle in a Uniform Magnetic Field. . . . . . . .... 44
Thomson's Experiment. . . . . . . . . . . . . . . . . . . . . . .... . . 48. ...
Work and Energy in Linear Motion of a Particle . . . . . ... . 51
Free Fall Towards the Sun from a Great Distance . . . . . .... 53
Momentum Conservation. . . . . . . . . . . . . . . . . . . . .. . .54. . . . . .
Inelastic Collisions . . . . . . . . . . . . ...... . . . . . .... . . 56
. . ....
Rocket Propulsion. . . . . . . . . . . . . . . . . . . . . . . . .... . .57 . . . . . .
Some Qualitative Remarks on Rocket Propulsion . .... .... 59
Ball Against a Wall . . . . . . . . . . . . . . . . . . . . . . ... . .61 .. . ... .
2.3 Problems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 .. ... . . . .
XIV Contents
3. Gravitational and Inertial Mass. . . . . . . . . . . . . . . . . . . . . . . . . .. 73

3.1 Gravitational Mass. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
3.2 Inertial Mass ........................................ 74
3.3 Proportionality Between Inertial and Gravitational Mass .. 75
3.4 Newton's Experiment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 76
The Satellite. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... . . 78
A n Elevator in Free Fall. . . . . . . . . . . . . . . . . . . . . .... . . 79
....
Three Balls. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 80
3.5 Problem. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... 80
4. The Galilei Transformation .... . ......................... 83

4.1 The Galilei Transformation . . . . . . . . . . . . . . . . . . . . . . . . . . .. 84
4.2 Galileo Speaks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 90
Velocity Transformation. . . . . . . . . . . . . . . . . . . . . . . .... . . 91 ..
4.3 Problems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 93
5. The Motion of the Earth ................................. 95

5.1 Examples. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 95
Vectors and the Rotation of a Rigid Body. . . . . . . . . . . . . . .. 95
A ngular Velocities in the Solar System . . . . . . . . . . . . . . . . .. 97
5.2 Problems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 99
6. Motion in Accelerated Reference Frames ................. 101

6.1 Newton's 2nd Law Within Accelerated Reference Frames .. 101
6.2 The Equivalence Principle of Mechanics ................. 108
6.3 The Einstein Box .......................... . .... . ..... 111
Balloon in Accelerated Frame .................... . ..... 114
Mass on an Oscillating Plate ... . ..... . .......... . ..... 115
Pendulum in an Elevator . ............................. 115
6.4 The Centrifugal Force ................ . .......... . ..... 116
Earth's Orbit Around the Sun . ........ . ...... . ......... 117
Grass on a Rotating Disk . ........................ . .... 117
The Variation of g with Latitude . ................ . ..... 118
6.5 Tidal Fields ......................... . . .............. 121
The Roche Limit . .................................... 126
6.6 The Coriolis Force ......... . .......................... 128
Coriolis Force on a Train . ............................. 133
Particle on a Frictionless Disc . ........................ 133
The Vertical Throw . .............................. . ... 134
6.7 Tidal Forces and Local Inertial Frames ........ . ......... 135
Global and Local Inertial Frames . ...................... 137
6.8 The Foucault Pendulum .......... . .................... 137
6.9 Newton's Bucket ..................................... 142
6.10 Review: Fictitious Forces .............................. 145
6.11 Problems ............................................ 146
Contents XV
7. The Problem of Motion ................................... 153

7.1 Kinematic and Dynamic Views of the Problem of Motion .. 153
7.2 Einstein Speaks ...................................... 155
7.3 Symmetry ........................................... 157
7.4 The Symmetry (Invariance) of Newton's 2nd Law ........ 157
7.5 Limited Absolute Space ............................... 159
7.6 The Asymmetry (Variance) of Newton's 2nd Law ......... 159
7.7 Critique of the Newtonian View ........................ 162
7.8 Concluding Remarks .................................. 163
8. Energy ................................................... 167

8.1 Work and Kinetic Energy .......... . .......... . ....... 167
8.2 Conservative Force Fields ..... . ....................... 169
8.3 Central Force Fields .................................. 170
8.4 Potential Energy and Conservation of Energy ..... . ...... 171
8.5 Calculation of Potential Energy ........................ 174
Constant Gravitational Field . .......... . ............... 175
Spring Force. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 175
Gravity Outside a Homogeneous Sphere ................. 175
8.6 The Gravitational Field Around a Homogeneous Sphere ... 176
8.6.1 The Field Around a Spherical Shell ............... 176
8.6.2 A Solid Sphere ................................. 179
8.7 Examples ........................................... 181
Particle on a Frictionless Curve . ....................... 181
String Force in the Pendulum . ......................... 182
The Gravitational Potential Outside the Earth ........... 183
Potential Energy Due to Electric Forces . ........... . .... 183
A Tunnel Through the Earth . .......................... 184
The Asymmetry of Nature . ....................... . .... 186
8.8 Review: Conservative Forces and Potential Energy ........ 188
8.9 Problems ............................................ 189
9. The Center-of-Mass Theorem ............................. 193

9.1 The Center of Mass ................................... 193
9.2 The Center-of-Mass Frame ............................ 197
9.3 Examples ........................... . ............... 199
Two Masses Connected with a Spring . .................. 199
Inelastic Collisions ............................... . ... 202
The Collision Approximation .......................... 204
Freely Falling Spring. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
The Wedge ... ....................................... 209
9.4 Review: Center of Mass and Center-of-Mass Theorems .... 212
9.5 Comments on the Conservation Theorems ............... 212
9.6 Problems ............................................ 213
XVI Contents
10. The Angular Momentum Theorem ........................ 219

10.1 The Angular Momentum Theorem for a Particle ......... 219
10.2 Conservation of Angular Momentum .................... 221
10.3 Torque and Angular Momentum Around an Axis ......... 223
10.4 The Angular Momentum Theorem
for a System of Particles .............................. 224
10.5 Center of Gravity .................................... 227
10.6 Angular Momentum Around the Center of Mass .......... 228
10.7 Review: Equations of Motion for a System of Particles .... 230
10.8 Examples of Conservation of Angular Momentum ........ 230
Particle in Circular Motion . ...................... . .... 230
Rotation of Galaxies, Solar Systems, etc . ........ ... ..... 231
11. Rotation of a Rigid Body ....................... . ......... 237

11.1 Equations of Motion ............. . ..... . .............. 237
11.2 The Rotation Vector .................................. 238
11.3 Kinetic Energy of a Rotating Disk ..................... 239
11.3.1 The Parallel Axis Theorem ..................... 240
11.3.2 The Perpendicular Axis Theorem ................ 242
11.4 Angular Momentum of an Arbitrary Rigid Body
in Rotation Around a Fixed Axis ...................... 243
11.4.1 The Parallel Axis Theorem in General Form ....... 244
11.5 Calculation of the Moment of Inertia for Simple Bodies '" 245
11.5.1 Homogenous Thin Rod ............... . ......... 245
11.5.2 Circular Disk . .. .......................... . .... 246
11.5.3 Thin Spherical Shell ..................... ... .... 247
11.5.4 Homogenous (Solid) Sphere, Mass M and Radius R 248
11.5.5 Rectangular Plate .............................. 249
11.6 Equation of Motion for a Rigid Body
Rotating Around a Fixed Axis ................... .. .... 250
11.6.1 Conservation of Angular Momentum .............. 253
11. 7 Work and Power in the Rotation of a Rigid Body
Around a Fixed Axis ............. .. ......... ... ...... 254
11.7.1 Torsion Pendulum ...... ... ..................... 255
11 .8 The Angular Momentum Theorem
Referred to Various Points ..................... . ....... 256
11.9 Examples .......... .. ........... ..... .... . .... . ..... 258
Rotating Cylinder .. ... ......... . ..................... 258
Falling Cylinder . ...... ... ................. .. ......... 260
The Atwood Machine . .... ........ . . . ................. 262
The Physical Pendulum .. . .......... . ..... . ... .. ...... 263
The Rod ............. ..... .... ... .... ....... .... ..... 264
11.10 Review: Linear Motion and Rotation About a Fixed Axis .. 266
11.11 Problems ... ......... .. .......... . ............. .. .... 267
Contents XVII
12. The Laws of Motion .................... .. . ........... . .... 279

12.1 Review: Classical Mechanics ....... . . .. .......... . . . .... 279
12.2 Remarks on the Three Conservation Theorems . .. . . .. ... . . 280
12.3 Examples . . . . . ..... .. . . ...... . .. . . . ..... . .. .. .. . ..... 280
Conservation of Angular Momentum . ... . ..... . . . .. . . .... 280
Rotating Rod ..... . . . ............. . ................... 284
Man on Disk . ...... . ... .. .... .. . . . . . .... .. . . . . . . ...... 286
The Sprinkler . .... . . . .. .. ..... . . . ... . ..... .... .. . . . . . . 287
Rolling . . .. ... . ... .. .. .. . . ...... . .. . .. . . ... . .. . . . . ... . 289
Yo- Yo on the Floor . ........ . ..... . .. .. . . ....... .. . . . .. 295
Rolling Over an Edge . ............. . ............ .. . . ... 297
Determinism and Predictability . . ... . . ............. . . . ... 299
12.4 Problems ...... . . .. . .... .. ... . . .. .. ... ... . . .. .. . .. . ... 300
13. The General Motion of a Rigid Body .. . .......... . . .. ..... 313

13.1 Inertia in Rotational Motion ...... . . . ................... 313
The Dumbbell . . ... .... . . . .. . ....... . . .. ..... .. ... . .. .. 314
Flywheel on an Axis .. .. .. .. . ..... ... . . . ..... ... . . . .. : . 316
Precession of a Gyroscope .. . .. .. . . .. .. ....... .. ... . . . .. 318
13.2 The Inertia Tensor ..... . . . ...... . .......... .. . ... . .... 321
The Dumbbell Revisited . . .. ............................ 324
13. 3 Euler's Equations ...... ... ........ ... .... . ... . . .. . . . .. 326
13.3.1 D erivation of Euler 's Equations ...... .. .. . . . ... . .. 328
13.4 Kinetic Energy .... . .. . ... .. .... .. .. .. .. . .. . . .. .. . ..... 330
13.5 Determination of the Principal Coordinate System ...... . .. 331
Rotating Dumbbell . ................. . ......... . . . . ..... 333
Flywheel . . : . . .......... . ..... . .. .. . . . ........ . .. . .... 334
The Gyroscope . .. .... .. .. . . ... .. . . . .. . . ... . . . . . .. . .... 335
Gyroscope Supported at the Center of Mass . . .. .. .. ....... 340
The Earth as a Gyroscope .. . .... .. ... . ..... .. .. . . . . .... 341
13.6 Problems ................ . ........... . .... . ... . ....... 342
14. The Motion of the Planets . . . .. . .... .. . . . ..... . ..... .. .... . 345
14.1 Tycho Brahe ... . . .. . . .. . . .. ...... . . .. ... . . ... . .. .. .. . . 345
14.2 Kepler and the Orbit of Mars ... ..... . . .... . . ... . ... .... 346
14.2.1 The Length of a Martian Year .................... 347
14.2.2 The Orbit of the Planet Mars ......... ......... ... 349
14.2.3 Determination of Absolute Dist ance
in t he Solar Syst em . . ... .... . . .. ...... . . . . . .. .... 351
14.3 Conic Sections ... .. . .. . . . . ..... .... .. ... .... . .. .... ... 352
14. 4 Newton's Law of Gravity Derived from Kepler's Laws . ..... 355
14.5 The Kepler Problem ................ .. . . ....... . ....... 359
14.5.1 Derivation of Kepler 's 3rd Law
from Newton's L aw of Gravity . .. .. . .. . .. . .. .... . . 364
XVIII Contents
14.6 The Effective Potential .............................. 366

14.7 The Two-Body Problem ............................. 367
14.7.1 The Two-Body Problem and Kepler's 3rd Law ..... 370
14.8 Double Stars:
The Motion of the Heliocentric Reference Frame ......... 370
14.9 Review: Kepler Motion .............................. 372
14.10 Examples ......................................... 373
Planetary Orbits and Initial Conditions ................ 373
Shape and Size of Planetary Orbits .................... 374
Motion Near the Surface of the Earth . ................. 376
Velocities in an Elliptical Orbit ....................... 377
Hohman Orbit to Mars . ............................. 378
The Face of the Moon (Spin-Orbit Locking) ............. 381
14.11 Problems .......................................... 385
15. Harmonic Oscillators . ................................... 389

15.1 Small Oscillations ................................... 389
15.2 Energy in Harmonic Oscillators ....................... 391
15.3 Free Damped Oscillations ............................ 392
15.3.1 Weakly Damped Oscillations ................... 393
15.3.2 Strongly Damped Oscillations .................. 394
15.3.3 Critical Damping ............................. 394
15.4 Energy in Free, Weakly Damped Oscillations ............ 395
15.5 Forced Oscillations .................................. 396
15.6 The Forced Damped Harmonic Oscillator ............... 398
15.7 Frequency Characteristics ............................ 400
15.7.1 w« Wo: A Low Driving Frequency .............. 400
15.7.2 w» Wo: A High Driving Frequency .............. 400
15.7.3 w ~ Wo: Resonance ............................ 401
15.8 Power Absorption ................................... 402
15.9 The Q- Value of a Weakly Damped
Harmonic Oscillator ................................. 403
15.10 The Lorentz Curve .................................. 405
15.11 Complex Numbers .................................. 406
15.12 Problems .......................................... 408
16. Remarks on Nonlinearity and Chaos . .................... 411

16.1 Determinism vs Predictability ......................... 411
16.2 Linear and Nonlinear Differential Equations ............. 412
Superposition ...................................... 413
16.3 Phase Space ....................................... 414
The Simple Harmonic Oscillator ...................... 415
Phase Space of the Pendulum . ........................ 416
Bifurcation in a Nonlinear Model . ..................... 421
Contents XIX
16.4 A Forced, Damped Nonlinear Oscillator ................ 424

16.5 Liapunov Exponents .......... . ..................... 427
16.6 Chaos in the Solar System ........................... 429
16.7 Problems .......................................... 431
Appendix. Vectors and Vector Calculus ...................... 433
Selected References ..................................... . ... 439
Answers to Problems ... .. .... . ................... . ......... 441
Index ......... . ............................................ 449

1. The Foundation of Classical Mechanics
1.1 Principia
It was an important event in the history of physics when Sir Isaac Newton
in 1687 published his book Philosophiae Naturalis Principia Mathematica.
In this famous work we find a masterly synthesis of the concepts of motion
and force. The Newtonian formulation of the laws of motion has, with su-
perior strength and vitality, survived more than 300 years. Although certain
fundamental aspects of Newtonian physics have been revised in this century,
Newton's principles still find widespread use both in the basic and in the ap-
plied sciences. Just one example: the launching of artificial satellites is based
directly on these principles.
Newton's work forms the basis upon which theoretical physics as well as
modern engineering science rests. We shall here, without going into details of
the historical development, discuss some of the most important steps in the
evolution of the Newtonian world picture.
1.2 Prerequisites for Newton
There are two main lines of inquiry that lead towards and form the basis
for Newton's efforts: one based on motions here on Earth, and one based on
the motions of bodies in the heavens. The first line is associated with the
name C alileo Calilei (1564-1642) ; the other with Nicolas Copernicus (1473-
1543), Tycho Brahe (1546-1601) and Johannes Kepler (1571-1630). Before
Calileo, the "impetus" concept dominated the thinking on motion. This idea
is somewhat related to the later concept of momentum, but although it was
thought that a body was given a certain impetus when it was thrown, the
body was thought to spend its impetus during its flight. When all the initial
impetus was spent, the body would stop and - if it was not supported - fall
down.
As for the free fall, it had been held since Aristotle (384- 322 B.C.) that
a heavy object falls faster than a light object. From immediate everyday
experience, both the impetus concept and the idea that a heavy object falls
faster than a light one, are not unreasonable.
J. M. Knudsen et al., Elements of Newtonian Mechanics

© Springer-Verlag Berlin Heidelberg 2000
2 1. The Foundation of Classical Mechanics
It was Galileo who cleared away these misconceptions and in this way
erected one of the pillars on which the work of Newton rests. Let us here
state the laws that Galileo formulated:
(1) The law of inertia: If a body is left to itself without influence from
other bodies it will continue a uniform linear motion if it initially had
one, and will remain at rest if it initially was so.
(2) The laws of free fall:
(a) All freely falling objects will, when starting from rest, fall an equal
distance in the same time.
(b) The distance fallen, 5, is proportional to the square of the time of
fall: 5 = ~gt2 where 9 is a constant (g is the acceleration). According
to Galileo the constant acceleration 9 is independent of the nature or
composition of the falling object.
What was so revolutionary about Galileo's way of thinking; why is his
work today considered as the beginning of modern natural science? There
are two basic features which separate Galileo's work from earlier attempts to
formulate laws of motion.
First, there is the systematic use of experiments to decide what is true
and what is false. Before Galileo - and in particular in the Greek tradition
- the view was that the laws of nature could be obtained by pure specula-
tion. Galileo showed that experiments are of fundamental importance for our
comprehension of the laws of nature.
The other revolutionary feature of Galileo's work was that he took the
bold step of extrapolating to a "pure" or idealized motion by systematically
disregarding features of the motion that need separate analysis. Galileo re-
alized that friction stems from the surroundings and is not a fundamental
feature of the motion. In other words, no momentum (or "impetus") is lost
in motions where no friction is present - e.g. in the motion of the celestial
bodies through interplanetary space no friction is present. In free fall on the
Earth we can also - in a first approximation - disregard air resistance, if
we focus on small, heavy bodies and let them fall through a short distance
so that they do not achieve large velocities. The basis for the other main
line of thought leading up to Newton's work was laid down even before the
birth of Galileo, by the Polish astronomer Copernicus. In his great work De
Revolutionibus Orbium Coelestium (1543) he replaced a geocentric model of
the world with a heliocentric model.
The idea of a heliocentric system, i.e., a system with the Sun at the
center, had been put forward in antiquity, but it had been suppressed by the
Ptolemean teaching which was dominant up to the time of Copernicus. In
the Ptolemean world model the Earth is at the center of the universe. The
Earth is at rest, while the celestial sphere turns once every 24 hours around
an axis which intersects the celestial sphere near the North Star. This world
model could explain several features of the apparent motion of the Sun, the
Moon and the planets. With complicated models of uniform circular motions
1.2 Prerequisites for Newton 3
where circles roll on other circles (producing the so-called epicycles) many
features of the motions of celestial bodies could be quantitatively predicted.
Copernicus removed the Earth from its central position in the universe. In
the world system of Copernicus, the Sun is at the center of the universe. The
planets move in circles around the Sun. The planes of the planetary orbits
have slightly different positions relative to each other. One of the planets is
the Earth, which rotates around its axis once every 24 hours and revolves
around the Sun once every year. The Moon on the other hand moves in a
circle around the Earth. The stars are suns like our own Sun, and they are
assumed to be at rest in the universe. Perhaps Copernicus did not achieve
any significant simplification compared to the Ptolemean system, but he had
in a decisive way shaken the belief in an immovable Earth, and the axioms
of Copernicus show a deep understanding of motion in the planetary system.
One can say that Copernicus gave a qualitative description of motions in the
solar system.
While speculations on the merits of the two world systems were going on,
the Danish astronomer Tycho Brahe chose a different path, closely related
to that of Galileo. Brahe was of the opinion that the conflict should be re-
solved not by philosophical speculations, but through accurate observations to
produce a set of measurements that could form a basis from which to pro-
ceed. To this end, he himself built and designed the necessary instruments,
and from his observatories on the then Danish island of Hven he measured
- through more than 20 years - the positions of planets and fixed stars with
an accuracy of about 1 minute of arc. After a controversy with the Danish
king, Brahe set up residence in Prague as "Imperial Mathematician". Here
he worked briefly with the young Johannes Kepler. They met just at the turn
of the century, and when Brahe died nearly two years later, Kepler took over
the entire set of observations left by Brahe, and proved himself worthy of this
legacy. The confidence Kepler had in the precision of Brahe's measurements
was, as we shall see, a decisive factor in his work.
Kepler built on the heliocentric system of Copernicus and decided from
the measurements of Brahe to find the orbit (around the Sun) of the planet
Mars. He was assuming a circular orbit, but little by little he realized that he
had to incorporate an eccentric orbit in which the planet moves in an irregular
fashion. In his main work Astronomica Nova (1609) Kepler describes how the
theory of circular orbits gives rather good agreement with the observations
of Brahe. But then suddenly, without any transition, the following crucial
statement appears: "Who should have thought it possible? This hypothesis,
which agrees so well with the observed oppositions is nevertheless wrong" .
What had happened was that Kepler, by considering some of the most
accurate of Brahe's observations, had come across a deviation of about 8
minutes of arc from the positions computed on the assumptions of a circu-
lar orbit. At this turning point Kepler proves himself worthy of the legacy
of Brahe. "On these 8 minutes of arc I shall build a world" was Kepler's
comment. Let Kepler speak for himself:
But for us, who by divine grace has been given an accurate observer
like Tycho Brahe, for us it is only fitting to appreciate this divine
gift and put it to use ... From this point on I shall follow my own
ideas toward this goal. For if I had thought that I could ignore these
8 minutes of arc , I would have been patching my hypothesis. But as
it is not permissible to ignore them, these 8 minutes of arc show the
way to a complete reformation of astronomy; they are the building
blocks to a large part of this work.
We see here one of the summits of quantitative science - and we shall see
another equally grand example when we describe the work of Newton, towards
the end of this chapter. It is on this background one must see the life work
of Kepler. By peculiar routes, but governed by an almost infallible intuition,
Kepler produced, after having given up the idea of circular orbits, his three
famous laws of planetary motion:
(1) The orbit of a planet relative to the Sun lies in a fixed plane containing
the Sun, and each planet moves around the Sun in an elliptical orbit with
the Sun in one focus.
(2) The radius vector from the Sun to the planet sweeps out equal areas in
equal amounts of time.
(3) The square of the period of revolution of a planet is proportional to the
third power of the greatest semi axis of the ellipse. If, therefore, T denotes
the period of revolution and a the greater semi axis,
1 a3
T2 = Ca3 or T2 = C (1.1 )
where C has the same value for all the planets.
Kepler found the third law in 1618 and published it in Harmonicu8 Mundi
(1619). An enormous leap forward had been achieved in the short span of
years from 1601 to 1618. In these three elegant laws Kepler summarized all
of the enormous amount of data left to him by Brahe. We have here the first
quantitative kinematic description of the solar system, a description which
precisely tells how the planets move, but without pointing to a cause for that
motion.
Even though Kepler perhaps has in mind some of the mystic visions which
his work also contains, we must acknowledge the justifications of his cry of
joy after all these years of work:
What I dimly perceived 25 years ago before I had discovered the
5 regular solids between the celestial orbits ... ; what I 16 years ago
declared the end goal of my research; what made me spend the best
years of my life in studying astronomy, what made me join Tycho
1.3 The Masterpiece 5
Brahe and live in Prague - that I have now, by the grace of God ...
finally uncovered. After having sensed the first inkling of a dawn 18
months ago, daylight 3 months ago, but only a few days ago the
full daylight of a wonderful vision - then nothing can stop me now. I
revel in joyous bliss. I challenge all mortals with this open confession:
I have robbed the golden chalice of the Egyptians for with them to
make a tabernacle for my God far from the lands of Egypt. If you
forgive me I will rejoice. If you are angry, I can bear it. I have thrown
the dice and written the book either for my contemporaries or for
posterity. It makes no difference to me. I can wait a hundred years
for a reader when God has waited six thousand years for a witness.
Kepler did not have to wait a hundred years for a reader.
1.3 The Masterpiece

Armed with the enormous empirical material which is condensed in the laws
of Kepler and Galileo, Isaac Newton (1642-1727) managed to discover his
famous laws of motion and the general law of gravitational interaction. In
order to continue the historical review in this chapter we postpone a more
detailed discussion of the laws of Newton to Chapter 2. We are assuming that
the reader already has an elementary knowledge of these laws, so we will here
only briefly state Newton's three laws of motion:
(1) Law of inertia. A body which is not acted upon by any force will either
be at rest or in a state of uniform linear motion.
(2) Law of acceleration. The mass m of a body, times its acceleration a
equals the net force F on the body: ma = F .
(3) Law of action and reaction. If body A acts on body B with a force,
then body B will act on body A with an equal but oppositely directed
force.
The acceleration a of a body is a vector, as is the force F. A brief discussion
of vectors and vector calculus is provided in the Appendix of this book. Bold
face variables are used to designate vector quantities.
Newton arrived at the general law of gravity through an ingenious analysis
of Kepler's three laws. This analysis - and its result - is of great importance
in the history of the exact sciences. We shall here demonstrate the main
features of Newton's method. We restrict ourselves to the simple case of
uniform circular motion, so that the physical argumentation will stand out
as clearly as possible. But the method can, as we shall see later (Chapter 14),
easily be generalized to the more realistic case of an elliptical orbit.
The fundamental law of mechanics connects the force F acting on a body
with the acceleration of that body produced by the force:
ma=F. (1.2)
Newton realized that Kepler's three laws permit a calculation of the ac-
celeration that the planets undergo in their motion around the Sun. If we
assume the validity of (1.2) and compute the acceleration of a planet from
Kepler's laws, we can determine the force that acts on the planet. The deter-
mination of the force takes place in a sequence of steps.
(1) Use of Kepler's 1st law. Since a circular motion is a special case of
elliptical motion (having both foci in the center), the assumption of cir-
cular orbits with the Sun at the center is at least not in conflict with
Kepler's first law. Incidentally, this assumption gives a "reasonable" ap-
proximation to the actual orbits of the planets, which are nearly circular
orbits.
(2) Use of Kepler's 2nd law. Kepler's second law (the law of areas) de-
mands that the circle is traced out with constant magnitude of the ve-
locity, i.e., the motion must be a uniform circular motion. The analytical
expression for the acceleration in a uniform circular motion was first
derived by Christian Huygens (1629-1695). In Example 1.2 below, the
acceleration in a uniform circular motion is determined by vector anal-
ysis. The result is: the acceleration is directed towards the center (the
Sun) and has the magnitude
v2
a=- , (1.3)
r
where v is the constant magnitude of the velocity and r is the radius of
the orbit.
(3) Use of Kepler's 3rd law. In order to use the third law we must intro-
duce the connection between v, r, and the period of revolution, T. We
have, for a uniform circular motion,
v=r·
21lT
(1.4)
By eliminating v, we get from (1.3) and (1.4) that

41f2r
a = --;:y2 . (1.5)
From Kepler's third law we have:

r3
T2 =c (1.6)
where C is the same constant for all planets. Inserting T2 from (1.6) into
(1.5) , we find:
41f 2 C
a= - - 2 - . (1.7)
r
We see that the acceleration of a planet depends only on its distance from
the Sun, since the acceleration is inversely proportional to the square of the
distance of the planet from the Sun. The constant C does not depend on
the mass of the planet since C is the same for all planets. However, C may
depend on the mass of the Sun. We shall later demonstrate that (1.7) is valid
also for elliptical motions (where r is the time dependent distance between
the Sun and planet, see Chapter 14). From Kepler's three empirical laws we
have computed the acceleration of a planet as a function of its distance from
the Sun.
(4) Use of Newton's 2nd law. If we now finally calculate the magnitude
of the force on the planet, by inserting the acceleration (1. 7) into (1.2)
we have, with m the mass of the planet,
411" 2 C
F=m-- (1.8)
r2
The force is attractive, i.e., directed towards the Sun, and has a magnitude
inversely proportional to the square of the distance to the Sun.
By means of Kepler's laws and Newton's second law we have thus calcu-
lated the force acting on a planet. At this point Newton took a decisive step
by realizing that the acceleration, as expressed in Equation (1.7) has a very
important property in common with the acceleration occurring in free fall
near the surface of the Earth. In both cases the acceleration is independent
of the mass of the accelerating object.
If we calculate the magnitude of the gravitational force on an object of
mass m near the surface of the Earth, we find:
F=mg, (1.9)
where 9 is the acceleration of gravity near the surface of the Earth. The force
is - just as the force calculated above in (1.8) is - proportional to the mass
of the object. Starting from this observation, Newton took the bold step of
postulating that the two forces - the force pulling an object (say, an apple!)
toward the Earth, and the force holding a planet in orbit around the Sun
- are of the same physical nature. Both forces are expressions of gravita-
tional attraction. Newton connected the laws for falling bodies here on Earth
(found by Galileo) and the laws for the motion of celestial bodies (found by
Kepler). The final quantitative test, which is the touchstone for any physical
theory, remained. This quantitative test is of fundamental importance for the
development of Western civilization.
Newton chose to test his theory on the motion of the Moon. Before we
retrace the calculations of Newton, there is one problem we must describe
more closely: can we use the law of gravity at all, near the surface of the
Earth? Obviously, as shown in Figure 1.1, the different parts of the Earth
pull in different directions on a given particle A near the surface of the Earth,
and it is not clear what the net force on the particle will be.
8 1 . The Foundation of Classical Mechanics
Fig. 1.1. The gravitational interaction between a particle A and a spherically

symmetric body B. The body B acts on the particle A as if all of the mass of the
body was concentrated in the center
However, in Principia, Newton showed the important theorem that the

gravitational force from a body with a spherically symmetric distribution of
mass is the same as it would be if the entire mass of the body was concentrated
in the center. We will prove this theorem later in the course (see Section 8.6).
We return now to Newton's calculation. As mentioned, Newton did not
test his law of gravity on the motions of the planets around the Sun, but
on the motion of the Moon around the Earth. The reason was, that the
distance from the Earth to the Moon was known from direct measurements
(triangulations). The central body is now the Earth, and the Moon takes the
place of a planet. Let the orbital period of the Moon be T , and let r be the
radius in the circular orbit of the Moon. The radius of the Earth is denoted
p.
If the gravitational force on a body near the surface of the Earth has the
same origin as the force with which the Earth acts on the Moon, we must
have:
(1.10)
(compare with (1. 7)). G' has the same value for all objects acted upon by the
gravitational force of the Earth. G' depends on the mass of the Earth, just
as G in (1.7) depends on the mass of the Sun. We can therefore calculate the
constant G' by observing the motion of the Moon. From Kepler's third law,
(1.11)
Inserting this into (1.10) we get
(1.12)
T is the so-called sidereal month, i.e., the time between two consecutive
positions of the Moon when the line connecting the Earth and the Moon has
the same direction with respect to the fixed stars.
Note. From the direct measurements the quantities on the right-hand side of
(1.12) are known (and were known by Newton).
T 27 days 7 hours 43 minutes

r 60.1 p
P 6.37 X 106 m
Inserting these values into (1.12) we find:
gcalculated = 9.8 m s-2 .
This means that,

(i) by observing the Moon and analyzing its motion, and
(ii) by assuming gravity to be the common cause, both of the motion of the
Moon around the Earth, and of the fall of an apple to the ground,
we have computed the acceleration which gravity imposes on an object near
the surface of the Earth.
From direct measurements of the quantity 9 - for example from experi-
ments with pendulae and watches - we have
gmeasured = 9.8 m s-2 .
This beautiful agreement is rightfully considered a high point in the history

of the exact sciences. One can say that these considerations and calculations
represent the first piece of theoretical physics, the birth of the method of
natural science. This method has changed the conditions of living for man on
our planet; both spiritual and material conditions have changed so rapidly
in the three centuries since Newton that we can declare that it is as if an
entirely new species came into being on our planet in the years 1543-1687.
History tells us that when Newton first performed these calculations, the
deviations were so large that he considered his theory to be contrary to
the facts and rejected it. Some years later new measurements showed that
the astronomers had used a false value for the distance to the Moon. When
Newton learned about this, he redid the calculations and found the agreement
between the measured and the calculated value.
What remains is to find the final formulation of the law of general gravi-
tational attraction. It is difficult to assert how Newton arrived at this; that
feat is a combination of systematic analysis and intuitive insight. A possible
route is the following: we saw that the Sun acts on a planet with a force F
directed toward the Sun and with the magnitude
47r 2 C
F=m--
r2
(1.13)
where C depends on the mass of the Sun. If a law of general gravitational

attraction exists, the planet must act on the Sun with a force F' of magnitude
p' = M47r 2 e (1.14)

r2 '
where M is the mass of the Sun and e depends on the mass of the planet.
If we assume the validity of Newton's third law (action = reaction), then
the force F with which the Sun acts on the planet must be equal in magnitude
(but oppositely directed) to the force F' with which the planet acts on the
Sun. We then have
(1.15)
or,
e e
(1.16)
M m
Let us define a constant G from the common value of elM and elm:
e e G
(1.17)
M m 47r 2 '
from which
47r 2 e = GM, (1.18)

47r 2 e = Gm. (1.19)
If we insert (1.18) and (1.19) into (1.13) and (1.14), respectively, Newton's
law of attraction of mass assumes the following symmetrical form:
P=G mM (1.20)
2 r '
where G is a universal constant. The numerical value of G depends only on

the choice of units. The currently accepted value is
G = 6.668 X 10- 11 N m 2 kg- 2 .
Newton generalized the above result to: two arbitrary particles attract each
other with a force proportional to the mass of each of the two particles and
inversely proportional to the square of the distance between the particles.
The constant of proportionality is the universal constant G.
For over 200 years this generalization has resisted innumerable tests. Its
area of validity is enormous. It is gravity which determines the fall of the
apple toward the ground, guides the Moon in its orbit around the Earth, and
the Earth and the planets in their orbits around the Sun. Gravity directs
the motion of stars in the galaxy, and each galaxy acts with gravitational
forces on other galaxies. It is in this general form - that any object in the
universe attracts all other objects in the universe - that the gravitational law
of Newton is seen as an important step forward. In its original form (1.8), it
is derived from the laws of Kepler and is merely a short and precise (although
surprisingly simple) way to express these laws. As is well known, Newton's
law of gravity was modified in the beginning of this century, by Einstein's

general theory of relativity.
Newton discovered his laws through a combination of intuition and sys-
tematic analysis of observed facts. These facts are elegantly and precisely ex-
pressed in the laws of Galileo and Kepler. One can ponder how the "detached"
measurements of planetary positions by Brahe - through the enormous com-
putational effort and unique intuition of Kepler - became the foundation of
Newton's laws of motion and in this way was decisive in creating one of the
pillars of modern science. Below follow some examples of the use of Newton's
laws applied to the study of uniform circular motion in the gravitational field
of a spherical distribution of mass.
Example 1.1. The Acceleration of Gravity. Calculate the acceleration

of gravity g at the surface of the Earth given the mass of the Earth M and
radius p. Assume that the Earth has a spherical distribution of mass.
Solution. Consider a particle with mass m placed near the surface of the
Earth. The gravitational force F = IF I on the particle is directed toward the
center of the Earth and has the magnitude F = mg. We shall now use the
fact that the gravitational force outside the Earth is the same as if all the
mass of the Earth was collected in the center. For this reason we can write
F as:
Therefore
mM
mg=G-2- ,
p
M
g=G 2 ,
p
By inserting numerical values (from the table on the inside cover of this
book), we obtain
9 = 9.8 ms- 2 .
As is well known, g is sometimes called the gravitational field strength at the

surface of the Earth. The vector g is directed toward the center of the Earth.
We have:
g F/m;
g i g 1 is therefore equal to the force per mass;
9 9.8 Nkg- 1 = 9.8 ms- 2 .
Remember: a force of 1 Newton (1 N) gives to a mass of 1 kg an acceleration
of 1 ms- 2, i.e., 1 N = 1 kgms- 2.
Question. Compute the gravitational field strength on the surface of the
Moon. [Answer: 1.6 N kg- 1 = 1.6 m s-2.] Compute the acceleration of gravity
on the surface of the planet Mars. [Answer: 3.7 Nkg- 1 = 3.7 ms-2 .] 6.
Example 1.2. Circular Motion
[See the Appendix for a review of vectors.] Let us consider a point P in

uniform circular motion. This is described by the radius vector r given by:
r = r(t) = r cos(wt)i + r sin(wt)j
Here, i and j are time independent unit vectors, forming an orthogonal coor-
dinate system (i· j = 0, i 2 = j2 = 1), w is independent of time and is called
the angular velocity. r = 1r I·
(1) Find the velocity v(t) of P and the acceleration a(t) of P.
(2) Show directly that v(t) is always perpendicular to r(t) (i.e., the velocity
has no component along the radius). Show that v(t) is always perpen-
dicular to a(t).
(3) Show that
4n 2
a(t) = - T2 r(t),
where T is the period of revolution (i.e., the acceleration is directed to-
wards the center).
(4) Show that

v = Ivl = rw,
and that
v2
a= lal = rw 2 = -r .
Note. Sometimes the word "speed" is used to designate the magnitude of the
velocity; i.e., the speed is equal to Iv I.
Solution.
(1) v(t) = dr/dt = -rwsin(wt)i + rw cos(wt)j,
a(t) = d 2 r/dt 2 = -rw 2 cos(wt)i - rw 2 sin(wt)j.
(2) v(t) . r(t) = -r 2w sin(wt) cos(wt)i . i + r 2w cos(wt) sin(wt)j . j = 0 since
i . i = j . j = 1. In the same manner, v(t) . a(t) = O.
(3) a(t) = -w 2(rcos(wt)i + rsin(wt)j) = _(41f2/T2)r(t) = -w 2r(t) since
wT = 21f.
(4) v=(v;+v~)1/2=rw, a=rw 2 =v 2 /r.
Question. A satellite moves in a circular orbit around the Earth. The period of
revolution is 98 minutes and the height over the surface of the Earth is 500 km.
From this, calculate the mass M of the Earth. [Answer: M = 5.5 X 10 24 kg.
The accepted value is 5.98 x 10 24 kg.] The mass of a planet can be inferred
from the motion of its satellites. Using the artificial satellites in orbit around
the Earth, we have improved our knowledge of the mass distribution of the
Earth. I'::.
Example 1.3. Communication Satellite. A satellite moves in a circular

orbit in the equatorial plane of the Earth. The satellite has such a height
h above the surface of the Earth that it remains vertically above the same
point of the Earth (communication satellite).
(1) Find h.
(2) What is the acceleration of the satellite? With respect to what?
Solution.
(1) The period T is apparently 24 hours. Let the mass of the satellite be m
and that of the Earth be M. The distance of the satellite from the center
of the Earth is r. The equation of motion for the satellite is Newton's
second law with gravity as the force:
mM
ma= G-2-.
r
Since
we find
GMT
r3 = _ __
2
47r 2
Inserting numerical values, one finds:
r = 4.22 x 10 7 m.
The radius of the Earth is: p = 6.37 X 10 6 m. Therefore:
h = r - p Rj 35.8 X 10 6 m = 35800 km.

Another procedure is to use Kepler 's third law directly, together with our
knowledge of the orbital period of the Moon and its distance from the
Earth:
moon: Tm 27.3 days,

rm 60.1 p, where p = radius of the Earth.
satellite: Ts 1 day,
rs p+h.
We then have
(60.1p)3
(27.3)2 =} h = 5.6p.
(2) The acceleration of the satellite with respect to a coordinate system fixed
to the surface of the Earth is of course zero, and the velocity with respect
to such a coordinate system also has the constant value zero. The accel-
eration with respect to a coordinate system with its origin at the center
of the Earth and axes fixed in relation to the fixed stars, is:
GM 9 9
a
(p + h)2 (1+h/p)2 (6.6)2
0.23 ms- 2 .
The difficult problems suggested in question 2 of this example will be dis-

cussed in the following chapters.
Question. A satellite moves in a circular orbit close to the surface of a planet.
The orbital period is T and the mass density of the planet is d. The planet is
assumed to have a spherical distribution of mass. Show that dT 2 is a universal
constant. f'::,
Example 1.4. Horizontal Throw. When deriving the law of gravitation

Newton compared the fall near the Earth (of an apple!) with the fall of the
Moon toward the Earth. Newton assumed that there is no essential difference
between a fall observed near the surface of the Earth and the motion of the
Moon around the Earth. This point was considered carefully by Newton.
Newton imagined an object thrown with horizontal initial velocity from a

tall mountain. With larger and larger initial velocity, the object will reach the
Earth further and further from the mountain. See the figure . Newton posed
to himself the problem of finding whether it is possible to throw the object
with such a large initial velocity that the object although falling under the
influence of gravity, never comes closer to the Earth.
Let us assume that the object is thrown with the horizontal velocity Vo.
We will investigate what condition Vo must satisfy for the object never to
reach the Earth in spite of the fact that it falls according to the law of free
fall, y = ~ gt 2 , where y is the vertical distance fallen. Consider the figure
below.
In 1 second the object falls the distance Yo = ~gt2 = ~g(1 second)2. The
figure is grossly exaggerated; in reality Yo « p, since p is roughly the radius
of the Earth, and Yo = ~g(1 second)2 ~ 5 m.
While the object has fallen Yo = ~g(1 second)2 , it has moved the distance
Vo x (1 second) forward in the inertial motion. From the figure,
Let (vo x l) = d,
p2 + d2 P
222
+ Yo + PYo·
Using Y5 « 2pyo, we find
d2 2pyo·
With d = (vo x 1), and Yo = ~9t2 = ~9(1)2, we obtain
2 1
Vo = 2P'29,
or
Vo =,;pg.
With this value of Vo as the initial velocity, the body will continue uni-
formly along a circle. By inserting the numerical values,
p 6.4 X 106 m,
9 9.8 ms- 2 ,
we get
Vo;:::O 8 X 10 3 ms- I .
Thus: if we could throw from the top of the highest mountain on Earth a
rock with a horizontal initial velocity of about 8 km s -1, the surface of the
Earth would "curve away" from the orbit of the rock in such a way that it,
although falling toward Earth, would never approach it.
The same result can of course be inferred from direct application of New-
ton's second law and the law of gravity. The object must under the influence
of the gravitational field of the Earth describe a uniform circular motion with
radius p where p;:::o radius of the Earth.
Question. Show directly from Newton's second law and the law of gravity
that Vo = ypg and calculate the period of revolution. [Answer: T = 1 hour
24 min.]
We will conclude this example by calculating the acceleration of the Moon
toward the center of the Earth, when the motion of the Moon is considered
as a ceaseless fall towards Earth. We shall employ the above results. In the
figure, p now stands for the radius of the lunar orbit, and we put p = r.
We let S denote the arc along which the Moon moves during 1 second. S
is almost a chord. Yo denotes the amount the Moon has fallen in 1 second.
We have Yo « S (see below). From the results above we infer that
S2
Yo = - [note: S ;:::0 vo(1 s)].
2r
The distance S through which the Moon moves in 1 second, can be calculated
from the values:
radius of lunar orbit: r 3.8 X 10 8 m,

lunar period: T 2.36 x 10 6 s.
We find:
21fr
S = T(1 second) ::::::: 1011 m.
We can now calculate the distance Yo through which the Moon falls in one
second.
(1011)2 -3
Yo = 2 x 3.8 X 108 = 1.34 x 10 m.
We see that the condition Yo « S is consistent. From kinematics (i.e. Galileo's
law of free fall) we know that
1
y = -at 2
2
In our case a is the acceleration in the fall of the Moon. In fact,
We can now determine a:
a = 2Yo(1) -2 = 2.68 X 10- 3 m s-2 .
From Galileo's law of free fall and from an analysis of the motion of the Moon,
we have calculated the acceleration in the fall of the Moon. (This acceleration
can also be found as: a::::::: 9.8/(60.1)2::::::: 2.7 x 10- 3 ms- 2. Why?).
Question. Calculate directly from the law of gravity the field strength in
the gravitational field of the Earth at the position of the Moon, or in other
words: calculate the acceleration in the fall of the Moon directly from the law
of gravity. [Answer: a = (GM/r 2 ) = 2.68x 10- 3 Nkg- 1 = 2.68x 10- 3 ms- 2 ]
6
Since the Moon falls freely in the gravitational field of the Earth, there is
no fundamental difference between the fall of the Moon and the free falls we
observe daily at the surface of the Earth. The Moon has simply "been given"
an initial velocity perpendicular to the gravitational field of the Earth and of
such magnitude that in spite of perpetually falling towards the center of the
Earth, the Moon never comes closer to the surface of the Earth.
The contribution of Newton to the understanding of gravitational phe-
nomena must especially be seen in the light of his inferral of the connection
between the mutual attraction of celestial bodies and the fall of bodies near
the Earth. This connection is by no means trivial. The Aristotelean ideas in
their time postulated a fundamental difference between the motion of heav-
enly bodies and bodies on the Earth.
Example 1.5. The Gravitational Constant. Newton did not know the
value of the gravitational constant G. If one assumes an approximately spheri-
cal Earth of radius p, one gets for the force on a particle of mass m,
Mm
mg=G-2- ,
p
where M is the mass of the Earth. Since he knew p, Newton could find the
product GM, by measuring the acceleration of gravity.
From this he could find the force on a particle at an arbitrary distance r
from the center of the Earth:
F = G Mm = G Mmp2 = mg
r2 p2 r2 r'
(!!.)2
In a famous passage in Principia Newton estimates the average density of
the Earth: "it is probable that the quantity of the whole matter of the Earth
may be five or six times greater than if it consisted of water" .
If Newton had used a value of, say, 5.5 gcm- 3 = 5.5 x 103 kgm- 3
for the average density of the Earth, he could - knowing the radius to
be p :::::: 6371 km - have calculated G. What value would he have found?
[G = (3g)j(47rpd), where d = density]
From laboratory measurements, H. Cavendish (1731-1810) made an in-
dependent determination of G. The currently accepted value of Gis:
G = 6.67 X 10- 11 Nm 2 kg- 2 .

Our knowledge of the masses of the objects in the solar system rests entirely
on the determination of G. 6.
Example 1.6. String Force
An object of mass m hangs at rest in the laboratory. The object is fastened

to the end of a string, the other end of which is fixed to the ceiling of the
laboratory. The acceleration relative to the laboratory of the object is nil.
From this it follows that the sum of the forces acting on the object is nil.
The string force S which acts on the object has the direction as shown on
the figure, and the magnitude is IS I = mg. This force is a result of a small
stretching of the string under the weight of the object. If the prolongation of
the string is so small that we can neglect it, the string is called unstretchable.
l.3 The Masterpiece 19
Question. A boy whirls a string of length 2 m around in an (almost) horizontal

plane. At the end of the string a mass M = 200 g is fastened. The time of
revolution is T = 0.4 s. Determine the magnitude F of the string force and
compare F to the magnitude of the gravitational force on the mass.
Solution.
0.2 x 2 x 4 X /T 2
42 ~ 98.7 N.
o.
F
F grav = 0.2 x 9.8 ~ 2 N; thus, -- ~50.
Fgrav
Example 1.7. Forces and Tension
'////
L - x
Consider a homogeneous rod of mass M and length L. The rod hangs freely
in the gravitational field of the Earth (see the figure). Consider"the hatched
segment of the rod. This segment has length x. Imagine a cross section divid-
ing the two parts of the rod. The upper part of the rod (L - x) acts on the
lower part (x) with contact forces evenly distributed over the cross section
of the rod. The tension in the rod is defined as the force per unit area in the
cross section. The tension in a particular point of an elastic body is described
by a tensor and is treated in the theory of elasticity. The important point
here is: at the cross section the segment L - x acts on the hatched segment
with a vertical force. This force has the magnitude:
F= Mgx
L .
(the segment x has acceleration 0 and is being pulled downwards by gravity).

At the cross section the segment x acts on the segment L - x with a downward
force. Again, this force has the magnitude
F= Mgx
L .
If our system is a freely hanging string instead of a rod, the quantity
T(x) = Mgx
L
is often - particularly in theoretical mechanics - called the string tension.
T(x) is not a tension, but a force. Usually, confusion is avoided since in
(almost) all cases, the quantity in question in problems of mechanics is the
force with which a string acts on a body. 6
Example 1.8. Dimensional Analysis
'-- --- M
A particle of mass M is fastened to one end of a string. The length of the

string is L, and we neglect the mass of the string. This system, which is
called the mathematical pendulum, can perform oscillations around the ver-
tical equilibrium position. The acceleration of gravity is g. What physical
quantities could the period T of the oscillation depend on?
The period might depend on the mass M of the particle, the length L
of the string, the magnitude 9 of the gravitational field, and perhaps on the
maximal angle () of deflection. The angle () is a dimensionless number (arc
length divided by L).
The period T has the dimension of time (s), M has the dimension of mass
(kg), L has the dimension of length (m), and 9 has the dimension of distance
per time squared (ms- 2 ). We assume that these three quantitites, raised to
various powers, enter the formula for the period:
Or, in terms of units,
s mX(kg)Y (~r
mX+Z(kg)Ys-2z.
From this:
x+ z 0,
y 0,
-2z 1,
1.4 Concluding Remarks 21
which leads to x = 1/2, Y = 0, Z = -1/2. We conclude that
T=C/f,
where C is a constant (a pure number) with no dimensions associated. The
value of C cannot be determined by this type of analysis.
As we will see in Section 3.4, for small oscillations about the equlibrium
position the period is independent of the maximal angle, and C has the value
27r. For large deflections, the value of C depends on the maximal angle of
deflection. Since this angle, e, is a dimensionless number, the general formula
above still holds. !::::.
1.4 Concluding Remarks
The main purpose of this book is to describe the problem of motion as this
problem was formulated by Newton. We also seek to clarify the content of
the modified version of Newtonian mechanics which is normally used today.
In order now to suggest some fundamental problems in the physics of
motion we will formulate a few questions, and let them stand unanswered.
The questions will be discussed in detail in the following chapters.
(1) Relative to which coordinate system should we measure the velocities
and accelerations that appear in Newtonian theory?
(2) Relative to what do the galaxies rotate? The figure following the title
page of this book shows a galaxy. A galaxy is a collection of hundreds of
billions of stars held together by mutual gravitational forces, just like the
planetary system is held together by gravitational forces. Looking at the
picture, one can sense the rotation, but is it, as Newton thought, a rota-
tion with respect to "absolute space"? And if so, what is this "absolute
space"?
(3) We say that the Earth rotates once every 24 hours. With respect to what
does the Earth rotate? Is it relative to the distant stars? Or, is it relative
to "absolute space"?
An answer to such - and similar - questions demands a deep understand-
ing of what it really means to say that a body moves. We shall see that in
the form in which we usually employ Newtonian mechanics, it is not easy to
gain insight into this question.
1.5 Problems
Problem 1.1.
A boy whirls a string of length 2 m around in an (almost) horizontal plane.

At the end of the string a mass M2 = 0.1 kg is fastened, and 1 m from the
boy's hand another mass Ml = 0.1 kg is fastened. The boy increases the
angular velocity w in the motion of the string. The hand of the boy and the
two masses will, when w is sufficiently large, be almost on a straight line.
When the string force reaches the magnitude 327 N, the string will break.
(1) Find the magnitude of the angular velocity when the string breaks, and
find where the break occurs.
(2) Is it possible to whirl the two masses in such a way that the two masses
on the string are exactly on a horizontal line?
Problem 1.2. The planet Mars has two moons, Phobos and Deimos.
(1) Phobos has an orbital radius of 9.4 x 103 km and a period of 7 h 39 min.
Calculate the mass of Mars from these data.
(2) Assume that the Earth and Mars move in circular orbits around the Sun.
Take the radius of the orbit of Mars around the Sun as 1.52 times the
orbital radius of the Earth. Calculate the length of the Martian year.
Problem 1.3.
"
M
1.5 Problems 23
A body of mass M is hanging in equilibrium in the laboratory, as shown on

the figure. The two strings have the same length. Determine the string force
S in the two strings.
Problem 1.4.
A heavy particle with mass m is attached to a rigid rod by means of two thin,
"unstretchable"strings, each oflength l. As the strings are unstretchable, one
can ignore their stretching during the motion. The entire system rotates with
constant angular velocity w about an axis along the rod, and both strings are
taut. The acceleration due to gravity is g.
(1) Calculate the string force in each of the strings.
(2) Find the condition under which the lower string is taut.
Problem 1.5.
A particle of mass m performs a uniform circular motion on a frictionless

table. The particle is attached to a cord that passes through a hole in the
table. The other end of the cord is attached to another particle with mass
M > m. The length of the cord on the table is T. Find the magnitude of the
velocity of m such that the system is in equilibrium.
Problem 1.6.
m
...... ....
A heavy particle with mass m is tied to one end of a string, which has
the length l. The other end of the string is attached to the ceiling of the
laboratory. The particle performs a uniform circular motion in a horizontal
plane. The angular velocity is w. The string is at an angle e to the vertical.
This system is called the conical pendulum. Ignore the various forms of friction
and the mass of the string.
(1) Find the magnitude of the string force S expressed in terms of m, l, and
w.
(2) Express e as a function of g, l, and w. Examine the result for w ---t 00 .
For the conical pendulum we have to assume that w > V7ifl. In the derivation
of the result of question (1) it is necessary to divide by sin e.
(3) State the value of sin e for w = V7ifl.
Problem 1.7.
A cone of revolution has sides that form an angle of 30° to the vertical. One
end of a thin cord with length l is attached to the apex of the cone. At the
other end of the cord a heavy particle of mass m is attached.
The particle performs a uniform circular motion with angular velocity w
in such a way that the particle stays in contact with the cone at all times.
1.5 Problems 25
The mass of the string is ignored as well as friction between the particle
and the cone, i.e., the reaction force from the cone - acting on the particle -
is perpendicular to the surface of the cone.
(1) Find the string force S and the force R with which the surface of the
cone acts on the particle (R is the reaction force).
(2) Find the largest value of w for which the particle can remain in contact
with the surface of the cone.
Problem 1.8. If you are not familiar with vectors, we suggest that you read
the Appendix carefully. A vector v has in a given coordinate system the
component Vy = 2. The vector v forms an angle of 45° with the positive
x-axis. The projection of the vector on the xz-plane forms an angle of 60°
with the positive z-axis. Find
(1) the components Vx and v z ,
(2) the length v == Iv I of the vector,
(3) the angle between v and the positive y-axis,
(4) the angle between v and the positive z-axis.
Problem 1.9. Find the angle between any two space diagonals of a cube.
Hint: find in a suitable coordinate system the components of the vectors
that represent the two body diagonals, and use the dot product (see the
Appendix).
Problem 1.10. A particle moves on a straight line with the constant velocity
v = (1,1,2) ms-I. At the time t = 0 it passes the origin. A person P is
situated at the point (1 , 2, 3) m. Find the time at which the particle is
nearest P and find the vector d from P to the particle at that time (see the
Appendix).
Problem 1.11. See the Appendix. In a given coordinate system two vectors
a and b have the following components:
Show that the components of their vector product are given by:
(a x b)x ayb z - azby ,

(a x b)y azb x - axb z ,
(a x b)z axby - ayb x .
(Hint: use the rules for vector products, and a = axi + ayj + azk , where i,j,
and k are unit vectors along the x, y, and z axis, respectively.)
Problem 1.12. At the time t = 0 a pendulum of length r = 0.7 m is started

from rest at an angle of 30° from the vertical.
(1) What is the component of the acceleration along a tangent, called the
tangential acceleration aT, and the component of the acceleration along
the radius, called the centripetal acceleration aN, at the time t = 0 ?
(2) What is the angular velocity wand the angular acceleration w = jj at
t = O?
(3) The mass of the pendulum bob is 0.5 kg. What is the tension in the string
(the string force) at t = 0 ?
2. Newton's Five Laws
As mentioned, we will here assume some knowledge of elementary mechanics

and the concepts within that theory. As the reader progresses through the
book, he or she will find these concepts treated in substantial detail.
2.1 Newton's Laws of Motion
Newton introduced the basic laws of mechanics in an axiomatic form with

associated lemmas and definitions. We reproduce the laws here in forms that
closely resemble Newton's original version.
Newton's 1st law (the law of inertia):
A body remains in its state of rest or in uniform linear motion as long as no

external forces act to change that state.
This law puts the state of rest and the uniform linear motion on an equal foot-
ing. No external force is necessary to maintain uniform motion. The motion
continues unchanged due to a property of matter we call inertia.
Newton made this law precise by introducing definitions of the concepts
momentum and mass. Momentum is the product of the velocity of a body
and the amount of matter the body contains. That amount of matter is called
the mass of the body and is a measure of its inertia. Instead of "mass" we
shall occasionally use the more precise term "inertial mass". The inertial
mass then, is a measure of that property of an object which makes the object
resist changes in its state of motion.
We can write
p=mv, (2.1)
where m is the inertial mass of the particle, v its velocity, and p the mo-
mentum. By "=" we mean "equal to, by definition". Mathematically, we can
express the law of inertia as
no external force =} p = constant vector. (2.2)

28 2. Newton's Five Laws
In this way the law of inertia coincides with the law of conservation of mo-
mentum for a particle.
It is by no means simple to grasp the physical content of the law of
inertia. We have (almost by definition) no experience with objects that are
not subject to some external influence. We can imagine placing an object,
not acted upon by any force, in an otherwise empty astronomical space. Can
we later by any sort of observation or experiment decide if that object is at
rest, or in a state of uniform motion? The answer to - or rather an analysis
of - this question is a central theme in this book.
Newton's 2nd law:
The change in the momentum of a body is proportional to the force that acts
on the body and takes place in the direction of that external force.
By change one means here the time derivative of the momentum. The law
can be expressed as
dp
F, (2.3)
dt
where F is the impressed force. If we assume that the mass of the body is
constant, Newton's second law takes the form
ma=F, (2.4)
where a == dv / dt == v is the acceleration of the body.

In a given (orthogonal) cartesian coordinate system, Newton's second law
can be written in component form (see the Appendix):
d2 x
m dt 2 = Fx ,
d2 y
m dt 2 = F y ,
d2 z
m dt 2 = Fz .
In the form (2.4) the law states that an impressed force F causes an acceler-
ation a which is directly proportional to that force; the constant of propor-
tionality IF I / Ia I is the inertial mass m of the object.
The whole concept of force has been the subject of much debate since it
was introduced by Newton. Let us here note the following: the second law
[(2.3) or (2.4)] should not be considered as the definition of the concept of
force. An essential feature of the law is that the force acting on the particle
is supplied by a force law separate from (2.3) or (2.4). One example of such
a force law is the law of gravity. Other forms are shown throughout the
examples in this chapter.
2.1 Newton's Laws of Motion 29
The second-order differential equation that results when some force law
is supplied to the second law is called the equation of motion for the particle.
An important property of the concept of a force acting on a given object
is that the force has its origin in some other material body.
Thus, Newton introduced the concept of mechanical force as the cause
of the acceleration of an object. This causal description of the motion of an
object constitutes what we call dynamics.
In this way a connection is set up between acceleration (a kinematic quan-
tity) and the force (a dynamical quantity). In Chapter 7 we shall see that this
coupling of a dynamical and a kinematic quantity was of enormous impor-
tance for the understanding of the problem of motion, i.e., for the realization
of what it means to say that a given body is in motion.
Newton's 3rd law (the law of action and reaction):
If a given body A acts on another body B with a force, then B will also act
on A with a force equal in magnitude but opposite in direction.
This law (postulate) is almost self evident when we consider contact forces
in equilibrium. The fact that the law of action and reaction also holds for
actions at a distance, is demonstrated for instance by the existence of tides:
the Earth keeps the Moon in orbit with a gravitational field, but the Moon
acts back on the Earth with a force which (among other things) is the cause
of ebb and flood in the oceans.
Water
Fig. 2.1. Illustration of Newton's third law
If the law of action and reaction did not hold, one could assemble systems
that perpetually increased their velocity through the action of internal forces.
Newton tested the law for a special configuration of objects. On a wooden
plate he attached a magnet and a piece of iron. The plate floats on a calm
surface of water. The magnet acts on the iron with a force directed to the
left (see Figure 2.1) and the iron acts on the magnet with a force towards
the right. The two forces are apparently equal in magnitude and the system
remains at rest.
It is important to realize that Newton's first three laws are without con-
tent unless a coordinate system is specified, relative to which velocities and
accelerations are to be measured. To the three laws Newton therefore added
two postulates. We quote here from the translation into English of Newton's
original Latin text [translation by Andrew Motte (1729), the emphasis is
ours].
Newton's 4th law (the postulate of absolute time):

Absolute, true, and mathematical time, of itself, and from its own
nature, flows equally without relation to anything external, and by an-
other name is called duration: relative, apparent and common time,
is some sensible and external (whether accurate or unequable) mea-
sure of duration by the means of motion, which is commonly used
instead of true time; such as an hour, a day, a month, a year.
Newton's 5th law (the postulate of absolute space):
Absolute space, in its own nature, without relation to anything exter-
nal, remains always similar and immovable. Relative space is some
movable dimension or measure of the absolute spaces; which our
senses determine by its position to bodies; and which is commonly
taken for immovable space; such is the dimension of a subterranous,
an arial, or celestial space, determined by its position in respect of
the Earth. Absolute and relative space is the same in figure and mag-
nitude; but they do not always remain numerically the same. For if
the Earth, for instance, moves, a space of our air, which relatively
and in respect of the Earth remains always the same, will at one time
be in one part of the absolute space into which the air passes; at an-
other time it will be in another part of the same, and so, absolutely
understood, it will be continually changed.
Newton was well aware of the fact that if his three laws were to have a real
content, he had to explicitly point to a coordinate system relative to which
all velocities and accelerations ultimately should be referred.
Newton did not choose a material body, say the Sun or the "fixed" stars,
as the reference for motion. As the basic system of reference, Newton picked
"absolute space" or "astronomical space". Newton has often been criticized
for the two postulates above. In particular the phrase "without relation to
anything external" has been widely discussed. For, what exists without rela-
tion to anything external, cannot be verified by any experiment.
A deep understanding of the postulates of absolute time and absolute
space will follow only from experience with concrete examples of motion. The
following chapters are a means to that end.
In Chapter 7 we will give a more unified discussion of the five fundamen-
tal postulates in the Newtonian world picture. By then the reader should
2.2 Integration of the Equation of Motion 31
understand both the power and the limitation of the Newtonian mechanics.
We shall see that Newtonian mechanics - in spite of its enormous success -
contains internal weaknesses.
For now we will stick to a presentation of Newtonian mechanics as it is in
practical use today. Later we can enter a discussion of what it is all about.
2.2 Integration of the Equation of Motion
Newtonian dynamics is a powerful tool for investigations of Nature. In order

that you fully realize this, it is of paramount importance that you actively
command essential elements of the theory. That means, you must be able,
in an independent manner, to apply the theory to specific examples. In this
section we will work out a number of such examples of the integration of
Newton's equation of motion. We will consider the motion of a particle. By
the word "particle" we mean a body whose spatial extensions are unimportant
for the problem at hand. We shall furthermore assume that the coordinate
systems referred to are reference frames in which the law of inertia holds.
Such reference frames are called inertial frames. Referred to inertial frames,
Newton's laws apply in their simple classical form. How to find inertial frames
at all is one of the major themes of this book.
Example 2.1. Constant Force
m K
o
• x
A particle of mass m moves along the x-axis of a coordinate system under

the action of a constant force K = IK I in the positive x-direction. Analyze
the motion; i.e., find the position of the particle as a function of time.
The equation of motion is obtained by inserting the force into Newton's
second law:
(2.5)
where K is constant. Let us write (2.5) as
m~ (dX)_K
dt dt - ,
or, since the velocity v of the particle by definition is given as v = dx/dt,

m dv = K . (2.6)
dt
The formal integration of equation (2.6),
v= J dV =
-dt
dt
JK
-dt
m'
gives
v= J~dt.
Loosely speaking: we multiply (2.6) on both sides with dt, which gives:
mdv = Kdt
or
K
dv = -dt. (2.7)
m
Integration of (2.7) gives:
or,
K
v(t) = - t + Cl , (2.8)
m
where Cl is a constant of integration. Equation (2.8) can be written in the
form
dx K
- = -t+Cl· (2.9)
dt m
We multiply both sides with dt:
K
dx = -tdt + c1dt . (2.10)
m
Integration of (2.10) gives
x = J~ + J tdt Cl dt ,
or
1K 2
X =--
2m
t + cl t + c2 , (2.11)
where C2 is a constant of integration.
Note. We have integrated twice. Equation (2.11) thus contains two constants
of integration. This is a reflection of the fact that Newton's second law is a
second-order differential equation. The integration of the equation of motion
(2.5) has not provided a complete solution of the problem from a physical
point of view. A complete solution requires specification of x as a function
of time. The polynomial (2.11) contains two unknown constants Cl and C2.
The force K determines only the acceleration of the particle, i.e., x is not
completely determined as a function of time by the force alone. In a specific
problem the two constants of integration, Cl and C2, are determined by the
initial conditions: the position and the velocity at the initial time, say, t = O.
Part of the strength of differential equations as they are applied to physical
problems is that they are insensitive to these arbitrary parts of the motion.
Let us assume that at the time t = 0 the velocity of the particle is Vo and
the position is Xo. By inserting these initial conditions into (2.8) and (2.11)
we find from (2.8) that Cl = Vo , and from (2.11) that Cz = Xo.
The complete solution is therefore:
1K z
X= --t +vot+xo.
2m
If the initial time is t = to rather than t = 0, we get
1K z
x = --(t - to)
2m
+ vo(t - to) + Xo .
This example applies to motions like the free fall (in vacuum) near the surface
of the Earth if the total vertical extension is so small that the variation with
height of the gravitational force can be disregarded.
Questions.
(1) A locomotive is driving with a velocity of 100 kmh- 1 on a straight rail-
road track. The mass M of the engine is 10 tons. At the time t = 0 the
engine begins to brake by blocking the wheels. This causes a frictional
force F to appear. We assume F to be constant.
(a) Write the equation of motion for the train at times t > O.
(b) Suppose the train comes to a halt after 400 m. Determine the mag-
nitude of F and the time tl at which the train stops.
[Answers: (a): Mdv/dt = -F; (b): note that x = ~ (-F/M)ti + voir
and v = 0 = -(F/M)tr + Vo ~ F = 9645 N; tl = 28.8 s].
(2) A stone is thrown vertically upwards in the gravitational field of the
Earth. A point P is at height h above the starting point of the stone.
The stone passes P on its way upward, 2 seconds after it was thrown,
and 4 seconds after it was thrown the stone passes P on its downward
fall. Calculate h and the initial velocity Vo of the stone. Disregard air
resistance. [Answers: h = 39.2 m; Vo = 29.4 ms- I .] What is the velocity
of the stone when it reaches its point of departure?
Example 2.2. The Harmonic Oscillator. A mass m is connected to one

end of a spring. The other end of the spring is attached to a wall. The mass
slides without friction on a horizontal table. The gravitational force on the
mass is therefore balanced by the reaction force from the table. If we compress
or stretch the spring, it will act on the mass with a horizontal force. It is
possible to construct the spring in such a way that the force acting on the
mass is proportional to the displacement of the mass and directed towards
Wall
x
Equlibrium position
the point of equilibrium. Such a force is often called an elastic force. Let us
choose an x-axis with its origin at the equilibrium position of the mass. We
can now write an analytical expression for the horizontal force acting on the
mass:
Fx = -kx, (2.12)
where k is a constant determined by the physical properties of the spring.
The force constant k has units N m -1; the minus sign indicates that the
force is directed against the displacement, towards the point of equilibrium.
The expression (2.12) is known as Hooke's law. We neglect the mass of the
spring. The equation of motion now follows by inserting (2.12) as the force
in Newton's second law:
(2.13)
In principle there are two more equations of motion, one for the y-direction
and one for the z-direction. However, we do not write these two (trivial)
equations, as the problem at hand is, from the point of view of physics, a
one-dimensional problem.
(1) What conditions must the constants A, w, and e satisfy for

x = Acos(wt + e) (2.14)
to be a solution to (2.13)7 (A, w, and e are constants, i.e., independent
of t).
(2) What is the physical meaning of A7
(3) Find the period T of oscillation, i.e., find the time between two consecu-
tive extremal positions.
(4) Show graphically in an (x, t) diagram the motion of the mass, using the
following initial conditions: the mass is pulled to the distance x = a from
the equilibrium point and released at t = 0 with no initial velocity.
Solution.
(1)
For X = A cos(wt + 0) we have:
~~ = -Awsin(wt + 0) ,
d2 x
dt 2 = -Aw 2 cos(wt + 0) .
For x = Acos(wt + 0) to be a solution to (2.13), we must have, for all
values of t,
m( _Aw2) cos(wt + 0) = -kA cos(wt + 0) ,
or
w={f. (2.15)
Therefore, for w given by (2.15) and for any value of A and 0, (2.14)
will satisfy the equation of motion (2.13). A and 0 are two constants of
integration (compare with Cl and C2 in Example 2.1). The integration
constants are determined by the initial conditions. Equation (2.14) is the
complete solution of (2.13). Physically this means that with a suitable
choice of A and 0, any motion of the mass can be represented by (2.14).
(2) A is the amplitude of the oscillating mass, i.e., the maximal deviation
from the point of equilibrium. The angle 0 is called the phase angle.
(3) T must satisfy the condition
cos(wt + 0) = cos[w(t + T) + OJ.

Since the function cos is periodic with the period 21l', we have
wT=21l',
or
21l' fm
T = -:;; = 21l'V k . (2.16)
Note. The motion described by (2.14) is called a harmonic oscillation

and the system is an example of a harmonic oscillator. We have the
important result: the period T in a harmonic oscillation is independent
of the amplitude. The frequency of the oscillation, i.e., the number of
oscillations per second is: v = liT and consequently w = 21l'v.
(4) x(O) = AcosO = a
v(O) = (~~) 1,=0 = -Aw sin(O) = 0 =} 0 = 0, (2.17)
i.e.,
cos(e) 1,
A a,
x a cos(wt).
The amplitude A = a is here set equal to 1:
····1· ··········;···········\··········,····· ······[·· .........:..... .

........ ...
:........
.... ; ............ ··· ·····1··········.;·····
. . ·
.
.. . .
. .
.
: . . .. ..
o ••••••• 1••••••••••: ••••••
;.••••••.••••
..:••••••••••••••••••• • ...;.·i· ................
••••••••• ;...... ,...............:......... .
i T / 4: T /2 : : 21t: :
-1 .... l. ..........:... ..: ........~ .: .~....({} .. :..... :
o t 2 4 8 10
Note that (2.13) is often written as:
d2x 2
dt 2 = -w x, (2.18)
where
2 k
w = -.
(2 .19)
m
Here w is a characteristic constant for the harmonic oscillator and is
called the cyclic frequency, the angular frequency, or the eigen frequency.
Question. Consider a particle with the mass 100 g = 0.1 kg. At time t =
o s,the particle is at the origin (x = 0) and the particle has the velocity
v = dx/dt =0 .5 ms-I. The period of oscillation is T = 2 s. Find the force
e
constant k, the phase at t = 0, and the amplitude A of the motion. [Answer:
e
k = 1r 2 /10 N m-\ = -1r /2; A = 1/21r m.] 6,
Example 2.3. Mass on a Spring in the Gravitational Field of Earth
We introduce a vertical (downwards) y-axis with its origin at the extension

of the spring when no mass is attached to it. When the mass has the position
y, the total force acting on the mass is given by
F = -ky + mg , (2.20)
where k is the spring constant. The equilibrium point of the mass is not
y = 0, since (2.20) shows that the force acting on the mass at y = 0 is not
zero, in fact F(y = 0) = mg. Let us determine the equilibrium position Yo,
i.e., let us determine for which value of y, F(y) = O. From (2.20) we have:
o -kyo + mg, (2.21 )

mg
(2.22)
Yo
T'
The equation of motion for the mass as it moves along the vertical y-axis is:
d2 y
m dt 2 = -ky + mg. (2.23)
Equation (2.23) can be written in the form
d2 y k
dt 2 + m (y - Yo) = O. (2.24)
In order to solve this differential equation, we introduce the new variable
x = y - Yo. (2.25)
From (2.25) we get

(2.26)
By inserting (2.26) and (2.25) into (2.24) we find that the problem has been
reduced to that of solving the differential equation
d 2x k
dt 2 + mX = 0, (2.27)
which is identical to (2.13) in example 2.2. The complete solution to (2.27)

is again
x = Acos(wt + 8). (2.28)
Re-introducing y as variable, we can write
y - Yo = Acos(wt + 8). (2.29)
Equation (2.29) shows that the mass will oscillate around the equilibrium
position given by (2.22), with the period
T= 27rj!!f.
Gravity in this case causes just a displacement of the equilibrium position
to Yo. This system therefore is equally well suited for the study of harmonic
oscillations as that of Example 2.2. 6
Example 2.4. Frictional Force: Sphere Falling Through a Liquid.

In the previous examples, the force has been dependent "at most" on the
position of the particle. We will now consider a case where the force depends
on the velocity of the particle. Consider the following physical system:
A small heavy ball (say, a lead shot) is sinking through a liquid. At not too
large velocities, the frictional force f with which the liquid acts on the ball
will be proportional to the velocity of the ball. Let us start the ball off in the
position x = 0 at time t = 0 and with velocity v = O. The mass of the ball is
m. Note, we neglect the motion of the fluid.
The gravitational force on the ball is in the positive x-direction and has
the magnitude mg. Because ofthe finite volume of the ball there is a buoyant
upward force on the ball with magnitude V pg, where V is the volume of the
ball and p the density of the liquid. These two forces, gravity and buoyancy,
are constant. Let us denote the sum of these by k (see the figure). The
magnitude of k is then
Ik I = Img - V pg I .
We now assume that the sphere is also acted upon by a frictional force pro-
portional to the magnitude of the velocity and directed against the motion,
that is:
f = -av,
where a is some positive constant.
Find the velocity of the particle as a function of time.
Solution. The equation of motion for the particle is
dv
m - =k-av (2.30)
dt
The velocity of the particle has no component perpendicular to the x-axis at
t = 0, and as all the forces act along the x-axis, the motion will take place
along the x-axis. We can thus write (2.30) as
dv
m dt = k - avo (2.31 )
Without integrating (2.31) we note that the velocity is going to increase until
the two terms on the right side balance. The velocity will asymptotically
approach a terminal velocity: Vt = k/a. For v = kia, we have dv/dt = o.
Let us integrate (2.31):
dv = ~dt. (2.32)
k/a - v m
We multiply both sides of the equation by -1:
d( -v) = -~dt. (2.33)
k/a - v m
Since k/a is a constant, we can write (2.33) as :
d(k/a - v) = -~dt. (2.34)
k/a - v m
Integrating both sides of (2.33) results in:
in (~ - v) = - : t + Cl , (2.35)
or
~- v = exp(cl) exp ( - : ) = C2 exp ( - : ) , (2.36)
where C2 == exp(cd is an integration constant.

Note. In (2.35) we have taken the logarithm of a quantity which has the
dimension of velocity. This is not a proper equation, and we should perhaps
write (2.36) directly.
We now use the initial conditions to determine the constant C2. The rele-
vant initial condition is:
v = 0 at t = 0 ;
consequently C2 = k/a. This means that we can write (2.36) as
v = ~ [1 - exp ( - : t)] . (2.37)
Note that v --> k/a for t --> 00.

Let us plot vasa function of time.
Question. What is the slope of the curve at t = O? [Answer: kim, use (2.31)].
Find x as a function of time by integrating (2.37):
Example 2.5. Frictional Force: Solid Against Solid. When a (dry)

object slides across a (dry) surface, the frictional force is to a good approxi-
mation independent of the relative velocity. If the object is held against the
surface, which reacts by a normal force of magnitude N, we can write the
frictional force as:
V
Ffr = -/lN R
since it is always pointing against the direction of motion. The constant /l
is called the coefficient of friction. The magnitude of the frictional force is
Ffr = /IN.
A wooden block with mass M is projected upwards with an initial velocity
Vo along an inclined plane.
/////////////////////////////
The block starts from the point A. We assume that the block is started in
such a way that it moves along a straight line up the inclined plane. During a
time T the block moves up the inclined plane, then stops. (It may then move
downwards again, see Problem 2.5.)
(1) Find the time T.
(2) Find the distance D which the block moves up the inclined plane before
it stops.
Answer. Draw all the forces acting on the body.
x
N = Mgcose; F = fLMgcose.
The equation of motion is:
Fe" FIr
~~
-Mgsine -fLMgcose,
- 9 (sin e + fL cos e) t + Vo ,
1 .
x -2g(sme+fLCOSe)t2 +vot.
(1)
T = Vo
9 [sin(e) + fLcos(e)]
(2)
D =~ v5
2 9 [sin( e) + fL cos( e)]
Example 2.6. The Atwood Machine
Two equal masses M are fastened to each end of a string that passes over a
pulley. On top of one of the masses we place a smaller mass m. The system
is then left to move under the influence of gravity. In this discussion of the
so-called Atwood machine, we will make the following assumptions:
(a) We shall ignore the mass of the string. This assumption means that the
string force is constant along all the free segments of the string. If the
string force was not constant along the string, a string segment, assumed
to have zero mass, would acquire an infinite acceleration.
(b) We shall neglect the mass of the pulley. This assumption implies that the
string force is the same on both sides of the pulley.
The tension T in the string arises from a small stretching of the string (sim-
ilar to that of a spring). In a dynamical problem such as the present, where
the system moves under the influence of gravity, the string tensions are de-
termined from the equations of motion for the massive objects.
The coordinate x (see the figure) determines the position of the mass
M + m. The position of the other mass is then determined by the length of
the string which we assume unchanged. That means the small change in the
length of the string is neglected.
The mass M + m has an acceleration downwards; the other mass has
the same acceleration, but upward, since we assume that the string is taut
during the entire motion. The acceleration of M is taken as positive when M
accelerates upwards. We shall often use the shorthand notation i; == dx/dt,
x == d 2 x/dt 2 .
Let us use this occasion to point to some straightforward rules which one
should always keep in mind when applying Newton's second law:
1. Decide to which body you apply Newton's second law.

2. Draw all forces acting upon that body.
3. Write the equation of motion for that body.
By adhering to these rules we get the equation of motion for the mass M + m:
(M + m)x = (M + m)g - T. (2.38)
Similarly, the equation of motion for Mis:
Mx=T-Mg. (2.39)
From (2.38) and (2.39) we find:

.. m
x = 2M +m g · (2.40)
The acceleration x is thus constant, and less than g.

By combining (2.39) and (2.40) one finds:
T = M 9 ( 1 + 2M: m) . (2.41 )
Example 2.7. Force in Harmonic Motion. In general one can say that
in Newtonian mechanics there are two types of problems:
(1) We know the forces and determine the motion

(2) We know the motion and determine the force.
We shall now consider an example of case (2). A particle with mass m is

moving along the x-axis. The motion is determined by the equation:
x = A sin(wt), (2.42)
where A and ware constants.

Find the acceleration x and the force Fx acting on the particle.
Solution.
-Aw 2 sin(wt), (2.43)
mx = -mAw 2 sin(wt), (2.44)
-kx, where k = mw 2 . (2.45)
We see that true harmonic motion can only be produced by a force of this
form, Le., by a force that is strictly proportional to the distance from the point
of equilibrium. On the other hand, harmonic motion is a good approximation
to many types of oscillations around a point of equilibrium, corresponding to
the approximation of a force law F(x) with its tangent at the origin.
Question. Coin on a plate in harmonic motion.
A coin with mass 20 g rests on a plate. The plate is acted on by an external

force in such a way that the plate performs a harmonic motion in the vertical
direction in the gravitational field of the Earth, described by the equation
x = A cos(wt). The amplitude A is 20 cm and the oscillation period is T = 1
second.
(1) Find the maximum and minimum values of the force R (the reaction
force) with which the plate acts on the coin throughout a period.
(2) Find the smallest value of the period for which the coin will remain
in contact with the plate throughout the motion (for fixed amplitude
A = 20 em.).
Answers.
(1)
Rmax mAw 2 + mg = 0.354 Nj

Rmin mg - mAw 2 = 0.0381 N.
(2)
Tmin = 27rH = 0.9 s.
Example 2.8. Charged Particle in a Uniform Magnetic Field. We

will study the motion of an electrically charged particle in a magnetic field
which is spatially uniform and constant in time. We denote the mass and the
charge of the particle by m and q respectively.
We shall take as an experimentally established fact that the force F on a
charged particle moving with velocity v in a magnetic field B is:
F = q(v x B). (2.46)
This force is called the Lorentz Force after the Dutch physicist H.A. Lorentz
(1853-1928).
The equation of motion for the particle is thus:
dv
mill = q(v x B). (2.47)
For simplicity we shall start by considering the case where the initial velocity
Vo of the particle is perpendicular to the direction of the magnetic field (see
figure below - here we have assumed that q is positive and B is perpendic-
ular to the paper and pointing out of the paper). From the vector equation
(2.47) we see that the acceleration of the particle is perpendicular to both
the velocity v and the magnetic field vector B. From this we can deduce that
if the initial velocity Vo is perpendicular to B, the particle will never develop
a velocity component parallel to B. The motion will remain in a plane per-
pendicular to B. Since the acceleration is also perpendicular to v, the vector
}-----------x
B (out of page)
v can never change magnitude, only direction. From this we conclude that
the magnitude of the force and therefore the magnitude of the acceleration
is constant. The only motion that can satisfy these conditions is a uniform
circular motion (with speed Vo = Iv I). The radius, r e , in the circular motion
is determined by writing the equation of motion in the following way:
(2.48)
where we have used that Vo is perpendicular to B, i.e.,

mvo
re = qB . (2.49)
This radius is called the gyro radius or the cyclotron radius (since it denotes
the radius of the particle orbit in a cyclotron accelerator).
The angular velocity, We, is given by
_ Vo _ qB
We - - --. (2.50)
re m
We see that the angular velocity We - and therefore the number of cycles
per second - is independent of both the initial velocity and the radius in the
orbit, and depends only on the strength of the magnetic field and the ratio
of charge to mass of the particle (this also follows from (2.49): since re is
proportional to vo, the angular frequency is a constant).
We shall now write a complete solution to the equation of motion. We set
up coordinates in such a way that the direction of the B field is along the
z-axis, which gives B the coordinates (0,0, B). In the usual way, i, j, and k
denote unit vectors along the x, y, and z axes, respectively. Still assuming v
is perpendicular to B, the vector v x B is given by (see Appendix)
v xB = (vyB, -vxB, 0). (2.51 )
We can now write the vector equation (2.46) in terms of three coordinate
equations
(2.52)
(2.53)
(2.54)
From (2.54): the component of v parallel to the magnetic field is not changed
by the field and is therefore constant (zero in our case). Equations (2.52) and
(2.53) can be written as
dv x qB
- -:;;;:Vy , (2.55)
dt
dv y qB
- - Vx · (2.56)
dt m
or, if we introduce We from (2.50):

dv x
(2.57)
dt
dvy
(2.58)
dt
We have here an example of coupled differential equations: the change of Vx
per unit of time is proportional to Vy and the change of Vy per unit of time
is proportional to V x .
In order to integrate this set of coupled differential equations we proceed
in much the same manner as with two algebraic equations of two unknowns.
If we differentiate (2.57) with respect to time, and substitute in this the
expression for dVy/dt from (2.58), we eliminate Vy from the equations. Now
we have a (second-order) differential equation, in Vx only:
d 2 vx 2
dt 2 + We Vx = o. (2.59)
This equation is recognized as the equation which governs the harmonic os-
cillator (although here the unknown quantity is the velocity component along
the x-axis), so the complete solution can be written with constants of inte-
gration, Vo and to:
vx(t) = Vo cos [we (t - to)] . (2.60)
Equation (2.57) now gives
1 dv x =-vosmwet-to·
vy(t)=--d . [( )] (2.61 )
We t
For simplicity we choose to = O. We can then integrate (2.60) and (2.61)

again (using Vx = dx/dt and Vy = dy/dt), and get, with the constants of
integration xo and yo:
Vo .
x-xo = - sm(we t ), (2.62)
We
Vo
y - Yo - cos (Wet) , (2.63)
We
displaying a uniform circular motion with center (xo, Yo) and radius
Vo mvo
re = - = --. (2.64)
We qB
The slightly more general case, where initially there is a velocity component
VOz parallel to the direction of the magnetic field, can now easily be dealt with.
We have already noted that this component of the velocity is not changed
by the presence of the magnetic field, so the z component of the motion is a
uniform motion:
The resulting orbit of the particle in this case is a helix with axis along the
magnetic field and this helix is traced out with a constant magnitude of the
velocity:
Vo = 1 Vo 1 = (v5x + v5y + v5z)1/2 .
We can easily find the pitch h of the helix, defined as the z-distance gained
through one revolution (Le., in the time Tc = 27r /w c)
(2.65)
B = (O,O,B)
Question.
(1) Find n, the number of cycles per second, for (a) an electron, (b) a proton,
both in a uniform magnetic field of strength 1 Wb m - 2 = 1 V s m - 2 =
1 T = 1 Tesla.
(2) Find the cyclotron radius for (a) an electron, (b) a proton, both moving
with a velocity of 1 x 106 ms- 1 in a uniform magnetic field of strength
1 T.
The proton has charge q =1 e 1= 1.6 x 10- 19 C.
[Answers: (la) ne = 2.8 x 1010 s-1; (lb) np = 1.5 x 107 s-1; (2a) re =
5.7 x 10- 6 m; (2b) rp ~ 10- 2 m = 1 cm.]
We now return to the first figure. (vo z = 0). Make a sketch of the motion
(use the "right-hand rule") if the particle:
(1) starts with velocity -Vo and q is positive,
(2) starts with velocity +vo and q is negative,
(3) starts with velocity -Vo and q is negative,
(4) starts with velocity +vo and q is negative but B --+ -B.
Note that if both the magnetic field and the charge change signs, the or-
bit is unchanged! Consider for comparison a uniform circular motion in a
gravitational field:
(2.66)
---1------'r=------I.c:> m
What orbit would the particle describe if we suddenly changed Vo to -vo?

!::,
Example 2.9. Thomson's Experiment. Consider an electron-beam tube

where electrons move in crossed electric and magnetic fields. A beam of elec-
trons is accelerated and then passed between two condenser plates.
Screen
n-nnn-~F~ v •
0 B
Let the electric field E between the condenser plates be parallel to the plane
of the paper and let the magnet be placed so that the magnetic field B is
perpendicular to the paper and pointing out of the paper.
In 1897 Sir J.J. Thomson measured the ratio of the charge q of the elec-
tron to the mass m of the electron by studying the motion of electrons in
a highly evacuated glass tube (no collisions with air molecules). The sketch
above shows a modern version of J.J. Thomson's apparatus. The electron
is negatively charged (how can one find out?). Let us write the charge q as
- 1e I· If the electric field has the direction shown in the sketch it will deflect
the particle downwards. As the electrons have a velocity and pass through
a magnetic field they will be under the influence of the Lorentz force too. If
the magnetic field has the direction shown in the sketch, the Lorentz force
will deflect the electrons upwards!
The resultant force on an electron is:
F = qE + q(v x B) = - 1 e 1 (E + v x B).
By a suitable choice of strengths of E and B the electric and magnetic force on
an electron may cancel each other. For, since E, B, and v are perpendicular
to each other the force F will be zero if
-leIE=-lelvB, (2.67)
E=vB. (2.68)
The ideas behind Thomson's procedure are the following:
(1) The position of the spot with both E and B equal to zero is measured
(that means the position of the undeflected beam).
(2) The position of the deflected beam with B = 0 and a known value Eo of
the electric field is measured.
(3) A magnetic field is now applied to the beam until the spot again reaches
the undeflected position.
Let us call the corresponding value of the magnetic field Bo. Assume that the
length of the condensor plate is l and the velocity of the electron is v = 1v I.
Show that the deflection l1y of an electron in a purely electric field (step 2)
Screen
1
.-4-.
2
v
-
is given by:
(2.69)
where l1y is measured at the far edge of the deflecting plates. l1y is not
directly measurable, but it may be calculated easily from the measured dis-
placement of the spot on the fluorescent screen. Thus: l1y, Eo, and l are
known. If we can find v we can calculate the ratio 1 e 11m.
The purpose of step 3 is to find v. If the magnetic field is increased until

the spot reaches the undeflected position again, we have
Eo
v = Bo . (2.70)
Substituting (2.70) into (2.69) we find
2L1y E o
B 2 l2 . (2.71)
m o
Thomson found
~ = 1.7 X 1011 Ckg- 1 .
m
The value quoted today is
~ = 1.75888 X 1011 Ckg- 1 .

m
Questions.
(1) What is the velocity of the beam of electrons when no deflection of the
electrons is produced by the simultaneous influence of the electric field
Eo = 3 X 10 5 V m -1 ,
and the magnetic field

Bo = 10- 2 T,
which are both normal to the beam and to each other?
(2) What would be the radius of the orbit of the electron if the electric field
was removed?
(3) In the calculation of (2.69) above we have not taken into account the fact
that gravity also deflects the electrons downwards. Let us assume that
Eo has the value Eo = 3 X 105 Vm- 1 as in the question above and that
the velocity of the electrons is the same found above. Let l = 1. 7 cm. Is it
a good approximation to neglect the gravitational force on the electron?
[Answers: (1) v = 3 X 107 ms- 1 ; (2) r = 1.7 cm; (3) the deflection due to the
electric field is: L1YE = 8.5 mm.The deflection due to the gravitational field
is:
L1Yg 1.6 X 10- 15 mm,

L1YE
5.3 x 10 15 .
L1Yg
We see that we may safely neglect the gravitational force.]
Example 2.10. Work and Energy in Lin€3T Motion of a Particle. We

will in a few examples consider more general methods for integration of the
equations of motion.
Note. We shall discuss here the important law of energy conservation for
motion in one dimension. When we have gained more experience, this law
will be treated in the general case of three dimensions (see Chapter 8).
•
m F
Consider a particle moving along the x-axis under the influence of some
impressed force F. We restrict ourselves to motions along the x-axis, so all
quantities are scalars. The equation of motion is
d2 x
m dt 2 = F,
or
dv
m dt = F. (2.72)
We multiply (2.72) by v = dx/dt on both sides:

dv dx
m dt v = F dt '
or
mvdv = Fdx. (2.73)
The right-hand side of (2.73) is now written Fdx = dW; the quantity dW is
called the work done by the force F over the segment dx.
The left-hand side of (2.73) can be written as d (~mv2), where the quan-
tity ~mv2 = T is called the kinetic energy of the particle. With this, we can
write (2.73) as
(2.74)
or
d (T) = d (W). (2.75)
The change in the kinetic energy of the particle over the

segment dx equals the work done by the force F.
This theorem, which makes a statement about kinetic energy, holds for all
forces without exception. It is based on Newton's second law only.
We now restrict ourselves to the special case where the force F depends
on position only, F = F(x). (In particular, F is not dependent on time).
When the force depends on the position coordinate x only, we can introduce
a new quantity, the potential energy U = U(x), defined through
dU(x) == -dW = -F(x)dx. (2.76)
The change in the potential energy U (x) over the segment dx

is equal to minus the work done by the force F.
J
In integral form:
U(x) =- F(x)dx (2.77)
The potential energy, U(x) of a particle in the force field F(x) is through
(2.77) determined up to an additive constant; this means for instance that
we can choose for which point Xo we want U(xo) = o.
With the definitions above, we have
dT = -dU(x),
or, upon integrating,
T+U(x)=E. (2.78)
The constant E is the value of a quantity called the total mechanical energy
of the system. The equation (2.78) is called a first integral to the equation of
motion since we have - through T - found dx / dt = v from the equation of
motion containing d 2 x/dt 2 . The number E is a constant of integration which
can be found from the initial conditions. A force field, where the force only
depends on the position coordinate of the particle is called a conservative
force field, since we have the simple observation:
The total mechanical energy - i.e., the sum of kinetic and

potential energy - for a particle moving in a conservative
force field, is conserved (constant).
An example of a nonconservative force is a frictional force. Frictional forces

depend not on position alone but also on, say, direction of motion. The energy
theorem - which we have derived only for one dimensional motion - is of vast
importance in physics, and will be discussed in much more detail later. Here
we will show, through an example, how the theorem can be used in the
integration of the equation of motion. We may write the energy conservation
theorem in the form
( dx)2
dt
2
= m [E - U(x)]. (2.79)
Example 2.11. Free Fall Towards the Sun from a Great Distance
@M
~ ~--~~--
-- ...... 1.....
IAU
K
•
m
)
x
~.~------------xO----------~~~
A comet K is found at some time at rest with respect to the Sun. The comet
has the distance Xo = 50000 AU from the Sun (1 AU = 1 astronomical unit =
the average distance of the Earth from the Sun = 1.495 x lOll m). We assume
now that the comet moves only in response to the gravitational pull of the
Sun, and we wish to determine the time passed from the comet begins its fall
until it is a distance of 1 AU from the Sun. Since the comet has no initial
velocity and in particular no initial velocity component perpendicular to the
direction toward the Sun, the comet will move on a straight line connecting it
to the center of the Sun. The mass of the Sun is considered to be so large that
the Sun does not move in response to the (reaction) force from the comet.
The mass of the Sun is considered in this sense to be infinite.
The equation of motion for the comet is then
dv GMm
m--
dt
= ------
x2 '
We multiply by v and integrate:

dv GMmdx
m dt v x 2 dt'
lov mvdv l x
Xo
_ G~mdx.
x
One finds, since v(xo) = 0,
~mv2(x) = GMm _ GMm .
2 x Xo
This result could have been found as a consequence of the energy theorem,
since the potential energy of the comet in the gravitational field of the Sun
is given by
U(x) =- J GMm GMm GMm

---dx= - - - + C = - - - .
~ x x
The choice of C = 0 implies U(oo) = O.
The energy theorem then implies
1 2 GMm GMm
-mv (x) - - - =0- - - .
2 x Xo
Using this result we have as a first integral to the equation of motion:

dx = ± [2GM(.!. _ ~)] 1/2

dt x Xo
We will use the minus sign of the square root since the dx / dt, with the chosen
orientation of the x-axis, is negative. We find for the time t lapsed between
l
Xo and x:
t =_ 1 x
(~ _ ~) -1/2 dx'
J2GM Xo x' Xo .
This integral is somewhat cumbersome. To obtain an order of magnitude esti-
mate, we discard the small term l/xo under the root. Then, after integrating,
t= ~ 1 (x 3/2_ x3/2) .
3J2GM 0
We want the time of fall from Xo = 50000 AU to the position x = 1 AU.

Since Xo » x, we get
t ';:j -2 1
x
3/2
3J2GM 0 ,
and finally, inserting values, t ';:j 4 X 10 13 S ';:j 1.3 X 106 years.
Question. Calculate the velocity a particle must have vertically upwards at
the surface of the Earth in order just to escape the gravitational field of the
Earth. Disregard air resistance and interaction with masses other than the
Earth. This velocity is called the escape velocity.
[Answer: 11.2 km S-l. Compare this to the velocity of a race car, a supersonic
jet and an air molecule!] !::,
Example 2.12. Momentum Conservation. Consider the equation of

motion for a particle:
Multiply both sides with dt:
mdv = dp = Fdt,
where dp is an increment of the momentum p mv of the particle. In
integral form:
or, in words:
The integral of the force over time equals the change of the
momentum of the particle during that time.
The integral of the force over time is called the impulse of the force. We
see that the impulse of the force, i.e., J Fx dt, is equal to the change in the
momentum. In Example 2.10 we saw that the work of the force, i.e, J Fxdx,
is equal to the change in kinetic energy. These two rules apply to all types of
forces since their derivation is based only on Newton's second law.
Consider two interacting particles. The particles may interact through an
electric or gravitational field, or they may be in direct contact or interact
via a spring attached to each of them. If we assume that the interactions
obey Newton's third law, we can derive the important result that the total
momentum of the two particles is conserved.
The proof is simple. For particle 1 we have
dpl
dt =F l2 .
Fl2 is the force exerted on particle 1 by particle 2. For particle 2 we have:
By addition:
d
dt (Pl + P2) = FI2 + F2I = 0,
where we have used Newton's third law. After integration, we find
P == Pl + P2 = constant vector,
i.e., the total momentum of the system is a constant of the motion. The
argument may be extended to a system of many interacting particles.
For an isolated (closed) system, i.e., a system where only

internal forces act, the total momentum is conserved.
Consider a collision between two isolated particles.
•
m M
• ~ x
v u
The two particles are constrained to move on a straight line: the x-axis.
The two particles collide. Whatever forces act during the collision, the total
momentum remains constant - the only assumption about the forces is that
they obey Newton's third law. The velocities before the collision are v and
u, and after the collision they are VI and UI. We therefore have
mVI+Mul=mv+Mu. (2.80)
We call a collision between the two particles elastic, if the total kinetic energy
is conserved in the collision (equivalently, no heat is produced in the collision;
the interaction forces are conservative). For the case of an elastic collision,
we have
12121212
-mvI + -Mu l = -mv + -Mu . (2.81 )
2 2 2 2
The set of equations (2.80) and (2.81) permit a calculation of the velocities
VI and UI, i.e., the velocities after the collision. See Problem 2.20. !:::,.
Example 2.13. Inelastic Collisions. The law of conservation of momen-

tum for a closed system of particles is truely independent of the forces acting
between different parts of the system. By the phrase "a closed system" we
mean a system of particles that does not interact with other physical systems.
,,~------------------------------------I
,
,,
: • v..
,
,, ,, :
.--_____ --...J~~*-----, ____ ,
As a simple example of the application of conservation of momentum we

consider a bullet (mass m, velocity v) that is shot into a wooden block of mass
M resting on a horizontal table. The bullet is stopped by - and embedded
into - the block. The system (bullet + block) is not an isolated system. It is
under the influence of gravity, the reaction from the table, and air resistance.
But if we consider the system for a very short time interval around the time
the bullet hits the block, these external forces cannot change the momentum
of the system much. If we neglect air resistance and consider the table as
frictionless, no external forces are acting on the system (block + bullet) in
the horizontal direction. The total horizontal component of the momentum
is then the same before, during, and after the collision.
(1) Let m be a gun bullet with mass 5 g and velocity v = 300 ms-I. Let
M = 3 kg. Find the velocity U of the system (m + M) just after the bullet
is stopped.
(2) Let us assume that the gun was fired such that the muzzle was 1 m from
the block. While the bullet travels from the muzzle to the block, it is
under the influence of gravity. Find the downwards velocity Vy that the
bullet gains before it hits the block and find the distance Sy it has fallen
below the height of the muzzle.
(3) Just before the collision the system thus has a small vertical momentum
mvy downwards. What happens to that momentum during the collision?
[Answers: (1) u = 0.50 ms- i , (2) Vy = 3.3 cms- i ; By = 0.055 mm.] Note the
order of magnitudes: Vy = gL1t, By = !g(L1t)2, where L1t is the short time
the bullet has fallen. I':-.
Example 2.14. Rocket Propulsion
- u v
The theorem of conservation of momentum can for certain systems be the

key to the integration of the equation of motion.
Let us assume that a rocket moves in force-free space. We describe the
motion with respect to an inertial frame I. At time t = 0, the rocket has
velocity o. Let m(t) be the mass of the hull plus the amount of fuel present
at time t. The engine throws out a constant amount of mass per unit time:
(2.82)
Note: fl > 0; the mass of the rocket decreases with time! We are assuming
that this mass is ejected in the direction of negative x with the constant
velocity Vrel relative to the rocket. If we denote the velocity of the rocket
relative to I by v(t), the mass ejected from the rocket at time t will have a
velocity
u(t) = v(t) - Vrel (2.83)
with respect to I. All quantities are scalars since the entire motion is one
dimensional. Let us consider the hull plus the mass of fuel present at time t.
The system is not acted upon by external forces, so we can use the theorem
of conservation of momentum. At the time t, the momentum of the system is
p(t) = m(t)v(t). (2.84)
The total momentum of the same system at time t + dt is
p(t + dt) = (m + dm)(v + dv) + fldt(u + du), (2.85)

because the amount of material expelled is JLdt. Note: in the last term of
equation (2.85) we might as well write u instead of u+du; the du is multiplied
by dt and differentials of second order disappear in the limiting process.
The theorem of conservation of momentum tells us that
(m + dm)(v + dv) + JLdt(u + du) = mv. (2.86)

Discarding differentials of second order, we find:
mdv + vdm + uJLdt 0, (2.87)

mdv + vdm - udm O. (2.88)
The differential equation for vasa function of m becomes (using (2.83)):
dv
m dm + Vrel = O. (2.89)
From (2.89) we obtain:

dm
dv = -Vrel-
m
and since v = 0 for m = m(O), we get
m(o)]
v(t) = vre1ln [ m(t) . (2.90)
We assume that the rocket starts with velocity zero and that the engine works
from t = 0 to t = T. The final velocity is then:
v(T)
m(o)]
= vre1ln [ m(T) .
Note.
(1) The final velocity depends logarithmically on the ratio between the initial
and final mass.
(2) Given m(T), it is crucial to have a high value of Vrel.
Even if the payload is only 1/10 %, i.e., m(O) = 1000m(T), we find v(T) =
vre1 ln(1000) = 6.9v re l; less than ten times the exhaust velocity.
This simple fact, based on the most fundamental mechanical laws shows
the difficulty in designing rockets to carry people out of the solar system.
Colonizing the galaxy will not be easy! Einstein's special theory of relativity
does not change this situation in any fundamental way.
Another numerical example: if the exhaust gasses have a temperature of
about 4000DC, the molecules have a velocity of the order of 2000 m S-l. If we
want a final velocity of the rocket of v(T) = 8 kms-I, we can find m(T) as
a ratio of m(O):
m(T) = m(O)exp [ - V(T)]
- .
Vrel
or
m(T) = exp [_ V(T)] ,
m(O) Vrel
m(T) = m(O) e- 4 = m(O) x 0.018.
This corresponds to a payload of about 2 % of the initial mass.
Warning. Consider again the rocket head and the fuel in the rocket head at
the time t. The equation of motion at the time t for this total mass is
dp = 0
dt .
From this one might be tempted to conclude
d
dt (mv) = 0,
or
or
This is not correct! If the above equations were correct, the acceleration of
the rocket head would be independent of the exhaust velocity Vrel. The above
equations neglect the fact that the part of m(t) which appears in the fuel
exhaust also carries a momentum. As shown above:
dp p(t + dt) - p(t)
dt dt
p(t) m(t)v(t) = mv ,
p(t + dt) (m + dm)(v + dv) + I1dt(v - Vrel).
The first term on the right-hand side is the rate of change in the momentum of
the rocket head. The second term is the change in momentum of the exhaust
fuel per unit time at the instant considered. If an external force of magnitude
F acts (e.g. gravity) we find
d
dt (mv) + I1(V - Vrel) = F.
Example 2.15. Some Qualitative Remarks on Rocket Propulsion.

Let us consider a rocket head containing oxygen plus kerosene. When the
oxygen and the kerosene are allowed to interact a very rapid chemical reaction
(an explosion) takes place. A lot of potential energy of the electrons inside
the molecules of the oxygen and kerosene is converted to kinetic energy of
the combustion products.
Oxygen
and
Kerosene
Let us sketch very roughly how a rocket works. After the explosion a lot
of kinetic energy is going into the molecules of the combustion products.
These molecules move in all directions. On average an equal number moves
in each direction. Let us in an extremely rough approximation assume that
half the molecules move along straight lines to the left, and the other half
along straight lines to the right.
Let us assume that the cap 'A' is still on. We consider the two molecules
shown. Let their velocities be v, and their masses m . They are reflected at
the end walls.
In the reflection from A, molecule 1 gives the momentum 2mv to the

rocket head to the left (the rocket head has infinite mass compared to one
molecule). Molecule 2, in reflection from the end wall, gives a momentum
2mv to the rocket head to the right. All in all the rocket head receives no net
momentum.
---------~
-------------~
v
If A is not there, molecule 1 is not reflected but moves out in free space,
while molecule 2 gives the momentum 2mv to the rocket head and then leaves
the rocket head to go out into free space. The rocket head now receives a
momentum 2mv towards the right.
If a mass 2m is expelled to the left with velocity v relative to the rocket

head, the rocket head receives a momentum 2mv towards the right.
Note. It is not possible to convert all the disorganized thermal energy in the
molecules of the combustion products into useful macroscopic kinetic energy
of the rocket head. For this to be achieved, all of the molecules should (after
the explosion) move in the same direction (first towards the right). This
is evidently impossible. With this remark we touch upon one of the most
fundamental laws of physics: the second law of thermodynamics. £::,
Example 2.16. Ball Against a Wall
fixed
..
m v wall
@
A ball of mass m is thrown horizontally against a fixed wall, so that just

before it hits the wall the ball has velocity v. Let us assume that it rebounds
with the same speed. The ball is in contact with the wall for a short time ..1t.
(1) Let us first consider the ball as our system. Find the average value F of
the force by which the wall has acted on the ball during the collision?
The value of F should be expressed by m, v, and ..1t.
(2) Numerical example: v = 25 ms-l, m = 100 g, and ..1t = 0.02 s. Find F.
Compare F to the force on a mass of 0.1 kg in the gravitational field of
the Earth.
(3) Let us next consider as our system the ball plus the wall, the latter being
fixed to the Earth. What about conservation of momentum? It still holds;
the Earth picks up the ball's change in momentum. Estimate the change
in the velocity of the Earth due to the collision between the wall and the
ball.
Solution. The force must be in the direction perpendicular to and outwards

from the wall, and the average force can be determined from the equation of
motion for the ball:
dpx _ F
dt - x, (2.91 )
dpx = Fxdt, (2.92)
..1Px = 12 Fxdt = (Fx)..1t . (2.93)
From (2.93) we get

(Fx)f),t = 2mv , (2.94)

(Fx) = 2:V . (2.95)
Numerical example: v = 25 ms-I, m = 100 g, f),t = 0.02 s, (F) = 250 N.

Let us next consider as our system the ball plus the wall, which is fixed to
the Earth. What about conservation of momentum? It still holds. The Earth
picks up the change in the momentum of the ball.
~---------------------~ ,
,,
-----------------------
The velocity of the center of mass of the Earth after the collision, measured
in the reference frame where the Earth was at rest before the collision, is
determined by
Mu = 2mv, (2.96)
2mv
u= M . (2.97)
Inserting numbers,
M = 5.98 X 1024 kg ,
2 x 0.1 x 25
u= ~ 0.8 X 10- 24 ms- 1
5.98 x 10 24
which is a very small velocity indeed.
Question. Equation (2.96) is not exact. Why?
2.3 Problems
Problem 2.1. Calculate the escape velocity from the planet Mars, and for
the (terrestial) Moon. Compare with Example 2.11.
Problem 2.2. A man stands on top of a tall building. From the edge of the
building he throws a small heavy particle vertically upwards with an initial
velocity of 20 m s-l. The particle hits the ground at the foot of the building
5 seconds later. Neglect air resistance.
2.3 Problems 63
(1) How tall is the building?

(2) How far above the ground is the highest point of the trajectory of the
particle?
Problem 2.3.
The mass ml slides without friction down an inclined plane. ml = 400 g,

m2 = 200 g, and e = 30°. Neglect the mass of the string and of the pulleys.
Find the acceleration of m2 and determine the string forces.
Problem 2.4.
Two blocks, one of mass m and one of mass M, rest on a smooth horizontal
table. The blocks are in contact with each other. As shown on the sketch,
the system is now acted upon by a horizontal force F, perpendicular to the
surface of the block of mass m. Find the force by which the block of mass m
acts on the block of mass M.
Problem 2.5.
~
F
A
Mg
e
r~???/
The body A, which has mass M, rests on a fixed inclined plane. The body
is acted upon by three forces: Mg, N, and F, where F is the friction force.
The friction force is a reaction force. The maximal friction force which the
interface between the body A and the inclined plane can supply, is determined
by the coefficient of friction J.l and the magnitude of the normal force in the
following way: IF Imax = J.lN. If the body is in motion down the inclined
plane, the friction force assumes this value.
Determine for what values of the angle e the body will move down the
inclined plane when the coefficient of friction J.l is known. (Sometimes it is
necessary to distinguish between static friction and friction when the body
slides. We shall not discuss this technical point further).
Problem 2.6. A road bend is circular with a radius of r = 300 m. The

surface is horizontal. The coefficient of friction between car tires and dry
asphalt is J.l = 0.8. Between car tires and icy road, J.l = 0.2.
(1) Determine the maximal speed with which a car can drive through the
bend when the road is: (a) dry, and (b) icy.
A gramophone record rotates with an angular velocity corresponding to 45
rpm (revolutions per minute). A small bug crawls slowly outwards in the
radial direction. The coefficient of friction between bug and record is J.l =
0.1.
(2) Determine how far away from the center the bug will get before it begins
to slip.
Problem 2.7.
A large block with mass M is going to be pulled with a constant velocity

along a horizontal surface. The coefficient of friction between the surface and
the block is J.l.
(1) Find the angle cP = CPo that the direction of pull shall make with the
direction of motion for the pulling force to be as small as possible.
(2) Find this smallest value of F.
Problem 2.8. The mass M is pulled a distance d downwards from the point
of equilibrium, and released. Ignore the mass of the string and the mass of
the spring. The spring constant is denoted by k.
(1) Determine the maximum value d = d max for d, if we demand that the
string must be stretched during the entire oscillation. M = 0.1 kg, k =
20 Nm- 1 .
(2) How far is the spring extended from its unforced position when M is
hanging at rest?
(3) What is the magnitude and direction of the acceleration at the lowest
position for M when the system is oscillating and d = dmax ?
2.3 Problems 65
/~ -4- spring , spring constant k
I J"'II ",<Ch>, .,<,.,
Problem 2.9. A mass M is suspended from two springs attached in series

as shown on the sketch. The system is hanging in the gravitational field of
the Earth. One spring has the spring constant kl ' the other k z. We neglect
the mass of the springs. Find the period T in the vertical harmonic motion
of the mass.
/.
Problem 2.10.
A homogeneous rope with mass M and total length L hangs from the ceiling
of the laboratory.
(1) Find the magnitude of the force F(z) by which the top part of the rope
acts on the lower segment of the rope at the distance z from the free end
of the rope.
Now the same rope is swung in the horizontal plane in such a way that the
rope performs a uniform circular motion with constant angular velocity w.
Ignore gravity.
(2) Find the magnitude of the force F(r) by which the inner part of the rope
acts on the outer segment of the rope at the distance r from the center
of the circular motion.
Problem 2.11.
(1) With what fraction of the escape velocity from the surface of the Earth
must a rocket start in order to reach the Moon? The distance between
the Earth and the Moon is 60 times the radius of the Earth.
(2) Consider a celestial body belonging to the solar system. Let us assume
that this body falls from a large distance towards the Sun. Estimate the
magnitude of the largest relative velocity with which the Earth and the
body can meet. [Note: an order-of-magnitude estimate is desired, not an
exact calculation. When the space probe Giotto passed Halley's Comet,
the relative velocity of the two bodies was 69 km s -1 .J
Problem 2.12.
F is a constant horizontal force acting on the system shown in the sketch.

F is perpendicular to the end surface. The masses m1 and m2 can move
without friction relative to the mass M which in turn can move without
friction relative to the horizontal table.
Determine F so that the masses m1 and m2 are not moving relative to
M. Assume that m1 and m2 are initially at rest relative to M.
Problem 2.13. A painter with a mass of 90 kg sits in a chair near the wall of
the house he is painting. He holds on to a rope which goes over a pulley and
suspends his chair as shown on the figure. He wants to move upwards, and
2.3 Problems 67
pulls the rope with a force so large that the pressure he exerts on the chair
corresponds to the weight of a mass of 50 kg, (i.e., 50 kgx 9.8 ms- 2 = 490 N).
The mass of the chair itself is 15 kg. Disregard the mass of the pulley.
(1) Find the acceleration of painter and chair.

(2) With what force is the painter pulling on the rope (i.e., what is the string
force)?
Problem 2.14.
v
--------~-
- d
A lead bob has a mass of 1/4 kg and is placed at the edge of a smooth table.
A rifle bullet with the mass of 1 g is shot through the bob. The bullet has
horizontal velocity v just before it hits the bob. It moves through the bob
in a very short time and exits with velocity 0.5v . The table has the height
h = 1 m . The bob hits the floor a distance d = 30 cm from the edge of the
table. Find v.
Problem 2.15. A projectile is to be fired in the gravitational field g near

the surface of the Earth. Determine the angle e = emax which the aim must
form with the horizontal in order that the projectile will get as far away as
possible from the spot where it was fired. Ignore air resistance and consider
the gravitational field to be homogeneous.
Problem 2.16. A homogeneous, flexible chain of total mass M has the length
L. Initially, the chain hangs vertically in the gravitational field in such a way
that the lowest point of the chain just touches the horizontal plate of a spring
scale (see the figure).
Now the chain is allowed to fall freely in the gravitational field of the
Earth. When the chain has fallen the distance 5 a segment 5 is resting on
the scale.
What will the spring scale show when the chain has fallen the distance
5? What will the spring scale show at the instant when the entire chain has
reached the scale?
Problem 2.17.
The sketch shows two wooden blocks A and B. A has the mass m and B has
the mass M. At the time t = 0 B is at rest relative to the table and A is
started with a velocity Vo relative to B. Between the table and B there is no
friction . Between A and B there is a coefficient of friction p. The velocity of
A relative to B will decrease and B will start to move along the table.
After a certain time T the block A will be at rest relative to B.
(1) Find the final velocity of the two blocks relative to the table.
(2) Find the time T.
(3) Find the distance D that A has moved relative to B when the velocity of
A relative to B has become zero.
2.3 Problems 69
Problem 2.18.
Water
jet
A jet of water from a fire hose impinges horizontally on a wall as indicated in

the figure. The speed of the water is u before it strikes the wall (see figure)
and it is assumed that the water has no velocity component perpendicular
to the wall after it has hit the wall. The cross section of the water jet is a .
The density of the water is p.
(1) Find the force F the water exerts on the wall. The magnitude of F should
be expressed in terms of p, a, and u.
(2) Numerical example: p = 1 gcm- 3 = 103 kgm- 3 , a = 25 cm 2 = 2.5 x
10- 3 m 2 , u = 60 ms-l. Now compute F. Compare the result with the
force on a man, of mass 70 kg, in the gravitational field of the Earth.
(3) How far will a water jet from a hose reach before it hits the ground, if it
is directed horizontally 1 m above the ground and the muzzle velocity is
u = 60 ms- l ?
(4) A stream of water impinges on a block B as indicated on the sketch. At a
certain time the block has velocity v < u relative to the ground as shown.
The velocity of the water before it strikes the block is u relative to the
ground. The cross section of the jet is still a and the density p. Find the
force that the water jet exerts on the block B.
u
B
Problem 2.19.
A projectile is shot from a gun located at the origin 0, and the projectile is
aimed at the point P where an object is located. At the exact moment when
the projectile is fired the object at P starts to fall in the gravitational field
of the Earth. The projectile and the object will - as indicated - meet at the
point B. We neglect air resistance.
Show that this mid-air collision will take place, independently of the value
of the muzzle velocity of the projectile, the mass of the projectile and the
mass of the object at P. The muzzle velocity of the bullet is assumed to be
so large that the bullet will reach as far as the vertical line through P.
Problem 2.20.
------~--------------~--~
1 2
A collision is called central if the velocities of the particles both before and
after the collision lie on the same straight line. A collision is called elastic if
the sum of the kinetic energies of the colliding bodies is the same before and
after the collision. See Example 2.12.
A sphere of mass m and with velocity v collides in a central and elastic
collision with a sphere of mass M initially at rest in the laboratory.
(1) Find the velocities of the two spheres after the collision.
(2) Find the fraction f of the original kinetic energy of m that is transferred
to M and the fraction r of the momentum that is transferred. Show that
for M -+ 00, f -+ O.
(3) Show that for m = M, f = l.
For m = M we see that all the energy and momentum of the incoming particle
is transferred to the particle at rest. From the answer to question 3 it may
be understood why fast neutrons from a neutron source are slowed down
2.3 Problems 71
very little by a lead plate 50 cm thick, while a layer of paraffin nearly stops
neutrons completely. In a central collision of the neutrons with a hydrogen
nucleus (M = 1) in the paraffin the kinetic energy of the neutron (m = 1) is
transferred to the hydrogen nucleus.
What fraction of the energy of the neutron is transferred to a lead nucleus
in a central collision (the atomic mass of lead is about 207)7
Problem 2.21.
(1) A small mass m collides in a central and elastic collision with a large mass
M that has velocity Vo before the collision. What should the velocity v
of m be, before the collision, in order that all of the kinetic energy is
transferred to M7 What happens if M = m 7 If M < m 7
(2) A moving small sphere of mass m collides elastically (but not necessarily
centrally) with another sphere of the same mass, initially at rest. We
assume that the two spheres do not start to rotate during the collision.
Show that either the two spheres move perpendicular to each other after
the collision or the sphere that was moving before the collision is stopped.
[Hint: from the conservation laws it may be shown that Ul . U2 = 0 .J
Problem 2.22. A rocket has an initial total mass (including fuel) of m(O) =
4 x 103 kg. The velocity ofthe exhaust gas is Vrel = 3.0 km s-l = 3000 m s-l .
(1) Consider the rocket at t = 0, i.e., just as it starts. How much mass must
be expelled per second (how large must JL be) for the rocket to have
an initial acceleration g upwards in the gravitational field of the Earth
(g = 9.8 ms- 2 )7
Let us now consider another rocket of the same design ( m(O) = 4 x 10 3 kg,
Vrel= 3.0 km S-l). Let us furthermore assume that the rocket engine works in
such a way that the mass of the rocket head (hull + fuel) decreases according
to the formula m(t) = m(O) exp( -bt).
(2) Determine the value of b such that the rocket is able to just keep itself
hovering in the gravitational field of the Earth.
Problem 2.23.
v
F
o o ~
A hopper is at rest relative to the laboratory. Sand drops continuously down

onto a conveyor belt below the hopper and comes to rest relative to the
conveyor belt.
Assume that the amount of sand dropping down onto the belt per unit
time, f1 = dm/dt, is constant. The velocity of the conveyor belt relative to the
laboratory, v, is also constant. Find the force F needed to keep the conveyor
belt moving with constant speed v.
3. Gravitational and Inertial Mass
Since the time of Galileo's famous experiments with freely falling bodies it
has been known that all material bodies, at the same place on Earth, fall
with the same acceleration when air resistance is disregarded. Through many
experiments this has become a well documented experimental fact.
After Newton had formulated the dynamics laws for the motion of ma-
terial bodies, these observations of Galileo took on a deeper significance. In
Newtonian mechanics, the force acting on a body is considered to be the
cause of the acceleration of the body. The mass of the body is a quantitative
measure of the fact that the body resists acceleration. The kinematics laws
of the free fall, experimentally verified by Galileo, ought to be explainable
through the dynamics laws of Newton. Let us briefly recall some relevant
features of Newton's theory.
3.1 Gravitational Mass
+ g 'LUIULU'
~
Fig. 3.1.
A mass A hangs from a spring near the surface of the Earth. From Newton's
general law of gravity we know that the Earth acts on the mass with a
downward force. This force has a direction and magnitude as if all of the
mass of the Earth was concentrated in its center. The expression for the
magnitude of the force F is:
(3.1)

74 3. Gravitational and Inertial Mass
The terms are:

p is the radius of the Earth;
Mg is the gravitational mass of the Earth; Mg is a quantitative measure
of the property of the Earth which makes it attract - and be attracted
by - other material bodies;
mg is the gravitational mass of the body. Similarly, mg is a measure of
the property of the body causing it to attract and be attracted to other
material bodies.
The magnitude of the force F can be measured with the spring scale. Often
(3.1) is rewritten as
(3.2)
The more gravitational mass a body has, the bigger its "weight", i.e., the
bigger the force by which the Earth acts on the body. It is well known that
9 = (G:;g) = 9.8 Nkg- = 9.8 ms-

1 2 .
3.2 Inertial Mass
Wal
x
of equilibrium
Fig. 3.2.
As Figure 3.2 shows, we now place the body A on a smooth horizontal table.
The body A is attached to a spring, the other end of which is fastened to
a wall in the laboratory. In this way A can be made to perform horizontal
oscillations. We know from Chapter 2 (see Example 2.2), that the equation
of motion for the body is
(3.3)
Here, the terms are:

k is the spring constant (k has units N m -1);
3.3 Proportionality Between Inertial and Gravitational Mass 75
mi is the inertial mass of the body A. The inertial mass is a quantita-

tive measure of that property of a body which makes the body resist
acceleration.
As shown in Chapter 2 the period of oscillation of the harmonic motion
is given by T = 27rJmi/k. The greater the inertial mass of the body, the
longer the period of oscillation. The gravitational mass does not enter the
expression for the period of oscillation. The reaction force from the smooth
surface exactly cancels the vertical force by which the Earth attracts the
body.
3.3 Proportionality Between Inertial

and Gravitational Mass
We now let the body A fall freely in the gravitational field of the Earth.
The body starts off with zero initial velocity. See Figure 3.3. We write the
Fig. 3.3.
equation of motion for the body. Note the orientation of the z-axis. We have
d2z
mi dt 2 = -mg 9 , (3.4)
~:: = - (:: ) 9. (3.5)
According to Newton all bodies have the following two properties:

(1) they resist acceleration (relative to absolute space);
(2) they attract and are attracted to all other bodies in the universe.
As mentioned, it has been experimentally verified with great accuracy that

on any given place on the Earth all bodies fall (in vacuum) with the same
acceleration. This acceleration is g. In Newtonian mechanics, the observation
that all bodies fall with the same acceleration is interpreted as an observation
of the equality of inertial and gravitational mass, i.e., mj = mg.
In other words: that property of a body which conditions the body to re-
sists acceleration is - according to Newtonian mechanics - the same property
which causes the body to attract (and to be attracted to) other bodies.
A somewhat more precise expression of the experimental result mj =
mg is the statement that gravitational and inertial mass are proportional.
This statement is more correct for the following reason: Newton's second law
expresses proportionality between on one side force, and on the other side
mass times acceleration:
(3.6)
where kl is a constant of proportionality. The general law of gravity can
similarly be expressed as
(3.7)
where k2 is a constant. By setting kl = 1 (dimensionless) and k2 = G (units

for G: N m 2 kg- 2 ), mj and mg will get the same dimension. Both kinds of
mass should then be measured in kilograms and mj = mg .
A practical consequence of this remarkable equality (proportionality) be-
tween mj and mg is that one can compare by a balance measurement, not
only the weights but also the masses of two given bodies.
3.4 Newton's Experiment
Newton realized that the proportionality between the gravitational force on

a body and the inertial mass of that body, is of fundamental importance
for an understanding of the universal law of gravity. Newton carried out ex-
periments with different materials in order to investigate the proportionality
between gravitational mass and inertial mass. See Figure 3.4. A body with
gravitational mass mg hangs from a light string with length L as suggested
on Figure 3.4. When the body is drawn away from its point of equilibrium,
it will oscillate around the vertical, driven by the component of the force
of gravity not canceled by the string tension. The tangential acceleration is
given by at = LB. The instantaneous value of the tangential force is given by
(see Figure 3.4)
Ft = - Fg sin( 8) (3.8)
(the minus indicates the restoring nature of the force) . We then find
3.4 Newton's Experiment 77
Fig. 3.4.
(3.9)
For small values of 8 we use the approximation sin(8) ::::J 8. Equation (3.9)
now takes the form
(3.10)
This is a differential equation of the same type as discussed in Example 2.2.
We see that for small amplitudes, the mass will perform harmonic oscillations
around the equilibrium point. In fact, the period of oscillation can be read
directly from (3.10); it is
(3.11)
All of Newton's observations suggested that T is independent of the magni-

tude and material quality of the oscillating mass. This experiment convinced
Newton that indeed mj = mg. We quote from Principia:
But it has long ago been observed by others, that (allowance being
made for the small resistance of the air) all bodies descend through
equal spaces in equal times; and, by the help of pendulums, that
equality of times may be distinguished to great exactness. I tried the
thing in gold, silver, lead, glass, sand, common salt, wood, water,
and wheat. I provided two equal wooden boxes. I filled the one with
wood, and suspended an equal weight of gold (as exactly as I could)
in the center of oscillation of the other. The boxes, hung by equal
threads of 11 feet, made a couple of pendulums perfectly equal in
weight and figure, and equally exposed to the resistance of the air:
and , placing the one by the other, I observed them to play together
forwards and backwards for a long while, with equal vibrations. And
therefore the quantity of matter in the gold was to the quantity of

matter in the wood as the actions of the motive force upon the gold
to the action of the same upon all the wood; that is, as the weight
of the one to the weight of the other. And by these experiments, in
bodies of the same weight, I could have discovered a difference of
matter less than the thousands part of the whole.
Several researchers have improved the accuracy of this type of measure-
ments. Currently the proportionality between gravitational and inertial mass
is known to be valid to 1:10 10 .
We have thus seen that by mechanical experiments (e.g. by studying pen-

dulum oscillations, masses on springs, collisions of balls, etc.) performed in
a gravitational field, it is not possible to distinguish the two properties of
matter: gravitational mass and inertial mass. These two phenomena, inertia
and gravitational attraction which conceptually are so different in Newtonian
mechanics, appear to be of the same origin. Inertial mass mj describes the
property of matter which causes it to resist acceleration. The gravitational
(heavy) mass mg describes the property of matter which causes it to attract
other bodies.
It is unfortunate in a physical theory to introduce two quantities that are

conceptually different, but which cannot be distinguished by any experiment.
Einstein postulated that inertial and gravitational mass are expressions of

one and the same property of matter. His general theory of relativity resolves
the problem in a very elegant manner; we shall later briefly return to this
issue (see, for example, Sections 6.2 and 6.3).
Example 3.1. The Satellite
Satellite
particle
A small spherically symmetrical satellite moves in a circular orbit around

the Earth (see the figure). An astronaut inside the satellite carefully places a
3.4 Newton's Experiment 79
particle at rest in the center of the satellite. The particle will now maintain
that position relative to the satellite while it orbits the Earth. This result is
an expression of the exact proportionality between gravitational and inertial
mass. Let us consider the situation in more detail.
Mass of the satellite: inertial mil gravitational mg.
Mass of the particle: inertial J-Li, gravitational J-Lg.
Gravitational mass of the Earth: Mg.
For the satellite, we have
v2 GMg
mi- -- --2-mg' (3.12)
r r
For the particle:
v 2 _ GMg
J-Li--;: - --;:zJ-Lg . (3.13)
Since mi = mg and J-Li = J-Lg, we see that both the satellite and the particle
fall freely in the gravitational field of the Earth with the acceleration
v2 GM
a=-=--g (3.14)
r r2
directed towards the center of the Earth. The particle initially had zero ve-
locity relative to the satellite, and the particle and the satellite have the same
acceleration. Therefore the particle will never change its position relative to
the satellite. We say the particle is weightless relative to the satellite as the
satellite falls freely in the gravitational field of the Earth. 6,
Example 3.2. An Elevator in Free Fall
w/•.
chalk
A person stands in an elevator box. At a certain moment two events occur:

the cable breaks, and the person releases a piece of chalk. The chalk will
now maintain its position relative to the person (and the elevator) because
of the equality (proportionality) between gravitational and inertial mass. For
an observer on the Earth, the equation of motion for the three bodies -
person, chalk and elevator - all reduce to the same equation d 2 z/dt 2 = g,
because the masses are canceled in all the equations. The person and the
chalk appear weightless relative to the elevator as the elevator falls freely in
the gravitational field of the Earth. 1:::,
Example 3.3. Three Balls. Three balls are at rest on a horizontal board.
Pb Al Wood
One ball is made of Pb, one of Al, and the third is made of wood . The for ce
with which the Earth acts on each of these balls is different (see the figure) .
In spite of this, if the board is suddenly removed, the three balls will fall
with identical accelerations, remaining flush with each other - because of the
universal equality (proportionality) between gravitational and inertial mass.
3.5 Problem
Problem 3.1. The Concept of Weight. A man has the mass m = 70 kg .

He is standing on a spring scale in the gravitational field of the Earth. The
weight of the man is defined as the force by which the Earth pulls in the
man, i.e.,
Weight = W = mg = 70 kg x 9.8 ms- 2 = 686 N. (3.15)
Now, the scale will not show 686 N, but 70 kg. Sometimes in physics one
uses a unit kg* for weight. This, however, is not a good idea, because among
other things it confuses the concepts of force and mass. But:
(1) How many kg* would the man weigh on the Moon?
(2) How many kg* would the man weigh on Mars?
Let us now assume that the spring scale is placed inside an elevator box near
the surface of the Earth.
3.5 Problem 81
(3) How many kg* would the scale read if the elevator is accelerating upwards
in the gravitational field with an acceleration of a = 2 ms- 2 ?
(4) Same question when the elevator is accelerating downwards with a =
2 ms- 2 ?
4. Galilei Transformation
1. The Foundation of Classical Mechanics
1.1 Principia
Andromeda
.......;:.::.:.:.....
It was an important event in the history of physics when Sir Isaac Newton
·:·:::::::::::::fj~)~!!~j~ji~tt::::::::.; .
in 1687 published his book Philosophiae Naturalis Principia " Mathematica.
._ ::::=::::~:.:
..=....
In this famous work we find a masterly synthesis of/ / the concepts of motion
and force. The Newtonian formulation of the laws/ / of motion has, with su-
2 000 000 light years /
perior strength and vitality, survived more than/ /300 years. Although certain
fundamental aspects of Newtonian physics have /
/
/
been revised in this century,
Newton's principles still find widespread use/ both in the basic and in the ap-
/
Milky
plied sciences. Just oneWay //
example: the launching of artificial satellites is based
»""ll~wlw~ill~111§§dtt~111111!11~~1~1~W§§1@@~~::"-sun
directly on these principles.
Newton's work forms the basis upon which theoretical physics as well as
modern engineering science rests. We shall here, without going into details of
the historical development,
100 000 lightdiscuss
yearssome of the most important steps in the
evolution of the Newtonian world picture.
Fig. 4.1.
1.2 own
Our Prerequisites
galaxy is called fortheNewton
Milky Way. The galaxy closest "to us is the
Andromeda Galaxy. Light from the Andromeda reaching us today initiated
There
out therearesome
two 2main
000 000lines of inquiry
years thatthis
ago. About leadtime
towards
man was andatform the basis
the beginning
forhis
of Newton's efforts:through
long journey one based on motions here on Earth, and one based on
evolution.
the motions of bodies in the heavens. The first line is associated with the
name
Let us Cbriefly
alileo Calilei
restate (1564-1642)
Newtons first ; the other
three lawswith Nicolas Copernicus (1473-
of motion:
1543), Tycho Brahe (1546-1601) and Johannes Kepler (1571-1630). Before
1. Law ofthe
Calileo, inertia.
"impetus" concept dominated the thinking on motion. This idea
2. rndv/dt
is somewhat = related
F. to the later concept of momentum, but although it was
3. Law of action and reaction.
thought that a body was given a certain impetus when it was thrown, the
body
As wasasthought
soon to spend
these laws its impetus
are written we areduring
faced its flight.
with the When all the
following initial
question:
impetus was spent, the body would stop and - if it was not
relative to which coordinate system should we measure positions, velocities, supported - fall
down.
and accelerations? Without an answer to this question the laws loose their
As for the free fall, it had been held since Aristotle (384- 322 B.C.) that
meaning.
a heavy
Shouldobject falls faster
we measure thantoa a light
relative object.system
coordinate From immediate
fixed on theeveryday
surface
experience, both the impetus concept and the
of the Earth? We know that the Earth revolves once around idea that a heavyits object falls
axis every
faster than a light one, are not unreasonable.
24 hours and orbits the Sun in an approximately circular orbit every year.

84 4. The Galilei 1Tansformation
The tangential velocity of the Earth in its orbit around the Sun is v =
3 X 10 4 m s-l = 30 km S-l. The radius in that orbit is r = 1.5 x lOll m. Thus,
the Earth is accelerated towards the center of the Sun with an acceleration
of magnitude a = v 2 jr = 6 x 10- 3 ms - 2 .
A coordinate system at rest relative to the surface of the Earth (say a
laboratory) clearly is in a complicated state of motion relative to the Sun.
Should we instead measure the kinematical quantities relative to the Sun?
The center of the Sun moves in an approximately circular orbit around the
center of our galaxy with a velocity of 250 km s-l. The distance of the Sun
from the center of the galaxy is about 30000 light years (~ 2.8 x 10 20 m).
The acceleration of the Sun with respect to the center of the galaxy is then
a = v 2 jr = 2.2 x 10- 10 ms- 2 .
Is it the center of our galaxy or perhaps the center of the Andromeda
galaxy which should be taken as the center of our reference frame? Attempts
to answer the questions of which coordinate system should be the basis for
measurements of positions, velocities, and accelerations will throw light on the
basic strengths and weaknesses of the Newtonian world picture, and enable
us to understand the profound revision of the problem of motion which was
initiated with Einsteins special and general theory of relativity.
Newton answered the question himself by adding to his three basic laws
the two postulates: the postulate of absolute time and the postulate of ab-
solute space (see Chapter 2). These two postulates therefore are of crucial
importance to the understanding of Newton's mechanics.
We shall, however, later see that Newtonian mechanics - in spite of its
enormous success - contains an intrinsic weakness. To perceive this, we shall
for a while forget the postulate about absolute space and through the experi-
ence we gain in applying the equation of motion, try to find which coordinate
systems may be used for a description of the phenomena of motion in New-
tonian mechanics.
We begin our discussion with one of the basic tools for comparing mea-
surements taken by observers in uniform motion with respect to each other.
4.1 The Galilei Transformation
By an inertial frame we understand a reference frame within which Newton's

three laws of motion, in their simple classical form, are valid.
In a certain sense the law of inertia is contained within the second law.
When there is no force on a body its acceleration (relative to absolute space!)
is zero, from which it follows that the velocity is constant. Newton used the
law of inertia to explicitly state that no force is required to maintain uniform
motion.
4.1 The Galilei Transformation 85
The so-called heliocentric reference frame has its origin in the cen-
ter of the Sun and coordinate axes that are not rotating relative
to the fixed stars, i.e., the three coordinate axes always point to-
ward the same three fixed stars. Experience has shown that the
heliocentric reference frame to a high degree of accuracy can be
considered an inertial reference frame.
By "experience" we mean in this connection that the predictions based on the

assumption that the heliocentric reference frame is inertial, are borne out by
experiments and astronomical observations. The fixed stars seem motionless
simply because they are so distant. The star next nearest to the Earth is
about 4~ light years (~ 4 x 10 16 m) away. Compare this to the distance of
the star nearest to the Earth, the Sun (distance 150 x 106 km ~ 8.3 light
minutes).
Every reference frame moving with uniform linear motion relative to the
heliocentric frame, is itself an inertial frame. This observation goes hand in
hand with the law of inertia, which tells us that there is no fundamental
difference between rest and uniform translational motion.
Let us investigate this problem a little further.
I y' I'
y
I - - -__ u
x'
z z'
Fig. 4.2.
The Galilei transformation is a prescription for comparing coordinate values

and velocities from two inertial frames (see Figure 4.2). We are considering
two inertial frames I and I', where I' moves with the constant velocity u along
the x-axis of I. We choose the coordinate systems such that the origins and
the axes of the two frames coincide at t = O. You may think of the frame I
as the heliocentric reference frame.
Denote by (x', y', Zl) the coordinate values relative to I' of a given particle
and (x, y, z) are the coordinate values of the same particle relative to I. If
the two coordinate systems move relative to each other as in Figure 4.2, the
Galilei transformation is given by
86 4. The Galilei Transformation
x' x - ut, (4.1)

y' y, (4.2)
z' z, (4.3)
t' t, (4.4)
m' m, (4.5)
q' q, (4.6)
F' F. (4.7)
We have explicitly expressed the relations between time, mass, charge, and
force in the two frames. The first three equations can be condensed into
r' (t') = r - ut . (4.8)
Equation (4.8) states the relation between the position vector r' relative to I',
the position vector r relative to I, and the relative velocity u of the coordinate
systems.
Differentiating (4.8) we get, using t = t',
dr' dr dt dr
-=---u=--u. (4.9)
dt' dt dt' dt
Introducing v' == dr'jdt', and v == drjdt we get, from (4.9),
v = v' +u, (4.10)

where v' is the velocity of the particle in I' and v the velocity in 1. Clearly the
velocity of a particle is a relative quantity, having different values in different
coordinate frames. That a quantity is "relative" means that the quantity has
different values when observed from different reference frames. Mass, charge,
force, and time intervals are, according to the Galilei transformation, absolute
quantities. These quantities have values that are independent of the frame
from which they are observed.
We now assume that a force F is acting on the particle. Then, the equation
of motion in I is
(4.11)
Going back to (4.9) and differentiating again, we find
(4.12)
Note: Equation (4.12) is the essential feature of the Galilei transformation.

When u is constant, the acceleration of a particle is the same when observed
from the two frames. The acceleration is an absolute quantity according to
the Galilei transformation.
Since furthermore m' = m and F' = F, we conclude that if (4.11) is the
equation of motion as seen in I, then the equation of motion in I' is given by
4.1 The Galilei Transformation 87
d2 ,
m,_r_
dt'2 = F' (4.13)
Comparing (4.11) and (4.13) we see that Newton's second law is the same
("looks the same") in all inertial frames. A somewhat more precise expression
is
Newton's fundamental equation of motion is invari-
ant under a Galilei Transformation
This result is called the relativity principle of mechanics. The relativity

principle of mechanics is often formulated in the following manner: there exist
a three-fold infinity of equivalent coordinate frames, all in a state of uniform
linear motion with respect to one another, in which the mechanical laws are
valid in their simple classical form. Or: the fundamental mechanical laws of
motion are the same within all inertial frames.
The physical content of the relativity principle of mechanics is the fol-
lowing: it is impossible, from mechanical experiments alone performed com-
pletely within a given inertial frame, to decide if the frame is "at rest" or in
"uniform linear motion". By mechanical experiments we mean oscillations of
pendulums, collisions between balls, motion of masses connected to springs,
etc. The term "mechanical" is used to emphasize that we - for now - will
exclude experiments involving electromagnetic phenomena. The mechanical
laws of motion are not influenced by uniform translation. One can play tennis
on board a ship sailing with uniform velocity. As long as the ship follows a
straight course with constant velocity the ball will obey the same laws as it
does on land. On land the tennis ball follows the same rules at all seasons
without being influenced by the velocity of the Earth through space. Our
speed relative to the Sun is 30 km S-1 and the speed of the solar system
relative to the center of the Milky Way is 250 km s-1 .
Summing all this up, one can with some justification say that in the
form in which Newtonian mechanics is used today, it rests on the following
assumptions:
1. The law of inertia is valid in the so-called inertial frames.

2. mdv/dt = F.
3. Action = reaction.
4. Experiments show that the heliocentric reference frame
to a high degree of precision is an inertial frame.
5. The Galilei tranformation is valid (the postulate of ab-
solute time is introduced here).
From 4 and 5 we infer that all reference frames moving with a constant
velocity relative to the heliocentric reference frame are inertial frames. To
this we add: after having written the laws of motion (1, 2, and 3 above)
Newton was faced with the problem of finding a reference frame in which
the laws were valid. As one can tell from the postulate of absolute space,
Newton decided that a reference frame fixed in some material body (e.g.
the Sun) could not be the basic reference frame for the fundamental laws of
mechanics. By choosing a reference frame fixed in the Sun, which is, after all,
just one of innumerable stars in the universe, one runs the risk of postponing
the problem instead of solving it: perhaps it will turn out later that this
reference frame is accelerating after all. Instead of a material body as the
universal reference point, Newton chose the abstract concept: absolute space.
Similar remarks can be made with regard to the concept of time. If
Newton had chosen the rotation period of the Earth (the second!) as the
theoretical foundation for the descriptiopn of time, the law of inertia would
have been found not to hold, for there are small irregularities in the rotation
of the Earth around its axis. Again we see from the postulate of absolute
time that Newton chose an abstract concept. We shall consider the problem
further in Chapter 7.
In the following considerations we will stick to the observation firmly
rooted in experience, that to a high degree of appmximation the heliocentric
reference frame is an inerlial frame.
The concepts of "reference frame" and "coordinate system" are not quite
synonymous: if we change the spatial coordinates without reference to time
(say, by rotating the coordinate system) we do not consider this as a change
of reference frame. A given reference frame contains an infinity of different co-
ordinate systems. A given reference frame contains all the coordinate systems
at rest with respect to a given coordinate system.
We will illustrate the physical content of the relativity principle of me-
chanics by describing a simple experiment.
Let us first state as a fact we shall later prove: For motion "over short
distances" and in a "short time span" a laboratory at rest on the Earth can
be considered to be an inertial system. We shall later state precisely what is
meant by "short distances" and "short time spans" .
y
A y' B
I
o o
Fig. 4.3.
Consider two railroad cars on a horizontal and straight section of the

railroad (see Figure 4.3). A is at rest relative to the ground. B moves with the
4.1 The Galilei 'fransformation 89
constant velocity u relative to A. The cars are closed, so the observers cannot
look out. The observers 0 and 0' now carry out experiments with identical
pendulums inside the cars. Each observer measures the time of oscillation
of the pendulum, and its equilibrium position relative to the walls of the
cars. They can discuss their observations with each other, over the radio. We
disregard all trivial effects like vibrations from the rails, etc. and ask: can the
two observers decide from the pendulum experiments alone, who is at "rest"
and who is in "motion" with the constant speed u ? In particular, can they
determine the magnitude of u?
The anSwer to these questions is no. The mechanical laws are invariant
going from one inertial frame to another. The behavior of the pendulae is
completely the same in the two cars.
It is an experimentally verified fact that by mechanical experiments alone
- performed completely within a given inertial system - it is not possible
to decide if the coordinate system is "at rest" or it is in uniform motion
along a straight line. This experience finds its mathematical expression in
the invariance of Newton's second law under a Galilei transformation.
The important point in this connection is the experimental verification
and not that the Galilei transformation perhaps can be considered self evi-
dent.
Let us illustrate the relativity principle of mechanics by some further re-
marks. The fundamental law of mechanics is about changes in velocity. To
the description of any physical system One can add a constant overall veloc-
ity without changing the fundamental behavior of the system. For instance:
by studying how the planets move around the Sun, it is not possible to in-
fer if the solar system as a whole "moves through space". According to the
Newtonian laws there will be no effect of a translational motion of the Sun
through astronomical space. Newton remarks that "The motion of bodies
among themselves is the same in space, whether that space is itself at rest
relative to the absolute space or moving at a uniform velocity in a straight
line". The fundamental mechanical laws contain no reference to an absolute
velocity. The laws are independent of uniform translational motion.
There is an important objection to the entire concept of absolute space.
We get into trouble if we wish to determine Newton's absolute space by means
of mechanical experiments. Suppose we somehow had arrived at the conclu-
sion that a certain reference frame R was at rest with respect to Newton's
absolute space. As we have just seen, any frame in uniform linear motion with
respect to R can with equal right be claimed to be the one at rest with re-
spect to Newton's absolute space. The outcome of any experiment in the two
frames will be exactly the same. Neither frame can be preferred as far as the
fundamental laws of mechanics are concerned. A body at rest in one system
will, however, be a body in uniform linear motion in the other frame. If an
observer in one frame claimed that a particle was at rest in absolute space,
an observer in the other frame would claim that the particle was moving with
constant velocity along a straight line in absolute space.
The relativity principle of mechanics thus detracts some of the "absolute-
ness" from Newton's absolute space. Newton was aware of this but, as we
shall later see, he had his very good reasons to introduce the absolute space.
4.2 Galileo Speaks
The space and time coordinate transform used in this chapter justly bears
the name of the Galilei transformation. Galileo was fully aware of the fact
that the mechanical laws are independent of an even translatory motion. We
quote from the book by Galileo: Dialogue Concerning the Two Chief World
Systems:
Shut yourself up with some friend in the main cabin below decks on
some large ship, and have with you there some flies, butterflies, and
other small flying animals. Have a large bowl of water with some
fish in it; hang up a bottle that empties drop by drop into a wide
vessel beneath it. With the ship standing still, observe carefully how
the little animals fly with equal speeds to all sides of the cabin. The
fish swim indifferently in all directions; the drops fall into the vessel
beneath; and, in throwing something to your friend, you need throw
it no more strongly in one direction than another, the distances being
equal; jumping with your feet together, you pass equal spaces in every
direction.
When you have observed all these things carefully (though there is no
doubt that when the ship is standing still everything must happen
this way), have the ship proceed with any speed you like, so long
as the motion is uniform and not fluctuating this way and that.
You will discover not the least change in all the effects named, nor
could you tell from any of them whether the ship was moving or
standing still. In jumping, you will pass on the floor the same spaces
as before, nor will you make larger jumps toward the stern than
towards the prow even if the ship is moving quite rapidly, despite
the fact that during the time that you are in the air the floor under
you will be going in a direction opposite to your jump. In throwing
something to your companion, you will need no more force to get it
to him whether he is in the direction of the bow or the stern, with
yourself situated opposite. The droplets will fall as before into the
vessel beneath without dropping towards the stern, although while
the drops are in the air the ship runs many spans. The fish in their
water will swim toward the front of their bowl with no more effort
than towitrd the back, and will go with equal ease to bait placed
anywhere around the edges of the bowl. Finally the butterflies and
4.2 Galileo Speaks 91
the flies will continue their flights indifferently toward every side, nor
will it ever happen that they are concentrated toward the stern, as if
tired out with keeping up with the course of the ship, from which they
will have been separated during long intervals by keeping themselves
in the air. And if smoke is made by burning some incense, it will be
seen going up in the form of a little cloud, remaining still and moving
no more toward one side than the other.
The cause of all these correspondences of effects is the fact that the
ship's motion is common to all the things contained in it, and to the
air also. That is why I said you should be below decks; for if this took
place above in the open air, which would not follow the course of the
ship, more or less noticeable differences would be seen in some of the
effects noted. No doubt the smoke would fall as much behind as the
air itself. The flies likewise, and the butterflies, held back by the air,
would be unable to follow the ship's motion if they were separated
from it by a perceptible distance. But keeping themselves near it,
they would follow it without effort or hindrance; for the ship, being
an unbroken structure carries with it a part of the nearby air. For a
similar reason we sometimes, when riding horseback, see persistent
flies and butterflies following our horses, flying now to one part of
their bodies and now to another. But the difference would be small
as regards the falling drops, and as to the jumping and throwing it
would be quite imperceptible.
Sagredo: Although it did not occur to me to put these observations to
the test when I was voyaging, I am sure that they would take place in
the way you describe. In confirmation of this I remember having often
found myself in my cabin wondering whether the ship was moving
or standing still; and sometimes at a whim I have supposed it going
one way when its motion was the opposite. Still, I am satisfied so
far, and convinced of the worthlessness of all experiments brought
forth to prove the negative rather than the affirmative side as to the
motion of the Earth.
Example 4.1. Velocity Transformation. Let us. consider a particle P

moving with constant velocity v relative to an inertial frame I (see the figure).
Let us assume that v is in the (x, y) plane with an angle 0 to the x-axis.
The inertial frame I' moves with the constant velocity u along the x-
axis of I , and the coordinate axis of the two systems are parallel. Find the
magnitude of v' and the angle 0' it forms with the x' -axis.
y I y' I'
P~-- u
x x'
Solution. From the Galilei transformation:

v'x Vx - u, (4.14)
v'y vy , (4.15)
v'z Vz . (4.16)
In our case:
Vx vcos(O) , (4.17)
Vy vsin(O) , (4.18)
and
v'x v' cos(O') , (4.19)
v'y v' sin( 0') . (4.20)
Thus, (4.14) states that
v' cos(O') = v cos(O) - u , (4.21 )
and (4.15) that
v' sin( 0') = v sin( 0) . (4.22)
Dividing (4.22) with (4.21) we find
, sin( 0)
tan( 0 ) (4.23)
=
cos (0) -u /v.
Squaring (4.21) and (4.22), and adding them we find
Iv'l= v' = Jv 2 + u2 - 2uv cos(O). (4.24)

These formulas express the vectorial addition of v and -u. Geometrically we
have:
V' =v-u
=v+(-u)
4.3 Problems 93
4.3 Problems
Problem 4.1. A particle moves along the x-axis of frame I according to the
equation x(t) = 5t 2 + 2t + 4.
Find the position, velocity and acceleration of the particle, as functions
of t in the frame I'. Take l'to move with the velocity u = 2 m S-1 along the
x-axis of I.
Problem 4.2. A spring cannon stands in the chimney of a toy train. The
train drives along a straight section with constant velocity v. The spring
cannon shoots a heavy ball straight upwards from the chimney.
"""""~""
(1) Where does the ball land (disregard air resistance)?
(2) Describe qualitatively the trajectory of the ball in a coordinate system
fixed in the laboratory.
(3) Describe qualitatively the trajectory of the ball in a coordinate system
fixed on the train.
5. The Motion of the Earth
As a preparation for later chapters we shall now review the description of the
motion of the Earth relative to the fixed stars.
Through a few examples we first demonstrate the application of vector
calculus to the analysis of the motion of a rigid body. We then apply these
results to the motion of the Earth.
5.1 Examples
Example 5.1. Vectors and the Rotation of a Rigid Body
:A'
I
I
I
:A
I
I
I
A rigid body rotates with the constant angular velocity w around a fixed axis
A-A'. The angular velocity vector w is defined as a vector along the rotational
axis pointing in the direction where the rotation and w defines a right-hand
spiral. The magnitude of w is / w /. See the figure.
Let P be some point fixed on the rotating body.
(1) Show that the velocity of Pis
v=wxr . (5.1)
where r is a vector from some origin 0 on the rotation axis to P.

96 5. The Motion of the Earth
(2) Show that the acceleration a of the point P is
a = w x (w x r) = _w 2 p , (5.2)
where p is the projection of r on a plane perpendicular to w. See the
figure.
Solution. During the rotation, the point P performs a uniform circular motion
with the angular velocity w in a plane perpendicular to w. The radius in that
circular motion is p.
(1) The velocity v of P has the magnitude Iv I = w p and is in the direction
along the tangent to the circle. We obtain a vector in this direction by
taking the vectorial product (cross product) w x p, and the magnitude
of this vector is Iw x pi = w p. Thus:
v = w x p. (5.3)
Note: when we construct the vector product of two vectors we may trans-
late them to a common origin. The area of the parallelogram spanned by
wand p is the same as the area of the parallelogram spanned by wand
r (draw the vectors from a common origin) . This means that
Iwxrl=lwxpl·
Since the direction of w x rand w x p is also the same, we conclude
that
v = w x r = w x p. (5.4)
(2) The acceleration of P has the magnitude a = w 2 p (see Example 1.2) and
is directed toward the center of the circle, i.e., toward the axis of rotation.
Therefore
a = _w 2 p. (5.5)
Since wand v are mutually perpendicular, we find that the magnitude
of w x v is I w x v I = wv = w 2 p. The vector w x v then points in the
direction of - p . Therefore:
w x v = _w 2 P = a. (5.6)
Substituting (5.4) into (5.6) we get:
a = _w 2 p = W X (w x r). (5.7)
Alternatively, (5.7) can be seen as a result of direct differentiation of

(5.4) :
dv d dr
a dt = dt (w x r) = w x dt
w x v = w x (w x r).
5.1 Examples 97
If the rotation of the body is not with constant angular velocity, the
acceleration of P is not given by (5.7). But the identity
(5.8)
still holds since it is a purely geometrical formula, valid for any two
vectors rand w. See the figure below.
- - - OJ
Ar
I
I
I
Example 5.2. Angular Velocities in the Solar System
(1) Find the angular velocity Wl in the daily rotation of the Earth around its
axis. The angular velocity is to be taken with respect to the fixed stars,
i.e., with respect to the heliocentric reference frame.
(2) Find the angular velocity W2 in the annual motion of the Earth around
the Sun. W2 describes the motion of the center of Earth around the center
of the Sun. We take this motion to be a uniform circular motion.
(3) What is the angle between the two rotation vectors Wl and W2 ?
Solution.
axis
(1) Consider the daily rotation of the Earth. The time T from some point P
on the Earth is directed towards the Sun till P again is directly towards
the Sun (see the figure) is 24 hours, or 86400 seconds. During this period,
the Earth has moved relative to the center of the Sun. During the time
T, the point P has turned relative to the fixed stars, not only by 21l' but
by an angle
21l' 366.24
() = 21l' + 365.24 = 21l' 365.24 .
(Here we use 1 year = 365.24 solar days, 1 solar day = 24 hours.) The
angular velocity of the Earth relative to the fixed stars, i.e., the angular
velocity relative to the heliocentric reference frame, is therefore
() 21l' x 366.24 -1
T 86400 x 365.24 s
7.29 X 10- 5 s-1
A stellar day Tst is thus determined by:
Tst = 21l' /Wl = 23.93 hours ~ 23 hours 56 minutes.

(2) 1 year ~ 3.16 x 107 s. From this:
W2 ~
21l' s-
1
~ 1.99 X 10- 7 s-1 .
3.16 X 10 7
(3) About 23.5°.
Questions.
(1) Determine the velocity V2 and the acceleration a2 of the center of Earth
in the annual motion of the Earth around the Sun; V2 and a2 are to be
determined relative to the heliocentric reference frame.
(2) Determine the velocity VI and the acceleration al of a point P on the
surface of Earth at 60° latitude. VI and al are to be determined relative
to a coordinate system having the origin in the center of the Earth and
axis fixed relative to the fixed stars. Such a coordinate frame is known as
the geocentric reference frame. This frame is not in uniform translation
relative to the heliocentric reference frame, and the geocentric reference
frame is thus not an inertial system. The absolute geocentric reference
frame is falling freely in the gravitational field of the Sun.
(3) Consider a coordinate system with its origin at the center of the Earth
and axes fixed relative to the Earth. Such a coordinate system, in which
the Earth is at rest, is sometimes called the Earth system. Give a qual-
itative description of how this coordinate system moves relative to the
heliocentric reference frame.
5.2 Problems 99
Answers.
(1)
IV21 Rw2 = 3.0 X 10 4 ms- 1 = 30 kms- 1 .
Ia21 Rw~ = 5.9 X 10- 3 m s-2 (towards the center of the Sun).
(2)
IV11 Iw x rcos(600) 1= 836 km/hour
la11 Iw x v11 = 1.7 x 10- 2 ms- 2 (direction as on the figure)
Note. We do not sense the velocity V1 = 836 km/hour. This velocity is com-
parable to that of a jet airliner, inside which we do not sense the velocity
either. Our velocity relative to the center of the Sun is V2 = 30 km S-l and
our velocity relative to the center of the galaxy is about 250 km s-l. The
acceleration in these motions is very small so that at any given moment the
motion is almost a uniform translation.
Newton's laws of mechanics are of such a character that according to
these laws we do not feel a translational velocity relative to the fixed stars
("the absolute space" 7). On the other hand, as we shall see, an acceleration
relative to the stars (the heliocentric reference frame) will make itself felt
immediately. 6
5.2 Problems
Problem 5.1. A rigid body rotates with 240 revolutions per minute around
an axis pointing into the first octant as shown on the figure.
The axis forms an angle of 45° with the x-axis and 60° with the y-axis.
(1) What is the velocity of a point P having the position vector r = (0,1,2) m?
(2) What is the acceleration of P?
Problem 5.2. Calculate the acceleration of a point on the equator of the

Earth. The acceleration should be determined relative to the geocentric refer-
ence frame (origin in the center of the Earth, axes fixed relative to the stars) .
6. Motion In Accelerated Reference Frames
6.1 Newton's 2nd Law

Within Accelerated Reference Frames
This section contains substantial mathematical derivations. The results -
not the derivations - are crucial to an insight into fundamental aspects of the
Newtonian theory of motion. The main result can be expressed in a simple
and elegant manner, and the physical content of this important result is quite
transparent (see Equation (6.36)).
Our problem is the following. We wish to find the mathematical equations
that describe the motion of a particle in the coordinates of frame S, which
moves in an arbitrary way relative to the inertial frames. Expressed differ-
ently: we wish to find how Newton's second law transforms from an inertial
frame to a frame S whose origin 0 performs an arbitrary translatory motion
relative to I, while at the same time S is rotating around some axis through
O. In textbooks on geometry it is shown that the most general motion of a
rigid body (a reference frame) can always be resolved into a translation of a
point 0 and a rotation around an axis through O. This result will also be
apparent after the reading of this section.
Fig. 6.1. The coordinate frame S moves in an arbitrary way relative to the inertial
frame I
Look at Figure 6.1. The reference frame I is an inertial frame. In our

terminology at this stage this means simply that I is the heliocentric frame or

102 6. Motion in Accelerated Reference Frames
some frame moving with constant velocity relative to the heliocentric frame.
We shall later discuss the peculiar position of the heliocentric reference frame
and study in depth why this coordinate frame is a good approximation to an
inertial frame.
Consider a particle having the mass m and being influenced by a force F.
The equation of motion for the particle relative to lis, as we have seen,
ma=F.
We now seek the equation of motion expressed in the coordinates of the S
frame, where S is moving in an arbitrary way relative to 1.
One can think of S as an accelerated spaceship that, as we observe it,
rotates around its center. We wish to find the laws of motion for a particle
as they appear to a passenger of the spaceship. Planet Earth, for instance, is
just such a spaceship moving in a circular path and rotating relative to the
heliocentric reference frame.
In order to find the equation of motion as seen from S, we must find the
connection between the kinematic description of the motion as seen from I
and as seen from S.
To this end, we denote the vectorial position of the particle relative to I
by R and the vectorial position relative to S by r (see Figure 6.1).
We begin by finding the connection between the acceleration relative to
I, a = (d 2 R/dt 2 )I, and the acceleration relative to S, a r = (d 2 r/dt 2 )s. We
are assuming here, based on Newton's postulate of absolute time, that the
time t is common to both observers.
As seen from the inertial frame I the particle is at the position R =
(X, Y, Z) where X, Y and Z are the components of R along the axes of 1.
For the velocity Va and the acceleration a relative to I we have:
va == (~~)I CR.)I = (X, Y, Z), (6.1)
a == (~t~)I (R)I = (X, Y, .2'). (6.2)
We stress that the differentiations above are performed by an observer at rest

with respect to 1.
As seen from the coordinate frame S the particle is at the position r =
(x, y, z) where x, y, and z are components of r along the axes of S. For the
velocity v r and acceleration a r in S we have:
vr (~:) s
== (r)s = (x, y, i), (6.3)
ar == (~;~) s (f)s = (x, ii, z). (6.4)
Similarly, the differentiations here are performed by an observer at rest with

respect to S.
6.1 Newton's 2nd Law Within Accelerated Reference Frames 103
If we denote the three unit vectors alonf the cartesion axes of S by i, j,

and k respectively, we can express position, velocity, and acceleration as:
r xi + yj + zk, (6.5)
xi + yj + ik, (6.6)
xi + yj + ik. (6.7)
Here v rand a r are often called the relative velocity and the relative acceler-
ation, respectively. In contrast, Va (6.1) and a (6.2) are called the absolute
velocity and the absolute acceleration. This is not necessarily a good termi-
nology. For, what is absolute velocity?
The configuration of the coordinate system S relative to I is given by:
(1) the position of the origin 0 of S, denoted by its principal vector Ro in 1.
(2) the direction of the three unit vectors i, j, and k. These three unit vectors
determine the axes of S as they are seen from 1.
The position vector R for a particle can be written (from I):
R=Ro+r, (6.8)
or
R(t) = Ro(t) + x(t) i(t) + y(t)j(t) + z(t) k(t), (6.9)
where we have introduced the explicit time dependence (see Figure 6.1). By
differentiating (6.9) with respect to time (from I), we find
It = CRo + xi + yj + zk) + (xi + yj + ik). (6.10)
The vector It is the velocity of the particle with respect to I, the velocity
we denote Va' We have split the terms on the right-hand side into two parts.
The terms in the first parenthesis are denoted by the comoving velocity Vrn:
(6.11)
The comoving velocity does not contain x, y, or z. The vector Vrn denotes the
velocity relative to I that the particle would have if it was fixed in S, i.e., if
it was moving only because of the overall motion of S.
The second parenthesis in (6.10) is (xi + yj + ik), i.e., the velocity of the
particle referred to S, the relative velocity Yr' Let us write (6.10) as
(6.12)
If you differentiate (6.10) again (be sure to get all terms!), you will find
R = (Ro + xi + YJ + zk) + (xi + yj + ik) + 2(xi + yj + ik). (6.13)
We view the right-hand side of (6.13) as the sum of three terms, as indicated
by the parenthesis. Let us introduce a notation: the acceleration of the par-
ticle relative to I (the so-called absolute acceleration) is R == a. The first
parenthesis on the right-hand side of (6.13) is called the comoving accelera-

tion Wm:
Wm = (Ro + xi + YJ + zk). (6.14)

The comoving acceleration W m is the acceleration the particle would have
relative to I if the particle was at rest at its instantaneous position (x, y, z)
in S.
The second parenthesis is the acceleration within S, the relative acceler-
ation a r :
a r = (Xi + ijj + zk). (6.15)

The third term on the right-hand side of (6.13) is called the Coriolis acceler-
ation (after the French physicist G . Coriolis). We have:
WeD = 2(xi + yj + ik). (6.16)

The term is complicated because it contains time derivatives of both the
relative coordinates (x, y, z) and of the unit vectors i, j, and k. We can now
write (6.13) in the compact form:
a = a r + W m + WeD' (6.17)
In order to use this fundamental relation between the absolute acceleration a
and the relative acceleration a r it is necessary to derive an expression for the
dependence of W m and WeD on the motion of the reference frame S relative
to the inertial frame 1. We will first demonstrate that the comoving velocity
of the particle can be split into two velocities, a translation and a rotation
about an axis through the origin of S,
Vm = Ito +w x r, (6.18)
and thus determine w. Next we shall express Wm and WeD in terms of w.

In order to rewrite the expression (6.11) for the comoving velocity, we will
express i, j , and k in a more transparent manner. The calculations are ex-
tensive , but the result is simple.
We will use the fact that S moves as a rigid body relative to 1. The
mathematical formulation of this fact is that the three unit vectors, i, j, and
k, that are fixed in relation to S, remain mutually orthogonal and of unit
length no matter how S moves. Only the directions of the vectors relative to
I change with time. We therefore have the conditions:
i·i=j·j=k·k=1, (6.19)
and
i·j =i·k=j·k=O. (6.20)
By differentiating (6.19) and (6.20), we find that
i.i=j·j=k.k=O, (6.21)
and
i·j+i·j=i.k+i.k=j·k+j·k=O. (6.22)
Every vector can be written in terms of three orthogonal vectors. We write
the three vectors i,j, and k as:
ani + a12j + a13k,

j a21 i + a22j + a23 k , (6.23)
k a31 i + a32j + a33 k .
Note. The coefficients a nm may - as well as the vectors i, j, and k - depend
on time. Of the nine coefficients a nm , only three are independent. This is a
direct consequence of the six equations (6.21) and (6.22). From (6.21) and
(6.23):
an = a22 = a33 = O.
From (6.22):
Let us change the notation:
-W y , (6.24)
We can then write (6.23) as:
wzj - wyk,
j -wzi +wxk, (6.25)
k wyi - wxj .
These expressions can be made more transparent. With this in mind we

introduce the vector
(6.26)
By directly applying the definition of the cross product we find that i x j = k,
k x i = j, and j x k = i (the coordinate system is right handed) . By means
of these results and using the definition of w (Equation (6.26)) we see that,
for instance, w x i = (wxi + wyj + wzk) x i = -wyk + wzj. It is thus easily
shown that (6.23) can be written in the compact form:
w x i,
j w x j, (6.27)
k w x k.
By means of the vector w we can thus describe how the vectors i,j, and k
vary in time. Note: (6.27) expresses the fact that the coordinate system S is
in rotation. For instance: the rate of change of i, which is i, is perpendicular
to i (the cross product). The length of i does not change, only the direction.
The motion of S relative to the inertial frame I can be described by the rate
of change of the position vector Ro to the origin of S, and a specification of
the axis of rotation by means of the rotation vector w through the origin of
S. Compare with the motion of the Earth relative to the heliocentric system,
as described in Chapter 5.
By means of (6.27), we can now write (6.11) as:
Ro + x (w x i) + y (w x j) + z (w x k) (6.28)
ito + w x (xi + yj + zk)
ito+wxr. (6.29)
A particle at rest relative to S has the velocity Vm (the comoving velocity)

relative to I, as indicated by (6.29). This expression for the comoving velocity
shows directly that the velocity relative to I of a particle at rest in S can be
regarded as a superposition of two velocities: Ro (a translation of the origin
of S), and w x r (a rotation about the axis in the direction of the vector w
through the origin of S). Note: in general, w may depend on time.
The absolute velocity (6.12) of the particle becomes:
va=r+Ro+wxr. (6.30)
Applying (6.27) we can write the Coriolis acceleration (6.16) as
W co 2 ( xi+ yj + zk)
2 (x (w x i) + y (w x j) + z (w x k))
= 2(w x r) = 2(w x v r ). (6.31)
By differentiation of (6.27) with respect to time, we have that
i wxi+wxi
w x (w x i) + w xi, (6.32)
and similarly for j and k.

By use of (6.32) the comoving acceleration Wm, specified in equation
(6.14), may be written as
Wm = Ro + w x (w x r) + w x r, (6.33)
where Ro is the acceleration vector in I of the origin 0 of S.

The last term in (6.33) depends on the rate of change of w. This term will
disappear whenever S is rotating around an axis that is fixed in space and
with an angular velocity w == Iw I that is constant in time. This happens to
be approximately true for the motion of the Earth around its axis. There is
actually a small precession around the normal to the ecliptic (i.e., the plane
that contains the orbit of Earth) with a period of about 25800 years.
The second term in (6.33) is called the centripetal acceleration.
ill
Fig. 6.2. Illustrating the centripetal acceleration in an accelerated (rotating) ref-

erence frame. The vector p is defined in the text
The centripetal acceleration is often written in the form

w x (w x r) = _w 2 p , (6.34)
where the vector p is the projection of the radius vector r onto a plane per-
pendicular to w . The centripetal ("center seeking") acceleration is therefore
directed toward the rotation axis, and grows in magnitude as the distance
from the axis (see Figure 6.2).
We can now write the complete equation of motion for a particle as seen
from the frame S. Recall that the basic equation of motion relative to the
frame I is
ma=F.
We now assume that the impressed force F acting on the particle is the same
in S as it is in I. Using the results above, we can write:
m(a r + Wm + WeD) = F, (6.35)
or, writing out the explicit expressions derived for Wm and WeD' we have
d2 r -
m -2 = F - mRo + mw 2 p - 2m(w x r) - mew x r). (6.36)
dt
Here, r, r, and r == (d 2 r/dt 2 ) denote the position, velocity, and acceleration
of the particle as seen from the frame S. In other words, (6.36) is the equa-
tion of motion within the frame S. Ro is the translational acceleration of
S, and w is a vector which describes the rotation of S. The set Ro and w

completely describe the acceleration of S relative to 1. Equation (6.36) is the
basic equation of motion for the particle from within the reference frame S.
Equation (6.36) shows that we formally can claim that Newton's second
law is valid within accelerated frames, if we interpret the last four terms
on the right-hand side as forces acting on the particle in addition to F.
These terms are fictitious forces. In Newtonian mechanics, fictitious forces
are distinguished from "actual" forces by the fact that the fictitious forces
depend only on the motion (Ro and w}of the frame S, the position (p), and
the velocity (r) of the particle, and on the inertial mass of the particle.
The presence of an actual force F can always be related to an interaction
between the particle and another material body. For example, if an electrical
force acts on the particle (which must then be charged) there is an interaction
with other electrically charged particles. If a gravitational force acts, there
is an interaction with some body that has gravitational mass. The fictitious
forces cannot in any obvious way be related to an interaction with other
bodies.
By introducing the fictitious forces, an observer in S can still claim that
Newton's first and second laws are valid. Note, however, that the third law,
the law of action and reaction, is not valid for the fictitious forces; for there
is no other body upon which a reaction force can act.
Let us emphasize that the fictitious forces have important similarities to
gravitational forces: like these, the fictitious forces are proportional to the
mass of the body.
In Chapter 7 we will go deeper into the theoretical interpretation of equa-
tion (6.36). Here we shall first examine the physical consequences of (6.36)
in some special cases.
6.2 The Equivalence Principle of Mechanics
Let us first assume that w = 0, W = 0, and Ro = (a, 0,0) == a. These as-

sumptions mean that the reference frame S has a constant linear acceleration
a relative to the inertial frame I (see Figure 6.3).
Fig. 6.3.
Our fundamental equation (6.36) now assumes the form:

6.2 The Equivalence Principle of Mechanics 109
d2 r
m dt 2 = F - ma. (6.37)
Equation (6.37) is the equation of motion for the particle, seen from the frame
s.The total force F' on the particle, as evaluated in S, is given by
F' =F-ma.
If F = 0, the total force is given by the fictitious force alone
F' = -ma = m( -a).

We know that at the surface of the Earth where the gravitational field is
nearly homogeneous, the gravitational force on a particle is given by
F=mg,
where g is the acceleration due to gravity.
By comparing the last two equations, we see that the fictitious force acts
like a force of gravity, being directly proportional to the mass of the particle.
The minus sign reminds us that the fictitious force is directed opposite to the
acceleration of the reference frame S.
We collect the result (6.37) and the above considerations in the so-called
equivalence principle of mechanics, which we can formulate as:
The motion of a particle, as seen by an observer in a refer-

ence frame S which is accelerated with constant acceleration
(a) relative to the inertial frames, will be as if a uniform
gravitational field (-a) acts in addition to the "natural"
forces.
or, simply:
An acceleration to the right is equivalent to a uniform grav-

itational field to the left.
The equivalence principle of mechanics - sometimes called the equivalence

principle of Einstein, or the strong equivalence principle - has its origin in the
proportionality between gravitational and inertial mass. The mass appearing
on either side of (6.36) is inertial mass. As the inertial and gravitational
masses are equal, the fictitious force may be conceived as a gravitational
force (see Chapter 3).
The similarity between gravitational and fictitious forces seems somewhat
coincidental in the above considerations. It remained unnoticed for two hun-
dred years. It was to become a major foundation for the general theory of
A B
Fig. 6.4. Illustrating the equivalence principle. The railroad car B is accelerated
relative to the car A, which represents an inertial frame. The cars are closed, so
that the observers cannot look out
relativity put forth by Albert Einstein in 1915. The theory of general rela-
tivity is basically a theory of the phenomenon of gravitation.
Look at Figure 6.4. As in Chapter 4, for motion over "short distances" and
in "short time spans" a laboratory at rest on the surface of the Earth can
be considered to be an inertial frame. A and B are two railroad cars on the
surface of the Earth where the gravitational field is uniform. Car A is at rest
relative to the ground, standing on a horizontal and straight railroad track.
Car B has a constant acceleration relative to the ground. Both cars are closed
so that an observer cannot directly look .at the outside world. Compare this
with Chapter 4.
Equation (6.36) shows that Newton's second law is not invariant under
transformations to accelerated reference frames. This is a mathematical ex-
pression of the following experimental fact: an observer in the car B, i.e., an
observer at rest within the accelerated reference frame S can, by perform-
ing mechanical experiments completely within his railroad car, decide that
he is accelerated relative to the inertial systems. Specifically he can measure
the angle ¢ that a pendulum, hanging from the ceiling and being at rest,
forms with the end walls. Through this measurement he can determine the
acceleration a .
The pendulum has its equilibrium position parallel to the total gravita-
tional field inside B, which is g - a. We then have tan(¢) = lal/igi. From
this experiment alone, the observer in B is unable to tell how much of the ac-
celeration of gravity is caused by a (nearly) homogeneous field of the planet,
and how much is due to a linear acceleration relative to an inertial frame.
The observer in B can determine the magnitude of the total acceleration field
by measuring the time of oscillation of his pendulum, given by
L ] 1/2
T=27r [ -
geff
In particular he would find in the above case

1/2
L )
T = 27r (
J g2 + a 2
6.3 The Einstein Box 111
The equivalence prin cip le of mechanics telIs us that the observer in B is

unable to infer from mechanical experiments alone how much (if any part)
of a homogeneous field of gravity is caused by gravitational attraction, and
how much is caused by his own reference frame being accelerated relative to
the inertial frames.
We saw that the relativity principle of mechanics brought the idea of
absolute space into some doubt. It is fair to ask how much existence absolute
space has, when no experiment can decide if one is at rest or in uniform
motion relative to absolute space. Newton's reply was that the resistance that
alI bodies has against being accelerated must be interpreted as an action of
absolute space. Acceleration is a change of velocity relative to absolute space
and for this, says Newton, a force is needed.
Below we shall analyze the equivalence principle of mechanics in further
detail.
6.3 The Einstein Box
1
B
*
A
1"/\
Fig. 6.5. T he Einstein b ox
Einstein considered a closed elevator box S far from alI gravitating masses (see
Figure 6.5). The box is assumed to have a constant acceleration a relative to
the inertial frame I. We do not concern ourselves with how the box acquired
this acceleration or how it is maintainedj the important feature is that the
acceleration has the constant value a. Inside the box, an observer B is present.
Two masses, m and Mare hanging from equalIy long strings, as shown on
Figure 6.5. B has a balance and a spring scale so he can compare masses and
measure forces.
Let us first describe the motion of the masses as seen by an observer A

at rest in the inertial frame I, and then as seen by observer B who is at rest
in frame S, i.e., at rest in the elevator box.
The observer A in the inertial frame I. The two masses both have an
acceleration a. They must consequently be under the influence of some force.
This force must be a "real" force since we are describing their motion from
an inertial frame where no fictitious forces act. The masses are acted on by
the string forces Tm and T M . The motion is known, so from Newton's law
we infer the magnitudes of the string forces:
Tm = ma, (6.38)
(6.39)
The observer B in the noninertial frame S. By carrying out mechanical

experiments (e.g., by measuring the period of a pendulum) B finds that a
homogeneous gravitational field exists inside the box. This is a consequence
of the equivalence principle of mechanics. B will measure a gravitational field
gs = -a. The two masses remain at rest relative to B. The conclusion is: both
masses are under the influence of a (fictitious) gravitational field, downwards.
The force resulting from this (fictitious) field is equal but opposite to the
string force. Thus B finds no net force on each mass:
Tm + m(-a) 0, (6.40)
TM+M(-a) o. (6.41 )
Equation (6.38) and (6.40) and Equations (6.39) and (6.41) are formally
identical - but the interpretations are profoundly different .
Let us consider a new situation inside the elevator box. At some time
t = 0 the two strings holding m and M break. What happens? Let us again
describe the events as seen and interpreted by the two observers A and B.
The observer A in the inertial frame I. No forces act on the masses

m and M when the string tensions disappear. From Newton's s econd law,
the masses will now stop accelerating. The floor of the elevator box, however,
will (by assumption) continue to accelerate upwards with acceleration a . The
floor (a distance L below when the strings break) catches up with the two
masses at a time determined by
L = ~at2
2
or t = {2I.
V-;;
Note: this result is purely kinematic or, if you like, purely geometric. No
gravitational forces are involved in the discussion (for the geometry of the
problem, see Figure 6.5 and compare with the Galileo experiment as discussed
in Chapter 1.).
The observer B in the accelerated frame S. The two strings of equal

length break at time t = o. The two masses have the same height (L) above
the box floor, and start falling under the influence of the (fictitious) gravita-
tional field in the elevator. If B orients his coordinate system with the y-axis
along the direction of a, the fictitious field is in the negative y-direction. The
masses follow the equations of motion:
my -rna, (6.42)
My -Ma. (6.43)
From these two equations, B finds that the two different masses fall with the
same acceleration toward the box floor. By integrating these equations, B
finds that the masses hit the floor at a time given by
t= ff:
-
L
a
.
All masses inside the elevator box will, as seen by the observer B, fall with
the same acceleration, just as if they fell in a uniform gravitational field. As
seen by the inertial observer A, this is entirely obvious!
In the gravitational field of the Earth all masses - regardless of their mass
or composition - fall with the same acceleration g. As discussed in Chapter 3
this expresses the fact that the inertial and gravitational masses are equal. We
can now begin to see that a deep law of nature is hidden in the equivalence of
the inertial and the gravitational mass. Gravitation and inertia do not seem
to be separate properties of matter, but rather two different aspects of a more
fundamental and universal property of space and material particles.
Questions.
(1) What is the total gravitational field at the center of the freely falling
satellite and in the freely falling elevator box in Examples 3.1 and 3.2,
respectively?
[Answer: zero. The particle in the freely falling satellite and the man in the
freely falling elevator box are said to be weightless relative to the satellite
and the elevator box respectively.]
(2) On the spring scale in the Einstein box in Figure 6.5 is a 1 kg mass. The
scale is calibrated on the surface of Earth. What does the scale show if
the acceleration a has magnitude 3 m S-2 ?
[Answer: 0.306 kg]

We will discuss the equivalence principle of mechanics further in Chapter

7. Let us emphasize here that the perfect equivalence between "accelera-
tion fields" and "gravitation fields" only hold for uniform or, more precisely,
homogeneous fields. An easy way to realize this is to consider a box with
dimensions comparable to those of the Earth (see Figure 6.6).
I
I
I
I
I
Fig. 6.6. Freely falling elevator boxes. Each elevator box contains two equal masses.
The gravitational force from the Earth on each particle is suggested
The gravitational field from the Earth inside each of the freely falling boxes
is not homogeneous.
If two particles are released from rest inside the box, they will either (a)
separate or (b) approach each other as the box falls freely in the field of
the Earth. By observing the behavior of the particles, an observer inside the
box can thus, without looking outside, conclude that he falls freely in an
inhomogeneous gravitational field. His elevator box can not be used as an
inertial frame.
We shall return to a detailed study of this problem in Section 6.5 (tidal
forces).
Example 6.1. Balloon in Accelerated Frame
A +g B
I"Ql'Q,~mmJuQ;g,~,
Toy balloons are tied to the floors of the two railroad cars A and B (see the
figure). The car A is at rest (or in uniform motion relative to an inertial
frame). The balloon inside A is parallel to the end walls. The car B has the
acceleration a relative to A. Find the equilibrium direction of the string to
the balloon in B.
Solution. The string of a balloon is always parallel to the local acceleration
of gravity, g - a, in B. The string will form an angle e with the floor where e
can be found from tan(e) = Igi/ial. A balloon will always float "upwards"
in the local gravitational field. L.
Example 6.2. Mass on an Oscillating Plate
m p
o
x = XQ cos rot
A body of mass m rests on a horizontal plate P. The plate oscillates horizon-

tally, i.e., perpendicular to the gravitational field of the Earth. The amplitude
in the oscillations is Xo = 15 cm, and the period T is 1 second. The friction
between the body and the plate is so that the body just stays on the plate
without slipping. Find the coefficient of friction JL between the plate and the
body. (Note: the maximal friction force between the plate and body is JLN
where N is the magnitude of the normal force).
Solution. In the reference frame where the plate is at rest we must include a
fictitious gravitational field. This field is always opposite to the' acceleration
of the plate in the lab frame (an acceleration to the the right gives a gravity
field to the left). The maximal value of the acceleration is W2xo. The condition
for the body just not to slip is then
xow 2 7l' 2X o
JL= - - = 4 - - .
g T2g
With the numerical values in the problem, JL = 0.6. Try the experiment with
a coin on a book. L.
Example 6.3. Pendulum in an Elevator. A mathematical pendulum has

a length such that the period of small oscillations in the surface gravitational
field of the Earth is 1 second. Such a pendulum can be used as a watch. The
pendulum is hung in an elevator which has a constant upward acceleration of
a = 1 m s-2 in the gravitational field of the Earth. How many "Earth" seconds
will the clock gain per minute? Disregard changes in the external gravitational
field.
Solution. On Earth the period of small oscillations is
T = 27r/f = 1 s,
where L is the length of the string. From this we conclude that the length is
gT2
L = 47r 2 = 0.25 m.
Inside the elevator the period of oscillation is:
T' = 27rJ g+a

L = 0.95 s.
In one minute of "Earth" time the pendulum oscillates 60/0.95 = 63 times. It
therefore gains 3 seconds "elevator time" per minute of "Earth time". Note
that we had actually no need to compute the string length since we have
directly that
T' ~
T=Yg+a'
6.4 The Centrifugal Force
We will now examine the consequence of Equation (6.36) in a situation where

W = 0 and Ro = 0 and where the particle in question is at rest in the
accelerated reference frame S. With w = 0, S clearly has a constant angular
velocity relative to the inertial frame I (and consequently relative to all other
inertial frames). The Coriolis force of -2m(w x r) vanishes since the particle
is at rest in S (r = 0). Equation (6.36) then has the form:
d 2r
m dt 2 = F + mw 2 p ,
where p = I p I is the distance of the particle from the axis of rotation, and
w = Iw I is the angular velocity relative to the inertial frames. The term mw 2 p
is known as the centrifugal (center fleeing) force. It is familiar to anyone who
has been on a merry-go-round. A particle at rest in S at the distance p from
the rotation axis has the speed v = pw relative to the inertial frame 1. We
can therefore say that the centrifugal force has the magnitude mv 2 / p.
Let us first note that there has been considerable confusion about the
concept of centrifugal force. We emphasize here that the centrifugal force is
6.4 The Centrifugal Force 117
a fictitious force that acts on a given particle, when we study the motion of
this particle in a coordinate frame in rotation relative to the inertial frames.
No centrifugal forces arise within inertial frames.
Example 6.4. Earth's Orbit Around the Sun. Let us as a simple appli-
cation of the concept of centrifugal force study the orbit of the Earth around
the Sun.
s Earth
Fig. 6.7. Earth in its orbit around the Sun
The orbit of the Earth is approximately circular. Let us first describe the
motion of the Earth in the heliocentric reference frame I (see Figure 6.7).
The heliocentric reference frame is an inertial frame and therefore no fictitious
forces act. The motion is completely described by the equation
v2 Mm
m-=G-- (6.44)
r r2
We are ignoring forces from the Moon and from other planets.
Let us next describe the motion from a reference frame S with its origin in
the center of the Sun and an x-axis that always passes through the center of
the Earth. S is rotating with an angular velocity equal to the orbital angular
velocity of the Earth. Relative to S, the Earth is at rest. This means that
the sum of the forces acting on the Earth seen from the rotating frame S is
zero. The forces acting are: (1) the gravitational pull of the Sun, and (2) the
fictitious forces. The fictitious force is the centrifugal force.
In this case, p = r, where r is the distance to the Sun. The gravitational
force points in the negative x-direction, and we have:
Mm
- G -2- + mw 2 r = O. (6.45)
r
The gravitational force and the centrifugal force have the same magnitude
and opposite directions. The two equations (6.44) and (6.45) are formally
identical, but their interpretations are profoundly different. f':,.
Example 6.5. Grass on a Rotating Disk. Let us build the following

apparatus. On the rim of a horizontal disk of radius r we mount two walls
and fill the space between them with soil (see Figure 6.8). We sow a rapidly
growing variety of grass in the soil. The horizontal disk is now set in rotation
with the constant angular velocity w. After the grass has come up, we find that
it has grown in the directions shown on the figure. This can be understood
in terms of fictitious forces: the grass is growing in an accelerated reference
frame. Within this frame, the grass "feels" a gravitational field composed of
the gravitational field of the Earth, and the fictitious centrifugal field w 2 r.
The grass grows in a direction opposite to the total gravitational field.
2
rw -~~L_ _ _-.1.-__-.w~
,,
Fig. 6.8. Grass on a rotating disk
Note. The grass cannot distinguish how much of the gravitational field is due
to the pull of the Earth and how much is because the disk is rotating.
As we shall see below, 9 itself (the local acceleration due to gravity) is
composed of a contribution from mass attraction proper and a contribution
from the centrifugal field that arises from the rotation of the Earth relative
to the heliocentric frame.
Question. Let the radius of the above disk be 2 m. Determine the angular
velocity of the disk such that the angle the grass forms with the vertical is
2°. [Answer:tanB=rw 2 jg ~W::::::0.4s-1J !::::.
Example 6.6. The Variation of 9 with Latitude. As yet another example

we shall consider the variation of g with latitude. Consider a reference frame
with its origin at the center of the Earth and axes fixed relative to the Earth.
The reference frame thus rotates relative to the heliocentric reference frame.
We consider a particle at rest within this frame. The particle has mass m
and is hanging from a string as suggested in Figure 6.9. What forces act on
the particle as seen from an observer standing next to it?
The Earth is not an inertial frame. We have to add fictitious forces to
the "natural" forces acting on the particle. Let us ignore the acceleration of
the Earth towards the Sun, and consider only the rotation which the Earth
performs in the heliocentric frame with a period of 23 hours and 56 minutes.
We shall - initially assume that the Earth is a perfect sphere with radius
R. The basic gravitational field of Earth will therefore act on the particle with
a force P directed toward the center of the Earth. In addition to this, the
6.4 The Centrifugal Force 119
Fig. 6.9. The forces on a particle at rest relative to the Earth as seen in a frame
fixed relative to the Earth
particle will be acted on by a centrifugal force:

K = mw 2 p, (6.46)
where p is the vector shown on the figure. The particle is also acted upon by
a string force S. Since the particle is at rest, we must have:
P+S+K = O. (6.47)
This tells us that the string will not be pointing exactly towards the center of
the Earth. When we measure the acceleration due to gravity on the surface
of the Earth, e.g., by a pendulum experiment (T = 27fJL/g), it is the ac-
celeration produced by the sum of mass attraction and the centrifugal force
that we measure.
Let us estimate the order of magnitude of these effects. We are assuming
that the particle is at latitude ¢. We then have:
p = Ip I = R cos ¢ . (6.48)
Along the direction of the radius vector from the center of the Earth, the
centrifugal force has a component of magnitude:
(6.49)
This component will reduce the local value of the acceleration of a mass point
towards the center of the Earth.
The component of the centrifugal force along the direction perpendicular
to the radius vector is:
K 1. = mRw 2 cos ¢ sin ¢ . (6.50)
This component will change the direction of the vertical, defined to be the
direction of a string in which a heavy object is hanging at rest. Water at rest
on the Earth will form a surface perpendicular to the vertical.
The component KII as given by (6.49) gives a contribution to the change

with latitude of the acceleration due to "gravity". Let us estimate the order
of magnitude of these effects:
w = 7.3 X 10- 5 S-1 ,
and
R = 6.37 X 103 km.
We find:
Rw 2 = 3.4 cm S-2 .
The calculation shows a difference of 3.4 cms- 2 between the acceleration of

gravity at equator (</> = 0) and at the poles (</> = 1r/2). From experiments it
is found that
gpoies 983.2 cms- 2
gequator = 978.0 cms- 2
Note. Using the model of the Earth as a perfect sphere, we have calculated
a difference in the acceleration of gravity between the equator and the poles
having the magnitude 3.4 cms- 2. The observed difference is: L::.g = gp -
geq = 5.2 cm s-2. This deviation is due to the fact that the Earth is not
a perfect sphere. The diameter from pole to pole is about 0.3 % less than
the diameter through equator. This bulging at the equator is produced by
the centrifugal force acting in the reference frame in which the Earth is at
rest. The centrifugal force drives the mass of the Earth away from the axis of
rotation. Such a concrete effect as the bulging of the Earth then, is produced
by the "fictitious" centrifugal force. We shall later use the bulging of the
Earth to confront two fundamentally opposed viewpoints in the theory of
motion:
Isaac Newton: The Earth bulges because it rotates in absolute space.

Ernst Mach: The Earth bulges because it rotates relative to the fixed stars.
(fixed stars = remote matter in the universe)
These profoundly different viewpoints will be discussed in Chapter 7.

Note. If the Earth was liquid, the surface would everywhere be perpendicular
to the plumb line. The Earth has (perhaps) once been partially liquified. It is
a fact that the Earth is not spherical. We shall here not go into the interesting
problem of the exact shape of the Earth. We shall just remark that a nonrigid
spherical body set in rotation relative to the heliocentric system, will deviate
from a spherical shape. Tensions will arise within the sphere. Its final shape
will result from a balance between gravitation, tension and the centrifugal
force, in the rotating frame.
Why is the heliocentric frame so fundamental that a planetary body ro-
tating relative to this reference frame changes its shape? This is the most
profound question in all theories of motion. We shall often return to it. /:;
6.5 Tidal Fields 121
6.5 Tidal Fields

In this section we shall consider a phenomenon which must be included in
any discussion of the theory of motion: the tidal forces. As the name suggests,
the well known phenomenon of the tides in the oceans has its origin in these
force fields.
Tidal fields arise in reference frames (planets) that fall freely in an in-
homogeneous gravitational field. In a reference frame that falls freely in a
homogeneous gravitational field no tidal forces appear. The tidal forces are
- as we shall see - what remains of gravitation when a reference frame falls
freely in an inhomogeneous gravitational field.
The tidal forces are thus of crucial importance in the theory of gravitation,
and they are with some right called "the true gravitational fields". Consider
Figure 6.10. The motion of the Earth around the Sun can be perceived as
composed of:
(1) the motion of the center of the Earth around the Sun in the course of a
year, represented by WI;
(2) the rotation of the Earth around its axis once every 23 hours 56 minutes,
represented by W2.
Sun
Fig. 6.10. Motion of the Earth around the Sun. The vectors are not drawn to
proportion
Since 1 year:=:::! 3.16 x 107 s, we have:

2n -7 -1
WI = IWII = 3.16 x 10 7 s :=: :! 2.0 x 10 s ,
2n
W2 = IW21 = :=: :! 7.3 X 10- 5 S-I .
23 h 56 min
We shall now - by means of the principle of equivalence - consider the phe-
nomenon caused by the yearly motion of the Earth around the Sun, described
by WI. In other words we shall investigate the deviations from inertial mo-
tion relative to the Earth due to the fact that the Earth has a translational
acceleration towards the Sun.
Let us introduce the so-called absolute geocentric reference frame, i.e., a co-
ordinate frame with its origin at the center of the Earth and axes that do not
rotate relative to the heliocentric frame, or, to put it another way, the axes of
the absolute geocentric reference frame always point towards the same three
fixed stars (Note: most fixed stars are so far away that we cannot detect any
changes in the directions of the axes over a year).
The Earth is in rotation relative to the absolute geocentric frame. The
Earth is in free fall in the gravitational field of the Sun, with an acceleration
a. This is the acceleration which the Sun would impose on a particle placed
in the center of the Earth. All points in the absolute geocentric coordinate
frame thus have an acceleration a towards the center of the Sun.
y
- ma
Center
of the Sun
Fig. 6.11. The absolute geocentric coordinate system. The x-axis is chosen to be
directed towards the Sun, at the time considered
Consider the motion of a particle within the absolute geocentric reference

frame. Let the particle have mass m. Our reference frame has an acceleration
a, so according to the equivalence principle we must add a fictitious force
m( -a) (see Figure 6.11). The particle is also acted on by gravitational forces
from the Sun and from the Earth (we disregard the Moon in this discussion).
We shall neglect the effects from the gravitation of the Earth, since we wish
to focus our attention on the deviations from inertial motion caused by the
free fall of the Earth in the gravitational field of the Sun.
If the particle in question was at the center of the Earth, the gravitational
force from the Sun would be exactly canceled by the fictitious force (a particle
at the center of the Earth is weightless in the absolute geocentric reference
frame, see Example 3.1). If the particle is not in the center, the gravitational
force from the Sun will be different in magnitude from the magnitude of
the fictitious force m( -a). In this case there is not complete cancellation of
the two forces: gravitation from the Sun and the fictitious force. It is the
difference between these two forces that causes the tidal field.
We shall now go through a few calculations to estimate the order of mag-
nitude of these effects. The radius vector from the center of the Sun to the
center of the Earth is called Ro, and the radius vector from the center of the
Sun to the particle is called R (see Figure 6.11). We have:
GM Ro
a=- R5 IRol' (6.51)
where M is the Mass of the Sun.

The gravitational force from the Sun will cause an acceleration of the
particle by the amount
GMR
gs = - R2 R .
The total force on the particle as seen from the geocentric system is the sum
of the force of gravity and the fictitious force:
F = -ma+mgs ,
or
F = GmM (~~ - ; ) . (6.52)
The expression in (6.52) is called the tidal force field from the Sun; it is, along
with a similar field from the Moon, the cause of the tides in the oceans. The
corresponding acceleration is
b = GM (RO R)
R~ - R3 .
To simplify the computations we will consider only the points on the surface
of the Earth that are on a straight line between the center of the Earth and
the center of the Sun (see Figure 6.12).
The x-axis has its origin at the center of the Earth and a positive direction
towards the center of the Sun. For points on this axis:
(6.53)
The acceleration field is then:
b= GM Ro
Ro
(2.. _2..) = (1 _
R5 R2
-a R5)
R2
Note: a points towards the Sun, -a away from the Sun. For R > Ro the field
is directed away from the Sun; for R < Ro it is directed towards the Sun.
Since R = Ro - x we can write:
For points near the surface of the Earth, x « Ro. By expanding the expres-
sion in the parenthesis (in the small quantity x / R o), we get:
Here the two
!
fields cancel
Here the gravitational
field from the Sun
is larger
1
~----------------~
x
\:e,e 'he ","","'

field is larger
Fig. 6.12. The tidai field
b~ -a [1 - (1 + 2~) ] (6.54)
b~ a(2~). (6.55)
For x > O, the vector b is directed towards the Sun, for x < O the vector b
is directed away from the Sun (see Figure 6.12).
Let us denote the radius of the Earth by p. On the side of the Earth facing
the Sun, the tidal field at the surface has the magnitude
b=(~:), (6.56)
and is directed towards the Sun. On the side fac ing away from the Sun, the
tidal field has the same magnitude directed away from the Sun. The net
effect is to introduce a distortion of the ocean levels as indicated (but grossly
exaggerated) in Figure 6.13.
Fig. 6.13. The tidai distortion of the ocean surface Ievels
Since the Earth rotates with respect to the absolute geocentric reference
frame, we have ftood in the oceans about twice a day. It was agreat triumph
for Newtonian mechanics that it provided an explanation for the well known
fact that ftood and ebb appear twice each day.
At the center ofthe Earth the parameter b would be zero (compare (6.55)).
The magnitude of a is given by a = Rowf. Inserting values for Ro and Wl,
we get
b ~ 5.1 X 10- 7 ms- 2 .
By using similar considerations on the Earth-Moon system, one finds that

the magnitude of tidal acceleration from the Moon is about 2.5 x b. The tidal
effect from the Moon is thus more than twice as large as that from the Sun.
This is due to that fact that the effect varies with the distance R as 1/ R3.
Recall that
b = 2ap = 2GMp
Ro R~
Since the distance of the Moon from the Earth is much less than the distance
of the Sun from the Earth, the contribution from the Moon ( which has an
almost vanishing mass compared with the Sun) is nevertheless the greater of
the two.
Note. The magnitude of the gravitational field from the Sun at the center of
the Earth is
GM
Fs = R2 . (6.57)
o
The magnitude of the derivative of Fs with respect to distance is
dFs 1= 2GM
IdR o R~'
The magnitude of the tidal field at the surface of the Earth is
I
b = dFs
dRo P
I
= 2G M p
R~'
(6.58)
as we found above.
It is interesting that the gravitational field from the Sun near the Earth
is 175 times as strong as the gravitational field from the Moon. The tidal
force, however, is an expression of the fact that the tidal field - and the
corresponding acceleration - is due to the variation with distance of the
gravitational field from the Sun (and the Moon).
It is consequently the Moon that is the primary source of ebbs and floods
in the oceans. If the Earth was not rotating, there would permanently be
a small flood in the regions facing toward and away from the Moon, and a
small ebb in the plane perpendicular to the direction to the Moon. Since the
Earth is rotating, the tidal forces drive the ocean basins with a frequency of
about 12~ hour, so that this is the typical interval between floods. The extra
half hour is due to the motion of the Moon around the Earth.
We shall not go into the geophysical aspects of tides here, but just notice
that the amplitude of tides vary from less than one meter in the open ocean
regions, to several meters in coastal regions, the most prominent of which is
the Bay of Fundy in Canada. Here, the maximum difference of water level
between ebb and flood is 15.4 m.
The theoretical significance of tidal forces is the following. In a laboratory
on the surface of the Earth we make a very small error if we simultaneously
disregard
(1) the gravitational fields of the Sun and the Moon; and
(2) the translational acceleration of the Earth.
If the Earth was not rotating relative to the heliocentric reference frame, a
coordinate frame affixed to the Earth would nearly be an inertial frame. One
can say that such a frame is inertial apart from the tidal fields. But we have
also seen that the tidal fields, which ultimately are caused by the variations
in the gravitational fields can never be transformed away by letting the coor-
dinate system fall freely. The tidal forces are "the true gravitational forces" .
The "homogeneous" part of an arbitrary gravitational field can always be
transformed away by going to a reference frame that falls freely in the local
gravitational field.
We shall return to the relation between tidal forces and inertial frames in
Section 6.7.
Example 6.7. The Roche Limit
Consider the figure. P is a planet with mass M. It has a moon of mass m,

and this moon is assumed to be held together by its own field of gravity. The
moon is in free fall around the planet, e.g. in a circular orbit of radius D.
Will the tidal force tear the moon apart?
The radius of the moon is r, and the radius of the planet is R. A crude
estimate of the distance Do which the moon must have for the moon to be
held together under its own gravity can be made as follows.
The gravitational force from the moon on a particle of mass J.-L (e.g., a
rock on the surface) lying in the equatorial plane, is GmJ.-L/r2. The tidal force
on the particle is 2G M W / DJ. The moon can just hold on to the rock when
these forces balance:
GMJ.-L _ GmJ.-L
2 -D3 r - --2- (6.59)
o r
If the orbital radius of the moon is less than Do, the rock on the surface will
be lifted away by the tidal forces. Let us assume that the average density of
the planet is pp and the average density of the moon is PM. This gives:
M - 47f R3 (6.60)
- 3 Pp,
471" 3
m = 3r PM . (6.61)
We then find , by substituting (6.60) and (6.61) into (6.59) that the critic al
distance can be expressed as:
Do = R (2 ; : r/ 3
= 1.26 R (;: r/ 3
(6.62)
The orbital radius Do is known as the Roche limit. The calculation of the
Roche limit Do is, when one takes into account the deformation of the par-
ticipating bodies, rather complicated. One finds :
1/3
(
Do = 2.46 R ;: )
which is (6.62) except for the numeric al factor.

For the system of Earth- Moon, Do = 2.9RE, where RE is the radius of
the Earth. If the Moon came closer than 2.9RE, rocks would be lifted away
from the lunar surface. Now look at the picture below.
Most of Saturn's rings are inside the Roche limit for Saturn! Note how in
this case the so-called fictitious forces are preventing the rocks in the rings
of Saturn from assembling into a moon.
The rather dramatic statement that a body inside the Roche limit will be
"torn apart" by the tidal forces is true only if the body is held together by
its own field of gravity. Gravity does play an important role for example in
the binding energy of the Moon, as can be seen directly from the fact that
the Moon is spherical. A rock that you pick up from the ground is nowhere
near spherical; its binding energy is determined by chemical bonds. Gravity
plays no role in keeping cobblestones together.
Many of the satellites orbit ing the Earth are inside the Roche limit of the
Earth. These satellites are not torn apart by tidal forces, again because they
are held together by chemical bonds, not gravity.
In fact, you yourself are inside the Roche limit of the Earth, but you
are also held together by chemi cal bonds (essentially electrical interactions) .
When gravity is dominant in the binding energy of a given body, this body
will have the shape of a sphere. 6.
6.6 The Coriolis Force

We shall consider the Coriolis force acting on a particle that moves along the
surface of the Earth.
I
I
c::b
I z
Fig. 6.14. A reference frame fixed on the surface of Earth
The natural reference frame here is a frame that is fixed relative to the
surface of the Earth and which therefore is in a state of rotation relative
to the heliocentric reference frame. We shall use a frame with origin on the
surface of Earth (in stead of, say, the center of the Earth). Such a frame is
commonly called the laboratory frame, and it is the frame from which you,
standing still on the surface of the Earth, view the phenomena around you.
Look at Figure 6.14. The lab frame is oriented here in such a way that
the y-axis points towards the north and the x axis towards the east; the z
axis therefore points vertically upwards.
We briefly demonstrate that w describes the rotation vector relative to
the heliocentric frame of any coordinate frame fixed relative to the Earth, no
matter where the origin is located.
Consider the point P which is at rest relative to both a coordinate frame
with its origin at the point 0 and the frame with its origin at the point
Q. The "absolute" velocity of P, i.e., the velocity relative to the heliocentric
frame, is:
Vp = vo+ w x OP
vo+ w x (OQ + QP)
vo+ w x OQ + w x QP
vQ+ w x QP.
All coordinate frames at rest relative to a given frame share

the same rotation vector.
6.6 The Coriolis Force 129
Recall Equation (6.36). The Coriolis force on a particle with mass m,

moving with velocity v relative to the surface of the Earth, is
F = 2m(v x w), (6.63)
where w means the rotation vector relative to the heliocentric reference frame
(the fixed stars).
Let us first consider a particle that has a horizontal velocity, and is located
at latitude P. From Figure 6.14 we find that the rotation vector of the Earth
has components
w = (0, wcosP, wsinP),
in the lab frame.
The vector F has a horizontal component FH and a vertical component
Fv. The vertical component of w gives rise to the horizontal component of
the force:
FH = 2mvw sin P ,
and the horizontal component of w gives rise to the vertical component of
the force:
Fv = 2mvx w cos P .
In geophysical considerations it is often the horizontal component of the
Coriolis force that is of interest. This force influences wind and ocean currents.
Since in the northern hemisphere the vector w points out of the Earth, we
have the general rule that:
In the northern hemisphere where w points out of the

ground, a particle moving with a horizontal velocity will
be drawn to the right
This holds whether the particle in question is moving towards the north,
south, east, or west. Similarly:
In the southern hemisphere, where w points into the

ground, a particle moving with a horizontal velocity will
be drawn to the left.
These rules are equivalent to counting P as positive on the northern hemi-

sphere and negative on the southern hemisphere.
Let us now write the general expression for the Coriolis force on a particle
having an arbitrary velocity (relative to the lab frame):
v == r = (x,y,i).
The components of F are then:
2mw( -z cos tP + y sin tP), (6.64)

-2mwx sin tP, (6.65)
2mwxcostP. (6.66)
Consider the following situation. A long vertical pipe is fastened to the ground
in the northern hemisphere, and a particle can slide down the pipe with no
friction. If we release the particle from the top of the pipe, it will slide down
to the ground, its vertical motion being free fall.
The Coriolis force, however, will cause the particle to push against the
side of the pipe. Since no friction forces act between the particle and the pipe,
this will not slow the particle down, and the reaction force from the pipe will
ensure that the velocity is vertical throughout the fall, i.e. , x = y = o. From
(6.64), (6.65), and (6.66) we find that the Coriolis force then has only the
x-component
Fx = -2mwzcostP.
z
Note: is negative. The particle is pushed by the Coriolis force in the positive
x-direction, towards the east.
Let us estimate the magnitude of this force; suppose the pipe is 100 m
high, and disregard air resistance.
We assume that the mass of the particle is 0.1 kg and that tP = 45°. The
Coriolis force on the particle the instant it reaches the bottom of the pipe is
then
fn::I: 1
Fx = 2mwy 2gh J2 = 4.6 x 10 N.
-4
Compare this to the gravitational force on the particle from the Earth:
Fg = mg = 9.8 x 10- 1 N.
It is now clear that if we remove the pipe, the particle will not fall along
the vertical. To the lowest approximation it will receive a velocity component
toward the east. In turn, this velocity component will give rise to a Coriolis
force toward the south, but this effect is proportional to w 2 and is typically
much smaller than Fx .
Let us write the complete set of equations of motion for a particle in free
fall near the surface of the Earth. The equations of motion are in cartesian
coordinates relative to the lab frame. One finds:
mx 2mw( -z cos tP + Ysin tP), (6.67)

my - 2mwx sin tP , (6.68)
mz 2mwx cos tP - mg . (6.69)
(The acceleration due to gravity is assumed to be constant throughout the

fall. )
This system of coupled differential equations replaces the much simpler

system which would hold if there was no rotation of the Earth relative to the
heliocentric frame (Le., w = 0):
mx 0,
my 0,
mz -mg.
We shall now integrate the coupled system (6.67)-(6.69) to find an expression
for the motion of a particle near the surface of the Earth. Start by differenti-
ating (6.67) with respect to time, then substitute (6.68) for y and (6.69) for
z. This gives us
d3x 2 dx
dt 3 + 4w dt = 2wg cos P. (6.70)
This equation can be regarded as a second-order differential equation in x:

d 2x 2.
dt 2 + 4w x = 2wg cos P . (6.71)
The complete solution to (6.71) is, as we have seen,

. gcosP
x = --
2w
+ A cos[2w(t - to)], (6.72)
where A and to are constants of integration. Compare this with Example 2.3,
where we solved a differential equation of the same form as (6.71). Show that
(6.72) is a solution by substituting (6.72) into (6.71).
Now we are in a position to solve the coupled system completely. We
integrate (6.72) to find x(t). In turn, this solution can be used to find y(t)
and z(t) from (6.68) and (6.69).
We shall not develop the most general solution; instead we shall consider
some special cases of interest. Suppose the particle falls from height z = ho.
As initial conditions we then have:
(x, y, z) (0,0, h o),

(x, il, i) (0,0,0).
The constants of integration in (6.72) are then: A = -gcosP/2w and to = 0.

Thus:
gcosP
X = - - ( 1 - cos2wt). (6.73)
2w
Now we use the fact that we typically are considering falls of a duration t
such that t « l/w. Since w = 7.3 X 10- 5 S-1, we have that l/w = 1.3 X 10 4 s
~ 3 hours. This leads us to use the small argument expansion of cos:
Then:
or
i; :::::; gw( cos q»t 2 . (6.74)

We now integrate (6.74) again with the initial condition x o at t = 0,
resulting in
(6.75)
The result (6.75) shows how a particle falling mainly along the vertical will
deviate, mainly toward the east (see Figure 6.14). That direction can of course
be inferred directly from F = 2m( v x w).
ii:::::; -2gw 2 (cosq>sinq»t 2 .
By integrating twice and using the initial conditions iJ = 0 and y = 0 for

t = 0, we get
(6.76)
This shows that there is also a second-order (notice the factor w2 !) effect
resulting in a small deviation toward the south.
To illustrate the order of magnitude of these effects, let us take as an
example a particle falling at 45° north. Suppose the fall lasts 10 s, and dis-
regard air resistance. The total height through which the particle falls is
z:::::; 1/2gt 2 :::::; 500 m.
From (6.75): x :::::; 0.17 m. From (6.76): y :::::; -4.4 X 10- 5 m. Thus, the
y-deflection is about 4 000 times smaller than the x-deflection! From (6.75)
(and h = (1/2)gt 2 ) we thus conclude that in general, a particle in free fall
through the height h will be deflected a distance
1
d:::::; -wcosq>
3
ff-h
-
9
3
toward the east, by the Coriolis force.

That the deflection of the particle towards the south is much smaller than
the deflection towards the east is directly evident in the differential equations
(6.67) and (6.68). The acceleration towards the south is proportional to i;
while the acceleration towards the east contains the term i, which is by far
the largest component of velocity.
The eastward deviation has been seen in experiments where balls have
been dropped from towers and into mine shafts. Such experiments provide
direct evidence for the rotation of the Earth relative to the heliocentric frame.
Example 6.8. Coriolis Force on a Train. A train with mass M = 10 5 kg

is driving towards the north with a velocity of 108 km/hour. The track is
straight and is located around F = 60° northern latitude.
(1) Find the Coriolis force on the train.
(2) What would the Coriolis force be if the train was going towards the east?
Answer.
(1) F = (2Mvwsin4>,O,0). Fx = 379 N (the "weight" of the train, i.e., the
gravitational force on the train, is 9.8 x 105 N).
(2) F = (0, -2Mvwsin4>,2Mvwcos4», Fy = -379 N, Fz = 219 N.
Example 6.9. Particle on a Frictionless Disc
Friction-
less disk - - - -_ _ _ _ _~::...::-
of particle
as seen from the
rotating disk
A horizontal disk is rotating about a vertical axis with the angular velocity w.
Let us study the fictitious forces assigned to a particle that is moving relative
to the rotating disk. We put the particle carefully down at some distance
r from the axis. If we assume that the disk is perfectly smooth such that
no forces of friction act, the particle will remain at rest as seen in the lab
frame. As viewed from a frame fixed on the disk however, the particle will
perform a uniform circular motion. How is this motion explained in terms of
the fictitious forces?
As seen from the disk, two forces must be taken into account: the Coriolis
force and the centrifugal force. The centrifugal force is directed away from
the center along a radius vector and has the magnitude
(6.77)
The Coriolis force has magnitude
IKc 1= 12m(vrel x w) 1= 2mw 2r. (6.78)
Here, VreI denotes the velocity of the partide relative to the disk. It is per-
pendicular to w, and we have used v = rw.
The Coriolis force and the centrifugal force are oppositely directed (check
the direction of the Coriolis force by using the right-hand rule) , so the result-
ing force is directed toward the center of the disk and has the magnitude
2 V;el
Ktotal = mw r = m - . (6.79)
r
This is exactly the force necessary to sustain uniform circular motion. One
could say that the partide performs uniform circular mot ion relative to the
disk "because" is has the relative velocity v = wr as soon as it is placed on
the frictionless rotating disk.
Question. A partide placed on a real disk with friction will not move in a
circle but will spiral away from the axis. Why ?
Answer. The Coriolis for ce can no longer keep the partide in a circular orbit
since, with friction dragging the partide, VreI < wr. Since the centrifugal for ce
is independent of velocity, it now "wins" and drives the partide outwards.
6
Example 6.10. The Vertical Throw. At the time of Galileo an issue

was much disputed: will the velocity of the Earth have any influence on
the mot ion of a cannon ball shot vertically upwards? We now know that a
uniform translational mot ion of the Earth relative to the heliocentric frame
- no matter how large that velocity - would have no influence whatsoever on
the trajectory of the cannon ball as seen from a reference frame fixed on the
surface of the Earth. If the Earth had only a uniform velocity, the cannon
ball would falI directly back to the place from which it was fired . But is this
the case? No, for the Earth does not move in uniform translation, it is in
rotation. A frame fixed on the surface of the Earth is not inertial. Let us work
out the problem of the cannon balI .
•
. .:.
/
6.7 Tidal Forces and Local Inertial Frames 135
A particle is fired vertically upward from a point on the surface of the

Earth, at iPO northern latitude. Let us ignore resistance from the air, and
consider only small vertical extensions, h « p where p is the radius of the
Earth.
We will show that the particle touches down at a point west of the launch
point. We use the differential equations (6.67), (6.68), and (6.69), and ignore
the quantities i; and iJ in the equations. We then have, approximately,
x ~ -2wi cos iP ,
i ~ -g.
Let the time when the ball is launched be t = 0 and the initial velocity be
Vo. Then
i ~ Vo - gt.
Now substitute this into the expression for x:
x:::::: -2w(vo - gt) cosiP .
Setting i; = 0 for t = 0 gives us
i; = -2wvo(cosiP)t + wg(cosiP)t 2 .
Finally, with x = 0 for t = 0, we find

1
x = 3gw(cosiP)t3 - w(cosiP)vot 2 .
From the expression for i we find that the amount of time from when the
ball leaves the cannon until it returns is t = 2vo/ 9 = 2J2h/ 9 where h is the
maximal height the ball reaches (Le. , 1/2mv5 = mgh) . Inserting this value
for t in the above expression for x(t) we get the x-deviation when the ball
touches down:
4
d = -3wcosiPy 9 .
(8h3
From the sign one can see that ball will touch down west of the launch point
x=O. 6
6.7 Tidal Forces and Local Inertial Frames
In the Newtonian interpretation of classical mechanics (see Chapter 7) an

inertial frame is either absolute space or a reference frame moving with con-
stant velocity along a straight line relative to absolute space ("the astronomi-
cal space"). While Newton was of the opinion that "the astronomical inertial
frame" forms the background on which mechanical phenomena take place,
Einstein introduced - for use in general relativity - the so-called local inertial
frames.
Let us first describe an example of a local inertial frame. If the Earth was
not rotating, the laboratory would be close to an inertial frame. The Earth
falls freely in the local gravitational field, which is essentially a superposition
of the gravitational field of the Sun and the Moon. Because of the free fall we
feel only the tidal forces produced by the gravity of the Sun and the Moon.
A reference frame, falling freely in the local gravitational field, and

not rotating relative to the distant fixed stars is, apart from the tidal
fields, an inertial frame
This definition actually rests upon the principle of equivalence. Since no

gravitational field is completely homogeneous, one can say that the principle
of equivalence is only "valid in a point". A freely falling, nonrotating coor-
dinate frame is thus only an inertial frame in an infinitesimal neighborhood
of the origin; i.e., a local inertial frame. One can say that the tidal fields
reveal the presence of the real gravitational fields. The tidal fields cannot be
transformed away by any choice of reference frame.
The reasons the heliocentric reference frame can be considered as a very
good approximation to an inertial frame, are the following:
(1) it is in free fall toward the center of the galaxy (the Milky Way);
(2) it is not rotating relative to the fixed stars;
(3) the tidal fields within the heliocentric frame due to the gravitational field
of the galaxy are so small that we cannot measure them.
The laboratory frame is not an inertial frame because the Earth is ro-
tating relative to the heliocentric frame. For many practical applications in
mechanics, the freely falling laboratory frame can be considered an inertial
frame. This is true for motions over "short distances and short time inter-
vals" , i.e., for motions where we can disregard the Coriolis force. The decisive
criterion for a specific coordinate frame to be considered inertial, is that the
accuracy to be applied to the given mechanical problem makes the tidal forces
undetectable.
The relativity principle of mechanics we formulate as follows. There exist
a threefold infinity of reference frames, all moving uniformly along straight
lines relative to each other, within which the basic mechanical laws are valid in
their simple, classical form. Such reference frames are called inertial frames.
The basic mechanical laws are independent of the choice of reference
frame, as long as the reference frame is chosen among the inertial frames.
In this formulation of the relativity principle of mechanics one very much
needs to be given an example of just one inertial frame. If you have one, you
have them all.
It is important to note the following. The relativity principle of mechanics,
as it is discussed in Chapter 4, rests on Newton's laws. We shall see that while
6.8 The Foucault Pendulum 137
Newtonian mechanics has not survived entirely, the relativity principle has -
in an expanded form - become one of the pillars on which modern physics
rests. The relativity principle is thus of a more fundamental character than
the laws of Newtonian mechanics from which it originated.
Example 6.11. Global and Local Inertial Frames. Newton uses the ab-
solute space (astronomical space) as an inertial frame. This is what might be
called a global inertial frame, because the frame is inertial everywhere. Most
presentations of Newtonian mechanics use the heliocentric frame as a global
inertial frame. The heliocentric frame is nevertheless what Einstein calls a
local inertial frame. The reason that we can use the heliocentric reference
frame as an inertial frame even for the study of the motions of the planets is
that the tidal forces that appear in the frame, because of its free fall in the
gravitational field of the Milky Way, are so small that we cannot measure
them.
It is interesting, that in order to define an inertial frame it is necessary
to refer to the fixed stars. Apparently the global distribution of matter in
the universe is decisive to any attempt to create a logically sound science of
mechanics - or, is it ?
Go outside on a dark winter's night when the sky is clear, and look up
at the stars. Carry with you a piece of string and tie a stone to the string.
You can now make the following - and, to the understanding of mechanical
physics, absolutely crucial - observation.
Take hold of one end of the string and release the stone. If the stars and
the string seem at rest relative to one another, the string will hang along the
vertical. Now set the string in rotation, i.e., give the stone an acceleration
relative to the fixed stars. You will now discover that the string no longer
is vertical. If you whirl the stone around sufficiently fast, the string can be
almost horizontal. In the reference frame where the stone is at rest the fixed
stars are rotating. In this reference frame there is a centrifugal force that,
combined with the force of gravity and the string force, keeps the stone at
rest and the string (almost) horizontal.
Can it be a coincidence that a reference frame, where the fixed stars seem
at rest, is an inertial frame? 6.
6.8 The Foucault Pendulum
In the year 1851 the French physicist Foucault carried out an experiment
which demonstrated that a coordinate frame fixed on the surface of the Earth
is not an inertial frame. Foucault studied the motion of a pendulum consisting
of a spherical body of mass 28 kg hanging in a 69 m long metal wire. The
experiment was located underneath the huge Pantheon dome in Paris. The
suspension point of the pendulum was constructed in such a way that the
pendulum could pivot freely in all directions; the period of oscillation was
(T = 27rVL/g) ~ 17 s.
Foucault observed that the plane of oscillation for the pendulum slowly
drifted around relative to the walls of the building (the "lab frame"). A full
rotation took about 32 hours. Let us set up the equations of motion for the
pendulum (see Figure 6.15).
string force
... ...
... ...
... ... y
... ...
...... ...
x ... ....
Fig. 6.15.
The pendulum is affected by the Coriolis force, and this is the cause of
the precession - slow rotation - of the plane of oscillation. Let us introduce
coordinates with the origin in the equilibrium point of the pendulum, the
z-axis in the vertical direction, through the point of suspension. Since we are
interested in the precession of the plane of oscillation, we project the motion
into the xy-plane. Let us limit the motion to oscillations of small angular
amplitude ( x / L « 1 and y / L « 1 where L is the length of the pendulum
string). For such a motion, z is small compared with x and iJ, so we disregard
z.
Let us write the equations of motion in the chosen coordinates. Denote
the magnitude of the string force by T. Then the components of the string
force (recall that x/L« 1 and y/L« 1 ) are (see Figure 6.15):
Tx ~ -T~
L'
y
Ty ~ -T-
L'
Tz ~ mg.
The components of the Earth rotation vector w in the lab frame are (with <P
being the latitude of the experiment, see Figure 6.16):
w = (-wcos<P,O,wsin<p)
Fig. 6.16.
The Coriolis force is:
K = -2m(w x r) = 2m(v x w).

Since
v = (x,y,i) ~ (x,y,O),
we get
v xw = (yw sin <1>, -xw sin <1>, yw cos <1» ,
and the x and y equations of motions become:

x
mx -T Z + 2mywsin<1>,
my -Tf - 2mxwsin<1> .
Assuming T = mg,
.. g
x+-x 2w sin <1>y ,
L
y.. + -y
g -2wsin<1>x. (6.80)
L
These equations are not trivial to solve. We shall here use some intu-
itive and physically reasonable assumptions about the solution and then a
posteriori verify that they are consistent.
Consider Figure 6.17. If the Coriolis force was not present, the pendulum
would oscillate in a plane, with angular frequency (for small oscillations)
J
given by 0: = g / L. For the angular frequency 0: we would find that 0: » w,
where w is the angular frequency of the rotation of the Earth.
The presence of the Cariolis force causes the plane of oscillation to turn
slowly, with an angular velocity f3, which is of the same order of magnitude
as w (except very near the equator where f3 vanishes completely, see below).
We have: I f31 ~ w « 0:.
21t
z
I
I Y
,, I
I
I
X , I
, I
't,
,
Fig. 6.17.
We consequently look for a motion of the pendulum of form
x cos at cos (3t ,

y cos at sin (3t . (6.81)
This set of equations describes a harmonic oscillator that oscillates with an-
gular frequency a along a line whose direction angle e turns with the angular
frequency B= (3. Differentiating twice and rearranging, one gets, after some
calculations:
x + (a 2 - (32)x -2(3y,
ii + (a 2 - (32)y 2j3± .
Comparing with the equation of motion (6.80), we find that the set (6.81)
will be a solution if
(3 = -wsinp,
and
a 2 - (32 = f.
We may thus consider the motion of the pendulum to be a constant rota-
tion of the plane of oscillation, superimposed on the harmonic oscillation of
the pendulum with angular frequency a ~ vi
g/ L. The plane of oscillation
turns with the angular velocity (3 = -w sin P where P is the latitude of the
geographical location. The period in the precessions is:
TF = ~ = ~ = 23 h 56 min
I (3 I w sin P sin P
For Paris, sin <p ~ 0.75, and consequently TF ~ 32 hours. Note: f3 is negative;
the plane of the Foucault pendulum turns clockwise.
Let us remind ourselves of the basic reason for the precession of the plane
of oscillation. Look at this sketch:
North
.,.
f
South
Fig. 6.18.
If the pendulum starts out to the south of the point of equilibrium, it travels
towards the north, but is deflected to the right. As it returns it is again
deflected towards the right of its velocity; the net result is that the plane of
oscillation (here the average tangent to the curve) will drift slowly clockwise.
Consider the following illustration:
-1{Polaris
: (the north star)
1\
I
"* "* ... - I .. '"
"* "*
Fig. 6.19.
Assume that we perform the Foucault experiment on the North Pole of the
Earth. In this case, the plane in which the pendulum swings turns clockwise
over the ground, one full turn each 23 hand 56 minutes (one meridian day).
Assume it is the winter period at the pole. Look up at the star Polaris, also
known as the North Star, and the surrounding stars. You will then observe
the stars majestically turning, keeping perfect step with the precession of the
pendulum plane. The pendulum seems locked by the stars, while the Earth
turns counterclockwise underneath.
We have placed this experiment on the North Pole for the following reason:
in the northern hemisphere the suspension point of the pendulum remains
fixed relative to the distant stars if and only if the point of suspension is
directly above the Pole. Placed elsewhere the point of suspension would rotate
with the Earth about the polar axis and this produces a counter rotation.
It is this counter rotation that makes the rate of turning for the oscillation
plane at other latitudes different from one meridian day (Paris: T;:::: 32 h).
At the equator a Foucault pendulum plane will not precess at all. Returning
to the experiment at the North Pole:
(1) is it absolute space (astronomical space) that holds the plane of oscillation
fixed while the Earth turns below; or
(2) is it the fixed stars - the totality of distant matter in the universe - that
interacts with the mass of the pendulum in such a way that the pendulum
swings in a plane that remains fixed relative to the stars?
We shall return to the problem in Chapter 7.
6.9 Newton's Bucket
It was particularly in the centrifugal force that Newton found his strongest
arguments for the existence of an absolute "astronomical space" as a sort of
stage upon which mechanical phenomena act. The unfortunate problem with
absolute space is, as we have seen, that such a space can only be found to
within a uniform translational motion. Does this observation make the whole
idea of absolute space meaningless? Not quite, as we shall see. Newton gives
a simple counter-argument based on a rotating bucket.
Bucket at rest Bucket in rotation

Fig. 6.20.
6.9 Newton's Bucket 143
A bucket of water rotates with the angular velocity w relative to the iner-
tial frames. The water is forced away from the center by the centrifugal force
(because it is constrained by friction to follow the bucket). Let us determine
the exact shape of the surface. In the co-rotating frame (where the water is
at rest), we consider a small surface element, of mass m.
Fig. 6.21.
The surface element is at rest, so the force Fo on it from the surround-

ing fluid must be perpendicular to the surface (otherwise, the fluid element
would move in response to a tangential tension). Fo must match the external
(gravitational) forces on the element. Projecting Fo on to the vertical and
horizontal directions we find (see Figure 6.21):
Fo coscfJ mg,
Fo sincfJ mrw 2 ,
or
rw 2
tancfJ
9
OJ
~~~--------~r
Fig. 6.22.
We seek the shape of the curve formed by the intersection between the
surface and a vertical plane through the center of the bucket. We introduce
a coordinate system as shown in Figure 6.22, and wish to determine the
function z = f (r). The slope of the tangent to the curve is at any point given
by
dz w2
tan</>= - = - r .
dr 9
Integrating this equation (with z = 0 for r = 0), we get
J = :2 J
dz rdr =} z = ~ w; r2 .
The surface is consequently a parabola of revolution.
When the bucket rotates relative to the inertial frames (i.e., relative to
the fixed stars), we observe the parabola of revolution. When the bucket is
not rotating, the surface is planar. In both cases there is no relative motion
between the water and the bucket.
Is it then the relative motion between water and the fixed stars which
causes this effect? Or is it the rotation of the water through "absolute space" ,
within which the stars seem to rest, that is decisive?
There can be no question that Newton was of the opinion that it is rotation
relative to absolute space that changes the shape of the surface of the water.
A change in the state of motion relative to absolute space requires a force.
If we make the bold thought experiment of keeping the bucket at rest and
rotating all the fixed stars around it, Newton would predict no effect on the
water, since the water would not be in rotation relative to absolute space.
Let us quote from Principia:
The effects by which absolute and relative motions are distinguished
from one another, are centrifugal forces, or those forces in circular
motion which produce a tendency of recession from the axis. For in
a circular motion which is purely relative no such forces exist; but in
true and absolute circular motion they do exist, and are greater or
less according to the quantity of the (absolute) motion.
For instance. If a bucket, suspended by a long cord, is so often turned
about that the cord is strongly twisted, then is filled with water, and
held at rest together with the water; and afterwards by the action
of a second force, it is suddenly set whirling about the contrary way,
and continues, while the cord is untwisting itself, for some time in
this motion, the surface of the water will at first be level, just as it
was before the vessel began to move; but subsequently, the vessel, by
gradually communicating its motion to the water, will make it begin
sensibly to rotate, and the water will recede little by little from the
middle and rise up at the sides of the vessel, its surface assuming a
concave form. (This experiment I have made myself).
6.10 Review: Fictitious Forces 145
At first, when the relative mot ion of the water in the vessel was the
greatest, that motion produced no tendency whatever of recession
from the axis; the water made no endeavor to move towards the
circumference, by ris ing at the sides of the vessel, but remained level,
and for that reason its true circular mot ion had not yet begun. But
afterwards, when the relative mot ion of the water had decreased, the
rising of the water at the sides of the vessel indicated an endeavor
to recede from the axis; and this endeavor revealed the real circular
mot ion of the water, continually increasing, tiU it had reached its
greatest point, when relatively the water was at rest in the vessel...
It is indeed a matter of great difficulty to discover and effectually to
distinguish the true from the apparent mot ion of particular bodies;
for the parts of that immovable space in which bodies actually move,
do not come under observation of our senses.
Read the last four lines again and think about them for the rest of your
life.
6.10' Review: Fiditious Forces

force F. These for~ţs Cil'e called fictitious .forces. They m3;Y be.distin-
guish.ed from r.eal forct;)s by, the,fact that the flctitiousforces are inde-
pendent Of tne phyşical sttuatipn of the.partiqle considered. They. :· a~e
dependent only on the.motion of the reference frame . an:d ţhe. position
aud velocit:y cgQrdlnates of the particle.
The . t~rmmU)2 p := -,-m ["i )( (WXi r)Jds c<l;lJed .. th~ . centrifugal fOfce.
The term -2m(wx r) is called theCori9lis forc~. The oţher terms haveno
~pecific nam~s; By the iutroduetiou of th~se fictitious forces the observer
in S mfLY say thât the firstand seqoud law of NeWtqu are correct . Tn~
third Ja;w, the law of actiol1 and reaction, does not hold for fietitious
f6rc~s. Tnere i~ uo other b.6dy on which there isa re~ctfou force.
The fictitiouş f()rces. have ou(')imp<;lftapt. property in cOlnmon~ith
gravitationaI forces: they are alI propqrtional ta the mass of ~he'particle.
Forces like. gravitational forces,electfical forces, elastic forces (string
fow~s, ţensions in springs) and,,so' on, 'aresQmetimes - ta distinguish
them: frorp. fictitious forces'""" oalled natural ..rorces.
6.11 Problems
Problem 6.1. A glass tube is filled with water and contains a piece of lead
and a piece of cork (see the figure below). The tube is placed on a horizontal
table and is accelerated toward the left by a blow from a hammer. How will
the lead and the cork pieces move relative to the tube? Give a qualitative
answer.
Problem 6.2. A wooden block B with the mass m is positioned on the front
end of a car, as shown in the figure. The coefficient of friction between the
car and the block is f..L. The car has an acceleration of a (towards the right)
relative to the inertial frames . Find the smallest value of lai that will make
the block stick to the car.
a
6.11 Problems 147
Problem 6.3. A satellite moves in an almost circular orbit around the Earth.
A frictional force from the atmosphere acts on the satellite. This force acts
against the mot ion of the satellite. Inside the satellite a small mass m is tied
to a string, and the other end of the string is fastened to the satellite. You
pick the point inside the satellite in which to fasten the string such that the
mass can perform small pendulum oscillations.
(1) In which direction is the equilibrium position of the string? Make a small
sketch showing the direction of the string and the point of attachment
inside the satellite.
(2) The mass is now set into small oscillations. Which quantities determine
the period T of oscillation? We are assuming that the satellite does not
rotate relative to the Earth and that T is much smaller than the orbital
period of the satellite.
Problem 6.4.
A man stands on a spring scale on the equator of the Earth. What should
the length of a day be if the scale showed the man to be weightless?
Problem 6.5.
~~--~,
,,' '''''', ,
I
,, , ,
I \
,
I \
I I
\
, I
I ,
I ,
\ I
\ I
\ I
,, ,,
\ I
\ I
,, ,,
........ -._~'
Consider a projectile in a circular orbit near the surface of the Earth. Disre-
gard trivial effects such as air resistance, etc. (a) Find the period of revolution
for the projectile. (b) For a reference frame in which the projectile is at rest,
describe all forces acting on the projectile.
Problem 6.6. The so-called "Round Tower" in Copenhagen has a height of

36 m. A small, heavy sphere (a bullet) is dropped from the top of the Tower to
the ground below. The sphere falls freely (neglect air resistance). Determine
the distance and direction from the point directly beneath the drop point, to
the point where the bullet actually hits the ground.
Pro blem 6.7. A rotor is a hollow cylindrical cabin able to turn around the
vertical symmetry axis of the cylinder. A person goes into the rotor, closes
the door and stands against the wall. The rotor is slowly set in rotation. At a
certain angular velocity the floor is opened downwards. The passenger does
not fall out but remains held against the interior wall.
The radius of the cylinder is R, the acceleration of gravity is g and the
coefficient of friction between the passenger and the wall is J1 .
(1) Find the least angular velocity w such that the passenger does not fall
downwards.
(2) The coefficient of friction between the man and the rotor has the value
J1 = 0.4. The radius of the rotor is R = 3m. The angular velocity is as
found in 1). Find the velocity of the passenger relative to the ground in
this case. Consider the Earth to be an inertial frame for the motions in
this problem.
Problem 6.8. A planar disk S made of wood is rotating relative to the lab
frame, in a horizontal plane about the vertical axis through the center of the
disk. The angular velocity has the magnitude w. The direction of rotation is
counter-clockwise. The figure shows the disk seen from above.
A small beetle of mass m is walking on the disk along a diameter. The

constant velocity of the beetle relative to the disk is at all times v. [Note: the
angular velocity w of the disk is maintained constant by an external source
and is thus independent of the location of the beetle]. The acceleration of
gravity is g and the lab frame is considered to be an inertial frame.
6.11 Problems 149
Let us consider a coordinate frame xyz, fixed relative to the disk (see
the figure). This coordinate frame has origin in the center of the disk, and
we arrange the frame in such a way that the beetle is moving out along the
x-axis, and the z-axis is vertical. We call this frame 'the rotating frame'.
Assume first that the friction between the disk and beetle is so large that the
beetle does not slip on the disk.
(1) Find the magnitude and direction of the fictitious forces acting on the
beetle as seen in the rotating coordinate frame.
(2) Show on a sketch the direction of the force F of friction with which the
disk acts on the beetle.
(3) Now let the coefficient of friction between the disk and the beetle be
J-L. Determine the distance Xo from the center where the beetle starts to
slide.
Problem 6.9.
A man mounts a bathroom scale positioned on a skateboard such that it,

as shown in the figure, can roll without friction down an inclined plane. He
stands on the scale and reads off his weight as he is rolling down the inclined
plane.
What is the slope () of the inclined plane if the scale displays 45 kg during
the descent and the actual mass of the man is 60 kg? Will the man stand as
shown in the figure?
Problem 6.10.
A very small bob (a particle!) of mass m is situated on the inside of a funnel.

The funnel is rotating about a vertical axis with constant angular velocity
corresponding to v revolutions per second, i.e., W = 27rV . The walls of the
funnel are at an angle () to the horizontal.
The distance from the axis of rotation to the bob is r which is much bigger
than the dimensions of the bob. The acceleration of gravity is g.
(1) Assume first that there is no friction between the particle and the funnel.
Determine the angular velocity Wo such that the particle does not move
relative to the funnel.
Assume now that the coefficient of friction between the particle and the funnel
is 1-".
(2) Determine the smallest value of v for which the bob starts to move up-
wards.
(3) Show that there is an angle () = (}max for which the bob cannot slide
upwards if () > ()max, no matter how large w is. Determine ()max'
Problem 6.11. A coordinate frame is affixed to the surface ofthe Earth some
place on the southern hemisphere. The position has the latitude <P (southern
latitude). The coordinate axis have the following directions:
x-axis: positive towards the East
y-axis: positive towards the North
z-axis: vertical (i.e., along the local plum line).
The vector w describing the rotation of the Earth relative to the heliocentric
system is considered constant in direction and magnitude.
A particle is at time t = 0 located at the origin, and the particle has
the velocity v = (v, 0, 0). We consider motions only over distances small
compared to the radius of the Earth. Air resistance is ignored.
We shall only regard the horizontal motion of the particle, i.e., we let
i = 0 (one can imagine the xy-plane to be a smooth horizontal plane which
by its reaction force constrains motion to the plane) .
(1) Show on a sketch the position of the vector w in the coordinate frame
described, and work out the components of w in the frame.
(2) Write the equation of motion for the x-coordinate and for the y-coordinate
of the particle.
(3) Find the distance of the particle from the x-axis for x = 10 km when
v = 500 ms- 1 and <P = 45° (southern) lattitude. Disregard the iJ term in
the equation containing x.
Problem 6.12. The centrifugal force from the rotation of the Earth pro-
duces a change of the direction of a plumb line, from the direction of the
6.11 Problems 151
gravitational force from the gravitating Earth. Let gE be the acceleration

that gravity "proper" would give a partide near the surface of the Earth.
Assume that the Earth is spherical. Show that the deviation of the plum line
from the direction of gravity "proper" at lowest order is given by an angle 0',
where
w2 p w2 p
o' = -- sin 2<[> ~ --.sin 2<[>
2gE 2g
p = radius of the Earth,
<[> = lattitude,
w = angular velocity of the Earth.
Determine the largest magnitude of 0' . [The deviation is towards the south
on the northern hemisphere.J
Problem 6.13.
A cubic box B is made of wood and is filied with day. The mass of the box
plus day is M . The box rests on a frictionless horizontal table. Athin plate A
of mass m is placed on top of the box B as shown in the figure . The coefficient
of friction between A and B is J.L.
A smali buliet with mass n is shot through B. Just before the buliet
hits B the buliet has velocity v. After the buliet has passed through B the
buliet has velocity v /2. Assume that the buliet moves along a horizontalline
perpendicular to the end faces of B and passes through the centre of mass of
B plus A.
It takes the buliet the time Llt to pass through B. We assume that the
horizontal component of the force by which the buliet acts on the box is
constant (independent of the velocity of the bulIet) . Find the smaliest value
of the time Llt such that the plate A does not move relative to B.
Problem 6.14. Consider a large horizontal frictionless area on the surface

of the Earth, at rjJ0 northern lattitude. Through an impulse a partide is set
in motion with a velocity v and then left to move freely. Ignore ali forms
of friction . Find the period in the uniform circular motion of the partide.
Compare with the period of precession for the Focault pendulum.
Problem 6.15.
A particle is constrained to move in a horizontal plane and without friction,

near the surface of the Earth. The location is 60 deg north, and the velocity
of the particle is 16 m/s.
Consider the motion of the particle in a time interval of 10 s.
1. Calculate approximatly how far (in em) the particle after 10 s has devi-
ated from the initial direction, because of the rotation of the Earth.
[You may take the Coriolis force to be constant in magnitude and direc-
tion throughout the short time interval.]
7. The Problem of Motion
In this chapter we will try to summarize some of the ideas and opinions on
which Newtonian physics rests. These views have been substantially revised
in the 20th century, largely due to the works of Einstein in his special and
general theory of relativity.
It is outside the scope of this book to give a presentation of the works of
Einstein. We shall merely outline the directions in which these theories point.
7.1 Kinematic and Dynamic Views

of the Problem of Motion
It was an important event in the history of physics when Kepler published

his famous three laws early in the 17th century. In these three laws the efforts
of many centuries of study of the kinematical aspects of motion in the solar
system culminated.
Fig. 7.1.
At the time of Leibniz and Newton the relativity of motion was a topic of
discussion. Leibniz emphasized that it only made sense to discuss the motion
of a body relative to other material bodies. According to Leibniz, motion is
change of position of one body relative to another material body. He pointed
out the inconsistency in declaring one particular body to be at rest since the
"rest" of a material body can only be referenced to other bodies.
Let us label these views with the term kinematic relativity. It states that
motion as a kinematic process, a change of spatial distances, is a relative

154 7. The Problem of Motion
concept. The basic feature of this viewpoint is that the state of motion of a
body can only be inferred by a comparison with other bodies. No observation
of a single body can reveal whether the body is in motion or not.
The cosmologies of Ptolemy and Kepler are kinematically equivalent. The
epicycles traced out by the planets in the model of Ptolemy are kinematically
equivalent to the ellipses in Kepler's model.
In the kinematical description of the motion of bodies we are concerned
only with describing the motions as they are observed, not with any under-
lying cause.
Galileo began to look for causes of motion and thus founded dynamics.
Newton used the ideas of Galileo to analyze the motion of celestial bodies
and through an ingenious analysis of Kepler's laws (see Chapter 1) Newton
arrived at the concept of mechanical force as the cause of motion for the
planets.
This dynamical view of motions in astronomical space in turn necessitated
precise assumptions of space and time to be introduced into mechanics. Such
axioms appear first in the works of Newton, as explicit definitions (see Chap-
ter 2).
Newton's discovery of the quantitative connection between the forces
causing the motion and a kinematical quantity, the acceleration, enabled him
to use force (dynamical quantity) as a measure of acceleration (kinematical
quantity).
Newton was convinced that there was such a thing as absolute motion
i.e., the motion of a body can be inferred without any comparison to other
bodies. According to Newton it is not true that all observable phenomena
are the same no matter which one of two bodies is considered to be at rest,
since there are differences as soon as dynamical phenomena are included in
the observations (centrifugal forces, Cariolis forces).
As mentioned above, Newton had to make very precise assumptions about
time and space when formulating his laws. Without these assumptions even
the simplest law of mechanics, the law of inertia, is without meaning.
Newton arrived at the conviction that an empirical reference frame, at-
tached to material bodies - e.g. the Sun - could never be taken as a valid
base for the law of inertia. This law with its close connection to the Euclidian
concept of space where the basic element is a straight line, seems a natural
starting point for dynamics in astronomical space. It is actually through the
law of inertia that the element of Euclidian space, the line, manifest itself
when we are far from reference bodies. Similar remarks can be made for
time, the uniform progress of which is expressed in purely inertial motion. If
one chose to take the rotational period of the Earth as the basis of time, the
law of inertia would not hold, because there are small but finite variations in
the rotation of the Earth.
7.2 Einstein Speaks 155
Through these and similar considerations Newton came to the conclusion

that an absolute space exists relative to which all kinematical quantities, in
particular acceleration, had to be measured.
7.2 Einstein Speaks
It may seem peculiar that Newton, both in the axiom of absolute space and in
the axiom of absolute time, states that these two phenomena exist "without
relation to anything external" . Newton himself often emphasized that he only
wished to investigate phenomena that could be verified by observation. But
that which exist "without relation to anything external" cannot be verified
by observation. We shall return to this issue.
Newton has often been criticized for the phrase "without relation to any-
thing external". In this connection it is instructive to quote from a speech
made by Einstein in 1927, at the bi-centennial for Newton's death:
Theoretical physics has grown from the Newtonian framework, which
provided stability and intellectual guidance for natural science for 200
years.
Newton's basic principles were so satisfactory from a logical point of
view that motivation for questioning them could only arise from the
strict demands of empiricism. Before I discuss this however, I wish
to emphasize that Newton himself, far more than the generations of
learned scientist following after him, was aware of the inherent weak-
nesses in his own intellectual construction. This feature has always
aroused my deepest admiration, and I would like to dwell upon it for
a few moments.
I. Newton's efforts to manufacture his system as tightly bound to
experience as possible are ubiquitous. In spite of this he neverthe-
less introduced the concepts of absolute space and absolute time.
But actually Newton is here particularly consistent. He had realized
that observable geometrical quantities (the distances between mate-
rial points) and their way through time did not characterize motion
in all its physical aspects. He demonstrated this in his famous ex-
periment with the rotating bucket. There must consequently besides
masses and spatial variables be something else that determines mo-
tion. This "something" is expressed in his notion of the absolute
space. He realizes that space must possess a physical reality if the
equation of motion are to make sense, a reality on equal footing with
material points and their distances.
The clear recognition of this reveals both the wisdom of Newton
and the weak points of his theory. For the logical structure of the
latter would undoubtedly be more satisfactory without this shadowy
concept; then only direct observables (mass-points, distances) would

enter the laws.
II. Forces acting directly and immediately at a distance such as those
introduced to describe gravity are not familiar from everyday pro-
cesses. Newton meets this objection by stressing that his law of grav-
itational interaction is not meant to be the final explanation but is
a rule derived from experience and observation.
III. Newton's theory contained no explanation of the remarkable fact
that the inertia and the weight of a body are determined by the same
quantity, the mass. Newton was quite aware of this puzzle.
None of these three points are severe logical stumbling blocks for
the theory. In a sense they just express an unfulfilled desire in the
struggle of the scientific mind to obtain a complete conceptual picture
of natural phenomena.
In this quotation we have highlighted a few lines. As a comment to those
lines we note the following ideas .
A • •• • B •
• • •
g •
g •
A A
Fig. 7.2.
In Figure 7.2, the two physical systems - particles with given masses
and mutual distances - are assumed to be kinematically equivalent at time
t = 0, each within its reference frame. Since the two systems are identical,
the particles have the same interactions. This does not, however, imply that
the systems will evolve in the same manner within their respective reference
frames. It will be the case if the two frames are in relative uniform linear
motion. But if frame B is accelerated relative to frame A, the two systems will
evolve differently within their frames. There must consequently, in addition
to masses and spatial variables, be something else that determines motion.
This "something else" was to Newton the effect of absolute space.
Before we go on to the description of the properties of the Newtonian
equations of motion, we will briefly review what it means for a physical law
to be symmetrical under a given transformation.
7.3 Symmetry 157
7.3 Symmetry
We will use Weyl's definition:

I. A body is symmetrical under some operation when it has the same ap-
pearance before and after the operation. Sometimes this property is called
invariance under the operation in question. An often quoted example is
a vase, symmetrical with respect to rotations around some axis A, as in
Figure 7.3.
A A
Fig. 7.3.
If the vase has a spot, it is asymmetrical under any rotation where the
angle of rotation differs from an integer multiple of 360 0 •
II. A physical law is symmetrical under some transformation when the math-
ematical expression of the law looks the same before and after the trans-
formation. This is often formulated as the physical law being invariant
under the given transformation.
7.4 The Symmetry (Invariance) of Newton's 2nd Law
I y'
I'
y
~ ____________~X'
z z'
Fig. 7.4.
We demonstrated the invariance of Newton's second law under a Galilei trans-

formation in Chapter 4, but it bears repeating here. A Galilei transformation
is a relation between two inertial frames, I and I', where the frame I' moves
relative to I with a constant velocity u (see Chapter 4). Time is common to
the two frames. We have:
r' r - ut,
t' t.
To find the relation between the velocity as seen from frame I and the velocity
relative to I', we use the chain rule:
dr' dr dt dt
-=---u-
dt' dt dt' dt' '
or
v' = v - u,
as t' = t. We assume now that a force F is acting on the particle. Then, the
equation of motion in I is
d2 r
m dt 2 = F. (7.1)
Differentiating the velocity relation further, we find:

d 2 r' d2r
dt,2 dt 2
Note: here is the essential feature of the Galilei transformation. When u is
constant, the acceleration of a particle is the same when observed from the
two frames, and acceleration becomes an absolute quantity.
Since furthermore m' = m and F' = F, we conclude that if (7.1) is the
equation of motion as seen in I, then the equation of motion in I' is given by
d2 ,
,_r__ F'
m d 2 t' - . (7.2)
Comparing (7.1) and (7.2) we see that Newton's second law is the same
("looks the same") in any two frames connected by a Galilei transformation.
In other words, Newton's second law is invariant under a Galilei transforma-
tion.
7.6 The Asymmetry (Variance) of Newton's 2nd Law 159
7.5 Limited Absolute Space
The relativity principle of mechanics is the starting point for all the following
considerations.
This principle is intimately connected with the concept of absolute space;
in fact the relativity principle of mechanics sets limits on the physical reality
of absolute space.
What do we mean by physical reality? A concept has a physical reality if
and only if the concept is associated with phenomena that can be verified by
a physical measurement.
We have here defined reality as the concept is commonly understood in
physics. It follows that the concept of a point fixed in absolute space does
not correspond to any physical reality.
If an observer claims that some particle is at rest relative to absolute
space, another observer can with equal right claim that the particle is in
uniform motion with respect to absolute space. No physical measurement,
mechanical or otherwise, can decide the issue. Physically, it thus becomes a
meaningless issue.
Because of the relativity principle of mechanics, Newton's absolute space
loses some of its physical reality.
We frequently use the words "reference frame" in place of the word
"space". We can thus formulate the relativity principle of mechanics as fol-
lows:
There is an infinite number of equivalent reference frames - the so-called
inertial frames - that are mutually in uniform linear motion. Referred to
these frames the mechanical laws are valid in their simple classical form.
Notice how closely connected space and mechanical laws are. Is it space
that impresses the mechanical laws on the material bodies, or is it the me-
chanicallaws that give rise to the entire concept of space?
7.6 The Asymmetry (Variance) of Newton's 2nd Law
We have seen that a single point in Newton's absolute space is a concept

without physical reality. One may ask: what then remains of Newton's abso-
lute space? The answer - from within the Newtonian world picture - is this:
The resistance a body offers against acceleration, must be interpreted as an
effect of absolute space. Any change in velocity relative to absolute space
requires a force.
Reference frames that are accelerated relative to absolute space - and
consequently relative to all inertial frames - are not equivalent to inertial
frames or to each other. We can describe any motion relative to such frames,
but Newton's second law is not invariant under a transformation to such
frames.
From Chapter 6 we recall the following results on the transformation

of Newton's second law when we go from an inertial frame I to a frame S
accelerated relative to I.
x
z
Fig. 7.5.
A particle of mass m is acted upon by a force F and moves through space.

We can regard its motion either from the frame I or from the frame S. The
motion described from I, obeys the simple law:
mR=F
The accelerated frame S is in a state of motion characterized by the transla-
tional acceleration Ro of its origin, and a rotation vector w. The equation of
motion of the particle, referred to S, takes the form:
mr = F - mRo + mw 2 p - 2mw x r- mw x r. (7.3)
Here, r is the vector from the origin in S to the particle. F, m, and the coor-
dinate t are assumed to be invariant quantities. One can see that Newton's
second law is variant (asymmetrical) with respect to a transformation to an
accelerated frame. This asymmetry means that we can by mechanical exper-
iments alone decide if we are accelerated with respect to Newton's absolute
space, or with respect to the inertial frames.
Put differently: Newton argued from (7.3) that accelerated motion is abso-
lute, and that this type of motion can be inferred without relation to anything
external, in particular without refererence to other bodies. Let us assume that
Ro = 0 and that w = o. Then (7.3) is:
mr = F + mw 2 p - 2mw x r. (7.4)
It was particularly in the second term (the centrifugal force) that Newton
found support for his doctrine of absolute space.
Newton devised the following procedure for the determination of absolute
space. Consider two heavy bodies, each of mass M, connected with a thin
elastic rod. This system is placed in empty space far from all galaxies. Now
measure the tension in the rod. If there is no tension in the rod, the rod is
7.6 The Asymmetry (Variance) of Newton's 2nd Law 161
M
Fig. 7.6.
nonrotating relative to absolute space. If there is a tension T in the rod, we

must have:
T = Mw 2 (L/2)
where L is the length of the rod. We have neglected the mass of the elastic rod.
In principle therefore, we can measure Iwi, i.e., the angular velocity relative
to absolute space by measuring the tension in the rod. Without relating the
system to anything external, we have measured a rotational velocity relative
to absolute space.
Nature has equipped us with a very basic example of a rotating system:
Earth itself. The Earth is not an exact sphere. If the Earth had been at rest
in absolute space it would, according to Newton, be spherical because only
its own gravity acts on its masses. The fact that the Earth has the shape of
an ellipse of revolution is proof, according to Newton, that the Earth rotates
relative to absolute space.
Centrifugal Centrifugal
force in the force in the
rotating rotating
frame frame
Fig. 7.7. The Earth as a body in rotation relative to the inertial frames
The centrifugal forces, acting in the rotating reference frame, are driving
the masses away from the axis of rotation. Without referring to external
bodies (e.g. stars) one can therefore deduce that the Earth is rotating relative
to absolute space. The procedure is obvious: measure the distance from pole
to equator. Then measure the circumference around equator and divide the
result by four. The two results are not equal: the distance from the center of
Earth to a pole is less than the distance from the center to the equator. This
is an indication of rotation relative to absolute space.
It is important to be aware of the following, with respect to the Newtonian
viewpoints. The above phenomena do not, according to Newton, arise because
of rotation with respect to other masses (e.g. the fixed stars), but because
of rotation with respect to absolute space. If the Earth was at rest, and the
fixed stars revolved around it once every 23 hours 56 minutes, no centrifugal
forces would act. The Earth would not be oblate, and the acceleration of
gravity would not vary with latitude. Similarly, a Foucault pendulum would
not show any drift.
The kinematical description is the same whether one regards the Earth
as being stationary and the starry sky as rotating, or vice versa. According
to Newton, the dynamical laws of motion give rise to an observable difference
between the two situations.
Newton's idea of absolute space is based therefore on concrete facts. Fic-
titious forces are well known from everyday life here on Earth. The Earth is
not spherical, but ellipsoida1. Observing the solar system, we find that plan-
ets trace out orbits where gravitational attraction is balanced by centrifugal
forces. Remote binary star systems show that the law of centrifugal forces
in rotating reference frames are valid thousands of light years away. Appar-
ently the presence of centrifugal forces is a universal phenomenon. Newton
concluded that fictitious forces result from a direct action of space itself.
7.7 Critique of the Newtonian View
One of the first to seriously criticize the above conclusions was Ernst Mach,
late in the 19th century. In his work The Science of Mechanics (1883) Mach
puts forward the point of view that mechanical experience can teach us noth-
ing about absolute space. He insisted that only relative motions are verifiable
and only these therefore have physical relevance. Newton's "proof" of abso-
lute space based on dynamical evidence must therefore be an illusion.
The Newtonian point of view rests fundamentally on the assumption that
Earth would not be ellipsoidal and the acceleration of gravity not less at equa-
tor than at the poles if it was the fixed stars that rotated in absolute space.
Mach points out that this assumption exceeds the verifiable and criticizes
Newton for having introduced nonverifiable elements into his mechanics.
Mach attempted to construct a mechanics without nonverifiable concepts.
He was convinced that fictitious forces were directly related to interactions
with the entire mass distribution in the universe. His attempt to construct
a new mechanics failed . One reason for this is that he ignored an important
property of fictitious forces: they are proportional to the mass of the body.
Gravitational forces have this property as well. There thus seems to be a
connection between the inertial mass (i.e., the property of a body that it
resists acceleration) and the gravitational mass (i.e., the property of a body
that causes it to attract other bodies) .
Furthermore Mach seems to have ignored the new developments in the
theories of electromagnetic phenomena in his efforts to construct a new me-
chanics. The relativistic theories of electromagnetic phenomena opened the
possibility of eliminating the dogma of absolute time.
The new mechanics was constructed by Albert Einstein.
7.8 Concluding Remarks

Before we conclude our discussion of the Newtonian world picture we wish
to restate the following observations.
Consider two railroad cars A and B. The cars are on a horizontal track;
A is at rest and B has a constant velocity v relative to A.
The invariance of Newton's second law under a transformation between
the frame where A is at rest and the frame where B is at rest means that
all mechanical experiments will take the same course inside the two cars. A
pendulum hanging from the ceiling will have a vertical equilibrium position
and the period of its oscillation will be the same inside each car. In other
words: by studying only mechanical phenomena inside the car, an observer
in B cannot tell whether he is at rest or in uniform motion relative to A.
Note: the fact that the basic mechanical laws are the same in the two
frames rests on firm experimental ground. One thing is to claim that all
motion is relative and base this claim on purely kinematical considerations,
e.g. that all motion is a change of distance relative to other bodies, and from
this conclude that motion is relative. It is a completely different matter to
verify by experiment that pendulum motion is the same in the two cars.
The two cars are found to be not only kinematically but also dynamically
equivalent. Is all motion then relative?
Suppose a car C has an acceleration a relative to car A. The two cars
(A and C) are of course again kinematically equivalent, but Newton's second
law is variant (asymmetrical) under transformation to a system accelerated
relative to an inertial frame. The pendulum in C will have an equilibrium
position not parallel to the end wall, and its period of oscillation changes by
a factor Jg/ J g2 + a2.
The equations of motion in the three reference frames are:

In A: mR= F
In B: mR.= F
In C: mR.= F-ma
"A spot has appeared on the vase"
Is acceleration absolute after all? One way out of the asymmetry is, as
Mach did, to claim that C is accelerated relative to the average mass distri-
bution of the universe (or the fixed stars as it is usually put) and that these
distant bodies directly give rise to the fictitious forces. In other words the
fixed stars are the direct cause of the fictitous forces that appear in acceler-
ated reference frames.
It seems to have been Mach's belief that fictitious forces are acceleration-
dependent actions of the distant stars. These actions should then be perceived
on an equal footing with gravitation or electromagnetic forces . If the car C
was at rest and the fixed stars were accelerating ta the left, the equilibrium
position of the pendulum would - according to Mach - be as shown, not
parallel to the end wall. According to Newton it would be parallel to the end
wall.
In the modern formulation: the fictitious forces are, according to Mach,
acceleration dependent interactions with a field generated by the mass distri-
bution of the universe. This should be compared with the velocity-dependent
interactions between a moving charge and a magnetic field of distant currents.
A uniform translational velocity relative to the stars cannot be perceived. An
acceleration relative to the stars can be detected. Why this difference between
the first and the second derivative with respect to time?
The view - that distant masses determine the inertial frames and thus
the local inertial properties of matter - has been called Mach's principle.
Mach's principle has played a certain part in cosmology and it influenced
Einstein in his construction of the general theory of relativity.
The about lQll stars of our galaxy are in the shape of a flat disc (see
Figure 7.8) .
x
Fig. 7.8. The galaxy with the heliocentric reference frame
If there is some truth to Mach's principle, our galaxy participates in the

determination of the inertia of a partide. The galaxy constitutes only a small
fraction of the total mass in the universe, but it is very much doser than
the rest. Since it is highly anisotropic in shape, one might imagine that the
inertia of a partide would be anisotropic. For instance, for the coordinate
frame shown, we might have:
mxx Fx,
myY Fy,
mzz Fz ,
Experiments using nuclear magnetic resonance have found that the iner-
tial mass of a nucleus to a very high degree of accuracy is isotropic, i.e., it is
independent of the direction of acceleration. This experimental observation
has been interpreted by Mach principle proponents as simply stating that
the inertial influence of our galaxy is vanishing compared to that of all other
masses in the universe. We shall not dwell on the details of this discussion,
and only point to the following considerations.
If one, from Mach's principle, believes that the overall mass distribution
of the universe determines local inertial mass, it is logical to wonder if there
are other connections between the global structure of the universe and basic
physical laws here on Earth. Would, for instance, the charge, mass or spin of
an electron change if the number of (or distribution of) protons and neutrons
in the universe changed? Cosmology is a science in rapid evolution. Perhaps
we shall some day know the answer.
We conclude this section with the following remark. The objective im-
portance of a discovery and the subjective interpretation it is given by the
discoverer can be in contrast to one another. Newtonian dynamics became
the cornerstone of theoretical physics. Newton's own interpretation has not
survived. Another example is the Schrodinger equation. This equation became
a cornerstone of quantum mechanics, but Schrodinger's own interpretation
of his famous equation has not survived.
8. Energy
We introduce the concepts of work, kinetic energy, and potential energy in

a general setting. The fundamental importance of these concepts arise from
the fact that they permit a formulation of a conservation law for mechanical
energy.
8.1 Work and Kinetic Energy
Consider a particle moving under the influence of a force F and described

from an inertial frame. The force F in question is arbitrary; it can depend
on the position r of the particle, or on other variables in the problem. As the
particle moves from the point A to the point B along some path (see Figure
8.1), work W AB is done by the force. This work is defined through the scalar
product and according to the curve integral:
W AB == i B
F·dr (8.1)
Recall that a· b = I a I I b I cos(8). The integral is understood to be

taken along the orbit of the particle. If the displacement dr is so small that
the force can be considered constant throughout the displacement, the work
performed along the displacement is the differential dW = F . dr. Note that
Fig. 8.1. Definition of work done by a force. A particle moves along the indicated
curve from A to B and a force F acts on the particle

168 8. Energy
if the force vector is everywhere perpendicular to the tangent vector of the

curve, no work is done by the force along the curve.
Assume that F denotes the vectorial sum of all forces acting on the parti-
cle. We may use Newton's second law to rewrite the defining equation (8.1),
using m as the mass of the particle:
l B lB
A
F . dr =
A
dv . dr =
m-
dt
lB
A
dv . vdt
m-
dt
{B mv. dv = {B d (~mv2) = ~mv~ - ~mvl .

JA JA 2 2 2
The quantity T = !mv 2
is called the kinetic energy of the particle. Using
this notation, we conclude that
WAB = i B
F· dr = TB - TA. (8.2)
The work done by all forces acting on a particle

equals the increase in kinetic energy of the particle
Equation (8.2) can be expressed in differential form:
dT = F· dr, (8.3)
or
dT dr
-=F·-=F·v. (8.4)
dt dt
Equations (8.2)-(8.4) are valid for all types of forces; the results are based
only on Newton's second law. The scalar product F . v is called the power
done by the force F.
According to (8.4) the rate of change of the kinetic energy is the scalar
product between force and velocity. If the particle happens to be acted upon
by a force which is always perpendicular to the velocity vector (e.g., the
Lorentz force on a charged particle in a magnetic field) there is no change in
the kinetic energy.
It is interesting to compare the set of equations (8.2)-(8.3) for the change
in kinetic energy to the set of equations describing the change in momentum:
dP = Fdt; ..1P = ltB

tA
F dt; P = mv .
Expressed in words the two sets of equations state that

- the work done by the force J F . dr gives the change in kinetic energy of
the particle;
J
- the impulse of the force F dt gives the change in momentum of the
particle.
8.2 Conservative Force Fields 169
8.2 Conservative Force Fields
If we consider how to calculate the amount of work done directly from the
definition (8.1) and Figure 8.1, a fundamental problem is immediately evi-
dent: in order to calculate the work done by the force it is necessary to know
the orbit of the particle. Since the basic problem of dynamics is exactly to
compute the orbit of the particle when the force is known, one must solve
the entire problem first in order to compute the work done by the force. It
may seem then, that the work integral J F . dr is not a useful concept at all.
In fact, as we shall see, the work integral is of immense value, in particular
when the motion is caused by so-called conservative forces.
A force field is called conservative if the force vector F of the field
J:
depends only on the position r of the particle and the work integral
F . dr is independent of the path of integration, depending only
on the initial point A and the final point B, of the path.
Fig. 8.2. Defining a conservative force field. The particle is moved from A to B
along the path 1 or along the path 2
Figure 8.2 shows two points: A and B, and two different paths, 1 and 2,
along which the particle can be moved (by you, for instance!) from A to B.
The condition that the force field is conservative is:
[B F . dr = [B F . dr ,
along 1 along 2
where path 1 and path 2 are arbitrary, and A and B can be any pair of points
in the region of space where the force field acts.
An equivalent formulation of this condition is
170 8. Energy
F· dr F· dr 0,
along 1 along 2
f
or
F ·dr = O.
This is the most common definition of a conservative force field: The work
integral around any closed curve in the field, is zero.
The force of friction is an example of a force that is not conservative. The
frictional force always points against the direction of motion. If we drag a
block around a closed curve on a table, the force of friction will do a finite
amount of work on the block. In this case j F . dr of- 0 so friction is not
conservative. Friction is not a fundamental interaction. Friction is ultimately
a consequence of electromagnetic interactions between molecules in the block
and molecules in the table. In dragging the block around the closed curve a
certain amount of heat has been produced, i.e., the chaotic thermal motion
of the atoms in the block and the table has - through frictional coupling -
been increased. See Example 8.9.
8.3 Central Force Fields

A central field of force is a field where the force F(r) is everywhere directed
towards (or away from) a center 0, and where the magnitude of the force
depends only on the distance r of the particle from O.
We proceed to show that all central fields of force are conservative. In
Figure 8.3 we have chosen two points, A and B in the field, and drawn two
arbitrary paths between them, path 1 and path 2. The field is conservative if
the work integrals along such two arbitrary paths are equal.
As an aid to the computation we have drawn two dashed arcs centered
in 0, and with radii rand r + dr, respectively. These two arcs bound small
segments drl on path 1 where the force is F l , and dr2 on path 2 where
the force is F 2. Since the two points have the same distance r from 0, the
magnitude ofthe forces is the same, F = IFli = IF21. The angles between drl
and Fl and dr2 and F2 are denoted (h and B2 as in Figure 8.3; consequently
and thus
Fl . drl = F2 . dr2 = Fdr .
The contribution to the work integral from the two segments is the same,
and consequently
F ·dr F ·dr.
along 1 along 2
8.4 Potential Energy and Conservation of Energy 171
r+dr
o
Fig. 8.3. A central force field is conservative
The work integral is independent of the path, and the field is thus conserva-
tive. In particular this means that the work done by the force along a closed
loop is zero:
i B
F . dr + ~A F . dr = f Fdr = O.
along 1 along 2
The gravitational force on the Earth from the Sun is a central force. If
we take the center of the Sun to be a fixed point in an inertial frame, the
gravitational force field around the Sun is a central field. The electrostatic
force on one proton from another proton is also a central force, a repulsive
force.
8.4 Potential Energy and Conservation of Energy

For a particle in a conservative force field, we can introduce the concept of
potential energy. The potential energy U = U(r) is a scalar function of the
position vector r of the particle.
Choose an arbitrary point P in the force field as the reference point for
U. We define the value of U at this point to be zero, U(P) = O. The potential
energy at any other point A is now defined through the equation:
U(A) == - LA F . dr. (8.5)
In words: the potential energy at the point A is equal to minus the work
that the force field does on the particle when the particle is moved from the
reference point P to the field point A.
172 8. Energy
Note that we do not have to specify along which path the particle was
brought from P to A. The field is assumed to be conservative, so the work
is independent of the path. This is precisely the reason why potential energy
makes sense only for conservative force fields.
Furthermore, we can now establish the important theorem about con-
servation of mechanical energy in a conservative force field. Consider two
arbitrary points A and B in the field (Figure 8.4). We have:
i A
F . dr = -U(A) and i B
F . dr = -U(B) .
B
P •
•
A
•
Fig. 8.4.
From this:
iBF.dr= iPF.dr+ iBF.dr=U(A)-U(B).
Furthermore, from the theorem (8.2) on the work integral we have:
l
A
B
F . dr =
1 2
-mvB - -mvA
2 2
1 2
.
The conclusion is that for two arbitrary points A and B:
~mv~ - ~mvi = U(A) - U(B)

2 2
or, with the terms in a different order,
1 2 1 2
"2mvB + U(B) = "2mVA + U(A).
We see, that the sum of the kinetic energy T and the potential energy U
does not change when the force causes the particle to move from A to B. This
sum T + U is called the total mechanical energy of the particle, or just the
mechanical energy. Since A and B were arbitrary, the mechanical energy is a
constant of the motion. Or equivalently, the mechanical energy of a particle
is conserved for motion in the conservative force field:
8.4 Potential Energy and Conservation of Energy 173
IT + U = Eo = constant
Under the influence of a conservative force field, the particle moves in such
a way that the sum of the kinetic and potential energy has a constant value.
The value of this constant depends only on the initial conditions.
Actually, all fundamental forces in nature are conservative. The notion
of conservative forces has been introduced to exclude frictional forces. The
frictional force, which is not a fundamental force, converts mechanical kinetic
energy into heat. As is well known, heat is basically motion of the atoms con-
stituting the warm body in question. If we could keep track of each and every
atom, the law of conservation of energy would be generally true, and there
would be no need for the introduction of friction forces, or nonconservative
forces .
The theorem about conservation of mechanical energy is of immense im-
portance in physics.
A Mathematical Note. In order to define the potential energy in (8.5) we

considered the force F to be known as a function of position. In some cases
it may happen that it is the potential energy function U(r) which is known.
The question is then, what force field F has given rise to U(r) .
According to (8.5), the potential energy results from an integration of
the force. Since differentiation and integration are opposite operations, the
force can be found by differentiation of the potential energy function. Let us
first consider the one-dimensional case (two examples of this are in the next
subsection); here, (8.5) has the form:
i X
F(x')dx' = -U(x) .
From calculus we then know that
F(x) = _ d~~x) .
This result can be generalized to the three-dimensional case. Here, the force
F is a vector with three components (say, in a cartesian coordinate system)
Fx, Fy, and F z , and the potential energy is (in general) a function of all the
spatial coordinates, x, y, and z. The component Fx is then found by differ-
entiating U with respect to x, considering y and z to be constants through
the differentiation. This operation is known as partial differentiation and is
denoted by the symbol au lax:
Fx = _ aU(x, y, z)
ax
Similarly, one has for the other components:
F __ aU(x,y,z)
y - ay
174 8. Energy
and
Fz = _ fJU(x, y, z)
fJz
The vector
fJU au fJU
grad U = ( fJx'
ay' fJz)
is known as the gradient vector for U. It points in the direction where the
change in U per unit of length is largest. In compact notation, F = -grad U.
For a one-dimensional motion of a particle the force is conservative if it is
a function of position only. The question arises as to whether a corresponding
statement is true of three-dimensional motion. Phrased differently: if the force
acting on a particle is a function of the position coordinates only, is there then
always a function U(x, y, z) that can serve as the potential energy function?
The answer is no. The components of the force have to satisfy certain
criteria if a potential function is to exist.
Assume that a potential function U does exist. Then we have:
fJFy fJ 2 U
ax - fJxfJy .
The order of differentiation can be interchanged, and the two expressions are
equal. That means:
fJFy
(8.6)
fJz
These expressions are the necessary conditions on Fx, Fy, and Fz for a po-
tential energy function to exist. These conditions express the condition that
dW = Fxdx + Fydy + Fzdz

is a total (exact) differential.
In books on calculus it is shown that these conditions are also sufficient:
if (8.6) holds at all points in space, the force components are derivable from
some potential function U(x, y, z).
8.5 Calculation of Potential Energy
When using the energy theorem, it is important to master the techniques of

calculating potential energy. Below are a few examples of the calculation of
J: J:
U in simple situations.
Advice: use the definition (8.5): U(A) = - F·dr, or -U(A) = F·dr,
Le., use the work the field does on the particle and not the work "you" may
happen to do "in moving the particle". We have not introduced that work at
all.
8.5 Calculation of Potential Energy 175
Example 8.1. Constant Gravitational Field. We choose a z-axis with

direction upwards and origin at the surface of the Earth. We choose U = 0
for z = o. The force on a particle is parallel to the z-axis, and near the surface
of the Earth it has the constant value: F(z) = -mg.
/ / / / / / / /;// / / / / //
The potential energy at a height h above the surface is then found from:
foh (-mg)dz' = -U(h),
which gives
U(h) = mgh.
Example 8.2. Spring Force
The horizontal x-axis has its origin at the equilibrium point of the spring.
We choose U = 0 at x = o. When the spring is stretched or compressed to a
position x away from the equilibrium position it acts with a force F = -kx
on the mass, where k is the spring constant. We then have:
foX (-kx')dx' = -U(x),

or
1
U(x) = 2kx2 .
Example 8.3. Gravity Outside a Homogeneous Sphere. We are as-

suming here that the gravitational field around a homogeneous sphere is
176 8. Energy
known to be the same as if all the mass of the sphere were assembled in the
center. This theorem will be proven in the next section.
r m
The force depends only on the distance r of the partide from the center. The
value is
Mm
F(r) = - G2- .
r
We set U = O at infinity. We then find:
-U(r) = - l00
r Mm
GTzdr' .
From this we get
U(r) = [_G Mm ] r = _G Mm .
r ' 00 r
8.6 The Gravitational Field

Around a Homogeneous Sphere
We shall now demonstrate the important and often used fact that outside
a homogeneous spherical distribution of mass the gravitational field is the
same as if the entire mass was concentrated in the center of the sphere.
8.6.1 The Field Around a Spherical Shell
Let us begin by consider ing a thin spherical shell with its center in the point
O, a radius R, and a total mass M s (Figure 8.5). We are going to calculate
the potential energy of a partide of mass m at the distance r from the center
of the shell. That the shell is "thin" means that the thickness of the shell is
negligible relative to the other length scales in the problem, R and r . Since
we are dealing with a sphere only the distance of the mass from the point O
matters.
8.6 The Gravitational Field Around a Homogeneous Sphere 177
r m
Fig. 8.5. Calculating the potential energy outside a thin spherical shell
A small surface element on the shell can be considered a point mass dMs .
The distance to m is rl, and we know that dMs contributes to the potential
energy of m with an amount
dU(r) = _C dMsm .
rl
This follows from Example 8.3 . The reference point for U is chosen as in
Example 8.3, Le., U(oo) = O.
Those points of the spherical shell that all have the distance rl from m
form a circle perpendicular to the line through m and 0, and with the center
on that line. The angle between r and the vector to some point on the circle
is denoted by G. The points on the shell with distance to m between rl and
rl + drl constitute a ring, as indicated on Figure 8.5. Corresponding to the
edges of the ring are the angles G and G + dG. The width of the ring is then
RdG and the radius is R sin G. The total area of the ring is 27r R2 sin GdG.
Denote by u the mass per unit of area of the shell:
Ms
u = 47rR2 .
The amount of mass in the ring is the area of the ring times u. The contri-
bution of the ring to the potential energy of m at rl is
dU(r) = _Cu27rR2 sin(G)dG· m .

(8.7)
rl
We can now integrate the contributions from all the different rings that make
up the shell. The variables rl and G are related by the cosine relation:
ri = R2 + r2 - 2rRcos(G),
which gives rise to the relation
This is inserted into (8.7) and one gets:

178 8. Energy
U(1') = -G
a27rRm
l'
l r-R
r +R
d1'l
m
= -G47rR 2 a- .
l'
Finally, using the definition of a, we arrive at:

Msm
U (l' ) = -G--, l' >R.
l'
The potential energy of the point mass m outside the spherical shell depends
on l' as if the entire mass of the shell was concentrated at the center!
It is also of interest to know the potential inside the shell, i.e., for points
with l' < R. The integration procedure is the same as before, only the lower
limit is changed (see Figure 8.6):
U(1') = -G a27rRm
l'
l r +R
R-r
d1'l = -G47rRam .
From this,
U(1') = _GM~m, l' < R.
Inside the shell then, the potential energy is constant, maintaining the value
which the outside potential assumes at the surface of the shell, l' = R.
Fig. 8.6.
The force F on the particle is derived from the potential energy by differ-
entiation with respect to the spatial coordinates. Expressed in terms of the
unit radial vector e r = r 11', we have:
dU
F = -Trer'
Using the two expressions for U inside and outside the shell, we get:
F-{
- -G(M
0 s ml1' 2 )e r
1'<Rj
l' > R.
Inside the shell, the particle does not feel any gravitational pull at all, and
outside the shell the force is the same as it would be between two point
masses, Ms and m respectively, separated by 1'. Figure 8.7 illustrates the
dependence of U and F on 1'.
8.6 The Gravitational Field Around a Homogeneous Sphere 179
U
R r
IFI R
Fig. 8.7. The gravitational field around a spherical shell with radius R. Potential
energy and magnitude of force as functions of distance from the center
8.6.2 A Solid Sphere
We now return to the original problem: the gravitational field around a ho-
mogeneous sphere. We divide (as the layers of an onion) the sphere into
concentric shells all of which are so thin that the above calculation applies.
Denote by mi the mass of the ith shell and let there be N shells total. If
the test particle is outside the sphere, r > R, the ith shell contributes to the
potential energy of the particle by an amount
r >R,
and the total potential energy is

N N
U(r) = L Ui(r) = - L C mim
r
;
i=1 i=1
and since 2::[:1 mi = M, we get

Mm
U(r) = - C - , r>R. (8.8)
r
Again, this is the potential energy the test particle would have if the entire
mass of the sphere was concentrated at its center.
It is instructive to compute the potential energy inside the sphere as well.
We can think of the particle moving through the surface of the sphere into
the interior, perhaps moving in a narrow shaft. If we know the force F which
acts on the particle at some distance r1 from the center, we also know the
potential energy as a function of distance:
180 8. Energy
J: F(r - 1) . dri = U(R) - U(r) .
In order to find the force at rl, we divide the sphere into two regions. One is
the shell with inner radius rl and outer radius R at the surface of the sphere.
The rest is the small sphere with radius ri. The particle is imagined to be
located at the surface of the small sphere with radius ri. The region outside
the small sphere, i.e., the region with rl < r < R, may be thought of as
composed of a number of thin shells, all of which gives rise to no force on a
particle within them. The only force on the particle arises from the matter
in the small sphere with radius ri.
The matter is distributed homogeneously, so the inner sphere has the
mass in proportion to its volume: (rI/ R)3 . M. The force F(rr) is then
rl)3 Mm
F(rl) = -G ( R rr Mm
e r = -G R3 rl, rl < R .
The work integral can now be calculated:
r
JR F(rr) . drl = _G
Mm
R3
r
JR rIdrl = _G Mm (r2 _ R2)
2R3 .
This expression equals U(R) - U(r). We know from the previous calculation
that U(R) = -GMmjR, so
Mm
-GIl - U(r) = -G Mm
2R
(r2 )
R2 - 1 ,
i.e.,
U(r) = -G Mm r2) .
2R ( 3 - R2 (8.9)
Figure 8.8 shows the radial dependence of U and 1F I.

Note, that the quantities U and F computed above are the potential en-
ergy and force respectively on a test particle with mass m. If one is interested
in characterizing the field around a spherical shell or a solid sphere, it is cus-
tomary to introduce the gravitational potential iP and the field strength g(r)
by dividing out the mass m of the test particle:
iP(r) == U(r) ,
m
and
F
g(r) == - .
m
8.7 Examples 181
u
•
R r
IFI
Fig. 8.8. The gravitational field around a homogeneous solid sphere with radius
R. Potential energy and magnitude of force as functions of distance from center
8.7 Examples
Example 8.4. Particle on a Frictionless Curve in the Gravitational

Field of the Earth
I
(2)
Remember: a frictionless curve means that the reaction force R (the force
of constraint) is perpendicular to the curve at any point, which again means
that R· v = O. The reaction force performs no work on the particle when the
particle moves along the curve. The velocity of the particle when it reaches
point (2) is independent of the shape of the curve and dependent only on the
initial velocity and the height h through which the particle has fallen. If we
start the particle in (1) with velocity zero, its velocity in (2) is determined
by
182 8. Energy
1 2
2mv = mgh, v = y'2gh .
Example B.5. String Force in the Pendulum
I· L
··············................1........... ~...
mg
A pendulum has mass m and string length L. Find the string force at the
lowest point of the orbit if the pendulum is started with velocity zero from a
point where () = 60° . (The string is assumed to be massless and taut at all
times). Note: the string force T performs no work, T· v = 0 at every point.
Solution.
/
,,
60' " ,,
,,
,,
,,
,,
,
T 'p/
""""
••••
./ L-TL cos 60'
" ............... . ........ i
v ,.:.;~-..
mg
The motion of m is not uniform circular motion. The particle has both a
centripetal and a tangential acceleration. The string force gives only a con-
tribution to the centripetal acceleration.
At the lowest point of the orbit we have
v2
my =T-mg.
From conservation of energy,
8.7 Examples 183
Combining the above two equations, we get
T = 2mg.
Questions. Are there any points on the orbit where the tangential acceleration
is zero? Are there any points on the orbit where the centripetal acceleration
is zero? l:,.
Example 8.6. The Gravitational Potential Outside the Earth. The

binding energy of a particle in the gravitational field of the Earth is the
amount of energy that must be supplied to the particle in order that the
particle is just able to escape the gravitational field of the Earth.
(1) What is the binding energy of a satelite with mass 100 kg at rest at the
surface of the Earth. (Let us in this question assume that the Earth is a
homogeneous nonrotating sphere.)
(2) Assume now that the satelite is at rest at the equator of the rotating
Earth. Would the kinetic energy of the satelite (relative to the absolute
geocentric frame) make any difference in its binding energy to the Earth?
Solution.
(1)
rOO mM GM
w= JR G-;:zdr=m R
9
=6.3x10 J.
(2) The kinetic energy, relative to the absolute geocentric frame, of the satel-
lite rotating along with the Earth (at the equator) is:
where T = 23 hours 56 min, and R = 6.37 X 106 m. We find:

T = 1.1 X 107 J = 0.011 X 109 J.
As long as we want our binding energy with only one significant digit
the kinetic energy of the satelite at rest relative to the Earth makes no
contribution to the binding energy of the satelite.
Example 8.7. Potential Energy Due to Electric Forces. What is the

ionization energy of the hydrogen atom in its ground state? An equivivalent
184 8. Energy
question is: what is the binding energy of the electron in the ground state of
the hydrogen atom?
Solution. In the ground state the average distance between the proton and
the electron has the value a = 0.53 x 10- 10 m. Classical mechanics may be
used in the stationary states. In the classical picture of the ground state, the
electron executes uniform circular motion around the proton. The binding
energy is thus:
w
w
Inserting numerical values we get: W = 13.6 eV. (1 eV = 1.6 x 10- 19 J;
q = 1.6 X 10- 19 C; 1/(47rfo) ~ 9 x 109 Nm 2 C- 2 .)
Note: the ratio of the electrical and the gravitational attraction between
the proton and the electron in the hydrogen atom is:
Fel L 1 ~ 2.3 X 1039 ~ 1040 .

FG 47rfo Gmpme
This is one of the dimensionless numbers that a complete theory combining
electrical and gravitational forces should be able to explain. We do not at
present understand how such an enormous dimensionless number may appear.
Does it have something to do with cosmology?
There are about (10 40 )2 = 1080 protons in the observable universe. The
universe is believed to be about 4.7 x 10 17 s old. A second, however, is not
a fundamental time scale. As a fundamental time scale, one could take the
time it takes light to move the distance known as the "classical electron
diameter": if an electron is considered a classical particle (as opposed to
quantum mechanical), a diameter of 4.5 x 10- 13 em can be associated with
the electron. The time for light to cross this distance is thus 4.5 x 10- 13 /3 X
10 10 = 1.5 X 10- 23 s. Measured in this unit of time, the age of the universe
is approximately 10 40 !
If you want to dig deeper you may consult The Feynman Lectures Vol I,
Chapter 7, section 7-7. The question is certainly interesting and fundamental:
Are there any relations between the properties of a single particle (proton,
neutron, electron, and so on) and the rest of the Universe? The entire problem
is also connected to Mach's principle, discussed in Chapter 7. 6
Example 8.8. A Thnnel Through the Earth. Assume that a very narrow
straight tunnel has been drilled through the Earth. The tunnel passes through
the center of the Earth and has been drilled along the axis of rotation of the
Earth. (No fictitious forces act on the particle in the tunnel.) The Earth is
assumed to be a sphere with a homogeneous mass distribution and radius R.
8.7 Examples 185
(1) Show that a particle P in the tunnel may perform harmonic oscillations.
Neglect frictional effects.
(2) Let us assume that the particle is dropped with initial velocity zero at
one end of the tunnel, e.g., at B. How long will it take until the particle
for the first time appears at A ?
Solution.
(1) The tunnel is very narrow! The force on the particle is of magnitude
F = G M mr / R3 and always directed towards the center of the Earth. The
particle thus has an equilibrium position at the center of the Earth and
- as F is proportional to the displacement r - the particle will perform
harmonic oscillations about the equilibrium position.
(2) The mass performs harmonic oscillations according to the equation of
motion:
or
d2r
m - = -kr
dt 2
The period of oscillation is
T = 27r/f;.
The first time the particle emerges at A it has gone through one half
oscillation. The time is
t ~ = 7rJ ~ = 7rj¥
42.2 minutes.
Note . If a particle "falls along" the surface of the Earth (we neglect fric-
tional effects) we have:
186 8. Energy
-----
A
mM
G R2 =mg,
v yfRg.
The time it takes the particle to go from B to A is:
= 1rR = 1r {ii = 42.2

t
v V-g minutes.
A harmonic motion is equivalent to the projection of a uniform circular mo-

tion onto a diameter. 6.
Example 8.9. The Asymmetry of Nature
Consider a block of mass M on a horizontal table. The coefficient of friction

between the block and the table is fL. The block is now given an initial ve-
locity Vo. The block then slides across the table and stops after a distance D
determined by:
1 2
2Mvo =fLMgD.
The original macroscopic kinetic energy T = ~ M v5 has - through the work

of the frictional forces - been converted into microscopic motion of the atoms
in the block and the table.
The main outcome of the experiment is that the original macroscopic and
"ordered" kinetic energy has been converted into "disordered" kinetic energy,
also called heat.
What about the reverse process? Would it be possible that the block and
the table cooled, and the energy thus liberated was converted into macro-
scopic ordered kinetic energy, i.e., would it be possible that the block at rest
8.7 Examples 187
suddenly started to move back, while the temperature of the table dropped?
Such a process would certainly not contradict the energy conservation the-
orem in its widest sense. If we kept track of each and every single atom in
the block and the table, we would find that their kinetic energy had dropped
by exactly the amount of energy the block had gained. Why do we then feel
with certainty that the block will not start to move? This question touches
one of the most fundamental theorems of theoretical physics: the second law
of thermodynamics. We cannot of course formulate the law at this point. We
may state that energy in the form of disordered (thermal) energy never spon-
taneusly (i.e., by itself) converts completely into ordered macroscopic kinetic
energy. In other words: energy located on "many degrees of freedom" never
collects spontaneusly on "few degrees of freedom". In due time you will learn
- loosley speaking - that:
- ordered motion in its entirety can always be transferred into disordered
(thermal) motion;
- disordered (thermal) motion will never spontaneously convert completely
into ordered (macroscopic) motion.
In other words: work can be converted into heat, but conversion of heat into
work needs deeper considerations. Some disordered (thermal) motion can be
converted into ordered motion, i.e., macroscopic kinetic energy. For example:
the combustion energy in a rocket engine is disordered (thermal) motion.
Some of this thermal energy is converted into macroscopic kinetic energy in
the majestic liftoff of the rocket! But you will find that not all of the energy
liberated in the combustion is converted into macroscopic kinetic energy.
A large part of our civilization rests on the conversion of heat into work
via combustion engines of various sorts. Even the engine inside our bodies
- among other tasks - converts disordered kinetic energy (gained from the
oxidation of reduced carbon compounds) into macroscopic kinetic energy
when we move.
Actually, the entire Universe is expanding, and the energy in the Universe
is continually being spread into more and more degrees of freedom. Life itself
is a by-product of this process. Gravity and the second law of thermodynamics
govern how the world works on the large scale - but it takes some time and
effort to understand the law of gravity and its connection to the second law
of thermodynamics. !:::,.
Remark. It is sometimes stated from government agencies that we should
have as a goal to conserve energy. Taken litterally, the statement is without
meaning: In any process, be it fire, wind, combustion, chemical reactions, or
whatever, the energy is the same before, during, and after the process.
What is it then that we want to conserve? It is evidently not the energy.
You will learn about this in detail, when you study thermodynamics. The
essential point is that the constant energy of the Universe is distributed in a
different manner before and after a natural process.
188 8. Energy
8.8 Review:Conservative Forces . and Potential

The kill(~tic €)uergy T of a partide is giv;en by T == ~mv2 .
1:
F· dr = 1 2
~kmg A along B
An equivalent defiriition is: .f F . dr = O along a.riy dosed path irI the
field.
For a conşetvative forcefield One tan introduce the potential energy
U(x, y, z) ofthe partide,defineHby:
i~?];'Y'z) F.. dr = =U(x,y,i\.

}p
Here -U(x, y, z) :::the work performed by the forces of the field
when the partide is moved from ~n. arbitrarily chosen 2i€lro position
(P) to the point (x, y, z) in the·field.
In a conservative force fieldthe eonservation law for meehanical en-
"ergy is valid.:
In words: aparticle moving un"cler the influente of a eonserv(ttive

a
foree moves in sueh way thaţ the sum of its kinetic energy and
itspotential energy hasaconstant value, determined by the ini.tial
eonditions.
8.9 Problems 189
8.9 Problems
Problem 8.1. A water jet from a fire hose has a muzzle velocity u = 25
m S-l. How high would the water jet go if it were directed vertically upwards
from the surface of the Earth?
Problem 8.2. Calculate the escape velocity for the planet Jupiter. In 1994
fragments of the comet Shoemaker-Levy hit the surface of Jupiter with a
velocity of about 60 km s-l .
Problem 8.3.
I·
A
A small particle of mass m slides without friction on a large sphere B. The

sphere B is fastened to a horizontal table in the gravitational field of the
Earth, such that B cannot move. To begin with the mass is at rest at the
highest point A of the sphere. We now give the mass a very small push and it
starts to slide down the sphere B. Find the angle e at which the mass leaves
the sphere B (neglect the kinetic energy the mass has when it starts to move).
Problem 8.4. Consider Example 8.8.

(1) With what velocity Vo should a projectile be started at the center of the
Earth to reach the surface of the Earth?
(2) How much time to would it take a particle to drop through a straight fric-
tionless tunnel drilled through the Earth between Copenhagen and Los
Angeles? (Show first that the particle may perform harmonic oscillations
in the tunnel.)
Problem 8.5. A cube of glass (B) has a spherical cavity of radius r. A

particle P of mass m is placed at the bottom point (0) of the hollow sphere.
We assume that P can slide without friction on the glass surface.
(1) Let us first assume that B is at rest on a table in the gravitational field
of the Earth. Assume that P is started from 0 with a horizontal velocity
Vo. Find the reaction R from the glass cube on the particle as a function
190 8. Energy
~g B
a
of B, where B is the angle the radius vector of the particle forms with the
vertical through O. (Note: the mass of the glass cube is assumed to be
much greater than m, so that the motion of P does not cause B to move.)
Also, determine Vo such that the particle just gets around the sphere.
(2) We now give the cube an acceleration a to the right along the table.
Assuming a = g..;3 , find the new equilibrium position of P.
(3) Let us start the particle at the point 0 with velocity zero relative to the
glass cube (the glass cube remains accelerated to the right with acceler-
ation a = g..;3). Find the reaction force F from the glass cube on the
particle P, when P passes through its new equilibrium point D.
(4) Let us assume that the acceleration a is equal to g. Where is the equilib-
rium position now?
Problem 8.6. A particle P with mass m can slide down an inclined plane
and into a circular loop with radius R.
(1) The particle starts at a height h above the top of the loop. The value of
h is such that the particle can just move around the circular loop. Find
h.
(2) Assume that the particle instead starts at a height 2h above the top of
the loop. Find in this case the force F with which the loop acts on the
particle P when the particle is at the height R above the lowest point of
the loop.
8.9 Problems 191
Problem 8.7.
vi ~g
p/
m" ---.nm
J;;;;;;;;;;
1 A
;;;.
A particle P of mass m is fastened to the end of a string with length L.

The particle starts from rest, in a position where the string is taut, and the
direction of the string forms an angle of 60° with the vertical. In the lowest
point of the circular arc, the particle P hits the body A.
Before the collision the body A is at rest on a frictionless horizontal table.
The body A has mass nm where n is some positive number.
The central collision between P and A is completely elastic. The accelera-
tion of gravity is g, and we neglect air resistance. The known quantitites are
thus n, m, L, and g.
(1) Just before the collision the particle P has kinetic energy To. Find To.
(2) Determine the velocity of the body A and the particle P just after the
collision. Furthermore, determine that fraction q of the kinetic energy To
of the particle that is transferred to the body A.
(3) Examine the limiting value of q as n ~ 00.
After the collision the particle P swings back.
(4) Determine cos 8 0 , where 8 0 is the angle between the string and the
vertical, when P is at the highest point of its orbit after the collision.
(5) Investigate the expression for cos 8 0 for n ~ 00. Comment on the result.
Problem 8.8.
ti
L
I
h :
I
jj---------------------
I
I
I
I
I
: section 1 section 2
I
I
A roller coaster is contained in a vertical plane and has the shape shown on
the figure. Section 2 of the roller coaster is symmetrical about the line L, and
192 8. Energy
has the shape of a sine curve, y = asin(x7l'/l). The dimensions of the car are
so small that the car can be considered a point particle. The track is assumed
to be frictionless, and the car starts at the point marked A (see the figure).
Determine the maximum value that h can have if we require that the car
does not leave the track at the top point in section 2.
Problem 8.9.
(1) Let us assume that the Sun is a homogeneous sphere with mass M and
radius R. Show that the gravitational self energy of the Sun is
3GM 2
U=----
5 R
(U may also be called the gravitational binding energy of the Sun, i.e.,
the energy that is needed to move all the particles of the Sun away to
infinity.)
(2) The Sun radiates the power P = 4 X 10 26 J s-1. Assume that the radiated
energy per second stems from the liberation of gravitational energy. From
the so-called virial theorem it is known that half the liberated gravita-
tional energy would go into an increase of the temperature of the Sun;
the other half would be radiated into space.
Estimate, on the above assumption, the order of magnitude of the age of
the Sun (this is known to be about 4.6 x 109 years).
9. The Center-of-Mass Theorem
In this chapter we investigate the motion of the center of mass for a system
of particles, acted upon by external forces. As an important special case we
shall consider the motion of the center of mass of the system of particles when
no external forces act on the system.
9.1 The Center of Mass

Consider a system consisting of N point masses (Figure 9.1). The system is
described relative to some inertial frame 1.
Fig. 9.1. Definition of the center of mass of a system of particles. The system could
be a rigid body. The center of mass point is denoted eM
The N point masses may be mutually interacting, and they may even form a
so-called rigid body. If the forces of interaction are so strong that the relative
positions of the masses do not change during the motion, the system is called
a rigid body. The following derivations hold for rigid as well as nonrigid bodies.
The position vector RCM for the center of mass is defined through the
equation
N
MRcM = Lmiri, (9.1)

i=l

194 9. The Center-of-Mass Theorem
where mi and ri are the mass and the position vector of the ith particle, and
M = I:i mi is the total mass of the system. We use here and in the following
the symbol I:i to denote a sum over all the particles in the system.
Differentiating (9.1) with respect to time, we find
MRcM = Lm{ri = LPi :::::P. (9.2)
The total momentum of the system, I:i Pi, is denoted P. Equation (9.2)
is then the statement that:
The total momentum P of a system of particles is the
same as that of a particle with mass M moving with the
velocity of the center of mass.
The theorem is often put in the form:
P = MVCM. (9.3)
where VCM is the velocity of the center of mass (eM).
We shall now study the motion of the center of mass when external forces
are acting on the system of particles. Differentiating (9.2) once more with
respect to time we get
(9.4)
According to Newton's second law, mii\ is equal to the total force acting on
the ith particle.
The total force acting on the ith point mass may be divided into two
parts: the external forces and the internal forces. The internal forces originate
in all the other particles of the system; the external forces represent forces
originating outside the system being considered, e.g. gravitational forces.
Let us denote the force by which particle i acts on particle j by F ij .
Accordingly, the force with which particle j acts on particle i must be labeled
F ji . From Newton's third law these two forces are equal in magnitude but
opposite:
Fij = - Fji .
From this equation we conclude that the sum over all internal forces van-
ish; they cancel each other pairwise. What remains in (9.4) is the sum over
external forces, Le.:
MRcM = LFext .
With I:i F ext ::::: Fext, the equation can thus be written:
MRcM = Fext. (9.5)

The content of (9.5) is often expressed in the following way:
9.1 The Center of Mass 195
The center of mass of a system of particles - rigid or non-

rigid - moves as if the entire mass were concentrated in
that point, and all external forces act there.
This is the center-of-mass theorem. Sometimes called the center-of-gravity
theorem because the center of mass is also a center of gravity when the system
is in a homogeneous gravitational field (see Section 10.5).
We can write (9.5) in several equivalent ways, Le.:
M dvcM = F ext (9.6)

dt '
or
dP = F ext (9.7)
dt .
Newton's second law guides the motion of the center of mass just as it guides
the motion of a single mass point. This is the justification for the often used
approximation of an extended body as a mass point.
From (9.7) we obtain the important theorem of conservation of momentum
P = LiPi'
If no external forces act on a system of particles the total
momentum P of the system is a constant vector.
A system of particles on which no external forces act is often called a closed
system.
The two theorems, the theorem of conservation of momentum for a closed
system, and the center-of-mass theorem have the same physical content.
eM
•
Fig. 9.2. The rocket
As an application, consider a rocket initially at rest in an inertial frame, and

not acted upon by external forces. To begin with the rocket - and with it the
center of mass - is at rest. Now the rocket engine starts. The forces that act
during the combustion of the rocket fuel are all internal forces, and internal
forces cannot provide acceleration of the center of mass. The rocket head
moves forward, expelling gasses behind it; the center of mass of the system
remains at rest (see Figure 9.2).
You cannot lift yourself by your hair! Explain why you are able to jump.
From the moment a gymnast leaps off the floor, she can no longer guide
her center of mass. No matter how the body twists and turns, the center of
mass follows a parabola in the homogeneous gravitational field.
\
\
Fig. 9.3. A shell explodes in mid-flight. Fragments fly off in different directions.
The center of mass follows a parabola before, during and after the explosion. The
forces that cause the shell to explode are internal forces of the system
See Figure 9.3; if we ignore air resistance, the center of mass of a shell fired
from the ground will follow a parabola. Even if the shell explodes in flight, the
center of mass - often called eM - will continue to follow the same parabola.
The forces acting during the explosion are all internal forces for the system
and as such they cannot contribute to the acceleration of the center of mass.
A man standing still on a perfectly frictionless horizontal surface cannot
by walking move himself along the surface. The only forces acting on the
man are gravity and the reaction from the frictionless surface. Both forces
are vertical. No matter what motions the man performs he can not move his
center of mass in the horizontal direction. The eM can only move along a
vertical.
Question. Assume the man is on the frictionless icy surface of a frozen lake.
If the man carries an object in his pocket, can he get to the shore? How?
The reason you are able to walk on a floor is friction. Your walking makes the
floor exert a friction force on your foot, with a horizontal component. This
component accelerates your eM in the horizontal direction.
9.2 The Center-of-Mass Frame 197
9.2 The Center-of-Mass Frame
Let us assume that our system of particles is not subject to any external
forces. The center-of-mass point eM will then - in some inertial frame I -
move with the constant velocity
P
VCM = M'
The reference frame where the total momentum of the particle system is
zero is called the center-oj-mass Jrame. The center-of-mass frame can also be
characterized as the frame where the center of mass of the system in question
is at rest.
The origin of the center-of-mass coordinate frame is usually chosen to
coincide with the center of mass~ and it moves with the velocity VCM relative
to our chosen inertial frame.
The center of mass frame will in the following be abbreviated to the "eM
frame". This expression is widely used in elementary particle physics.
We shall demonstrate the following important theorem:
The kinetic energy of a particle system, as seen in an inertial frame, can

be written as a sum: the kinetic energy of the particles relative to the
eM frame, plus the kinetic energy of a particle with mass equal to the
total mass of the system and moving with the eM velocity.
Fig. 9.4. The center-of-mass frame
Measured in the inertial frame I, the total kinetic energy of the particle system
is:
T = L ~miV;'
i
Here, Vi is the speed of the ith particle relative to 1.

Consider now the motion of the ith particle relative to the CM frame. Let
the velocity of the ith particle relative to CM be Vri (r for relative). From
the Gallilei transformation follows:
Vi = V ri + V CM .
Consequently the kinetic energy relative to I can be written
T =L
,,1 "2mi(Vri + VCM)
2
,
i
or
from which
Now, Li miVri = O. (Why? Because this is how the CM frame was defined!)
The expression is reduced to
(9.8)
or
1 2
T = "2MVCM +Tr ,
where Tr == Li ~mi(v;i) means the kinetic energy of the particles relative

to the CM frame. This quantity is known as the relative kinetic energy. The
first term, ~Mv6M' is called the translational energy. Equation (9.8) is the
mathematical content of the above theorem. The result is sometimes called
Konig's theorem.
Fig. 9.5.
As an illustration of (9.8) we consider a system consisting of three masses

interconneted with springs (Figure 9.5). At some initial time the springs are
9.3 Examples 199
compressed and the entire system is given a velocity and released in the
gravitational field of the Earth. No matter how the individual particles move
relative to one another, the CM of the system will fall along a parabola. The
total kinetic energy of the system can be written as
1 2
T = '2MVCM + Tvibr,
where Tvibr means the kinetic energy associated with the vibrations of the
particles relative to one another, or more precisely relative to the CM of the
system. If, moreover, the system is in rotation around the CM, the relative
energy will contain an additional term Trot, so that Tr = Tvibr + Trot. We
shall return to the study of rotations in Chapter 11.
9.3 Examples
Example 9.1. Two Masses Connected with a Spring
fixed
wall
Two blocks A and B each have mass M. They are connected with a spring
and rest on a frictionless surface. The spring is compressed an amount L1l by
means of a string connected to A and B. The spring constant is k . Initially,
block A is touching a wall. At t = 0 the string breaks.
(1) Calculate the accelerations aA, aB, and aCM of block A, block B and the
center of mass, respectively, just after the string breaks, i.e., before block
B starts to move. Calculate the reaction force from the wall just after
the string breaks.
(2) Calculate the time t = h when block A begins to move (assuming this
happens as soon as the spring reaches its equilibrium length for the first
time).
(3) Calculate the magnitude P of the total momentum for the system and
the magnitude of the velocity of the center of mass VCM for t > h .
(4) Calculate the translational kinetic energy of the system for t > tl .
(5) Find the maximum compression of the spring in the harmonic oscillations
the system performs around the center of mass after A has left the wall,
and determine the period of these oscillations.
Solution.
(1)
k.1l k.1l
aA = 0; aB = M; aCM = 2M .
The external force which gives the CM its acceleration is the reaction
force from the wall. If the wall was not there CM would rerriain stationary
while the blocks oscillated on either side of it. R = k.1l .
(2)
tl = ~ = ~I¥.
(3) For t ~ tl no external horizontal forces act on the system. Consequently
P = I P I is constant, and P equals the momentum of the system for
t = tl' For t = t l , block A is still motionless and B has the momentum
M VB . The velocity VB may for instance be found from
1 2 1 2 (k
2MvB = 2k(.1l) =} VB = .1ly Ai .
From this: P = M.1lJk/M, and VCM = P/2M = (.1l/2) Jk/M.
(4)
1 1
= 2(2M)vCM = 4k (.1l) .
2 2
Ttrans
(5) The energy in the oscillations is Eose = ~k(.1l)2 - Ttrans = ik(.1l)2.

The maximum compression d of the spring occurs when the entire oscil-
latory energy is potential:
In the CM frame, both masses perform harmonic oscillations. Suppose one

mass at sorrie time is a distance x away from the equilibrium position and the
other mass a distance y from its equilibrium position. Since in the CM frame
the center of mass is stationary, we must have M x = My (had the masses
been different this equation would have been MAX = MBY; see below). Here,
we have x = y; the total stretching of the spring is therefore always given by
2x. In the CM frame, each block moves according to the equation of motion
Mi = -k(2x) ,
from which
9.3 Examples 201
Note carefully the following concerning the translational kinetic energy

Ttrans = (1/2)(2M)v6M of the system. From the center-of-mass theorem it
follows that in order for the velocity of the CM to change, an external force
must act. This implies that only an external force can change the transla-
tional kinetic energy. However, one cannot conclude that the change in Ttrans
is the work done by the external force. This would be true if the external
force de facto acted in the center of mass.
The external force which gives the CM its acceleration is the reaction
force from the wall. The work done by the external force on the system is
zero, since the point in which the force acts does not move. While the force
acts on the system, Ttrans changes from 0 to (1/4)k(.1l)2.
Internal forces can change the total kinetic energy, but they can never
change the total momentum of the system.
In the combustion of rocket fuel some of the chemical (potential) energy
of the fuel is converted into macroscopic kinetic energy of the rocket head.
Internal forces have also here produced mechanical kinetic energy.
Questions.
M m
Two cubic blocks A and B of the same size but different masses (M and m,
respectively) are connected with a spring of spring constant k (see the figure).
The mass of the spring is negligible.
Initially, the blocks are at rest on a horizontal frictionless table. A is now
hit by a blow from a hammer. If the blow acts with the force F(t) during the
short time interval .1t, the system receives the total momentum
l<1t F(t)dt = P.
Here F(t) is perpendicular to the end face of block A and a line through F(t)
goes through the center of mass of block A, along the spring and through
the center of mass of block B. The interval .1t is assumed to be so short that
the block A does not move appreciably during the blow. After the blow the
entire system begins to move along the table.
(1) Calculate the maximum compression .1l of the spring. Hint: find first the
initial kinetic energy of A; next, the translational kinetic energy of the
system; finally the energy in the oscillations.
(2) Calculate the period T in the oscillations of the system after the blow.
Hint: in the CM:
mx = My; L11 = x + y; mx = -kL1l; Mjj = -kL1l;
mx = -k (1 + :.) x; Mjj = -k (1 + ~) y.
Answer.
(1)
L11 = pJ (m +:)Mk .
(2)
mM
= 21r ~.
Y"k '
T
l-l=m+M'
where I-l is called the reduced mass of the system.
Example 9.2. Inelastic Collisions. We consider a completely inelastic

collision: A bullet (mass m, velocity v) is shot horizontally into a block of
wood with mass M. The block is at rest on a frictionless horizontal table. It··
is assumed that the bullet is stopped completely by the block and that the
system of the bullet plus block moves as a single body after the collision.
I
I
.
m
______
..
~~~~~:::;:;J
(1) Find the velocity u of the system after the collision.

(2) Find the mechanical kinetic energy Ta after the collision.
(3) Calculate the amount Q of initial kinetic energy which has been converted
into heat.
(4) Find the translational kinetic energy in the lab frame before and after
the collision.
(5) Find the relative kinetic energy Tr before the collision. Tr is the kinetic
energy in the CM frame.
9.3 Examples 203
Solution.
(1) No external forces act in the horizontal direction, so the initial momentum
is conserved:
mv = (M +m)u,
or mv
U= .
m+M
(2)
(3)
1 2 Mmv 2
Q = 2 mv - Ta = 2(m + M)
(4) No external horizontal forces act on the system. The velocity of the CM
is therefore the same before, during, and after the collision:
P mv
VCM = m +M = m +M .
From this we get:
1 p2
Ttrans = 2(m + M)V~M = 2(M + m) 2(M +m)
.......
m
, eM
x M
(5)
1 2 mMv 2
Tr = 2 mv - Ttrans = 2(m + M)
Note. The heat energy Q generated in the collision is equal to Tr ,i.e., the
kinetic energy in the motion relative to the CM. The overall translational
energy in the system cannot be converted into heat. This fact is closely related
to conservation of linear momentum, since
1 p2
Ttrans = 2(m + M)V~M = 2(M + m) ,
where P is the total momentum in the lab frame. Since P remains constant,
we see that the translational part of the kinetic energy is also constant. Of the
total amount of kinetic energy T available in the lab frame T = Ttrans + Tn

only Tr can be converted into other forms of energy (in this instance, heat)
whereas Ttrans remains as translational mechanical energy before, during, and
after the collision.
If we look at the collision from the CM frame, the CM is stationary, and
in particular there is no mechanical energy after the collision. This is the
definition of a perfectly inelastic collision! The kinetic energy Tr has been
completely converted into heat.
This simple consideration - and its relativistic version - is of enormous
importance in accelerator physics, where (in particle collisions) only the ki-
netic energy in the CM frame can be converted into new particles! l':,
Example 9.3. The Collision Approximation
/
.---------------------I /
---....u:
/
/
/
/
/
mV
-----..
/
/
/
M /
/
/
-------~~~~~~~r_n77r.n~~/·/
A box filled with clay (total mass M) rests on a frictionless horizontal table
in the lab frame. A spring is attached to one end of the box, and to a solid
vertical wall. A bullet of mass m is now shot horizontally into the box and is
stopped after a time ,1t. The time interval ,1t is assumed so short that the
box does not move appreciably during the time ,1t. The velocity of the bullet
just before it hits the box is horizontal and has the magnitude v. The spring
constant is k.
(1) Calculate the maximum compression d of the spring.

(2) Numerical example: M = 10 kg, m = 5 g, k = 40 Nm- I , v = 400 ms- I .
(3) What condition must ,1t satisfy if we can really ignore the motion of the
box while the bullet is being stopped?
Solution.
(1) While the bullet stops its motion relative to the box, nonconservative
forces act (heat is produced). Thus, during this process the theorem of
conservation of mechanical energy does not apply. We have, however,
assumed that the distance the box moves during the collision can be
ignored. This means that during the collision no horizontal external forces
act on the box (the spring has not yet begun to compress). Momentum
is therefore conserved in the collision. Denote the velocity of box plus
bullet just after the collision by u (see the figure) then:
9.3 Examples 205
mv = (M +m)u.
When the bullet has stopped its motion relative to the box, the noncon-
servative forces no longer act. From this point on, the energy theorem
may be used to determine the maximum compression of the spring:
or
mv
d= -.;'F-k :;::::(M====+=m:::;:)
The result may be found in a more cumbersome (but instructive) way
by computing the total translational kinetic energy T trans just before the
collision. It is this kinetic energy which is converted into compres.~on of
the spring. The kinetic energy Tr relative to eM is converted into heat.
The calculations proceed like this (see example 9.2):
mv M
VCM = M +m ~ v- VCM = v M +m .
T. = m ( Mv ) 2 +M ( mv ) 2 mM v 2
r 2 M + m 2 M +m 2(M + m) .
1 2 m 2v2 1
Ttrans = T - Tr = "2mv - Tr = 2(M + m) = "2(M + m)v~M .
1 mv
-.;r.k M:=;:=+=m=;::)
2
Ttrans = "2 kd ; d = 7;(
(2) d = 10 cm.
(3) The main point is that the momentum delivered to the box during the
collision by the bullet (mv) should be large compared to the momentum
given to the box by the spring during the time ..:1t. The momentum
delivered by the spring force to the box during the collision is given by
10r
Llt
p = F(t)dt,
where F(t) is the spring force as a function of time during the collision.
As we do not know the details of forces acting between the bullet and
the box we do not know the spring force as function of time either. Thus
we cannot compute P. But we can estimate the order of magnitude of P.
The distance the box moves towards the wall during the collision will be
of the order of magnitude
..:1x ~ u..:1t.
The spring force will thus be of the order of magnitude
F ~ k..:1x ~ ku..:1t .
The order of magnitude of P is then
This should be compared to the momentum delivered by the bullet
mv = (M +m)u.
The condition that must be satisfied (if we apply conservation of momen-
tum during the collision) is thus:
mv = (M + m)u » ku(L1t)2 ,
v+
that is
M m
L1t « k'
Note. The period T in the harmonic oscillations the box performs under the
action of the spring force is
T = 27r v
M+
k m .
In the quoted numerical example we find:
VM+m_~
k - 2 s.
The time L1t should thus be much less than 1/2 s. Is that reasonable? The
approximation in which we neglect L1x ~ uL1t is called the collision approxi-
mation.
Question. A ballistic pendulum can be used to determine the velocity of bul-
," ,
, ,
m
-.-.V < ....
" '
"i
lets. The pendulum consists of a wooden box filled with sand. The mass of
box plus sand is M. A bullet of mass m with a horizontal velocity v strikes
the pendulum and remains embedded in it. The time required for the bullet
to come to rest with respect to the box is L1t.
9.3 Examples 207
(1) M = 2 kg, m = 5 x 10- 3 kg and the maximum height h to which the

pendulum swings is given by h = 10 cm. Find the velocity v of the bullet
just before impact.
(2) How many percent of the original mechanical energy of the bullet is
converted into heat?
(3) State the condition Llt must satisfy in order that the collision approxi-
mation may be used.
Answer.
(1)
M+m ~
v= V2gh = 560 ms- l .
m
(2)
M
/= M+m
That is, more than 99.8% of the original initial kinetic energy is converted
into heat. If M were not free to move, all the original kinetic energy would
be converted into heat. What does it mean that M is not free to move?
It means that M is fastened to the Earth, e.g. by leaning against a wall.
The body that takes up the momentum of the bullet is now M plus the
whole Earth. Such a massive body may absorb the bullet without gaining
any kinetic energy (T = p2/2M, M = 00)
Example 9.4. Freely Falling Spring
I
t
f:::>B
1 k
~ At,
h
/7//%7777/71//.T/7/7/ /77;'l/.077,
A mechanical system consists of two equal nonelastic particles A and Beach

of mass m. A and B are connected by a spring with force constant k and
natural length t. The mass of the spring is negligible compared to m.
The system is allowed to fall freely through a height h in the gravitational
field of the Earth. It is assumed that h » t. During the free fall the direction
of the spring is vertical and the spring maintains its natural length.
The lower particle A now meets a fixed horizontal plane. The collision is
completely inelastic and A thus stays at rest during the collision. The particle
B does not stop until it has compressed the spring a certain distance d < l.
The spring will now expand and the system will jump upwards.
During the collision, Le., during the time interval where A is in contact
with the horizontal plane, we neglect the gravitational force.
(1) Find the kinetic energy T of the system just before the collision.
(2) Find d.
We now consider the system at the time t = to , when the spring again has
its natural length.
(3) Find the total kinetic energy To , the total momentum Po and the trans-
lational kinetic energy Ko of the system at the time t = to .
(4) From the time t = to the gravitational force acts again. The center of
mass of the system will rise to a height hI above the horizontal surface
before it stops. Find hI.
Solution.
(1) T = 2mgh.
(2) The kinetic energy of A is lost in the inelastic collision. The kinetic energy
of B is used to compress the spring:
1
2kd 2 = mgh,
d= Vrn,.
2 9h .
(3)
To mgh,
Po mVA +mvB =m·O+m· J2gh =m~,
Ko
1 2 P6 1
2(2m)vCM = 2(2m) = 2 mgh .
(4) The translational energy Ko is converted into potential energy in the

gravitational field of the Earth.
Questions.
(1) Obtain the last result hI = ih
without using energy conservation, Le.,
by direct use of the eM theorem.
(2) Show that the condition for neglecting gravity during the collision is:
mg« kd.
9.3 Examples 209
Example 9.5. The Wedge
7;};;7;77~
A wedge of mass M is at rest on a horizontal frictionless table. A particle P of

mass m is placed on the wedge. The figure shows a vertical section containing
P and the CM of the wedge. There is no friction between the particle and
the wedge. The height of the wedge is h.
The particle starts to slide down the wedge at t = 0, and the wedge
accelerates to the left (P slides along the straight line contained in the vertical
plane through the CM of the wedge and P).
(1) Find, by applications of the conservation laws, the velocity u of the wedge
relative to the table when P reaches the table.
(2) Find, by direct application of the equation of motion, the time interval
t it takes P to reach the table. Then find u by direct application of the
equation of motion.
Solution.
(1) We apply energy and momentum conservation in a reference frame fixed
relative to the table. The velocity of P relative to the wedge at the moment
P touches the table is v. The velocity of the wedge relative to the table
at the same instant is u.
Conservation of momentum in the horizontal direction gives
Mu = m( v cos B - u) . (9.9)
(There are no external forces in the horizontal direction acting on the
system). Conservation of energy (no heat is produced!) gives:
1 1 .
mgh = 2Mu 2 + 2m [(v cos e - u)2 + (vsme)2] (9.10)

2 2mgh
u = -----,--;-,,..,----,- (9.11)
(M + m) Cn"!:ts'rO - 1)
Check the dimensions!
Let us examine (9.11) for various limiting cases:
M -+ 00 =} u 2 -+ 0;
e -+ 0, =} u 2 -+ 0, because h -+ 0;
7r
e -+ =} u 2 -+ 0, because cos e -+ 0.
2
(2) While P is moving down the wedge, the wedge is accelerating to the left:
Seen from a reference frame fixed relative to the wedge, a fictitious force
acts on P. Let the acceleration of the wedge relative to the table be a, and
the acceleration of P relative to the wedge be u. Then (see the figure)
the equation of motion for P in its motion relative to the wedge is:
ma = mg sin e + ma cos e . (9.12)

The equation of motion for P perpendicular to the surface of the wedge
is:
0= N + masine - mgcose. (9.13)
The equation of motion of the wedge along the horizontal surface is:
Ma = Nsine. (9.14)
This provides us with three equations in three unknowns (a, a, and N).
From (9.13) and (9.14) we find a by eliminating N:
mgcosesine
a= . (9.15)
M +msin 2 e
(Check: for M -+ 00, a -+ 0, and for e -+ 7r /2 , a -+ 0.) From (9.12) we
then find a:
9.3 Examples 211
. () mg COS 2 () sin ()
0: = gsm + M + msin ()
2 . (9.16)
(Check: for () ~ 7r/2, 0: ~ g.)

The distance S the particle has moved relative to the wedge is given by
S=-/!-.
sm ()
(9.17)
We thus have, for T,
~o:T2 =
2
-/!-
sm ()
. (9.18)
From (9.18) we find:
T = J 2h
0: sin ()
. (9.19)
To find the velocity u, we use, since the accleration a is constant,
u=aT. (9.20)
Writing (9.20) as u 2 = a 2T2, and using (9.15) and (9.16), we find:

2 2mgh
u = -------;~;-:---..,.. (9.21 )
(M + m) (m~~sTII - 1)
Question. Find the reaction force R from the table on the wedge during the
motion.
Answer.
R=Mg(M+m). (9.22)
M + msin 2 ()
9.5 Comments on the Conservation Theorems
These theorems are not explicitly stated in Newton's Principia. One may
claim that the theorems of energy and momentum conservation indirectly
are contained in Newton's third law.
The rate of change of the momentum of a particle equals the force acting
on the particle. This force has its origin in some other particle on which the
reaction force acts. As an amount of momentum "flows" into a given particle,
an equal amount must - according to Newton's third law - "flow" out of (at
least) one other particle; here momentum is thought of as a vector quantity.
The overall amount of momentum in a Newtonian universe can never change.
If we similarly take into account every (elementary) particle in the uni-
verse, we can say that energy can neither be created nor destroyed, only
transferred from one place to another. The law of action and reaction en-
sures that if energy flows into one particle, the same amount is given up
elsewhere.
9.6 Problems 213
In mechanical physics we have restricted ourselves to considerations of

mechanical energy and we have found it to be conserved only in systems
with conservative forces. The reason is that we cannot keep track of the
detailed motion of all particles (atoms, electrons) in the universe or for that
matter in the smaller mechanical systems under consideration.
9.6 Problems
ci
Problem 9.1.
k :, fi,""
wall
m
.~ / ;' / /,,;' /;' / /
i'
~,
/~~
A block of mass m slides along a frictionless horizontal plane. The block has
velocity v. A massless spring with spring constant k is attached to the block
as shown on the figure above. The block collides with the fixed wall, and it
is assumed that the spring does not bend during the collision.
(1) Find the maximum compression of the spring.
Let us now assume that the block instead of hitting a fixed wall hits another
block of mass M initially at rest on the table. See the figure below.
(at rest)
ci"",,[,"'~'"
(2) What is now the maximum compression of the spring?
,I,
Problem 9.2.
.----.
m v
M
~ fixed
~ wall
A wood block of mass M rests on a horizontal frictionless table in the lab

frame. One end of the block rests against a solid vertical wall. A bullet of
mass m is shot horizontally into one end of the block. The bullet moves a
distance D into the block. Assume that the force acting on the bullet while
it is being stopped is independent of its velocity.
How far would the bullet move into the block if the experiment is repeated,
now without the wall, Le., if the block is free to slide along the table?
Problem 9.3.
___ ,u. v .. __ ~w
u u ~Jr-,-,'-~--''---M::-:------,,''l
A physical system consists of a wood block B of mass M. B is at rest on a

frictionless, horizontal table. A small vertical wall W is placed near one of
the ends of B. The wall is fastened to B. A spring with spring constant k
connects the wall to a smaller block C of mass m (see the figure).
There is no friction between C and B. Initially both C and B are at rest
relative to the lab frame. A pistol bullet of mass f.L is now shot into (and
absorbed by) C. Just prior to hitting C the bullet has velocity v parallel with
the spring.
We assume that the bullet is stopped in such a short time -1t that we
may neglect the distance that C has moved relative to B in the time -1t.
The mass of the wall and of the spring may be neglected.
(1) Find the maximum compression d of the spring, in terms of v, k, M, m,
and f.L.
(2) Find the time T it takes from the time the bullet is stopped till the spring
reaches for the first time its maximal compression.
(3) Find the maximum compression of the spring if the pistol bullet instead
of being shot into C is shot into B.
Problem 9.4.
~ ~ ~v.
"~",,, ", ",,:,
An old cannon is standing at rest on a horizontal plane in the gravitational
field of the Earth. The cannon can slide (recoil) on the ground, and the
coefficient of friction is f.L. The mass of the cannon is M and it is loaded with
a bullet of mass m. The cannon is fired, and the bullet leaves the cannon
with a velocity v relative to the cannon.
(1) Find the distance S that the cannon recoils. It is assumed that the time
-1t it takes the bullet to leave the muzzle is so short that the cannon has
moved a negligible distance in that time.
9.6 Problems 215
(2) Suppose that v = 600 ms-l, m = 1.2 kg, M = 300 kg, J.L = 0.1 , and that
the length of the muzzle is L = 2 m. Show that the order of magnitude
of the momentum Llp given to the cannon by the frictional force during
the time it takes the bullet to pass through the muzzle is small compared
to the recoil momentum imparted to the cannon from the firing.
Problem 9.5. A big sledge is standing at rest on a horizontal icy surface. A

rail is placed in such a way that the sledge can move only along a straight
line. It is assumed that the sledge moves without friction. At one end of the
sledge n persons stand. They all have the same mass.
These n persons can now leave the sledge in two different ways:
(a) All n persons run to the front of the sledge and jump off simultane-
ously, all with velocity v relative to the sledge. This velocity is along the
direction in which the sledge is constrained to move.
(b) One at a time, a person runs to the end of the sledge and jumps off with
velocity v relative to the sledge while the remaining persons stand still.
This process continues until they have all left the sledge.
Demonstrate which of the two methods imparts to the sledge the highest
velocity relative to the ice.
Problem 9.6.
......1 - - - - - 2 0 m - - - - -........
A sledge of mass 120 kg stands at rest on the horizontal surface of an icy

lake. A man of mass 70 kg stands at one end of the sledge so that initially
the distance from the man to the shore is 20 m. The man now walks 3 m
relative to the sledge, towards the shore. Then he stops. Assuming that the
sledge moves without friction on the ice, how far from the shore is the man
when he stops?
Problem 9.7.
A B
:?=77EJ±1:::L/J?77
Two identical wood blocks, A and B, both of mass M, are fastened to the
ends of a spring. The spring constant is k.
Both blocks are placed on a frictionless horizontal table. Initially, both

blocks are at rest and the spring has its unstretched length.
A bullet of mass m ( m « M ) is now fired into A with velocity v, and is
absorbed there. This causes A to begin moving in the direction of the spring,
towards B (see the figure). It is assumed that the time it takes the bullet to
come to rest is so short that A has not moved appreciably before the bullet
has come to rest relative to A.
After the bullet is stopped, we ignore all frictional effects. Furthermore,
we neglect the mass of the spring. Find the following quantities:
(1) the velocity u of A just after the bullet is stopped relative to A;
(2) the total momentum P of the system following the impact;
(3) the velocity VCM of the center of mass of the system after the impact;
(4) the translational kinetic energy K of the system, after the impact;
(5) The total mechanical energy E of the system, after the impact (the po-
tential energy due to gravity is constant and may be set equal to zero);
(6) The maximum compression ..:1l of the spring;
(7) The period T of the oscillations of the spring.
Problem 9.8.
The American physicist Benjamin Thomson (Count of Rumford, 1753-1814)

measured the velocity of cannon balls.
The cannon hangs horizontally in he gravitational field of the Earth (field
strength g, see the figure). After firing, the cannon swings like a pendulum.
The center of mass (CM) of the cannon reaches a height h. The value of his
measured. The unloaded mass of the cannon is M and the cannon ball has
mass m. We neglect the mass of the gunpowder. At the moment the cannon
ball leaves the barrel the velocity of the ball is horizontal, and the magnitude
of the velocity relative to the cannon is v.
We neglect the distance that the cannon has moved during the firing, i.e.,
during the time in which the cannon ball passes through the barrel. We also
neglect air resistance. Determine vasa function of M, m, h, and g.
Problem 9.9. Two blocks, both of mass m, are fastened to a spring as shown
in the figure. The length of the unforced spring is L, the spring constant is
k. We neglect the mass of the spring. The system is attached to the ceiling
9.6 Problems 217
(1) m
(2) m
of the laboratory with a string. The acceleration of gravity is g. Initially, the

system hangs at rest. At time t = 0 the string is cut, and the system falls
freely in the gravitational field. We neglect air resistance.
(1) Determine the position of the center of mass (eM) at times t > O. The
position should be given as the distance YCM from the eM position at
t = O.
(2) Determine the acceleration al of the top mass (1) and the acceleration
a2 of the bottom mass (2) just after the string has been cut.
We now refer the motion for t > 0 to a reference frame that has its origin in
the eM. Denote in this frame the position of the lower mass (2) by z.
(3) Determine z as a function of t (consider only values of t such that (2)

has not hit the floor).
Problem 9.10.
On a smooth inclined plane forming the angle B with the horizontal, two
bodies A and B with masses M and 2M respectively, are placed. See the
figure. The two bodies are connected with a spring, the mass of which can
be disregarded. The spring has spring constant k. The upper body, A, of
mass M, is tied to a peg at the upper end of the inclined plane. The string
is assumed to be massless and unstretchable. Initially, the system is at rest,
and the inclined plane is fastned to a horizontal floor, so that the inclined
plane cannot move. The laboratory is taken to be an inertial frame, and the
magnitude of the acceleration of gravity is g. Known quantities are, M, g, B,
and k.
218 9. The Center-of-Mass Problem
1. Determine the stretching .t1L of the spring.
At the time t = O the string is cut.
2. Determine the magnitude aA of the acceleration of body A and the mag-

nitude aB of the acceleration of body B immediately after the string is
cut.
After the string is cut, the system begins to slide (without friction). Let
us consider the system when the center of mass has moved a distance D down
the inclined plane. The distance D is measured along the inclined plane. See
the figure.
3. Determine the time t for the center of mass (CM) to move the distance
D, and find the translational kinetic energy, Ttrans , when the system
has moved the distance D.
While the system is in motion down the inclined plane, the two masses
oscillate harmonically relative to the CM.
4. Determine the period T of these harmonic oscillations.
Problem 9.11.
A homogeneous circular ring A with mass m is suspended by a string from

the laboratory ceiling. The ring is at rest. Two small heavy balls B and C
(both of mass M) are fitted on the ring, so that they are constrained to slide
along the ring. The movement of the balls is assumed to be frictionless. The
balls are started at the top of the ring with negligible initial velocity, and
they slide down A in different directions. See the sketch for a definition of ().
The acceleration of gravity is g. B and C may be regarded as point masses.
Determine the condition M must satisfy in order for the ring to rise in
the field of gravity. For a value of M satisfying this condition, indicate the
angle () = ()a at which the ring begins to rise.
10. The Angular Momentum Theorem
The previous chapters have dealt with problems of translational motion. We

now begin an investigation of motions where rotation is an important ingre-
dient. We start by associating with one particle the so-called angular momen-
tum, defined with respect to some reference poi~t often taken to be at rest
in an inertial frame. These considerations will then be extended to a system
of particles. Forming the angular momentum for a system of particles with
respect to the center of mass (CM), we shall see that the equation of motion
for this angular momentum has a simple form, even if CM is accelerating.
10.1 The Angular Momentum Theorem for a Particle

Consider a particle of mass m, moving with velocity vector v relative to some
inertial frame I (Figure 10.1). The origin of the coordinate frame is denoted
O. The particle has the momentum p = mv relative to the inertial frame.
We define the angular momentum L relative to 0 as:
L == r x p = r x mv . (10.1)
Fig. 10.1. The definition of angular momentum around the point 0
The symbol" x" between two vectors describes the vector cross product. Since
L depends on the reference point 0 one talks about "the angular momentum
with respect to 0" , or "the angular momentum around 0" .

220 10. The Angular Momentum Theorem
Differentiate L with respect to time:

dL d dr dp
- = - (r x p) = - x p +r x -.
dt dt dt dt
On the right-hand side, the first term equals zero since dr j dt = v, and the
two vectors are parallel. Thus:
Assume that the particle is accelerated by a force F. From Newton's

second law, dpjdt = F, so the equation of motion for L becomes
dL
dt =rxF.
The product N = r x F (Figure 10.2) is called the torque on the particle with
respect to (or around) the point O.
Fig. 10.2. The definition of torque N around the point 0
With this notation:
dL =N (10.2)
dt .
This is the content of the angular momentum theorem:
The rate of change of the angular momentum of a particle around
some point 0 equals the torque on the particle, with respect to
O.
10.2 Conservation of Angular Momentum 221
10.2 Conservation of Angular Momentum
Equation (10.2) implies the theorem of conservation of angular momentum:
particle with respect to °

If no torque acts on a particle, the angular momentum of that
is constant in time.
A classic example is the motion of a particle in a central force field.
A central force field is a field where the force on the particle is
always directed towards (or away from) a point fixed in an inertial
frame. This point is called the force center.
Fig. 10.3. A central force has no torque around the force center
Consider a particle being accelerated by a force of type F = f (r )e r , where

e r is a unit vector along the position vector r relative to the coordinate center
0, and f(r) is some function depending only on the magnitude of F. Such a
force has no torque around ° (Figure 10.3):
N == r x F = r x f(r)e r = o.
°
From the angular momentum theorem it then follows that the angular mo-
mentum of the particle around is conserved. The angular momentum as-
sociated with a particle moving in a central force field is a constant vector.
Angular momentum is a key concept in many physical problems because
several important models of physical systems involve central forces.
Consider a planet orbiting the Sun. Let us assume that the Sun is at rest in
an inertial frame and use the center of the Sun, 0, as origin of our coordinate
°
system. We have just seen that the angular momentum of the planet relative
to is a constant vector. From this it follows that the position vector and
the velocity vector always remain in the same plane (Figure 10.4). The orbits
of the planets around the Sun are planar.
We can now demonstrate that Kepler's second law follows from the fact
that angular momentum of a planet is conserved (see Figure 10.4). The in-
finitesimal increment in area marked by an increment dr along the orbit is:
or
r
Fig. 10.4. Planet orbiting the Sun. The orbit remains in a plane perpendicular to
the angular momentum vector L
1
dA ="2r x dr.
We have represented the area swept out by r as a vector A. The rate of

change of A is obtained by dividing by dt:
dA 1 dr 1
- = -r x - = -r x v.
dt 2 dt 2
Let m denote the mass of the planet. We then have:
dA rxmv L
dt 2m 2m
Since the (mass and the) angular momentum vector L is constant, the
area swept out by the radius vector of a planet changes with a constant rate.
This is Kepler's second law. Note how this follows simply from the nature
of the central force field. The result does not contain any information about
how the force varies with distance.
The planets in our solar system do not exactly follow Kepler's 2nd law.
The most important reason for this is that each planet is not only attracted
towards the Sun, but also perturbed by the gravitational field from the other
planets.
Additionally, even for a single planet and the Sun, the center of the Sun is
not fixed but orbits the common center of mass of the planet and the Sun. In
other words, the Sun is also accelerated towards the planet, and a reference
frame with its origin at the center of the Sun is not an inertial frame. A true
"heliocentric" frame should have its origin in the common center of mass for
the entire solar system.
The Sun has an enormous mass, 330000 times the mass of the Earth. The
center of mass of the solar system is quite close to the surface of the Sun.
The issue will be considered again in Chapter 14.
It is important to note that the concept of angular momentum is not solely
associated with closed orbits. Figure 10.5 illustrates motion with constant
angular momentum along a nonclosed orbit; a free particle of mass m moves
10.3 Torque and Angular Momentum Around an Axis 223
along a straight line with velocity v, as seen from an inertial frame. We

consider the angular momentum relative to some point 0 not on the line,
and use the notation of the figure.
o
Fig. 10.5. Angular momentum for a nonclosed orbit
There is no force acting on the particle, so the torque around any point (in
particular 0) is zero. The constant angular momentum around 0 is given by
La = r x mY,
La is perpendicular to the plane of the paper and points into the plane of the
paper. The magnitude of La is
La = ILal = mvd,
where d is the perpendicular distance from the line to O. The quantity d is
sometimes called the arm of the momentum.
10.3 Torque and Angular Momentum Around an Axis

We have defined torque and angular momentum relative to a point. One can
also consider torque and angular momentum with respect to an axis (a line).
This is simply the projection of the (torque or angular momentum) vector
on that line.
The torque N == r x F, in cartesian coordinates, is
Nx yFz - zFy ,
Ny zFx - xFz,
Nz xFy - yFx .
The projection of the torque N on, say the z-axis, is the component N z . This
is called the torque around the z-axis. Figure 10.6 shows intuitively how N z
emerges: The component Fy acts over the arm x and "creates a rotation in
the positive direction around the z-axis". The component Fx acts over the
arm y and "turns in the negative direction around the z-axis" .
For angular momentum we talk in a similar way about the angular mo-
mentum around some axis; this is the projection of L onto the axis.
r
o :./ y
-------------~~
Fx
x
Fig. 10.6. The torque N z of the force F around the z-axis

for a System of Particles
We shall now develop the angular momentum theorem for a system of parti-
cles. Consider first the total torque (with respect to 0) of the internal forces
for the system (see Figure 10.7). The force on particle i from particle j is
denoted F ij and the force on particle j from particle i is denoted F ji. We as-
sume that the force is directed along a line connecting the two particles. From
Newton's third law we then have that Fij = -Fji . Clearly, the contributions
to the torque around 0 from such a pair vanishes:
N ij rixFij+rjxFji
(ri - rj) x Fij
0,
since the vectors (ri - rj) and Fij are parallel. This is also obvious from
Figure 10.7 since the two forces have the same arm but the torques are in
opposite directions.
o
Fig. 10.7. The angular momentum theorem for a system of particles. There is no
net torque from internal forces
10.4 The Angular Momentum Theorem for a System of Particles 225
Similarly the contributions from the forces between any other pair of
particles vanish. The torque from internal forces is zero.
Now let us assume that in addition to the internal forces external forces
act on the system. We can write the angular momentum theorem for each
particle:
dL i _ N·
dt - t·
Here, Ni is the total torque around 0 from all forces acting on particle i,
external as well as internal. Summing over all particles:
"dL i =
~ dt
"N
~ t,
i i
or
The contribution to the total torque from the internal forces is, as we have
seen, zero. Thus the only contribution to the total torque is from the external
forces. Letting L tot == Ei L i , we then have the following result:
~Ltot = "Next
dt ~ t .
t
This is the angular momentum theorem for particle systems:

The time derivative of the total angular momentum for a particle
system relative to a point 0 fixed in an inertial frame, equals the
sum of the torques of the external forces, around O.
An important special case of the angular momentum theorem is obtained
when there are no external torques on the system: Lto t = constant vector.
When the sum of external torques around a point 0 (fixed in
an inertial frame) vanishes, the total angular momentum of the
particle system relative to the point 0 is a constant of the motion.
When no external forces act on a system of particles, we talk about a closed
system. Thus:
For a closed system the total angular momentum is conserved.
When we developed the center of mass theorem we saw that internal

forces can never change the total momentum of a system. We now have
the analogous result that internal forces can never change the total angular
momentum of a system.
As an application of the theorem of conservation of angular momentum
we consider the system in Figure 10.8. A wheel may turn without friction
around an axis A which passes through the center of the wheel. The axis is
fixed in the lab. A rope is passed over the wheel as shown. Near one end of
the rope is a monkey of mass M. In the other end a bob is attached, also of
mass M. Initially the system is at rest. Disregard the mass of the rope and
the mass of the wheel.
Ig
Fig. 10.B. The monkey
Question. The monkey now begins to climb up the rope with velocity v. How
will the bob move?
Answer. We consider the wheel, rope, bob, and monkey to be one system.
Initially the system is at rest, and there is no angular momentum around A.
The external forces on the system are: The reaction force on the pivot A and
the gravitational forces on the monkey and the bob. The torques around A
of the latter two add to zero. The reaction force acts in A, i.e., it has no arm
and therefore no torque. There is thus no external torque around the axis
A, and so the angular momentum around A will remain zero. The bob will
therefore also move upward with velocity v, cancelling the contribution to
the angular momentum around A from the monkey.
The forces between monkey and rope - complicated as they are - are
internal forces for the system, and internal forces can never generate any
amount of total angular momentum. The center of mass of the system moves
upwards. Where is the external force that produces the acceleration of the
eM? [Answer: the reaction force in A.]
As another example, consider Figure 10.9. A man stands on a potter's wheel

which can turn without friction around a vertical axis. He holds a stick at
the end of which a disk has been mounted so that it can turn without friction
around the stick. Initially the potter's wheel is at rest in the lab, the stick is
horizontal and the disk is spinning as indicated in the figure.
Now the man turns the stick into a vertical position. Since the potter's
wheel turns without friction, no external torque around a vertical axis can
arise on the system "man + potter's wheel + stick" . The vertical component
of the total angular momentum is conserved, and since it was zero initially
it will remain so. When the stick is turned to a vertical position the pot-
10.5 Center of Gravity 227
Fig. 10.9. The potter's wheel. Angular momentum about the vertical axis is con-
served
ter's wheel will start spinning in the opposite direction with such an angular
velocity that the vertical component of the total angular momentum still
vanishes.
The total angular momentum has, however, been changed since the initial
horizontal component has vanished. When the man turns the stick, he feels a
strong reaction force from the potter's wheel, preventing him from starting a
rotation around a horizontal axis. If he was floating in free space (!!) he would
indeed start spinning, both around a horizontal axis and around a vertical
axis as he attempts to turn the stick.
Since he is standing on the potter's wheel, strong reaction forces (from the
bearings that keep the wheel moving horizontally) prevent him from rotating.
These forces are external to the system (man + stick). They only influence
the horizontal components of the angular momentum vector.
10.5 Center of Gravity

Consider a body in the gravitational field of the Earth (Figure 10.10). We
assume that the extent of the body is such that the acceleration due to
gravity, g, can be considered as constant. The total torque of gravity around
some fixed point 0 is then
Mg
Fig. 10.10. The definition of the center of gravity
where mi is the mass of the volume element (mass point) with position vector
ri (from 0).
From the definition of the position vector RCM of the center of mass we
find:
Lmiri = MRcM,
where M is the total mass of the body. Thus,
No =RCM x Mg.
The total torque of gravity on a body can be calculated by assuming the
entire mass to be concentrated at the center of mass. This is the reason the
center of mass is often called the center of gravity.
In particular, the torque of gravity about the CM is zero, and thus a
body in a homogeneous gravitational field and supported at the CM is in
equilibrium for all positions of the body.
10.6 Angular Momentum Around the Center of Mass
x
o
Fig. 10.11. Defining the angular momentum around the eM
For a system of particles the total angular momentum with respect to some
fixed point 0 (Figure 10.11) is defined as:
The value of La depends on which point 0 is chosen as reference point.

Let RCM be the position vector of the center of mass. We can now rewrite
La in the following way: we add and subtract the vector
in the defining equation for La:
La = L(ri - RCM) x mivi +L RCM x mivi .

10.6 Angular Momentum Around the Center of Mass 229
The first term in the above expression is the total angular momentum in the
motion around the eM, i.e., the angular momentum computed with eM as
the reference point. See Section 9.2. The second term can be written RCM x P
where P = MVCM is the total momentum. Thus,
Lo = LCM + RCM X P.
In words: the total angular momentum around some point 0 is the angular
momentum around eM, LCM, plus "the angular momentum of CM" around
the point 0, ReM x P. The "angular momentum of CM" is an abbreviation
of "the angular momentum of a particle with the mass of the entire system,
moving with eM" .
Note that the angular momentum of CM (often called the orbital angular
momentum) depends on the choice of origin, O. LCM (often called the intrinsic
angular momentum or spin) is independent of the choice of origin, O.
The angular momentum theorem with respect to the origin of the inertial
frame is:
dL o _ Next
dt - 0 .
It is often useful to take the angular momentum relative to the center of

mass:
dLcM _ Next
~- CM'
There is an important difference between the above two equations. The first
is valid when the angular momentum and the torque are calculated relative
to a point 0 at rest in an inertial frame. The second is valid no matter
how CM moves. If the center of mass is accelerated we should include a
homogeneous fictitious gravitational field in the system where CM is at rest.
But a homogeneous gravitational field has no torque around CM. Therefore
the angular momentum theorem is valid in its simple form even if CM is
accelerated.
When the external torque about CM vanishes the spin~angular
momentum of the system of particles (the body) is a constant of
the motion.
If we assume that the solar system is isolated, i.e., if we disregard the
forces from the other masses in the galaxy, the total angular momentum of
all the planets in the system remains constant.
Consider our own planet, Earth. The Earth stays in rotation around an
axis through its center of mass, with an angular momentum which is almost
constant. This is due to the fact that most external forces acting on the Earth
(from the Sun, the Moon, and the other planets) have no (or a very small)
torque around the CM.
Later we shall see that due to external forces the Earth's angular momen-
tum vector, which currently points toward the North Star, slowly changes
direction. With a period of about 25800 years the rotational axis of Earth
sweeps out a cone with an interior angle of about 23.5°. In some 12000 years
the axis of the Earth will point toward the bright star Vega in the constella-
tion The Lyre. The phenomenon, which is further described in Chapter 14 is
called the precession of the Earth's axis.
10.8 Examples of Conservation of Angular Momentum

Example 10.1. Particle in Circular Motion. A particle with mass m is
tied to a string which goes through a hole in a smooth horizontal table. The
particle moves in uniform circular motion with a velocity Vo. The radius of
the circle is Ro. By pulling the string one can slowly diminish the radius of
the circular motion to a new value, R I .
10.8 Examples of Conservation of Angular Momentum 231
(1) Calculate the new velocity, VI.

(2) Calculate the increase LJ.T in the kinetic energy of the particle.
(3) Show that the work done by the string force equals the increase in kinetic
energy of the particle.
Solution.
(1) The string force has no torque around the center of the circle. Thus, the
angular momentum about the center is constant:
(2)
l Rl lRl
(3)
2
W = F . dr =- v
m-dr = -1 mv5 [R2 ]
Rg - 1 .
Ro Ro r 2 I
Example 10.2. Rotation of Galaxies, Solar Systems, etc
A particle moving in a central field has a constant angular momentum around

the force center. An important example of this type of motion is a planet
orbiting the Sun. In this example we shall discuss some of the implications
of conservation of angular momentum in gravitating systems.
The orbit of a planet lies in a plane perpendicular to the angular mo-
mentum vector. The motion is two dimensional and it is useful to describe
it in polar coordinates with the force center as the origin. We resolve the
velocity vector in two components: a radial component Vr and a transversal
component Vo (see the figure). The potential energy, U(r), of the planet in
the force field depends only on the distance r to the force center. The total
energy E is constant, and can be written in the following way:
1 2 1 2
E = U(r) + "2mVr + "2mvo
This expression is valid for any particle moving in a central force field.
For a planet of mass m in the gravitational field of the Sun, we have
U(r) = -GMm/r where M is the mass of the Sun and G is the gravitational
constant. The magnitude of the angular momentum about the force center
is L = mrv(} ; the radial component Vr of the velocity does not contribute to
the angular momentum about O. The "angular" part of the kinetic energy is
thus:
1 2 L2
-mv(} = - -
2 2mr 2 '
and the total energy can be written as
L2 1
E = U(r) + -2mr
- 2 + -mv;.
2
Since L is a constant of the motion, we can collect the first two terms in a
so-called effective potential energy Ueff , depending only on r:
L2
Ueff(r) = U(r) + 2mr 2 .
The second term is sometimes denoted the centrifugal potential energy.

The figure below shows Ueff as a function of r for the gravitational po-
tential. Ueff(r) is the sum of the two terms. The constraint of keeping L
constant appear as a repelling potential which adds a positive term to the
effective potential energy.
energy
For a bound particle (a planet!), the total energy E is negative. One sees
from the figure how for a given amount of total energy, the particle can never
come closer that rmin (perihelion) and never be further than rmax (aphelion)
from the force center. The value of r will in fact oscillate between these two
values.
As a further illustration we shall briefly describe the formation of a galaxy.
Consider a gas cloud of astronomical dimensions. The cloud has mass M and
total angular momentum L. No external torques act on the cloud, so L is a
constant vector. The gas cloud begins to contract under its own gravity. When
the volume decreases, the angular velocities increase, keeping L constant, and
the cloud now contains more kinetic energy. The increase in kinetic energy
comes from the work done by the gravitational forces as the particles fall
towards the center of the cloud.
The formation of galaxies is a complicated process, not understood in
details. The following qualitative arguments are to be considered a very crude
sketch.
Consider a particle with mass m, in the outer regions of the cloud, near
its "equator". The particle will have gravitational potential energy of order
U(r) = -GMm/r, where r here is a measure of the distance to the center of
the cloud. The particle will experience an effective potential energy of form:
12 GMm
Ueff(r) = 2mr2 - - r - ,
where I is the magnitude of the orbital angular momentum about the center
of the cloud. This expression for the potential energy is far from exact; since
the mass of the cloud is not uniformly distributed, the distance dependence
will be more complicated than the above. We shall, however, base our crude
arguments on the above form of Ueff.
If the particle is nearer to the center of the cloud it must interact with
other particles in the neighbourhood. Its orbit around the center of mass
(which would otherwise have been an ellipse) is now perturbed. Let us assume
that the particle ends up in an approximately circular orbit with a constant
distance to the rotational axis of the cloud.
If we further assume that the radius r of the orbit corresponds to a min-
imum of the effective potential energy, we can find r from the equation
0= dUeff(r) = _~ + GMm ,
dr mr 3 r2
i.e.,
r = GMm2
This value of the radius for the orbit of the particle is where the gravitational
attraction from the cloud just provides the centripital force for the circular
motion:
v2 GMm
m-=-- -
r r2 ·
In such orbits then, the radius depends on the initial value of the angular
momentum of the individual particle.
This is a crude picture of the evolution of the gas cloud in the direction
perpendicular to the rotational axis. The contraction parallel to the axis is
a different story. The cloud can collapse parallel to L without any minimum
parallel distance to the center of mass. The net result is that the cloud flattens
and forms a disk shaped galaxy.
When interstellar clouds collapse, as in our own Milky Way, the same
process happens although on a smaller scale. The contraction perpendicular
to the rotational axis is halted by centrifugal forces, but contraction continues

in the direction parallel to L. The result is a star surrounded by gas and dust
in an equatorial plane. In such a cloud of gas and dust surrounding the
newborn Sun, the planets Mercury, Venus, Earth, Mars, etc. were formed.
These processes took place some 4.6 x 109 years ago. The solar system is
shaped like a disk, perpendicular to the original angular momentum vector.
All the planets move in the same direction around the Sun, the original
direction of rotation. The solar system "remembers the process of birth",
i.e., the initial conditions.
The formation of planets around newborn stars is much studied, theoret-
ically (including computer experiments), and observationally. Figure 10.12
shows the cloud of gas and dust around the star ,B-pictoris. Planets are prob-
ably forming in this cloud, and perhaps life will emerge on one of these
planets. One of the most essential tasks of natural science in the 21st century
will be to determine whether the Earth is unique in the Milky Way. Related
to this is the problem of the origin of life and its cosmic abundance.
Note the decisive role conservation of angular momentum plays in the
formation of planets and thus of life itself. If the initial gas cloud has no
angular momentum the result will be a spherical star with no planets, and
no life.
From the point of view of mechanics we might ask; what is the original
gas cloud rotating relative to? Relative to astronomical space, or relative to
all the other masses in the universe? In modern physics, absolute space has
been abandoned, and rotation is thought of as rotation relative to the other
masses in the universe. It is worth reflecting upon the fact that uniform
rotation relative to the stars is felt , but uniform linear motion cannot be
detected by any physical experiment at all! 6.
Fig. 10.12. Cloud of dust surrounding the star ,B-Pictoris. Photograph © NASA,
Washington
11. Rotation of a Rigid Body
The general treatment of the motion of a rigid body is rather involved and
we shall in this book consider only certain special cases. The insights to
be gained can be applied in virtually all of theoretical physics, and have
important technological ramifications. Applications of the theorems derived
stretch from studies of the spin of the electron and rotating atomic nuclei, to
investigations of the motion of planets and galaxies.
11.1 Equations of Motion
A body is said to be rigid if the distances between the mass elements (the
particles of the body) remain constant throughout the motion. The most
general laws of motion for particle systems are of course true for rigid bodies
in particular:
In translational motion, the center-of-mass theorem holds:
Md:~M = LFext. (11.1 )
In rotational motion around the center of mass, the angular momentum the-
orem holds:
dLcM _ ""' Next (11.2)

dt - ~ CM ·
The simplest case of rotation of a rigid body is found when the body rotates
around an axis fixed in inertial space and fixed relative to the body. This
case will be given the most attention but we shall also briefly discuss the case
where the body rotates around an axis which changes its orientation in the
inertial frame in question and in relation to the rigid body.
In several important cases - e.g., the gyroscope - the body under consid-
eration will move with one point 0 fixed in the inertial frame. The equation
of motion for the rigid body in this case is:
dLo = ""' Next (11.3)

dt ~ 0

238 11. Rotation of a Rigid Body
In (11.3) we have used 0 as the point relative to which the angular momentum
of the body and the torque of the external forces will be calculated.
Before we move on to apply (11.1)-(11.3) we shall demonstrate an impor-
tant proporty of the vector that describes the rotation of a rigid body.
11.2 The Rotation Vector
Fig. 11.1.
From geometry it is known that the most general motion of a rigid body can
always be separated into a translational motion of a point A in the body,
and a rotation about an axis through A. We thus need six coordinates to
specify the position of a body relative to a given coordinate system: three
coordinates to pin down the point A, two coordinates to specify the direction
of the axis of rotation, and 1 coordinate to specify the angle through which
the body is rotated. We say that a rigid body has six degrees of freedom.
Let us for simplicity assume that the body shown in Figure 11.1 is a plane
disk. The velocity of an arbitrary point P in the body relative to the chosen
coordinate frame may be written as
(11.4)
where W A is the vector describing the rotation about the axis through A. The
vector r A is the translational velocity of A, and rAP is the vector from A to
P.
Let the rotation vector W A be perpendicular to the plane of the paper.
If we choose another point B in the body instead of A, we may describe the
motion of the body as a translation of the point B plus a rotation around
an axis through B. The rotation vector in this case is denoted WB, and the
velocity of the point Pis:
(11.5)
where
11.3 Kinetic Energy of a Rotating Disk 239
rB = r A + W A X r AB .
But then, inserting rB in (11.5):
(11.6)
Since the right-hand sides of (11.4) and (11.6) must be equal, we get:
or,
WB x rBP = WA x (rAP - rAB) = WA x rBP .
We conclude that WA = WB == w. The point B was arbitrary. Therefore the
motion of the rigid body relative to a chosen frame is given by specifying the
translational velocity of an arbitrary point A in the body, and the rotation
vector w. Both the magnitude and the direction of ware independent of the
choice of A. See also page 101.
11.3 Kinetic Energy of a Rotating Disk
OJ
x
Fig. 11.2. A disk rotating around a fixed axis
Let us start by considering a rigid body in the form of a thin disk confined
to the xy-plane. The disk is rotating around the z-axis, with rotation vector
W = (O ,O,w) .
A mass element mi in the disk at the position ri has the velocity
The magnitude of Vi is:

The total kinetic energy T of the rotating disk is found by summing over all
the mass elements:
The quantity in the paranthesis is denoted by I z , and called the moment of

inertia around the z-axis. The kinetic energy for the rotating disk is thus
1 2
T = 2Izw , (11. 7)
where
Before we proceed with the study of rigid bodies in rotation we need two
important results concerning the calculation of moments of inertia.
11.3.1 The Parallel Axis Theorem
The vectors in Figure 11.3 all lie within a disk in the xy-plane, and we seek
to calculate I z . The z-axis is perpendicular to the paper and pusses through
the point 0, and eM denotes the center of mass of the disk. The moment of
inertia about the z-axis may be written as
Iz L mir; = L mi (RCM + r~)2
The cross term 2RcM . 2:i mir~ vanishes because of the definition of the
center of mass. The second term is the moment of inertia ICM about an axis
11.3 Kinetic Energy of a Rotating Disk 241
through eM parallel to the z-axis. Thus, the moment of inertia about the
z-axis may be written as
Iz = ICM + MR~M'
where M = Li mi is the total mass of the disk. This is the parallel axis
theorem.
The moment of inertia about an arbitrary axis z equals the mo-
ment of inertia about an axis parallel to z through the center of
mass, plus the total mass M of the body times the square of the
distance d between the two axes:
The parallel axis theorem is used to determine moments of inertia about

axes that do not pass through eM or possibly not through the body at all.
Note that an immediate consequence of the parallel axis theorem is that the
moment of inertia about an axis through eM is smaller than about any other
axis parallel to this. We shall see later that the parallel axis theorem is valid
in general, not just for bodies with a planar mass distribution (a disk).
The kinetic energy for a disk rotating about the z-axis can - using the
parallel axis theorem - be written as
1212122
T = '~/zW = "2ICMW + "2 MR cMW ,
or, since VCM = RCMW,
1 2 1 2
T = "2ICMW + "2 MVCM = Trot + Ttrans . (11.8)
Equation (11.8) is a special case of Konig's theorem (see Section 9.2).

The angular momentum of the rotating disk around the origin 0 is:
Since ri and Vi are mutually perpendicular we find that Lo is along the z-axis
(the rotational axis) and has the magnitude
Lo = L z = I:: mirtw = Izw .

The parallel axis theorem shows that the magnitude of Lo can be written as:
Lo = Lz ICMw + MR~MW
ICMw + MvCMRcM·
For a planar mass distribution we thus have: the angular momentum about
some point 0 equals the angular momentum about eM plus "the angular
momentum of a particle with the mass M, located in and moving with the
eM". See Section 10.6.
11.3.2 The Perpendicular Axis Theorem
Fig. 11.4.
This theorem concerns moments of inertia about axes that are perpendicular
to each other. We still consider a thin disk (see Figure 11.4). In contrast
to the parallel axis theorem, this theorem is valid only for a planar mass
distribution. Consider the moment of inertia about the x-axis:
and similarly about the y-axis:
Furthermore:
or
(11.9)
This is the perpendicular axes theorem:
For a planar mass distribution (a disk), the moment of inertia

about an axis perpendicular to the disk equals the sum of the
moments of inertia about two mutually perpendicular axes in the
plane of the disk and intersecting where the perpendicular axis
passes through the disk.
We now leave planar mass distributions.

11.4 Angular Momentum of an Arbitrary Rigid Body 243
11.4 Angular Momentum of an Arbitrary Rigid Body

in Rotation Around a Fixed Axis
Consider a rigid body rotating about the z-axis with rotation vector w
(O,O,w), assumed to be constant in time and fixed relative to the body. The
angular momentum about the origin is
Lo = Lri x miVi.
We decompose ri into two vectors ai and hi (see Figure 11.5). The vector ai
is perpendicular to the rotation axis while hi is parallel to it (Figure 11.5).
Fig. 11.5. A rigid body in rotation about a fixed axis
Since
we have that
and
Since for three vectors A, B, and C in general,

A x (B x C) = B (A . C) - C (A . B) ,
we find that Lo may be written as
Lo L miw(ri· ai) - L miai(ri . w)

Lo thus has a component parallel to w (i.e. parallel to the z-axis) and a

component perpendicular to w. The component parallel to the z-axis is Izw,
where Iz is the moment of inertia about the z-axis.
Note that the moment of inertia is defined by means of the perpendicular
distance of the mass point from the axis of rotation, not the distance from
O. As we shall see in due time, the fact that Lo is not necessarily parallel to
w makes the general problem of rotation complicated. For the translational
motion of a body, the momentum is always parallel to v, P = MVCM.
The kinetic energy is
We find that
and consequently
11.4.1 The Parallel Axis Theorem in General Form
A body is rotating about the z-axis. Assume that we know the moment of
inertia of a body about an axis A parallel to the z-axis and through eM
(Figure 11. 6) .
z-axis A
Fig. 11.6. The parallel axis theorem for an arbitrary body
The kinetic energy of the body in its rotation about the axis A is then
1 2
TA = 2IcMw .
Note that w is the same for rotation about the z-axis as for rotation about
A. The motion of the body can be considered either as a pure rotation about
11.5 Calculation of the Moment of Inertia for Simple Bodies 245
the z-axis or as a translation of the CM plus a rotation about A. In both

views the rotation vector is w .
Using Konig's theorem, we may write the total kinetic energy of the body
in rotation about the z-axis as
1 2 1 2
T = '2 MVCM + '2IcMW .
The velocity of CM is VCM = wd (where d is the perpendicular distance from

the rotation axis through 0 to eM). Thus,
1 2 2 1 2
T '2 Md w + '2IcMw
'12 (Md
2 +IcM ) w2 .
If the motion is described as a pure rotation about the z-axis we have:

1 2
T = '2Izw .
Comparing these results we find:
I z = ICM+Md 2 .
Using Konig's theorem we have demonstrated the parallel axis theorem for
an arbitrary rigid body in rotation about a fixed axis.
11.5 Calculation of the Moment of Inertia

for Simple Bodies
11.5.1 Homogenous Thin Rod
dx
L
x
Fig. 11.7.
A homogeneous thin rod has length L and total mass M. A small segment
of length dx has the mass (Mj L)dx. The segment, at distance x from the
rotational axis A, will contribute to the moment of inertia with an amount
The total moment of inertia for the homogeneous rod is now found by inte-
grating over the range of x-values:
fA
M
= L 10
r x dx = "3 M £2 .
L
2
1
By applying the parallel axis theorem we find that for an axis through the
center (of mass) for the rod:
2
1 2 £ 1 2
fCM = -M £ - M (
- ) = -M £ .
3 2 12
The moment of inertia about the axis through the center could also have
been calculated as the sum of two moments of inertia for rods of mass M /2
and length £/2:
fCM = ~M (~)2 +~M (~)2 = ~M£2.

3 2 2 3 2 2 12
For a body of complicated shape the moment of inertia can be found by piec-
ing the body together from "simple" parts with known moments of inertia.
11.5.2 Circular Disk
1ilI-------!. A B
•
•
•
Fig. 11.8.
A homogeneous circular disk has mass M and radius R. The axis A of rotation
is perpendicular to the plane of the disk and passing through the center. A
ring of thickness dr at distance r from the center has the mass
dM = 21l'rdr M = M 2rdr
1l'R2 R2 .
The contribution of the ring to the moment of inertia about A for the disk is
dIA = (dM) r 2 = M 2rdr

R2 r 2 = 2M
R2 r 3dr.
Consequently the total moment of inertia for the homogeneous disk is
2M {R 3 1 2
IA = R2 Jo r dr = 2MR .
From the parallel axis theorem we find that the moment of inertia about an
axis perpendicular to the plane of the disk but at the edge of the disk (the
point B on the figure) is
IB = ~MR2 + MR2 = ~MR2 .

2 2
From the perpendicular axis theorem we can find the moment of inertia about
a diameter of the disk:
1 2 1 2
ID +ID = 2MR , or ID = "4MR .
A homogeneous cylinder can be considered as a stack of disks sharing the

same axis and adding their mass. Thus for a cylinder around its axis of
symmetry:
I = ~MR2.
2
If the cylinder has height h one finds for the moment of inertia about an axis
perpendicular to the cylinder axis and through CM:
I = ~M ( R2 + ~2)
A cylinder is a rod made up from disks!
11.5.3 Thin Spherical Shell
A homogeneous thin spherical shell has mass M and radius R. From the
symmetry of the shell follows that Ix = Iy = I z . We have (Fig. 11.9):
Ix f (y2 + z2) dM ,
Iy f + (z2 x 2) dM ,
Iz f + y2)
(x 2 dM .
The moment of inertia around an arbitrary diameter can thus be written as

x
Fig. 11.9.
or
11.5.4 Homogenous (Solid) Sphere, Mass M and Radius R
Fig. 11.10.
Inside the solid sphere consider first a spherical shell with radius r and thick-
ness dr . This shell has mass:
dM = 41lT 2 dr M = 3r 2 dr M
1. 7r R3 R3 '
3
The spherical shell contributes to the moment of inertia around a diameter

by the amount:
dI = ~ (3~~r M) r2 .
The total moment of inertia around a diameter for the entire homogeneous
ior r
sphere is then
2M R 4 2 2
1= R3 dr = '5 MR .
The moment of inertia for a homogeneous sphere around an axis which is
tangential to the surface of the sphere is
2 2 2 7 2
I tan = '5 M R + M R = '5 MR.
11.5.5 Rectangular Plate
b
O~
x
x
a dx
Fig. 11.11.
A homogeneous thin rectangular plate has sides a and b as shown on the

sketch. The entire mass of the plate is M. We calculate the moment of inertia
around the Z-axis which goes through the CM of the plate and is perpendic-
ular to the plate.
Consider the plate to be made up from rods of thickness dx (see the
sketch).
M M
dM = -b dx . b = -dx.
a a
The total moment of inertia of the plate with respect to the Z-axis is
Iz = ~ 1:~~2 (~~ + x2) dx = ~ (a 2 + b2) .

A different way of obtaining this result is the following. Since the plate can
be seen as a stack of rods each of length b, Ix = 112Mb 2 and similarly Iy =
112 Ma 2. From this, Iz = Ix + Iy = 112M (a 2 + b2).
We conclude this section with a table of some moments of inertia (Table
11.1).
Table 11.1. Moments of inertia for selected bodies

of total mass M Moment ofInertia
j~
a
thin rod
~::9
c:=:iJ circular disk
thin
spherical
shell
homogenous
sphere
Rectangular plate
1 f2 2)
12 M\a +b
11.6 Equation of Motion for a Rigid Body

Rotating Around a Fixed Axis
We are now in a position to apply the angular momentum theorem to the

study of the rotation of a rigid body around a fixed axis which we choose as
the z-axis (see Figure 11.12). The equation of motion is ' (regarding the lab
frame as an inertial frame)
11.6 Equation of Motion for a Rigid Body Rotating Around a Fixed Axis 251
dL o = "Next
dt ~ 0 ,
where the angular momentum Lo and the torque Not are both taken about
the point 0 which is on the axis of rotation. That the axis of rotation is fixed
means that it is stationary in the inertial frame we use (the lab frame) and
is fixed relative to the rigid body.
Fig. 11.12.
When considering rotation about a fixed axis, we can re-write the equa-
tions of motion in a simpler form.
Since the axis of rotation, A, is stationary in an inertial frame and relative
to the body, the inertial proporties of the rigid body with respect to A are
constant. We shall later see that it is just this fact which makes rotation
about a fixed axis a particularly simple problem.
In order to describe the motion we need to consider only the component
along the z-axis of the angular momentum vector, and the corresponding
component of the torque of the external forces along the z-axis. In many
instances those two vectors will have components only along the z-axis. In
particular this will be the case for bodies with a high degree of symmetry
(e.g. a cylinder or a disk) when the rotational axis coincides with the axis of
symmetry.
We now demonstrate that the equation of motion for a rigid body in
rotation about a fixed axis is a scalar equation. The vector w / Iw I is a unit
vector along the axis of rotation. The component of Lo along the axis of
rotation is:
LA = L'I~I=~[~rixmivi]'w
~ [~miri x (w x ri)]'w.
Using the identity A x (B x C) = B (A· C) - C (A· B), we get:
LA ~ [~mdr;W-(ri.W)ri]].W
~ [~mi [r?W 2 - r;w 2 COS 2 (8)]]
~ [~mir;w2 Sin (8)]. 2
The distance of the particle mi from the rotation axis is li ri sin(8)

=
We find
or
LA = JAW,
where h is the moment of inertia of the body with respect to the axis of
rotation. LA is called the angular momentum of the body with respect to
the rotation axis. The component along the rotation axis of the torque of the
external forces, denoted N A, is given by
The equation of motion describing the rotation of a rigid body around a

fixed axis A is then
dL A -N
dt - A,
or, since the moment of inertia about A is constant,
An important special case of the angular momentum theorem for rotation

around a fixed axis results when the torque of the external forces vanishes:
When the external torque around a given fixed axis A is zero, the
angular momentum about the axis A is a constant of the motion.
11.6 Equation of Motion for a Rigid Body Rotating Around a Fixed Axis 253
11.6.1 Conservation of Angular Momentum
Consider Figure 11.13. A man stands on a turntable which may turn, without
friction, about a vertical axis. The turntable (and the man along with it) is
now given an angular velocity WI about a vertical axis while the man holds
two heavy weights strecthed away from his body as in (1) on the sketch. While
in rotation the man pulls the weights towards his body, as in (2). We consider
the man plus the turntable as our system. As mentioned, the turntable rotates
without friction, i.e., the vertical axis is well oiled. No external torques act
about the vertical axis of the system. Therefore the angular momentum about
this axis is conserved.
(1) (2)
Fig. 11.13. Man on a turntable. Illustration of the conservation of angular mo-

mentum
Suppose the moment of inertia about the vertical axis initially is It. When
the man rotates with his arms stretched out, the angular momentum is
When the man pulls the weights towards his body he changes (decreases) the
moment of inertia to a new value, h , where 12 < It. The angular velocity
will increase in such a way that the angular momentum is kept constant:
so W2 > WI·
Consider the change in the mechanical kinetic energy. The kinetic energy
of a system rotating about an axis is
1 2 1
T= -lw = -Lw.
2 2
Consequently
1
-LWI
2 '
1
"2LW2 .
The angular momentum L is unchanged, so we have
T2 > TI since W2 > WI .
The mechanical kinetic energy of the system has increased. The man has
done work against the centrifugal force in the co-rotating frame. Compare
with Example 10.1.
Internal forces can produce a change in macroscopic mechanical kinetic
energy (compare with the rocket). However, internal forces cannot produce a
change in the total angular momentum.
(2)
Fig. 11.14.
Let us finally consider the body of the man without the arms as our
system (see Figure 11.14). The man's body rotates faster after he has pulled
the weights towards his body. The moment of inertia of his body is unchanged,
so the angular momentum of the body has increased. This increase is caused
by a torque. Question: How does this torque originate? The man is assumed
to be fixed to the turntable.
11. 7 Work and Power in the Rotation of a Rigid Body

Around a Fixed Axis
A rigid body may rotate around the z-axis. An external force F, is acting at
the point A which has position vector r (see Figure 11.15). Consider now an
. infinitesimal rotation of the body. In polar coordinates
x = rcos(8) , y = rsin(8).
11.7 Work and Power in the Rotation of a Rigid Body Around a Fixed Axis 255
Fig. 11.15. Work in the rotation of a rigid body around a fixed axis (the z-axis,
out of the plane of the paper)
We find that
dx -r sin( B)dB = -ydB ,

dy r cos( B)dB = xdB.
The work done by the force F when the body turns through the angle dB is
thus
dW Fx . dx + Fy . dy = (xFy - yFx)dB
or
where N z is the torque about the z-axis. The work equation for a rigid body
rotating around a fixed axis is then .
11.7.1 Torsion Pendulum
Fig. 11.16. The torsion pendulum
A torsion pendulum consists of a circular disk, suspended by a steel wire, as

shown in the sketch. When the disk turns, the steel wire will be twisted about
its axis. A torque is now acting to turn the disk back towards the equlibrium
position. When the angle 0 is small, the torque is, to a good approximation,
proportional to 0, i.e.,
N = -CO,
where C is called the torsion constant.
If the disk is turned away from its equlibrium position and released, it
will carry out a simple harmonic oscillation, governed by the equation
Here, I = ~ M R2 is the moment of inertia of the disk with respect to the

vertical axis through its center.
The angular frequency in the oscillation becomes
and consequently the period of the oscillation is
These relations permit a measurement of the torsion constant C for a given

steel wire.
As for the linear harmonic oscillator one can introduce potential energy:
U(O) =- 1()
00
NdO =-
1() -CO dO
1
= -C0 2
2
.

Referred to Various Points
We have demonstrated how the rotational motion of a body may be described

by means of the angular momentum theorem. Before we end this section, we
shall briefly discuss how to refer the angular momentum theorem to various
points inside or outside the body.
A rigid body moves under the influence of forces relative to some given
inertial frame I. An arbitrary particle P in the body is acted upon by external
and internal forces, Ff."t and F?t respectively.
We now consider a point Q, which moves with velocity vQ relative to the
inertial frame I. We wish to use Q as center for calculating the torque and
the angular momentum. The point Q may be inside or outside the body.
11.8 The Angular Momentum Theorem Referred to Various Points 257
Fig. 11.17.
The total torque of the forces about Q is
NQ = 2: (QP X Fpxt) ,
P
and the angular momentum about Q is
LQ = 2: (QP x mpvp) .
p
By differentiating LQ with respect to time we obtain
+ 2: (Q.p x mpvp )
p
Now,
p p
since the total torque of the internal forces vanishes.

Furthermore,
QP = Vp -vQ,
i.e.,
LQ = 2: (QP X Fpxt) + 2: (vp - vQ) x mpvp ,
p p
or, since L:p mpvp = MVCM,

The angular momentum theorem is valid in the simple form
LQ = NQ,
whenever vQ x VCM = 0, that is, only if one of the following conditions are
satisfied:
°
(1) vQ = (Q is a point at rest in the inertial frame);
(2) Q = eM (Q coincides with the center of mass);
(3) vQ is parallel to VCM.
11.9 Examples
Example 11.1. Rotating Cylinder. A circular disk with radius R may

turn without friction about a horizontal axis A. The mass of the disk is M
and its moment of inertia about the axis A is thus I = ~ M R2. A light string
is wound around the side of the disk. A mass m is tied to one end of the
string and supported. The string is taut. At the time t = 0, the support of
the mass m is released and the mass m begins to fall.
Find the angle e that the disk has turned as a function of the time t.
Solution.
The equation of motion for m is

11.9 Examples 259
d 2x
m dt 2 = mg - S. (11.10)
The equation of motion for the disk is
(11.11)
Since we are assuming that the string does not slip on .the disk we have the
following geometrical constraint on the motion:
. .. dw
x = RB + constant =>:i; = RB => x = RB = Rdt . (11.12)
Combining (11.10) and (11.11) with the geometrical constraint we find:
d2 B
(1 + mR2) dt 2 = mgR.
By integration (and the use of the initial conditions x = 0 and B = 0 for
t= 0) we find
1 = ~MR2.
2
Questions.
Assume that we pull with a constant horizontal force F in the string and that
there are frictional forces resisting the rotation of the disk about the fixed
horizontal axis A. We assume that the frictional forces can be represented
by a torque N about the axis A, where N at all times is proportional to
the instantanous angular velocity of the disk. We denote the constant of
proportion " i.e., N = ,W.
(1) The disk will approach a constant angular velocity wf' Determine wf'
(2) Determine the angular velocity W of the disk as a function of time. Initial
conditions: W = 0 at t = o.
(3) Determine the power P yielded by the force F when wf is reached.
Answers.
(1)
wf=-' ,
FR
(2)
where I is the moment of inertia about A.

(3)
F dx = FRw = F2R2
dt f 'Y'
dB F2R2
N- ='YwJ = - - .
dt 'Y
When W f is reached, the power yielded by F will be converted into heat by
the frictional forces. Check the dimensions (Js- 1 ). 6.
Example 11.2. Falling Cylinder. A light string is wound around a ho-

mogeneous cylinder, with mass M and radius r. The other end of the string
is attached to the ceiling of the laboratory. Initially the system is held at
rest with the string vertical (see the figure). At the time t = 0, the cylinder
is allowed to fall under the influence of gravity. The initial velocity of the
cylinder is 0. Disregard the mass of the string.
'~
S
~l
A
Mg
The string force acts on the cylinder in the point A, where the free end
of the string is a vertical tangent to the cylinder.
(1) Find the translational acceleration of the center of mass of the cylinder
and find the string force S.
(2) Determine the acceleration of the center of mass by using the work equa-
tion in the lab frame.
Answers .
. (1) The center-of-mass theorem: MdvcM/dt = L:Fext, shows that CM will
move along a vertical line, because the external forces have no component
in the horizontal direction. We find that
11.9 Examples 261
Mx = Mg - s. (11.13)
Note: S is not acting in the eM. This does not matter. The eM moves
as if all the external forces were acting in the eM. We apply the angular
momentum theorem about the symmetry axis through eM:
Mr2 ..
-2-0 = Sr. (11.14)
Since the string is taut throughout the motion there is the following
geometrical constraint:
x = rO + constant,
or
.. dw
x = rO = rdt- (11.15)
From (11.13)-(11.15) we get:

.. 2 1
x = '3 g , - -Mg
S- 3 .
(2) Note: the string force acts on a particle which is instantaneously at rest in
the lab frame. Loosely speaking, the particle participates in two motions,
a downward translation with velocity i = VCM and a rotation which
directs the particle upwards. The velocity in the rotation is Vrot = ril =
rilr = i. The total instantaneous velocity of the particle at A is thus
zero, seen from the lab frame (the acceleration of the particle at A is not
zero, or the particle would not move at all). The motion of the cylinder
can thus be seen as a pure rotation about the axis A. Such an axis is
called the instantanous axis of rotation. Since the string force acts on a
particle instantanously at rest, the string force performs no work. The
only force that does perform work is the force of gravity. We find
M gx. =
d {I
dt '2 M x
.2 + '12 . '12 M r 20'2} ,
or, since i = ril, Mgi = ~Mix. Thus, x = ~g
Remarks. The angular momentum theorem can be applied about two more
axes (points):
(1) about A (in instantaneous rest):
(Mr2 + ~Mr2) e Mgr,
(M r 2+-1Mr 2) -dw Mgr.

2 dt
[Note: the rotation vector w is the same.]
(2) about the point of suspension, 0 (at rest in the inertial frame).
1 2
Lo MVCMr+ 2Mr w
Mxr + ~Mr2w,
:t (MXr + ~Mr2w) = Mgr.

The two equations expressing the angular momentum theorem with ref-
erence to 0 and A are the same [£ = rB = r(dw/dt)J. When £ has been
determined from one of these equations, the string force 8 may be deter-
mined from the center-of-mass theorem.
Questions. Let us assume that the point 0 is not fixed. A person is holding
the string (a yo-yo!). The string is now brought vertically upwards in such
a way that the center of mass (eM) for the cylinder is not falling while the
string is unwinding.
Determine the string force and determine the work that has been per-
formed on the cylinder when it has reached the angular velocity w.
Answer. 8 = Mg; W = Mr 2w 2/4. f:::,
Example 11.3. The Atwood Machine
~g
We shall again consider the Atwood machine (see Example 2.6) , including
now the moment of inertia f of the disk. The string force is not the same on
either side of the disk. These string forces are internal forces for the system
as a whole, but external forces for the disk. The string forces provide the
torque necessary to give the disk an angular acceleration.
We have three bodies, and we write the equation of motion for each one:
(M +m)£ (M+m)g-8, (11.16)

M£ T-Mg, (11.17)
fa (8 - T) r, (11.18)
11.9 Examples 263
where a is the angular acceleration of the disk. In addition we have the

geometric constraint that
x = rO = ra. (11.19)
Note: among the forces acting on the disk only the string forces have a torque
about the axis O. The weight of the disk and and the reaction force from 0
have no torque about O.
From (11.16)-(11.19) we find
m
2M + m + I/r 2 9 ,
S (M +m)g (1- 2M +:::+I/r 2 )
T Mg (1 + 2M +:::+I/r 2 ) .
For I = 0 we recover S = T.
A different way of solving the same problem is to use the angular momen-
tum theorem about the axis O. The only external forces that have a torque
about 0 are the forces of gravity on the masses M and m + M.
The magnitude of the angular momentum about the axis 0 is
L = (M + m) ir + M ir + I ~
r
The angular momentum theorem now gives that
! (( M + m) ir + M ir + I~) = (M + m) gr - M gr .
From this equation x may be determined.

There is a third way to determine x: The rate of change of kinetic energy
for the system equals the power from the external forces (the strings remain
taut throughout the motion, so no energy is stored in the strings!). Therefore
From this equation x may be determined.

Question. Determine the reaction force R from the axis 0 when the system
is in motion. The mass of the disk is (of course) 2I/r2. [Answer: R = S +
T + 2Ig/r2.] 6
Example 11.4. The Physical Pendulum. A rigid body, which is able to

oscillate about a horizontal axis 0 that does not pass through the center of
mass, is called a physical pendulum (see the figure). Let eM be the center of
mass. When the line segment OCM forms an angle 8 with the vertical, the
force of gravity has a torque about the axis O. The moment of inertia about
o is denoted 1. The distance from 0 to CM is R.
Show that for small oscillations (sin(8) ~ 8) the physical pendulum will
exercise harmonic oscillations. Determine the period of oscillation.
Solution.
d 28
1 dt 2 = -MgRsin(8) ~ -MgR8.
The solutions to this equation are harmonic oscillations with a period of
T = 27fJ M:R'
For R -+ 0, T -+ 00. How can this be interpreted?
Often the so-called arm of inertia (k) is introduced. For a body of mass
M, able to rotate about a given axis 0, k is defined by 1 == Mk 2 . If this is
inserted into the expression for the period we get
[k2
T = 27fy gR .
Note. The period is independent of the mass. This is due to the fact that the
gravitational and inertial mass are equal (proportional).
Question. Consider larger amplitudes of the pendulum. Show that the angular
velocity B== w is given by
. 2MgR
82 = - 1 - [cos(8) - cos(80 )] ,
where 80 is the maximum angle of deflection.
Example 11.5. The Rod. A rod of mass M and length l can turn in a
horizontal plane and without friction around a vertical axis 0 through the
11.9 Examples 265
m _- -i.~v
... _________ ~_. If
center of mass. Initially the rod is at rest, as shown in the figure. A bullet
of mass m is now shot into one end of the rod and remains embedded in the
rod. The velocity of the bullet just before it hits the rod is v (see the figure).
Find the angular velocity w of the rod after the bullet is stopped.
Solution. The only external forces acting on the system are gravity and the
reaction force from the axle O. None of these external forces have any torque
around 0: L: Next = O. Thus, the angular momentum of the system around
o is conserved. The moment of inertia of the rod around 0 is denoted 1.
L before the collision = mv(l/2) (L is the angular momentum around 0).
L after the collision = Iw + m(l/2)2 w.
Conservation of angular momentum around 0 means that
mv(l/2) = Iw + m(l/2)2 w . (11.20)
Furthermore
(11.21)
(11.20) and (11.21) now give

6mv
w= .
Ml +3ml
Questions. A homogeneous rod of length 2l and mass M can turn without

friction in a horizontal plane and about a fixed vertical axis 0 through the
center of mass of the rod. Initially, the rod is at rest. A particle of mass m
collides in a short, completely elastic collision with one end of the rod. It is
assumed that the collision time Llt is so short that the rod does not turn
during the collision. The velocity of the particle (both before and after the
collision) is horizontal, and perpendicular to the original direction of the rod.
The magnitude of the velocity of m before the collision is u.
(1) Find the velocity of m and the angular velocity w of the rod after the
collision.
(2) It is assumed that the force F by which the particle acts on the rod is
constant during the collision. Find F. Numerical example: l = 0.30 m,
M = 0.3 kg, m = 0.04 kg, u = 20 ms- 1 , Llt = 10- 2 s.
Answers. Use conservation of angular momentum about O and - because the

collision is elastic - conservation of kinetic energy.
(1)
3m-M
v
3m+M u ,
6mu
w
(M + 3m)l .
Note: if M > 3m, the rod is so massive that it reftects m. For M = 3m
the mass m is stopped by the rod.
(2)
F = 2mMu
.6.t(M + 3m)
Ivi 8.6ms- 1 ,
w 38s- 1 ,
F 1.14 X 10 2 N.
11.10 Review: Comparison Between Linear Motion

andRotation About a Fixed Axis
The mathematical description of tlie rotation of a rigid body about a

fixed axis is equivalent to the mathematical description of the mot ion of
a particle alorig a straight line.
By a fixed axis we shall mean an axis that. is kept fixed relative to the
inertial frame used as refereuce, and fixed relative to the rotating body.
AII methods and results can, as shown in the table below, be taken over
directly from linear motion of a particle to rotation of a rigid body about
a fixed axis.
The angular momentum theorem may be applied to the following types

ofaxes:
(a) any fixed axis;

(b) any axis through the center of mass (CM) even if the axis is accel-
erated (homogeneous gravitational fie.lds, fictitious or not, have no
torquea15ouCCM);" <,
any axis that has a veloeity v Q parallel to the velocity of the center
of ma~s, v OM.
11.11 Problems 267
angular position:()
ang;ularvelocity: W =0
angular accel~ration:.a = W ==0
torque: N ... .
moment of inertia: 1
angular momentuin: L = 10 =:= Iw
Equation of mot ion;'
dL ==N
tit
We. furtherm()re want to call attentioJ;l to the following facts.

(1) Fora systein ~. rigid or nonrigip ~ on 'which uO e:xternal torques are
ading, angular momentum is conserved.
(2) Parallel axis theorem: the moment of inertia of abody of mass M
about an axis through an arbitrary point O is. equal to the sum of
the .moment of inertia about the parallel axis through the OM arid
MiJ,2, where d is the distante. between. CM and the axis through O.
11.11 Problems
Problem 11.1. A homogeneous disk of radius r == 0.20 m can oscillate as a

physical pendulum araund a horizontal axis O located 0.10 m (r /2) from the
center of mass of the disk. The disk is perpendicular to O. Find the period
for small oscillations of the disk.
Problem 11.2.
A heavy circular disk with radius R and mass M is fastened to a light stiff
rod. The mass of the rod is negligible compared to the mass of the disk. The
system can oscillate as a physical pendulum about a fixed horizontal axis A.
The distance from A to the center of the disk is L (see the figure).
(1) Determine the period for small oscillations when the disk is fastened to
the rod as shown, i.e., when the disk swings in the plane of the paper.
(2) Assume now that the plane of the disk is perpendicular to the plane of
the paper. Determine the period for small oscillations.
Problem 11.3.
Two weights, each of mass M are fastened to the ends of a rod of length 2l
and negligible mass. The dimensions of the weights are much smaller than
11.11 Problems 269
t. The rod can rotate without friction in the horizontal plane about a fixed
vertical axis 0 through the center of the rod.
A beetle of mass m sits at rest on one of the weights. The rod is set in
rotation about 0 with angular velocity WI. The system is then left to itself.
(1) Find the kinetic energy Tiof the system.
The beetle now crawls along the rod towards O. When it reaches 0, it stops.
(2) Find the angular velocity W2 and the kinetic energy T2 of the rod, when
the beetle has reached O.
The beetle then slides out along the rod until it hits one of the endpoints of
the rod, where it then remains without motion relative to the rod .
(3) Find the angular velocity W3 and the kinetic energy T3 of the system after
the beetle has come to rest.
(4) Let us assume that we can neglect the friction between the beetle and
the rod. Find the speed v of the beetle relative to the rod as the beetle
reached the endpoint of the rod.
Problem 11.4.
t
L
A thin homogeneous rod has length L and mass M. Initially the rod is vertical
and at rest in the laboratory. One end of the rod is in contact with the floor,
which is assumed to be frictionless. The acceleration of gravity is g.
The rod is now given a slight push and begins to fall in the gravitational
field . The initial value of the velocity of the CM is zero.
(1) Describe qualitatively the curve along which the CM moves.
Consider the rod at the moment when the CM of the rod is a distance y below
the initial position. The acute angle which the rod forms with the vertical is
denoted by 8.
(2) Write an expression for the kinetic energy T and the potential energy U
of the rod, as functions of M, L, g, 8, and e.
(3) Determine iJ as a function of L, g, and 8.
Problem 11.5.
A homogeneous thin rod has length L and mass M. Initially the rod lies
at rest on a horizontal frictionless table in the laboratory.
The rod is then hit with a hammer, at a point A which is at the distance
d from the center point 0 of the rod (see the sketch). The blow is in the
direction perpendicular to the rod. The blow transfers a momentum P to
the rod. It is assumed that the duration of the blow is so short that we may
neglect the motion of the rod during the blow. The known quantities are thus
L, M, d, and P.
(1) Determine the velocity of the center of mass (CM) of the rod after the
blow.
(2) Determine the kinetic energy of the rod after the blow.
We assume that there is a point C on the rod which - just after the blow -
is at rest relative to the table.
(3) Determine the distance OC. Under what conditions is C a point on the
rod?
Problem 11.6.
An Atwood machine consists of a cylinder that may turn around a fixed,

horizontal symmetry axis, O. The radius of the cylinder is r, and the moment
11.11 Problems 271
of inertia of the cylinder around 0 is I. A string passes over the cylinder and
the string holds a mass M at each end.
Initially the two masses are at rest. A small disk, with a hole in the center,
has mass m. This disk may slide without friction along the string. The small
mass m now falls freely in the gravitational field, through a height h. The
initial velocity of the disk is zero. The disk then hits one of the masses M in a
completely inelastic collision. The duration of the collision can be neglected.
The string is taut and the string cannot slide on the cylinder. The accel-
eration of gravity is g.
(1) Determine the velocity u of the total mass M + m just after the collision.
The velocity u should be given as a function of m, M, g, h, I, and r.
(2) Determine the amount Q of the mechanical kinetic energy that is con-
verted into heat during the collision. Q should be given as a function of
m, M, g, h, I, and r.
Problem 11.7.
A plane, homogeneous disk may slide on a frictionless horizontal table. A

light cord is fastened to the disk and wrapped many times around the rim of
the disk. The cord goes without friction through a small eyelet at the edge
of the table and is connected to a mass m (see the figure). The mass of the
disk is M and its radius is R. The system is started from rest and the cord
is assumed to be taut throughout the motion. Ignore the mass of the cord.
(1) Find the velocity v of the center of mass of the disk as a function of time
(before the disk hits the eyelet or the mass m hits the floor).
(2) Find the distance S the disk has moved, and the distance d the mass m
has been lowered when a length l of the cord has been unwound.
Problem 11.8.
L
A
3L
A thin, homogenous rod has length 4L and mass 4M. The rod is mounted
so that it can turn in a vertical plane about an axis A which passes through
a point located a distance L from one end of the rod. The oscillation plane
of the rod is perpendicular to A. See Figure 1. Ignore frictional forces and
air resistance. The acceleration of gravity is g. The laboratory is considered
an inertial frame. Known quantities are M, L, and ~.
1. Show, that the moment of inertia about the horizontal axis A is
2. Determine the period for small amplitude oscillations about the equilib-
rium position.
2L
Q
We now consider a situation where the rod is hit by a particle Q carrying

a horizontal momentum. See Figure 2. The rod hangs in its equilibrium
position before the collision. The particle hits the rod at a distance L from
the lower end of the rod. The particle goes through the rod in a time so short
that we may neglect the motion of the rod during the passage of the particle.
The particle loses an amount of momentum P during the collision. The
vector P is horizontal.
11.11 Problems 273
3. Determine the magnitude P = IPI when the rod is observed to swing

exactly up to a vertical position (the rod turns 180 0 about A).
After the 180 0 turn, the rod is at rest, but in an unstable equilibrium
position. The rod now swings back towards the stable equilibrium, and we
consider the instant where the rod passes through the horizontal position.
4. Determine the horizontal component of the force F with which the axis
A acts on the rod, as the rod passes through the horizontal position.
5. Determine the vertical component of the force F as the rod passes through
the horizontal position.
Problem 11.9.
1
L
A thin, homogeneous rod is mounted in such a way that it can turn in a

vertical plane about a fixed horizontal axis A. The axis A is perpendicular
to the plane of motion and A goes through one end point of the rod. The
rod has mass M and length L. The acceleration of gravity has magnitude g.
The laboratory frame is considered an inertial frame.
We first consider small oscillations about the equilibrium position, and
we wish to compare small oscillations of the rod with small oscillations of a
mathematical pendulum (a mathematical pendulum is a point mass hanging
from the end of an unstretchable, massless string).
1. Determine the string length L1 that a mathematical pendulum should

have, in order that the pendulum has the same oscillation period as the
rod.
We now consider a new situation. See the sketch below.

Initially, the rod hangs at rest. Then, at some distance x from the top
end of the rod (where the axis of revolution is) the rod is given a brief
blow with a hammer. We assume the the duration L1t of the blow is so brief
x
r
L
I
p
that the rod does not turn significantly during the time L1t. Thus, the rod
receives the momentum P from the hammer. The vector P is horizontal and
perpendicular to the axis of revolution. P lies in the vertical plane in which
the rod can move.
2. Determine the velocity V CM of the eM of the rod just after the blow has
been struck. The magnitude of V CM should be expressed as a function
of P, x, M, and L .
3. Determine x such that the horizontal component of the reaction force
from the axis A vanishes during the blow.
4. Assume that x has the value to be found in question 3. After the blow
the rod turns a maximum angle eo
from the vertical. Determine cos eo.
Problem 11.10.
o A
A thin, homogeneous, circular ring of mass M and radius R lies on a smooth,

horizontal table in the laboratory. The ring may rotate without friction about
a vertical axis through a point A on the ring. The axis is fixed with respect
to the table. See the sketch. A beetle of mass of mass m is crawling along
the ring with constant speed va with respect to the ring. The beetle begins
at the point A, and the ring is initially at rest. When the beetle crawls along
the ring, the ring rotates around the axis through the point A. Determine
the angular velocity w of the ring in its rotation about the axis, when the
beetle has crawled half way around the ring, i.e. when the beetle has reached
11.11 Problems 275
the point on the ring diametrically opposed to A. The angular velocity w

is regarded with respect to the laboratory, which is considered an inertial
frame. Express w in terms of m, M, Vo, and R .
Problem 11.11.
MF=======~======~M
A B
~~----2R
A mechanical system consists of two like bodies, A and B, each of mass M.

The bodies are connected with a rigid rod of length 2R. The extension of
each body is much smaller than R, and they can be taken in calculations as
point masses (particles). We disregard the mass of the rod.
The system may turn in a horizontal plane, about an axis O through the
centre point of the rod . The rotation about O is without friction , and we
disregard air resistance.
Initially, the system is at rest in the lab frame, which is taken to be an
inertial frame. The body A is now struck by a pistol bullet of mass m. The
collision is completly inelastic, and the pistol bullet comes to rest inside the
body A. Prior to the collision the bullet had the velocity v. The velocity v
is horizontal and perpendicular to the rod connecting A and B, and perpen-
dicular to the axis O. The bullet comes to rest insi de the body A in a time
Llt so short that the rod does not turn noticably during the collision.
Known quantities are thus: M, R, m, v == Ivi, and Llt .
1. Determine the angular velocity w in the rotation of the system after the
collision.
2. Determine the amount of heat generated by the collision.
During the brief collision, the rod acts on the axis O with a for ce F, which
is assumed to be constant.
3. Determine the magnitude and direction of the horizontal component of F.

Problem 11.12.
s 1
3R
A heavy, homogenous ball has mass M and radius R. The ball is attached to
a rigid rod S. One end of the rod S is mounted so that it can turn without
friction around a fixed horizontal axis A. The center of the ball can thus
move in a vertical plane perpendicular to the axis A, containing the rod S.
The distance from the axis A to the center of the ball is 3R.
We disregard the mass of S. The magnitude of the acceleration of gravity
is g. Known quantities are then R, g, and M.
1. The system can perform small oscillations around the stable equilibrium
position. Determine the period of these oscillations.
We now imagine the system displaced so that the rod S is horizontal.
---3R
11.11 Problems 277
Initially the system is supported while the rod is horizontal and at rest.
Then the support is removed, and the rod begins to turn without friction
about the axis A .
2. Determine the angular velocity w as a function of the angle e that the

system has turned about the axis A. See the above sketch.
3. Determine the angular acceleration a as a function of the angle e that
the system has turned about the axis A.
We now seek the force F with which he axis A acts on the system, when
it has turned the angle e.
4. Determine the magnitude and direction of the component Fl of F which

is parallel to the rod S.
5. Determine the magnitude and direction of the component F2 of F which
is perpendicular to the rod S.
12. The Laws of Motion
This chapter contains nothing new. Through a series of examples we wish

to illustrate the application of the basic theorems of mechanics. We start by
reviewing classical mechanics in the form of ten "laws" or theorems.

280 12. The Laws of Motion
12.2 Remarks on the Three Conservation Theorems

The formulation of the three conservation theorems (energy, momentum, and
angular momentum) is one of the main results of our development of classical
mechanics. We have seen how the conservation theorems follow from the basic
laws of Newtonian mechanics, in particular from a combination of Newton's
second law and Newton's third law (the law of action and reaction).
In the 20th century the world witnessed a revolut ion in the sciences, a rev-
revolution
olution which is comparable to that brought about by Brahe, Kepler, Galilei,
and Newton. The revolution has resulted in the recognition that Newtonian
mechanics has a limited range of applicability. The Newtonian laws are now
seen as an approximation, valid for macroscopic bodies, and for velocities
much less than the speed of light. Deeper and more precise laws are formu-
lated in the theories of quantum mechanics and relativity.
The inner dynamics of atoms, i.e., the motion of electrons around the
nucleus, and the motion of nucleons within the nucleus are governed by the
laws of quantum mechanics and special relativity.
A perhaps surprising point is the following: the three fundamental conser-
vation theorems, i.e., conservation of energy, conservation of momentum, and
conservation of angular momentum are valid also in the world of quantum
mechanics and relativity.
In these, the most fundamental theories of the structure of matter known
to us, the three conservation laws still hold.
The conservation theorems reveal some of the most essential aspects of
how nature works.
12.3 Examples
Example 12.1. Conservat ion of Angular Momentum. The Collision

Conservation
Approximation
CH
Cf) l"19
A
d
f A
i~ m
.--. v ._______ __
._______ __
eM
CM
s
S
The physical content of this example is closely related to the physical content
of Example 9.3.
A homogeneous sphere S, with mass M and radius R hangs at rest in
the gravitational field of the Earth (see the figure). The sphere S may turn
12.3 Examples 281
without friction about the horizontal axis A which is a tangent to the top
of the sphere and perpendicular to the plane of the paper. A small pistol
bullet with mass m is now shot into the sphere. The velocity of the bullet
is horizontal and perpendicular to the vertical plane that contains the axis
A and the center of mass (CM) of the sphere S. The pistol bullet comes to
rest inside the sphere S at a point which is directly underneath the point A,
a distance d from A.
We assume that m « M and that the time ..1t L1t in which the bullet comes
to rest relative to S is so small that we can disregard the angle through which
S turns in the time interval ..1t
L1t (the collision approximation).
(1) Determine the angular velocity Wo with which the sphere S begins to turn
about the axis A, i.e., determine Wo just after the collision is over.
(2) Determine the maximum value Bo of the angle B that the sphere S turns
after the collision.
..1t must satisfy in order that we may disregard
(3) Determine the condition L1t
the angle through which the sphere S turns during the collision.
(4) Determine the conditions for which the linear momentum P is conserved
in the collision.
Solution. We start with some general considerations.

If the sphere S had not been fastened to the axis A we should have been
able to compute
corn pute the velocity of CM after the collision. We could have used
the law of conservation of the momentum for a closed system. In the present
case, however, the sphere S is connected to the fixed axis A, and reaction
forces from the axis will in general act during the collision. Thus: in general
the momentum will not be conserved in the collision. We do not know the
reaction forces, but we do know the geometrical constraints on S, and that
the reaction forces determine these constraints.
By applying the theorem of conservation of angular momentum about the
axis A we circumvent the problem of the unknown reaction forces acting at
A. The reaction forces have no torques about the axis A.
..1t was not sufficiently short, the sphere S would turn
If the collision time L1t
a finite angle before the collision was over. When the CM is no longer verti-
cally underneath A, gravity will produce a torque about A, and the angular
momentum of the system is no longer conserved. Do not use conservation of
mechanical energy when heat is produced.
(1) The angular velocity Wo is determined from conservation of angular mo-
mentum during the collision:
mvd (h
(lA+ md2 )wo
)wo,,
h ICM + MR2 = ~MR2 + MR

MR 2 = = ~MR2
MR22 =
5 5
mvd
Wo (12.1)
(7 /5)M R2 + md 2
(2) When the collision is over there are no more nonconservative forces in
the system (we disregard friction in the pivot at A). We can then use the
theorem of conservation of mechanical energy. The kinetic energy with
which the sphere S begins to turn is:
T ="2 1~ (7
"5 M R +
= (iMR2 2
md 2) wo·
+md w5. 2) 2
We now assume that S, as a result of the collision, turns an angle 8()o0 away
from the equilibrium position. The center of mass of S is then raised an
amount
hhI1 = cos(80 )].
R[l - cos(()o)].
= R[1
Assuming the bullet to be placed as indicated on the figure, the bullet is

lifted an amount
h = d[1
h22 = d[l - cos(()o)].
cos(80 )].
cos(80 ):
We therefore have the following equation to determine cos(()o):
~ (iMR2
(~MR2 + md2) w5 = MgR [1 -
w5 = cos(()o)]
cos(80 )] + mgd [1 - cos(()o)]
cos(8 0 )] .. (12.2)
(mvd)2
(mvd)2
(8 ) -= 11 _
ccos (()) - --:---=--'-:--;-;i-----c--::-----:::T
os o - 2g(MR+md) (tMR2 (tMR2+md +md2) 2)
Check the dimensions!
(3) The condition for Llt to be ignorable is that the amount of angular mo-
mentum imparted to the sphere S by the bullet in the collision (Le., (i.e.,
mvd) is large compared to the angular momentum the sphere received
from the torque of gravity during Llt. In the estimates below we shall use
m«
that m «: M.
The magnitude of the angular momentum about A which the torque of
gravity supplies to the sphere S during the collision is:
L = llLlt
Llt
Ndt
Ndt = llLlt
Llt
MgRsin[8(t)]dt.
MgRsin[()(t)]dt.
We do not know in detail the forces acting between the bullet and the
sphere S during the collision. We are thus unable to calculate L exactly.
But we can estimate the order of magnitude of L. The angle through
which S turns during the collision has the order of magnitude,
~ woLlt,
()8 ::::::
because the angular velocity is of order Wo during the collision. Since 8()
is a small angle, we let sin( ()) ::::::
sin(8) R:; ().8. The torque of gravity is therefore of
magnitude
12.3 Examples 283
N ~ MgRwo.t1t.
L1t.
L1t. The order of magnitude of L is
This torque acts in the time interval .t1t.
therefore
L = MgRw
M gRwo(.t1t)2
(L1t)2..
The magnitude of L is to be compared to the angular momentum im-
parted to the sphere S by the bullet:
mvd~ hwo.
The condition for conservation of the angular momentum about A during
the collision is then:
J~;R .
or
L1t«
.t1t« =
= J~;
The period T of oscillation for the small oscillations which the sphere S
can perform under the influence
infl.uence of gravity is given by
=
T = 27ry0.
2rrV~
MgR 2rrvfiR
MiR == 27ry 59
59·.
L1t « T.
The condition may thus be expressed simply as .t1t
L1P in the horizontallinear

(4) We calculate the change .t1P horizontal linear momentum during
the collision, still assuming that m « M. The velocity of the center of
mass just after the collision is known:
VCM =
= Rw
Rwoo·
The horizontal component of the linear momentum just after the collision
is then
MRmvd
Pafter = MRwo
Pafter = MRw o ~ IA
fA
The linear momentum before the collision is
Ibefore =
Fbefore = mv.
Therefore the change in linear momentum in the collision is
L1P == Pafter
.t1P Pafter -- Ibefore =
Fbefore = mv [~~d - 1] .
We may conclude that in general the linear momentum is not conserved
---4 o. The finite change in the linear
L1t -+
during the collision, even if .t1t
momentum is produced by the external force on the system: the reaction
force from the axis A. The average value of the horizontal reaction force
from A during the collision is
From the above expression for LlP we see that LlP = O fA = 1

0 if M Rd/ lA
or,
d= lA fA = ~R
=~R
MR 5 '
that is, linear momentum is conserved when d == (7/5)R. When the pistol
= (7/5)R from the axis A, the
bullet hits the sphere S at a distance d =
axis is affected with a minimal amount.
Example 12.2. Rotating Rod
:1
I
: L
I
I
;;;;;;;;));;;;;;;;;;;;;;;;;;;;))))/J)J)/
;;;;7;;;;;7;77;7;771;;;;;;;;;;;;77;77;7~
An elevator is moving vertically upwards with the constant acceleration a in

the gravitational field g of the Earth. A horizontal axis A is fastened to the
bottom of the elevator. A homogeneous rod with mass M and length L can
turn without friction about A. Initially the rod is at rest and vertical. At a
certain time the rod begins to turn in the vertical plane perpendicular to the
axis A. The rod begins its motion with (near) zero angular velocity.
(1) Determine the angular velocity w and angular acceleration Qa as functions
wand
of the angle () formed with the vertical.
(2) Determine the force F, with which the axis A acts on the rod when the
rod forms the angle () with the vertical. The force F may be decomposed
into suitable components.
12.3 Examples 285
Solution.
(1) The magnitude of the gravitational field in the elevator is 9g + a. From
,
.................................
.................................t
}.h =
.................................. 1..
M(g + a) T
= L/2 (1 - cos(9))
costS))
conservat ion of energy theorem, we find w as a function of O:

the conservation B:
1(1) "2L [1[1-- cos(O)]

"2 3ML2 w 2 == M(g + a) 2" cos(B)]
3(g + a)[l - cos(O)]

cos (B)]
~
=} w=
w =
L
The angular acceleration can be determined from the angular momentum
theorem:
L.
M(g + a)~
a) "2 sin(O),
sm(B) ,
3g+a .
aQ =
= "2 -y- sm(
"2---Y- O) .
sm(B).
A
A different way of determining aQ is by differentiation:
dw
Q = dt .
a=
The angular velocity w is known. We find, by differentiating the expres-

sion for w 2 :
2 dw _ 3(g + a) . (O)(B) dO
dB
2w
w dt - L sm dt'
dB/dt == w,
or, since dBjdt
dw 3(g + a) sin( sin(B) O)
-d-t
dt == aQ == ---"''-----2-=-L--'..-.:..
2L ..
(2) The reaction force from the axis A gives, together with the effective grav-
ity M (g + a), the acceleration of the center of mass of the rod. The center
ofthe
of mass does not perform a uniform circular motionj
motion; the acceleration has
two components. A radial accelerat ion directed along the rod towards A,
acceleration
and a tangential acceleration perpendicular to the rod.
We decompose F into two components: F = FI Fl +F

+F22 where FI
Fl is along the
rod and FF22 is perpendicular to the rod. The component FI
Fl is calculated
as follows:
M(g + a) cos(B)
cos(O) - FI
Fl ,
5cos(O) -- 33
M( + a)) 5cos(B)
M(g
g+a 2
2
0 = 01)
(check for B O!)
The orthogonal component F
F22 is calculated as follows:
L
M-a M(g + a) sin(B) F2 ,,
sin(O) - F2
2
1 .
F2 4M(g + a) sm(B)
"4M(g sm(O)
F22 in the limit 0B == O

(check the magnitude of F 0 1).
!).
Example 12.3. Man on Disk
i
,
E J
A large horizontal circular and homogeneous disk with radius R and mass
0 through the center of
M can rotate without friction about a vertical axis O
12.3 Examples 287
mass of the disk. The disk is at rest and a man with mass m stands near the
edge of the disk.
The man now walks around the periphery of the disk until he returns to
his starting point on the disk. There he stops. Find the angle 6Bo0 which the
disk has turned relative to the ground, when the man stops.
Solution. The angular momentum about O
0 is conserved since there are no
line fixed relative

ta
to the disk
t = O
0 0
t > O
extern al torques about O. Internal forces cannot produce a total angular

external
momentum. As the man walks along the circumference of the disk he exerts
a force on the disk equal to, but oppositely directed to the force the disk
exerts on the man. The total angular momentum about O 0 therefore remains
zero:
L == R2
m
d¢l _ I dB
R 2 d4;
dt
=0
d6 =
dt '
°
=°
where I =
= (1/2)M RR22 is the moment of inertia of the disk about the axis O.
= 4;¢I = 0 at t == 0, gives:
Integration, using the initial values 6B =
2m4; = MB.
2m¢l=M6.
When the man has finished his walk along the periphery of the disk
4;0
¢lo + Bo
60 = 211"
21l"..
From this we find

6 _ 211"
21l"
Bo0= + M/2m .
- 1 +M/2m
Example 12.4. The Sprinkler. A garden sprinkler consists mainly of a

straight tube that can turn about a fixed vertical axis O. Through an opening
at O
0 in the middle of the tube, water is fed into the tube, exiting at the ends in
a direction perpendicular to the tube (and perpendicular to the axis through
0).
O).
Vo
Va t
11,-----:-----_
1,---1-~_
'1t~v,lvo
o
2L
We assume that the mass per unit time, J-L, jl, which passes through a cross
section of the tube, is independent of the motion of the tube. J-L jl is thus
constant in time. We furthermore assume that the water, at the instant it
velocity 'vo relative to the tube.
leaves the tube, has a constant velocity'va
In the motion about the axis O 0 the tube is subject to frictional forces. We
assume that these forces can be described by a torque N which is proportional
to the angular velocity of the tube. The constant of proportionality is denoted
total length of the tube is 2L and the moment of inertia of the tube
,. The totallength
0, when the tube is filled with water, is 1.
with respect to the axis O, I. We assume
the tube starts its motion at t = O.o.
Determine the angular velocity w(t) of the tube in its mot motion
ion about the
axis O.
Solution. When the angular velocity of the tube is w, the magnitude of the
Vo - Lw. The differential equation
velocity of the water relative to the earth is Va
governing the rotation of the tube about the axis O 0 is:
II~~
dw
dt
= 2jl (va - Lw) L -
2J-L(vo-Lw)L-,w, ,W , (12.3)
dw _ 2jlL 2 + , (w _ 2jl Lva ) (12.4)
dt 1 2jlL2 +,
Note that the angular acceleration vanishes when the tube has reached the
wI,
angular velocity w f' given by
2jlLva
2J-L Lvo
2J-LL2 +,
wf == 2jlL2
WI .
By integration of dw / dt we find:
wIf [1 - exp ( - 2J-LL

w == w (t) == w +,)]
2jlL21I2 +, t )]
Questions.
(1) Determine the final angular velocity wf WI for the case, == o. What is the
physical interpretation of this result?
12.3 Examples 289
(2) Assume that the water is suddenly turned off at a time after wf has been
reached. How long would it take for the tube to come to rest, for finite
values of , ? The mass of the empty tube is set to m, and we assume
of"(
that the tube is empty a very short time after the water has been turned
off.
Answers. (1) W = vo/L,
= vo/ L, (2) W == wfexp
wf exp (-3,t/mL
(-3"(t/mL 2 ).
Example 12.5. Rolling. When a body of circular cross section is rolling,

it is rotating about an axis while at the same time in translational motion.
Think about a ball
baU or a cylinder. Both in rolling and in sliding the body is
in contact with the supporting surface.
m
= 2 R ro
=
v =
= R m
ro
L
~_---lf-+:"CM
~_--If-+'CM
Pure rolling, i.e., rolling without sliding, is characterized by the fact that
part of the body in contact with the supporting surface is instantaneously
the pari
at rest relative to the surface. In pure rolling we have the foUowing
following reiat
relation
ion
between the translational velocity of the center of mass VCM, the radius R of
the circular cross section, and the angular velocity W in the rotation of the
body:
VCM = Rw. (12.5)
Loosely speaking, the point or line of contact participates in two motions,

one forwards and one backwards, both with speed VCM. Let us use an inertial
frame 1I at rest relative to the supporting surface. The parts of the rolling
body in contact with the surface thus has velocity zero relative to 1I (pure
rotation).
The motion of the body can be considered in two different ways:
(1) a translation of the center of mass, CM, plus a rotation about an axis
through CM; or
(2) a pure rotation about the instantaneous axis of rotation, which passes
through the point C of contact between the body and the surface.
Note. In a rotation about a fixed axis C, the axis C is characterized by the

fact that the velocity of each point of C is zero (relative to the chosen inertial
frame). The angular velocity w is independent of the axis chosen to describe

the motion.
The kinetic energy of a rolling body of revolution can be written in dif-
ferent forms. If we choose to describe the motion as in point (1) above, we
have:
(12.6)
If we choose to describe the rolling as in (2), Le.

i.e. as pure rotation about the
"contact axis" C, we find
1 2
T= "2Icw , (12.7)
where w has the same value in (12.6) and (12.7). From the parallel axis
theorem we have
Ie = M
le R2 +
MR2 +IICM
cM .
From this it can be seen that the right hand sides of (12.6) and (12.7) have
the same value.
Let us sketch a concrete case.
The figure shows a homogeneous symmetrical body with a circular cross

section (cylinder, sphere, cylindrical shell, spherical shell, etc.). The moment
of inertia about the axis of symmetry and perpendicular to the plane of the
paper is I. Let us denote this axis by A. The mass of the body is M and the
radius of the circular cross section is R.
If the body is placed at rest on an inclined plane, as shown in the figure,
and if there is no jriction
friction between the body and the inclined plane, the body
will slide down the inclined plane. For, in case of a frictionless inclined plane
there is no torque about the axis A through CM eM and consequently the body
will not begin a rotation.
Let us now assume that there is friction between the supporting surface
(the inclined plane) and the body. The force of friction is f (see the figure).
Other forces acting on the body are: the normal reaction N from the surface,
and the force of gravity, MMg.
g.
12.3 Examples 291
We write the equations of motion for the body.
center-oj-mass theorem. The accelerat

The center-of-mass eM along the inclined plane
ion of CM
acceleration
VCM = a. Thus
is VeM
Ma = J.
= Mgsin(B) - f. (12.8)
eM accelerates as if aU
Note: the CM eM.
all the external forces were acting in CM.
The angular momentum theorem. The angular acceleration about the axis A
is w = a. We have
Ia = JR.
fR. (12.9)
The axis A is accelerated, but the angular momentum theorem can be applied
about an accelerated axis without adding fictitious forces, when the axis
passes through the center of mass.
We stated that if fJ = O 0 the body will (if started from rest) slide down
the inclined plane, without rolling. This is a consequence of (12.9) which, for
fJ = O,
0, gives a = O.
Loosely speaking, the frictional forces acting from a surface on the body
e) are in such
at the point of contact (or in general line of contact, the axis C)
a direction that they "oppose sliding". Just as we speak of the ideal case of a
perfectly smooth (frictionless) surface, we speak in theoretical mechanics of
the ideal case of a "perfectly rough" surface. A perfectly rough surface is a
surface over which a body cannot slide, only roU.
roll.
Let us first assume that the inclined plane is perfectly rough so that the
body performs pure rolling down the inclined plane. For pure rolling we have
(differentiate (12.5) with respect to time) the foUowing
following relation between a
and a:
a=Ra. (12.10)
From (12.8)-(12.10)
. 1
a = gsm(B) 1 + I/MR2 (12.11
(12.11))
so-called arm of
We introduce the so-caUed oj inertia, k, which is defined by:
I Mk2,
Mk 2 , (12.12)
. 1
a sm( B) 1 + k2 / R2
gsm(B)
9 (12.13)
If a body slides down a frictionless inclined plane, the accelerat ion of the
acceleration
body will be gsin(B).
(eM)
Equation (12.13) shows that the acceleration of the center of mass (CM)
of a body that roUs
rolls down an inclined plane is less than the accelerat ion of
acceleration
the same body when it slides down a frictionless inclined plane with the same
slope.
slope.
Furthermore, the mass of the body does not appear in (12.13). The ac-
celeration is determined completely by the geometry of the particular body,
through the arm of inertia, k. AII
All bodies falI,
fall, regardless of their overalI
overall mass,
with the same acceleration.
Examples.
Hollow cylinder with thin walIs:
(1) HolIow walls:
1,
1
a 2gsin(B) =
= O.5gsin(B).
(2) Homogeneous cylinder:
I
1 ~MR2 ==
'= Mk 2 ,
2
1
2'
2
a ~gSin(B) ~~ O.667gsin(B).
3gsin(B)
(3) Homogeneous sphere:

12.3 Examples 293
~MR2
5 Mk22 ,
-== Mk
2
5'
a ~gSin(O) ~ 0.714 gsin(O).
Note. The radius R does not appear in these expressions. Consequently, alI all
hollow cylinders, regardless of their radius, rolI
roll with the same acceleration
down a rough inclined plane. Any sphere will rolI roll faster than any homoge-
neous cylinder, which in turn will roll
rolI faster than any holIow
hollow cylinder.
Consider the case where the surface is not completely rough, and ask the
question: what condition must the coefficient of friction J.J1L satisfy in order
that the body rolls without sliding down the inclined plane?
From (12.8)-(12.10), which describe pure rolling, we find
k2
1= k 2 + R2
f = Mgsin(O) k2 . (12.14)
This result means that il if a body is to roll without sliding, If must have the
value given by (12.14). The expression (12.14) shows that fis f is always directed
up along the inclined plane, independent of whether the body rolls upwards
or downwards. The friction force is a reaction force which tries to hold the
contact point at rest, Le., tries to prevent the body from sliding.
The question now is: can a given surface provide the friction necessary to
keep the body rolling without sliding?
The answer is: the maximal friction force that a surface, with coefficient
J.L can provide is
of friction J1
fmax
Imax = J1N
J.LN = J1Mgcos(O).
J.LMgcos(O). (12.15)
The condition for a body to roll without sliding is thus
k2
Mgsin(O)
M k 2 + R2 :S
9 sin(O) k2 ::; J.LMgcos(O)
J1M 9 cos(O) , (12.16)
(12.17)
Equation (12.17) expresses the fact that the surface must be so rough that
it can provide the friction force necessary for rolling without sliding. When
o~
O --; 90° a very large coefficient of friction is necessary for pure rolling to be
maintained.
If the condition given by (12.17) is not satisfied, the motion down the
inclined plane will be both rolling and sliding.
Consider again pure rolling. When a body rolls without sliding down an
inclined plane, the friction force does no work. The friction force acts on a
body particle which is instantaneously at rest. The power done by the friction
force therefore is zero.
Let us use the energy theorem to describe the motion of a body that
i.e., the work per
rolls without sliding down an inclined plane. The power, Le.,
time done by the normal force, is zero. (Why?) The gravitational field is
conservative, so
rot + Ttrans + U =
Trot
T = constant.
We wish to determine the translational velocity of the body when it has rolled
a distance d down the inclined plane. We assume that the body starts from
a state of rest.
The initial position is denoted (1) and the final position (2). We set the
potential energy equal to O a in the final position. Using the above equation,
we have
T rot (1)
Trot (l) + Ttrans(1) + U(1)
U(l) = T rot (2) + Ttrans(2) + U(2) ,
Trot
a + Oa + Mgdsin(O) =
O ~Iw2 + ~MV6M
~MV~M + O. (12.18)
Since VCM = Rw, we find that
V6M = 2gdsin(O) 1 + I~MR2 . (12.19)
Alternatively one can determine the velocity of the center of mass when the
body has rolled part way down the inclined plane by applying (12.13).
Rolling friction. Apart from what could be called sliding friction, a body
VCM
rolling on a level surface will experience rolling friction. We shall not deal
with the phenomenon in any detail, but merely suggest the underlying causes.
Consider a cylinder rolling without sliding on a horizontal surface. Because
of the elastic deformation of the body and the supporting surface, there is
a finite size surface of contact between the two bodies. When the cylinder
rolls, there will be an asymmetry in the contact forces (due to the elastic lag
distribution
forces). This asymmetry in the distri bution of the forces from the surface will
produce a slight torque in a direction to counteract the rotation. A detailed
12.3 Examples 295
knowledge of rolling friction - and material properties - is essential in the

manufacture of baII
ball bearings.
Question. A car is at rest on a horizontal road. The car now begins to accel-
erate along the road. The forces released by the combustion of the gasoline
are internal forces for the system (the car). What gives rise to the extern al
external
force that produces the acceleration of the center of mass of the car? t::,
/::,
Example 12.6. Yo-Yo on the Floor. A disk shaped body of revolution

(Yo-Yo) is provided with a deep groove in the median plane perpendicular to
the axis of symmetry A which passes through the center of mass of the body
(see the sketch). A string is wound around the cylindrical shaft, which has
radius r. The moment of inertia with respect to A is 1I and the total mass
of the body (the Yo-Yo) is M. The free end of the string is held by a hand
pulls lightly on the string with force F.
which puIIs
A·····
A --.-- ....
••.... ............
.•....••.•...
Yo-Yo
The coefficient of friction between the body and the horizontal surface (a
table) is J.L.
fl. Let us assume that the body is always in contact with the table
and that the body initially is at rest on the table.
(1) Determine the initial angular acceleration a of the Yo-Yo as a function
(), assuming that the Yo-Yo does not slide.
of 0,
(2) Determine the value 0() == 0 0 for which the body cannot perform pure
()o
rolling. To which side does the body roll for ()0 > ()o
00 and ()0 < ()o,
00 , assuming
no sliding?
(3) What condition must the coefficient of friction J.L
fL satisfy in order that the
motion, for given values of F = = IF I and (),
0, can be pure rolling?
Solution.
(1) Orientation: VCM is positive towards the right. The positive direction of
rotation is chosen in such a way that a positive value of w gives a positive
value of VCM .
dVcM
M dVCM Fsin(B)-f,
Fsin(8)-j, (12.20)
dt
la
Ia -Fr+fR.
-Fr+jR. (12.21 )
The geometric condition for pure rolling is

dVCM
cM
Ra=cit=a. (12.22)
From (12.20)-(12.22):
F
F .
a= + MR2 [Rsm(8) - r].J .
a = 1l+MR2[Rsm(B)-r (12.23)
This value of the angular acceleration can be found directly by using the
angular momentum theorem with respect to the axis C (the contact axis
- see the sketch).
Rsin(B). Using the angular momentum theorem

The segment CB equals Rsin(8).
with respect to C we get
(MR22 + 1)a
(MR I)a = rJ.
F[Rsin(B) - r].
= F[Rsin(8) (12.24)
This leads directly to (12.23).

12.3 Examples
Examples 297
eo is determined by the condition aa

(2) Bo = O:0:
Rsin(eo) - r
Rsin(B r O 0,,
sin(eo) = rr/R,
sin(Bo) / R,
e > eo =>
B> Bo ==? towards the right,
e < Bo
B eo ==?
=> towards the left.
The last two results can also be found by inspection of the figure below.
When the extension of F passes through C, the Yo-Yo cannot roU roll (rotate
about C). When the extension of F faUs falls to the right of C, Le., e < Bo,
i.e., B eo,
the body roUs rolls towards the left.
(3) The largest reaction force on the Yo-Yo that the surface can provide is
fmax = ţtN
Jmax J..LN = ţt[Mg
J..L[Mg - Fcos(B)].
Fcos(e)]. For pure rolling, we have from (12.21)
fJ == (Ia+Fr)/R
(Ia+Fr)/ R where aa is given by (12.23). The condition tobe satisfied
by ţtJ..L is therefore
[IF/(I + MR
MR2)] [Rsin(e) - r]
2 )] [Rsin(B) + Fr
J..L>
ţt>
- [
R Mg - Fcos(B)]
Fcos(e)] .
B) == M 9 the coefficient of friction must be infinite. How

Question. For F cos( e)
do you interpret this result? What happens for F cos( e) B) > M 9 ? 6.
/::,
Example 12.7. Rolling Over an Edge. A homogeneous circular disk with

mass M and radius R rolls w on a horizontal floor. The
roUs with angular velocity W
disk collides with the edge of a low, fixed table B. The edge of the table is
at a height h above the floor. We consider the case h < R. The collision is
regarded as completely inelastic and of short duration. Assume that the disk
does not slip on the edge.
Determine the least angular velocity which the disk must have if it is to
"roU up onto the table" .
"roll
Solution. The physical content of this example is very close to the physical
content of Example 12.1.
8
ro
A
B
////////////////////////////////////////////
When the disk collides with the edge A, the disk begins a rotation about
an axis through A, perpendicular to the plane of the paper. When the collision
time is so short that the disk has not turned appreciably during the collision
we can use the theorem of conservation of angular momentum about the
axis through A during the collision (the collision approximation). This can
be used to calculate the angular velocity W a with which the disk begins to
Wa
turn about A. When the collision is over, the forces that act on
On the disk are
conservative, and we may use the theorem of conservation of the mechanical
energy to determine the critical angular velocity Wo.
Wo.
We have
LA =
LA = LCM +RCM +
X MVCM.
RCM X MVCM .
Assume that the disk before the collision has just the critical angular velocity
Wo.
WooThen conservation
conservat ion of angular momentum about the axis through A means
that
Using
1 2
= "2~32 M
1 1M
'2 M RR2 ,, A
CM = "2
ICM I =
iA MRR22 ',
and =
VCM = Rwo, we find
W
a
=
= Wo (1 - ~~)
3R .
We can now determine Wo

Wo by using the energy theorem:
2y'(1/3)gh
2J(1/3)gh
Wo = R (1 - 2h/3R)
In the limits
(a) h ---*
---+ 0O =} Wo ---* 0
Wo ---+ O (which makes sense!),
(b) h ---*
---+ ~ R =} Wo ---*
---+ 00 (so this is a critical height of the table)
table)..
12.3 Examples 299
:%
'------.----1;::;
:% fixed
waH
;::;
;::;
;::;
;::;
:%
The cases h > R can only be realized by a special setup (in the example we
assumed h < R - see the sketch).
Questions.
(1) A hole C in a boom is at a height h == (3/2)R above the floor (see the
sketch). Find the total angular momentum of the disk about the point
C on the axis before the collision. Answer: Le Lc = o.O. For h > R, the
spin angular momentum and the orbital angular momentum are oppo-
= (3/2)R the two contributions to the total angular
sitely directed. At h =
momentum cancel.
(2) Consider again the original collision with the edge of the table. What
fraction of the original mechanical energy of the disk is "lost" as heat in
the collision? Answer:
f == ~~
3R
(1- ~~)
3R'
for h -> ~R we have f -> 1.
Example 12.8. Determinism and Predictability. Newton's equations of

motion have the mathematical property that, given a specific initial state of
the system (initial positions and velocities), the solution at any finite future
(or past) time is determined uniquely. A system of differential equations with
this property is said to be deterministic.
It is important to realize, however, that this does not necessarily mean
that almost identical initial states have almost identical solutions. lndeed,
Indeed,
many conceptually simple systems have the remarkable property that almost
identical initial states can have very different solutions.
,,
,
Three mutually gravitationalIy

gravitationally interacting masses, or a pendulum attached
to a pendulum (the double pendulum) in a gravitational field, are examples
of systems of this type, the complete solution of which cannot be provided
by present-day mathematics.
Thus, while deterministic in the above sense, systems can exhibit behavior
"chaos") making it impossible to predict the evolution of the system without
(("chaos")
infinitely precise knowledge of the initial state. Such systems are the subject
of much current investigation. See also Chapter 16. !:::.
6
12.4 Problems
Problem 12.1.
a
• •
~~ tt ~.
m v
• .
eM
CM
a
-A
'/ '/ '/////
A homogeneous, cubic wooden block of mass M is at rest on a horizontal,

frictionless table in the laboratory. The block has sides a. A pistol bullet of
mass m and horizontal velocity v is shot into the block at a height h = a/2
over the table. The velocity of the bullet is perpendicular ta
to the side of the
block (see the figure). The bullet is assumed to carne
come to rest in the center of
mass of the block. The block now begins to slide along the frictionless table
until the block is suddenly stopped in a completely inelastic collision with a
lip, A, at the edge of the table. From this point the motion is pure rotation
about A. The lip A is assumed to be so small that it can be considered an
axis, and in particular it provides no torque on the block while it is in rotation
about A. It is assumed that the collision time Llt is saso short that the block
does not turn appreciably about A during Llt.
(1) Determine the velocity (magnitude and direction) of the center of mass
CM just after the block has been stopped at A.
(2) Determine the magnitude of the maximum velocity VVm m == IVVmm I that the
pistol bullet may have if the block is not to fali
fall down from the table.
Problem 12.2. A rod of length 1 m stands upright an on a horizontal floor in

the gravitational field of the Earth. The rod now begins to falI
fall (from rest),
12.4 Problems 301
rotating around a horizontal axis. (The end of the rod in contact with the
floor does not slip on the floor.)
(1) Determine the velocity relative to the floor of the free end of the rod,
just before it hits the floor.
particle falling freely from rest through 1 m.
(2) Determine the velocity of a partide
Problem 12.3. A ring of mass M and radius r turns in a vertical plane

without friction about a horizontal axis through the center of the ring because
a mouse with mass m runs up the ring. The mouse remains on the same point
B in space (!). B is the end point of a radius vector that forms the angle 0()o0
with the vertical.
Determine the angular acceleration a0: of the ring.
Problem 12.4.
A homogeneous cylinder with circular cross section of radius r is at rest

small bulIet
on a frictionless, horizontal surface. A smalI bullet is shot horizontalIy
horizontally into
the cylinder at the height h above the surface, and perpendicularly to a
vertical plane through the cylinder axis. The bulIet
bullet passes along a straight
line through the cylinder in a very short time, and vertically above the CM
bullet looses the momentum P to the cylinder.
of the cylinder. The bulIet
Find h so that the mot ion of the cylinder after the "collision" is pure
motion
rolling, i.e., rolling without sliding.
Problem 12.5. A narrow straight tunnel has been drilled through the Earth
between Copenhagen and Los Angeles (see Problem 8.4). How long will it
take for a baII
ball to rolI
roll through the tunnel?
Problem 12.6. A homogeneous baII ball with mass M and radius R is set in
floor in such a way that the baII
motion on a horizontal floar ball initialIy
initially (i.e., at
time t == O) .
Vo.
0) has the translational velocity Vo We assume that the baII
ball begins
its motion as pure sliding. The coefficient of friction between the baIIball and
the floor is 1-". ball between the time t == 0O and t == tr is
{l. The motion of the baII
a combination of sliding and rolling. At time t = tr the ball
baII begins to rolI
roll
without sliding. In the time between t = O 0 and t = tr the baII
ball has moved a
distance D along the floor.
(1) Determine ttr.

r.
(2) Determine D.
(3) Determine the work W that the forces of friction perform in the time
from t == O
0 to t == tn
t" Le.
i.e. in the time where the ba11
ball was both rolling and
sliding.
(4) Show that W = J~r jlMgVc(t)dt,
pMgVc(t)dt, where Vc Ve is the velocity of the contact
point relative to the supporting surface (the floor).
Problem 12.7.
a
~
Initially, A is at
A homogeneous sphere K is at rest on a horizontal plate A. Initia11y,
rest in the laboratory. At t == O0 the plate A begins to accelerate with constant
acceleration a in the horizontal direction (see the figure). The acceleration of
gravity is g. The coefficient of friction between K and A is p. jl.
smallest value pjl = Pmin
Find the sma11est jlmin the coefficient of friction may have if
we demand that the sphere K will begin to ro11 roll without sliding relative to A,
when we give A the acceleration a.
Problem 12.8.
A thin homogeneous circular ring with mass M and radius R lies on a fric-
tionless horizontal table in the laboratory. The ring may turn without friction
around a vertical axis thraugh
araund through the point A on the perimeter of the ring. The
axis is fixed relative to the table (see the sketch).
A sma11
small beetle with mass m crawls around
araund the ring, moving with constant
speed Va
Vo relative to the ring. The ring is initia11y
initially at rest, and the beetle starts
out from the point A. As the beetle moves araund around the ring, the ring ratates
rotates
around the axis thraugh
araund through A.
12.4 Problems 303
Determine the angular velocity of the ring in its rotation around A when
the beetle reaches the point half way around the ring. The angular velocity w
is measured relative to the lab frame, and w should be expressed as a function
of m, Ad,
of~, vo, and R.
M, Vo,
Problem 12.9.
l A
t
g B
"l[""""""~:,,
A thin homogeneous rod of mass M Ad and length L is at rest and horizontal
in the gravitational field of the Earth. One end, B, of the rod rests on the
edge of a table. The other end A is supported by a vertical force (from a
hand). The hand is suddenly removed, so that the end point A is no longer
supported.
Determine the magnitude and direction of the force F by which the table
acts on the rod just after the hand is removed.
Problem 12.10.
A
•
v
v
...
-----~ -------
------~-------- --------~
2
------------.
A
A heavy homogeneous rod has mass Ad M and length L. The rod can turn
without friction about a fixed horizontal axis A which is perpendicular to
the rod (see the sketch). Initially the rod hangs at rest, vertically, in the
gravitational field of the Earth. The acceleration of gravity is g.
A bullet with mass m L1t through the rod near
~ passes in a very short time Llt
the lower end (so the distance from A A to the point of impact can be taken to
be L).
The direction of the velocity of the bullet - both before and after passing
through the rod - is horizontal, perpendicular to the rod and to the axis A.
Before the collision the velocity of the bullet is v, and after the collision it is
v /2 (see the sketch).
Llt is so small that the rod does not move as the bullet
We assume that ..1t
passes the rod. Ignore air resistance.
(1) Determine the angular velocity Wo of the rod just after the bullet has
passed through the rod. The angular velocity should be expressed as a
function of m, M, v, and L.
(2) Determine what the speed v of the bullet should be such that the rod
swings exactly 90° from its initial equilibrium position. Express v in terms
of m, M, g, and L.
ofm,
From its horizontal position the rod now swings back. As it does so, the axis
Fl.
through A acts on the rod with a reaction force F 1.
(3) Determine F F11 at the moment when the rod passes through its initial
vertical position. Express F 1 as a function of M and g.
We now return to the original situation and consider the rod as the bullet is
passing through the rod. During this time, ..1t,
Llt, there is a reaction force from
the axis A on the rod. The horizontal component is called F 2 •. Assume that
F22 is constant during the time ..1t.
F Llt.
(4) Find FF22 =1 FF22 1for the situation described in question (2). Express F
F22
in terms of ..1t,
Llt, M, g, and L.
(5) State the condition that ..1t
Llt must satisfy for the approximation used to
be valid.
Problem 12.11.
I
R
M L
M
p
P
13
CII !
A thin homogeneous rod R has length L and mass M. The rod lies at rest on
a frictionless horizontal table. A small piece of sticky day
clay (a partide),
particle), of the
same mass M strikes the rod, near the end, with velocity v. The velocity v
is perpendicular to the rod. The collision is completely inelastic, and after
the collision the partide
particle P sticks to the rod. After the collision the partide
particle
P and the rod move together as one body.
(1) Determine the velocity of the CM of the system before and after the
collision (take the system to be both P and R).
(2) Determine the angular velocity w in the rotation around the CM after
the collision. The angular velocity should be given as a function of v and
L.
12.4 Problems 305
L1T is lost in the collision

A certain amount of mechanical kinetic energy LlT
(converted into heat).
L1T as a function of M
(3) Determine LlT M and v.
(4) Assume now that the rod is fastened in such a way that it can turn
without friction in a horizontal plane around a vertical axis through the
point A. This axis is at a distance d from the point where P hits the rod
(see the sketch). We assume that the duration of the collision is so short
that the rod does not turn during the collision. Determine d in such a
way that there are no reaction forces from A during the collision (the
point A is then called the center of percussion).
calledthe
---..
p
Problem 12.12. A rod of mass M and length L is kept at rest with one end
on a frictionless horizontal floor and the other on a frictionless vertical wall.
The rod is in a vertical plane perpendicular to the wall. The inclination with
00' At a certain time the rod is left free
respect to the horizontal is initially ()o.
to move.
(1) Find iJ() as a function of O.().

(2) Find the critical
critic al angle OI
()1 for which the rod leaves the wall.
(3) Assume that 0()o0 = 90°90 0 (we then have to give the rod a slight push to set
.x
it in motion). Find the horizontal component XCM ofthe of the velocity of the
center of mass just before the center of mass reaches the floor.
Problem 12.13. A thin homogeneous rigid rod has length L and mass M.
One endpoint of the rod is fastened to a light, rigid steel wire, also of length L.
oflength
_L _ _ L- O
B M A
B
A B
~
The other end of the wire is fastened to a horizontal axis O in the laboratory.
The system can turn in a vertical plane. The steel wire cannot bend, and its
mass can be ignored relative to M.
Initially the system is held at rest in a horizontal position. It is then set
free to move under the influence of gravity. The acceleration of-gravity is g.
Ignore air resistance.
(1) At the moment the system begins to move, Le., while the rod is stiH
horizontal, the wire acts on the rod with a force F1' at the endpoint A.
Determine F 1.
(2) At the moment when the rod is vertical, the wire acts on the rod with a
force F 2 . Determine F 2 •
(3) Assume now that the steel wire breaks at the moment when the rod is
vertical. The lower point B of the rod is then at some height h above the
laboratory floor. When the rod hits the floor is has turned so that it is
exactly parallel to the floor. Determine h.
Problem 12.14. Find the angular velocity w(t) for the sprinkler in Example
12.4 by consider ing the problem from a reference frame in which the tube
is at rest. Use Newton's second law in an accelerated reference frame. Don't
forget the term m( r x w).
12.4 Problems 307
Problem 12.15.
The mechanical system sketched on the figure consists of two homogeneous

disks, mounted as shown, i.e. Le. with their plane sides in the same vertical
plane. The disks have masses M MIl and M 22 ,, and radii RI
Rl and R 22 .. Disk #1
can rotate without friction about a horizontal axis A through the center of
the disk and perpendicular to the disk. The axis A is fixed in the laboratory
which is considered an inertial frame.
An unstretchable string has been wound around both disks, connecting
them as shown. Disk #2 may falI fall as the string unwinds. The string is con-
sidered massless. The acceleration of gravity is g.
Initially, the system is at rest, and the segment of the string not wound
around the disks, is tight and vertical.
The system is now left to move under infl.uence influence of gravity. The known
quantities are M lI , M 22,, Rl,
R 1 , R2'
R 2 , and g.
1. Determine the total kinetic energy of the system when the center of the
lower disk has moved a distance x downwards in the field of gravity.
2. Determine the magnitude S of the string for ce during the motion of the
force
system.
Problem 12.16.
A thin string has been wound around a homogeneous circular disk A in such
a way that the disk can roll on a horizontal surface. The mass of the disk is
M, and its radius is R. The string goes around a small frictionless pully e
at a height 2R above the horizontal table B. To the free end of the string a
bob of mass M is attached.
The disk, the pully e, the bob and the string are aU in the same vertical
plane.
The system begins to move under the infl.uence of gravity. The string is
taut throughout the entire motion, and the mass of string and the mass of
the pully e can be disregarded. It is further assumed that the coefficient of
friction J.L is so large that the disk roUs without slipping on the horizontal
table. The magnitude of the gravitational acceleration is g.
1. Determine the magnitude of the force S with which the string acts on
the disk as long as the system moves under the infl.uence of gravity.
2. Determine the smallest value of J.L that permits the disk to roU without
slipping.
Problem 12.17.
s
12.4 Problems 309
A baU of mass M and radius R is at rest on a rough inclined plane as

shown on the figure. The inclined plane is at rest on a horizontal table in
the laboratory, which may be regarded as an inertial frame . The inclination
of the plane with respect to the horizontal is designated by B. The baU is
restrained by a thin horizontal piece of string S. The direction of the string
is a tangent to the baU at the top of the baU. The other end of the piece of
string is attached to the inclined plane. The magnitude of the acceleration
of gravity is designated by g. Known quantities are M, R, B, and g.
1. Determine the magnitude ofthe force F with which the piece of string
acts on the baU.
The coefficient of friction between the baU and the inclined plane is des-
ignated by /-L.
2. Determine the minimum value /-L = /-Lm of the coefficient of friction such
that the baU remains stationary.
It is now assumed that the coefficient of friction between the baU and the
inclined plane has the value /-L = /-Lm determined in the previous question. At
some given time the string is cut.
3. Determine what value the angle B must have for the baUe to begin to roU
without slipping, after the string is cut.
Problem 12.18.
A heavy homogeneous cylinder C, with a circular cross section of diameter 2r,

is loaded on the back of a truck. See the sketch. The distance from the center
of the cylinder to the end of the truck's body is L. The truck is initiaUy at
rest on a horizontal stretch of road . When t = O the truck begins to move to
the left at constant accelerat ion a. The cylinder then begins roUing without
slipping along the back of the truck, and subsequently the cylinder faUs to
the road. Air resistance may be ignored.
1. Determine the elapsed time T, from the starting of the truck's motion,
and to the time when the cylinder leaves the back of the truck. Assume
that the cylinder loses contact with the truck the instant the center of
mass of the cylinder is positioned vertically above the re ar edge of the
truck. Express T in terms of L and a.
2. Determine magnitude and direction of the horizontal velocity v of the

cylinder C with respect to the road just as the cylinder touches the road.
Express the magnitude v = Iv i in terms of L and a.
Assume that the size of the coefficient of friction f..L between the road and
the cylinder initia11y causes the cylinder to ro11 as we11 as slide along the road .
3. Determine the elapsed time T from the instant the cylinder touches the
road and until the cylinder starts rolling without slipping. Express T in
terms of L, a, f..L, and the acceleration of gravity g.
Problem 12.19.
A homogeneous sperical billard ba11 of radius R and mass M is initia11y at

rest on a horizontal table. By appropriately pushing the ba11 with a finger,
it is possible to make the baU move, satisfying these conditions: The center
of mass of the ba11 has a horizontal velocity Va while the ba11 has angular
velocity Wa about a horizontal axis, perpendicular to Va. See the sketch.
The velocity Va of the center of mass is in the drawing plane. The rotation
vector Wa points out from the drawing plane. This is sometimes referred to as
"backward spin" . The magnitude of the acceleration of gravity is designated
by g, and the coefficient of friction between the ba11 and the table top is f..L .
At t = 0, the baU starts moving to the right, both ro11ing and sliding.
1. Determine the magnitude of the velocity of the center of mass V(t) and
the magnitude of the angular velocity w(t) as functions of the time t as
long as the ba11 is both rolling and slipping.
Consider two scenarios with different values of the ratio ~. (Va = IVa 1, Wa
= IWal).
2. Determine the ratio lQ

Wo
such that the ba11 both ro11s and slides for a while,
and then comes to a halt. Express the ratio lQ
WQ
in terms of R.
12.4 Problems 311
Consider a different scenario. Assume, that the initial conditions satisfy

the following: The ball both rolls and slides until the velocity of the ball's
center of mass is zero. However, the ball stiU
still has a non-zero angular velocity
("backward spin"), and thus the ball begins to roll and slide back towards
its original position. Assume, that the ball wiU
will stop sliding after some time
and perform a pure rotary motion.
Yll.
3. Determine the ratio Wa
Va such that the ball's center of mass finally attains
Wo
¥
Va (pure rolling towards the left, see the sketch).
a constant speed ~Va
Problem 12.20.
1 0jO] ________
_________--:;
} 1Ig
!tI-_ ---
F
B
---'
..•.
••••,,/'
",'
A homogeneous, straight rod of length R and mass M may rotate without

friction about a fixed horizontal axis O. The axis O 0 is perpendicular to the
rod and goes through one of the rod's end points. A point mass B, also
of mass M, has been attached to the other end point of the rod. The rod
is initially at rest in the Earth's gravitational field. The magnitude of the
acceleration of gravity is g. The laboratory is considered to be an inertial
frame. At t = 0, the point mass B receives a collision impulse given by
{Llt
J == Ja
la Fdt
The duration Llt

L1t of the collision impulse is so short, that rotation of
the system consisting of the rod and the point mass is inconsiderable. The
force F is at any time perpendicular to the rod as well as the axis O. After
the collision, the rod begins rotating about O, 0, and the rod just reaches
a horizontal position before swinging back. Air resistance is ignored. The
known quantities are R, M, and g.
1. Determine the magnitude of the collision impulse, i.e., determine P = P.

2. Determine the magnitude and direction of the force K with which the
axis 0O acts on the rod. when the latter is horizontal.
3. Determine the magnitude and direction of the force G with which the
axis O acts on the rod the instant the rod reaches the vertical position
during its subsequent rotation.
4. Determine the conditions, .1t must satisfy to justify the employed colli-
sion approximation.
5. Determine the average < F v > of the horizontal component of the force
the axis O acts on the rod during the collision.
Problem 12.21.
A homogeneous, circular cylinder of radius r and mass m is rotating with

angular velocity Wo about the axisof the cylinder, which is horizontal. At
t = 0, the rotating cylinder is placed on a horizontal surface such that the
velocity of the cylinder's center of mass is zero with respect to the surface.
°
The coefficient of friction between the cylinder and the surface is f-1 and the
magnitude of the acceleration of gravity is g. From t = to t = tI, the
cylinder is both slipping and rolling. For t > tI, the cylinder is only rolling.
1. Draw a sketch depicting the forces acting on the cylinder on the surface.
2. Determine tI .
°
3. Determine the magnitude of the mechanical energy transformed into heat
between t = and t = tI.
13. The General Motion of a Rigid Body
In the previous chapters we have considered rotations of rigid bodies possess-

ing a high degree of symmetry. We have seen that the rotational motion of
such a body around an axis fixed in inertial space and fixed relative to the
body is as simple to describe as the motion of a single particle moving along
a straight line. The same equations govern the motion in these two cases,
when p is substituted for L, and v is substituted for w, etc.
Nevertheless we shall see in the following that, for certain types of ro-
tational motion around an axis fixed in inertial space, new and somewhat
surprising features may arise. As soon as we leave the case of rotation about
an axis at rest both relative to an inertial frame and relative to the rotating
body, the similarity between translational and rotational motions comes to
an end.
13.1 Inertia in Rotational Motion
The differences between the description of translational motion and the gen-
eral description of rotational motion, stem from the following facts.
For translational motion, the mass m (the "inertia") is a scalar quantity.
For instance,
p=mv.
This means that the velocity v and the momentum p are always parallel
vectors.
The corresponding equation for rotational motion, i.e., the connection
between Land w, is more complicated. The inertial properties of a rotating
body cannot be expressed as a scalar quantity alone, it must be described
by a quantity called a tensor. This tensor - called the inertia tensor - act in
the equations as a scalar only in the simple case of rotation about an axis
fixed in inertial space and fixed in the rotating body. For this special case,
LA = JAW in analogy to, say, Px = mv x ·
Since the connection between the angular momentum vector and the an-
gular velocity vector is in general expressed by means of a tensor, the angular
momentum vector will not necessarily be parallel to the rotation vector.

314 13. The General Motion of a Rigid Body
Before proceeding to the description of the inertia tensor we wish to illu-

minate the essential issue by briefly describing a few particularly transparent
cases of rotation about a fixed axis where Land ware not parallel, but where
the problem nevertheless can be solved without introducing the formal inertia
tensor. The following three examples are of this nature.
Example 13.1. The Dumbbell
We regard a system consisting of the following parts: two, identical, small,

heavy balls, each of mass m are connected with a thin rod R with total
length 2a (a dumbbell). We disregard the mass of the rod, and the balls are
considered so small that they can be treated as point masses.
The mid-point of the rod R is attached to the mid-point 0 of a horizontal
axis A which can rotate without friction in two bearings Band C. The rod R
forms the constant angle B with the axis A. The axis A has total length 2 l.
The system is now set in rotation. The axis A rotates in the bearings B
and C with the angular velocity w. The rotation vector w is thus directed
along A (see the sketch). Note: the axis of rotation is fixed in the inertial
system, but not relative to the body. The rod R moves on the surface of a
cone.
As our system we consider the axis A plus the rod R and the two balls. As
long as we disregard friction in the two bearings Band C (and air resistance),
the system will continue its rotation with angular velocity w. There are no
external torques with components along w.
Let us now analyse the motion. We first calculate the angular momentum
with respect to o. The sketch shows an instant where the rod R is in the
plane of the paper. The topmost ball (1) is moving out of the paper and
the lower ball (2) is moving into the paper. The point masses give the same
contribution to the angular momentum. The total angular momentum vector
is in the plane of the paper and has the direction indicated on the sketch.
During the entire motion L will lie in the plane spanned by wand the rod
R. The magnitude of the velocity relative to the lab frame of each of the two
masses is
v = [asin(B)]w.
The magnitude of L is
13.1 Inertia in Rotational Motion 315
L = 2amv = 2am[asin(O)]w = 2a 2 mwsin(O).

The angular momentum vector moves around on the surface of a cone with
its apex at O. The tip of L is on the above sketch moving into the plane of
the paper. The direction of dL/dt is thus into the plane of the paper.
We now calculate the magnitude of dL/dt.
Note. In this example the motion is known and from this we determine the
forces and torques necessary to maintain the motion.
dLI = IL I cos(O)w = 2ma w cos(0) sm(

Idt 2 . 0) .
2
In order to maintain this rate of change of the angular momentum vector,

it is necessary to have an external torque on the system. This torque comes
from the bearings. In each bearing there must be, because of the rotation, a
reaction force, Fr. The instantaneous directions are indicated on the sketch.
The torque No at 0 must have the same direction as dL/dt, that is,
perpendicular to the axis A and into the paper. No has the magnitude:
No = 2Fr l = 2ma2 w2 sin(O) cos(O),

from which
ma 2 w2
Fr = - l - sin(O) cos(O).
Since the center of mass for the system is at rest, we know from the center
of mass theorem that the sum of external forces on the system vanishes.
The total reaction force from the bearings thus becomes, counting all
reaction forces positive upwards:
Fr + mg in the left side, i.e., bearing B,
-Fr + mg in the right side, i.e., bearing C.
Apart from the reaction forces in the bearings, the force of gravity acts on
the balls. The total external force on the system is therefore
Fr + mg - Fr + mg - 2mg = o.
The reaction forces caused by the rotation change direction as they rotate
with the system. Note the following.
(1) There is no external torque with components along the rotational axis.
Thus, w remains constant. A possible friction in the bearings may provide
a torque about the axis A and thus brake the rotation.
(2) for e = 7r/2 the angular momentum vector L is parallel to A and dL/dt
vanishes. This case gives the smallest wear on the bearings; the system
is said to be dynamically balanced.
Example 13.2. Flywheel on an Axis
A heavy flywheel in the shape of a homogeneous circular disk has mass M

and radius R. The wheel is fastened to a horizontal axis A passing through
the center of mass of the wheel. The axis of symmetry for the wheel forms a
small angle () with the axis A. The flywheel is thus mounted out of alignment.
The rotation axis A is assumed to be turning without friction in two bearings,
one on each side of the wheel, both at a distance l from the center of mass of
the wheel. The angular velocity of the axis - and consequently of the wheel
- is Woo
Since we disregard friction in the bearings the wheel will continue to rotate
with angular velocity woo No external torques have components along woo We
shall see that the misalignment of the wheel causes a stress on the bearings
due to the rotation.
As our system we take the wheel and the axis A which is considered
massless. The reaction forces on the axis from the bearings are thus external
forces for the system.
We begin by determining L CM , Le., the angular momentum with respect
to the center of mass (see the figure below).
We split Wo into two components:

(along the normal to the plane of the flywheel); and
Wl
W2 (the remainder, in the plane of the flywheel).
The angular momentum vector LCM can also be split into two components,
corresponding to the two components of Woo
LCM (13.1)
where
1
IL11 = hwo cos( ()) = "2 M R2Wo COS( ()) , (13.2)
IL21 = 12wosin«()) = ~MR2wosin«()), (13.3)
LCM ~MR2wO cos 2«()) + ~ sin 2«()) . (13.4)
The moment of inertia about the normal to the disk of the flywheel is
different from the moment of inertia about a diameter of the disk. This is the
cause of LCM being not parallel to Wo. Had the two moments of inertia been
equal, LCM would have been parallel to woo Since the moment of inertia about
an axis perpendicular to the disk is the greater, LCM lies as shown on the
figure. The angle between Wo and LCM is denoted 0:. It may be determined
from the relation
L2 1
tan«() - 0:) = -L = - tan«()).
1 2
In order to determine the torques at the bearings, i.e., the external torques
on the system, we must first compute dLcM/dt. The motion is known, and
we are determining the forces.
The endpoint of LCM describes a uniform circular motion about the axis
A. At the instant shown on the above sketch, the tip of LCM is moving out of
the paper. The vector dLcM/dt is thus perpendicular to the paper, pointing
out of the paper.
In order to determine the magnitude of dLcM/dt we need to first deter-
mine the component L1. of LCM that is perpendicular to the axis A; we will
then have
From the sketch we can deduce the relation

L1. = L1 sin«()) - L2 cos«()).
Inserting the previously determined L1 and L 2, we find
dLcM I = 4MR
I~ 1 2Wo2 cos«()) sin«()).
In order to keep the wheel rotating, there must be an external torque

N. This torque stems from the reaction forces F1 and F2 at the bearings.
The directions are as shown on the figure above. The two forces have equal
magnitude (why?) but the directions are different ("right-hand rule"). The
magnitude F of the two forces is given by
1 .
2Ft 4M R2w6 cos«()) sm«()) ,
~ M R2w5 cos( ()) sin( ())
F
8 t
Because of the weight of the system there will also be reaction forces to
counteract the force of gravity. At the instant shown on the sketch, the total
reaction force on the bearings, with the upward direction as positive, is given
by
Right bearing: F + Mg/2,
Left bearing: -F + Mg/2.
The total force in the vertical direction is thus
Mg Mg
F+--F+--Mg=O.
2 2
This assures us that the center of mass has no net acceleration.
Note: the reaction forces due to the rotation change direction continuously,
as the axis rotates. The reaction forces caused by gravity are always in the
upwards direction.
When a flywheel is misaligned on its axis of rotation, the bearings will
wear faster than if the wheel is accurately mounted perpendicular to the axis
of rotation. A misaligned wheel is said to be dynamically imbalanced.
Question. Let us assume that the wheel rotates at 300 revolutions per minute.
Calculate F, when M = 1000 kg, R = 1 m, e = 1°, l = 0.2 m .
Answer. F ~ 1.08 X 104 N, the weight of the wheel is 9.8 x 103 N. !:::,.
Example 13.3. Precession of a Gyroscope. In this example we consider

the physics behind the phenomenon known as the precession of a gyroscope.
A gyroscope (also known as a top) is, in its simplest version, a wheel or
disk able to rotate about an axis, called the gyro-axis or the figure-axis, that
passes through the center of the disk (wheel).
o
A
c D
The figure shows a disk of mass M that can rotate about the axis A. We
ignore the mass of A. This axis is suspended so that it can pivot about the
origin 0 of the coordinate frame, assumed to be fixed (in the lab frame). Let
. us initially assume that the disk is not rotating about the axis A, and that
the axis is also supported in the other end point, P, so that the axis A is
along the (horizontal) x-axis. Then we remove the support at P.
The force of gravity, Mg, acting at the center of mass, has a torque N
about O. This torque points in the y-direction (i.e., into the paper). The disk
will begin to fall, initiating a rotation about the y-axis.
We now change the procedure. Before removing the support we set the
disk spinning with a large angular velocity around the axis A. We again
remove the support at P. One may then observe a peculiar phenomenon:
The disk will not fall as in the case described above, instead the axis A will
remain almost horizontal initiating a slow rotation (one says the gyroscope
is precessing) around the z-axis. On the figure the disk will precess around
the z-axis in the positive direction (the disk goes into the paper).
We wish to show that this behaviour is in accordance with the angular
momentum theorem. The disk D of radius r is rotating with a large angular
DO
1
0
B
Lo
~
A
C 0
Mg
velocity Wo about the axis A. The rotation vector Wo is directed such that
the rotation carries the point B out of the page. The distance from 0 to the
center of mass of D is l. We assume that we start the gyroscope off in such a
way that it carries out a so-called regular precession. We now show that such
a motion is in accordance with the angular momentum theorem:
dL o -N
dt - o·
Note. We do not integrate this equation. We simply verify that regular pre-
cession is a possible motion.
The magnitude of the torque of gravity about 0 is
No = lMg,
where No is directed into the plane of the paper.
When the gyroscope is in precession, the axis A continuously changes its
direction and this motion carries the center of mass for the disk D around
in a circular motion about the point O. This motion will add to the angular
momentum about O. If, however, the angular velocity of the disk is quite
large, the component of the angular momentum along A will be much greater
than the component due to the precession about O. In that case we can
assume as an approximation that Lo is parallel to A. The figure below shows
the precessing gyroscope as seen from above. The end point of Lo moves in
,
,,
I
I o
,,
\
,,
a uniform circular motion about O. The angular velocity in this motion is

denoted .no, and we wish to determine .no in accordance with the angular
momentum theorem.
For the magnitude of Lo we have
Lo = ILo I = I Wo = "21 M r 2wo,

and consequently the magnitude of dLo/dt is
dLOI = ILol no = "21 Mr 2wono.

Iill
The angular momentum theorem demands that
1 2
"2Mr wo.no Mgl,
.n _ 2gl
0- r2wo .
The center of mass of the disk D moves in uniform circular motion about
the point O. The centripetal force necessary to maintain this motion is the
reaction force from the axis C. Furthermore, the center of mass has no ac-
celeration in the vertical direction. From these two results we can determine
the total reaction force from the axis C (see the figure below).
For the reaction force from C one gets
Mg
13.2 The Inertia Tensor 321
c D
Mg
As a numerical example let us assume that we have a rapidly spinning top,

e.g., Wo = 200 s-l. Furthermore: l = 0.30 m, r = 0.15 m, M = 2 kg. One
then finds
no 1.3 s-1 (note: no « Wo ),
Fl Mtn5 = 1 N,
F2 Mg = 19.6 N.
13.2 The Inertia Tensor
We shall now examine the connection between the general state of motion of
a rigid body, and the quantities L, w, and the kinetic energy T - all measured
relative to an inertial frame. We shall restrict ourselves to the cases where one
point 0 of the body is fixed in the inertial frame, so that the instantaneous
motion of the body can be regarded as pure rotation around an axis through
O.
This assumption is not excessively restrictive since the general motion
of the body relative to an inertial frame can always be decomposed into
translation of the point 0 and a rotation about an axis through O.
If the decomposition of the motion is performed such that the translation
represents the motion of the center of mass, the rotation will represent the
motion of the body in the center of mass frame.
We thus consider a rigid body moving in such a way that the point 0 is
fixed in an inertial frame . The instantaneous motion of the body is charac-
terised by the rotation vector w. The total angular momentum with respect
to 0 is
(13.5)
For the vector product of three vectors we have, in general,
A x (B x C) = B (A . C) - C (A· B) .
Using this expression in (13.5) we obtain
Lo = L [r;w - ri (ri· w)] mi· (13.6)
Note: Lo is not necessarily parallel to w. Using ri = (Xi, Yi, Zi) and w =

(wx,wy,w z ) we may write the component Lx along the x-axis
or, with r; = xr + Y~ + z~,

(13.7)
Defining
- ""'
~ m·x·z·
t t t,
we write (13.7) as follows:
(13.8)
i.e., Lx depends linearly on all three components of w.

Using the same procedure to write the components of L along the Y and
z axes we obtain the following important connection between the vectors
L = (Lx,Ly,L z ) and w = (wx,wy,w z ):
Lx Ixxwx + Ixywy + Ixzwz ,

Ly Iyxwx + Iyywy + Iyzwz ,
(13.9)
For instance:
Iyy = L mi (x; + Zl) , moment of inertia around the y-axis,
Iyx = - L miYiXi, product of inertia. Note that Iyx = Ixy.

A rigid body is composed of a continuous distribution of mass. If, instead
of summation, we had used integration, the equations would have taken the
following form:
Lo = j r x (w x r) dm = j r x (w x r) pdV,
where p is the density.
Lo = j[r w-r(r.w)]PdV,
2
Ixx j (y2 + z2) pdV ,
Ixy - j xy pdV ,
etc.
Consider again (13.9) . From the definitions of the products of inertia:
Iij = I ji · Of the nine quantities lxx, I xy , ... only six are independent.
The nine quantities Iij are the components of a (sym-
metric) tensor of rank 2. This particular tensor is
called the tensor of inertia.
This description has the same geometric content as the following: The three
quantities (Lx, Ly, L z ) are the components of a vector (L).
The prescription for calculating the nine (six) components of the tensor
of inertia are given above. The connection between Land w may - in the
coordinate system used - be written by means of the matrix equation:
(13.10)
The nine (six) numbers in the above matrix represents - in the given coordi-
nate system - a physical quantity, which exists in its own right, independent of
the specific coordinate system chosen to represent the tensor (i.e, the physical
quantity). The nine (six) numbers describe the instantaneous inertial prop-
erties of the body in rotation around a fixed point 0 [w = w(t)]. Such inertial
properties exist of course independently of any particular coordinate system
we choose to describe the inertial properties.
For a given coordinate system and a given orientation of the rigid body
being studied there is a clear prescription for calculating the nine (six) com-
ponents of the matrix representing the tensor in that particular coordinate
system.
When we go from one coordinate system to another, the physical quantity
called the inertia tensor will not change, but the components of the matrix
representing the tensor will change. Compare with the three components
representing a given vector in a chosen coordinate system.
Equation (13.10) may be written:
L=lw
where I is a linear operator connecting two physically different vectors Land

w. The components of I in a given coordinate system has the dimensions of
mass times length squared.
If we write the angular momentum theorem ("the equation of motion")
it immediately becomes clear that the problem of motion is complicated:
:t (Iw) = Ngxt . (13.11 )
For instance, the x-component of (13.11) is

dL x
ill = N ox ,
or
(13.12)
As (13.12) shows, generally both the components of the vector wand the
components of the tensor I are time dependent, and this makes the solution
of the problem quite difficult.
There exist various methods for avoiding some of the difficulties, and in
the following sections we shall demonstrate a few of these methods. First,
however, we shall show directly how the idea of the tensor of inertia may be
used to solve the problem of the rotating dumbbell of Example 13.1.
Example 13.4. The Dumbbell Revisited. See Example 13.1. We shall

calculate here the angular momentum L for a rotating dumbbell using the
tensor of inertia.
The coordinate frame (xyz) is fixed in the laboratory (assumed to be an
inertial frame). The z-axis is along the vector w, while the x-axis is perpen-
dicular to the axis of rotation and in the plane of the paper. The y-axis is
then perpendicular to the paper and pointing outwards.
It is assumed that the rod R is in the plane of the paper at the time t = 0,
Le., in the xz-plane. The coordinates of particle (1) and particle (2) at any
time t ;:: 0 are then:
(1)
Xl asin(B)cos(wt) =:=rcos(wt),
Yl asin(8)sin(wt) =:= rsin(wt) ,
Zl acos(8)=:=d.
(2)
X2 -asin(8)cos(wt) =:= -rcos(wt) ,
Y2 -asin(B) sin(wt) =:= -rsin(wt) ,
z2 = -acos(B) =:= -d.
We have used the abbreviations
r=:=asin(8), and d=:=acos(B).

We now proceed to calculate the components of the tensor of inertia:
Ixx m (yr + zi) + m(y~ + z~)

2m (r2 sin 2(wt) + d2) ,
Ixy Iyx = -mXlYl - mX2Y2

-2mr2 sin (wt) cos(wt) ,
Ixz -2mdr cos(wt) .

The process is analogous for the other components. We obtain for the
tensor of inertia at the time t:
[
r2 sin 2(wt) + d 2
I = 2m _r2 sin(wt) cos(wt)
_r2 sin(wt) cos(wt)
r2 cos 2(wt) + d2
-rd cos( wt)
-rdsin(wt)
1.
-rdcos(wt) -rdsin(wt) r2
The components of the matrix representation for the tensor of inertia are thus
time dependent. The rotation vector in the (xyz) system has the components
w = (O,O,w).
The angular momentum vector is nOW determined:
L=lw.
We find
Lx -2mdr cos(wt)w,
Ly -2mdrsin(wt)w,
Lz 2mr 2 w.
From the angular momentum theorem applied in the inertial frame (the
laboratory frame) we can determine the torque that the bearing has to supply
in order to maintain the rotational motion:
dL 2
citx = 2mdrw sin(w(t) ,
dL y
cit = 2mdrw 2 cos(w(t) ,
dL z
cit=O.
Let us compare these with the results from Example 13.1.
The magnitude of the angular momentum vector is
L IL I = y'4m2 d2 r 2 w2 + 4m.2r 4 w2 ,
L 2ma 2 wsin(B).
The magnitude of dL/dt is
I~~ I = vi4m 2 d2 r 2 w4 = 2mw 2 a2 sin( B) cos( B) .

This is exactly what we found in Example 13.1.
13.3 Euler's Equations
The inertia tensor for a rigid body rotating about a point fixed in an inertial
frame can be written in matrix representation:
I
lxx lxy
= [ lyx lyy lyz
lxz 1 ,
lzx lzy lzz
where I is a symmetric tensor, i.e., lxy = lyx, lxz = lzx, and lyz = lzy.
In books on geometry it is shown that for any symmetric tensor it is
possible to find a coordinate system such that in this coordinate system the
matrix representing the tensor has elements only along the diagonal. For
13.3 Euler's Equations 327
the case of the inertia tensor this means that regardless of what shape a
given body has it is always possible to find a coordinate system in which the
products of inertia are zero. The inertia tensor is then said to be in diagonal
form.
A coordinate system in which the tensor of inertia is in diagonal form is
called a principal coordinate system for the body in question and with re-
spect to the fixed point O. The coordinate axes for the principal coordinate
system are called the principal axes. The three diagonal elements of the ma-
trix representing the inertia tensor are the so-called eigenvalues of the inertia
tensor.
If a principal coordinate system has been found, the inertia tensor ex-
pressed in that system is in diagonal form:
The eigenvalues It, h, and 13 are the moments of inertia around the three
principal axes. The quantities It, 12 , and 13 are called the principal moments
of inertia for the body, with respect to the point O.
In a principal coordinate system we have the following connection between
Land w:
In components this is
Lx Itw x ,
Ly = hwy,
Lz Iaw z ·
These results show that when the three principal moments of inertia are
identical, the vectors Land ware parallel.
The principal coordinate system reflects the mass-geometric properties of
the body. It is evident that a principal coordinate system offers advantages
in the calculations. But, the advantages have a price. A principal coordinate
system is fixed relative to the rotating body. Therefore a principal coordinate
system will in general not be an inertial system.
L. Euler (1707-1783) showed that it is possible to re-formulate the equa-
tions of motion for a rigid body rotating around a fixed point, in such a way
that we obtain certain advantages by using a principal coordinate system to
describe the motion of the body.
13.3.1 Derivation of Euler's Equations
Consider a rigid body rotating about the origin 0 of an inertial frame I (the
lab frame). In the figure below, the axes in I have been named X, Y, and Z.
The instantaneous rotation vector is w. Let us emphasise: the symbol w
always represents the rotation vector relative to the inertial (the laboratory)
frame. The symbol L is always used to describe the angular momentum about
the point 0 and relative to the inertial frame.
We now introduce a coordinate frame S fixed relative to the rigid body
and we choose S in such a way that S is a principal coordinate frame for the
body. In the figure the axes of S are labelled x, y and z, and are marked with
dashed lines.
Consider an arbitrary particle (a mass point) which is part of the rotating

body. Let the particle have position vector r. Of course the particle will always
have the velocity zero measured relative to S. Relative to the laboratory frame
I, however, the particle has the velocity r = w x r.
Consider now an arbitrary vector Q. The time derivative of Q measured
relative to the inertial frame XY Z is denoted
The time derivative of Q measured relative to the rotating coordinate frame

(the principal frame) S is denoted
Suppose Q is constant in the principal system. This is for instance the case
just considered, Q = r. We find
If Q varies with time, even relative to the principal frame S, we evidently

have
13.3 Euler's Equations 329
In operator form we thus have the following connection between differentia-

tion with respect to time relative to the two coordinate frames I and S:
Note: for the rotation vector w we have
The same result is true also for position vectors for points on the line
containing the instantaneous rotation vector w.
We are now able to derive the Euler equations for a rigid body rotating
about a point fixed in an inertial frame. We shall use the principal coordinate
frame S, rotating with the body. In the coordinate frame S the components
of the inertia tensor are constant (i.e., independent of time).
The angular momentum theorem relative to the inertial frame I (the lab-
oratory) is
( dLO) =N .
dt 0
I
From the operator equation given above the angular momentum theorem can
be written in the form
( d~o ) s + w x Lo = No,
where the vectors are taken to be represented by their components in the
principal coordinate frame S.
The advantage of this method is that in the coordinate frame S the com-
ponents of the inertia tensor are independent of time. This means that
( ~o ) s = I ( ~~ ) s == Iw .
The vector equation of motion, relative to the frame S, becomes
Iw + (w x L) = N,
with L = (Ixwx, Iywy, Izw z ), where Ix is the moment of inertia about the
x-axis in the principal coordinate frame S, and analogously for Iy and I z .
Written in components along the axis in the principal frame S the equa-
tions are
Ixwx - (Iy - Iz)wywz = N x ,
Iywy - (Iz - Ix)wzw x = Ny ,
Izw z - (Ix - Iy)wxwy = N z .
In this form the equations of motion are known as the Euler equations.
It is important to be completely aware of the following point: Euler's
equations are a set of coupled differential equations that describe how the
rotation vector moves relative to the principal frame S. The rotation vec-
tor describes the instantaneous motion of the body relative to the inertial
frame (the lab frame). Determining w(t) for a rotating body corresponds to
determining v = v(t) for the motion of a particle (CM).
At the risk of repeating ourselves, but to avoid misunderstandings:
Euler equations describe the motion of the body relative to

the inertial frame 1. The vectors that describe the motion
in I are represented by their components in the principal
(body) frame.
Note that Euler's equations for a homogeneous sphere rotating about CM

has the simple form:
I dwx Nx ,
dt
I dwy Ny,
dt
I dwz Nz ·
dt
In this very special case the equations have the same form in the inertial
frame as in any coordinate frame fixed in the body. In the description of the
rotation of a homogeneous sphere there is no need to introduce a coordinate
frame fixed relative to the sphere, since the inertia tensor is isotropic ( Ix =
Iy = Iz = (2/5)M R2) in both cases.
13.4 Kinetic Energy
Sometimes it is useful to use a principal coordinate system when the problem

is to calculate the kinetic energy T of a rotating rigid body, measured relative
to the laboratory frame. The kinetic energy is:
T l = l:l-m·v·
l: -m·v·
2"
2
2"
. (w x r·)
,
i i
1
-w·
2 """'
~,
r· x m·v·
"
i
1
-w·L
2
1
-w·lw.
2
13.5 Determination of the Principal Coordinate System 331
Vectors and tensors may be represented by their components in an arbi-

trary coordinate system.
Let us choose to express both wand I in a principal coordinate system:
or
T = 2'1 (Ixwx
2 + Iywy2 + Izw z2)
Again, T is the kinetic energy measured relative to the inertial system (the
laboratory), but expressed through components of wand components of I
relative to the principal coordinate system (which is fixed in the body).
13.5 Determination of the Principal Coordinate System
There is a systematic way to determine the principal coordinate system for

a body with a given point 0 as its origin. This procedure may be found in
books on geometry or in textbooks on analytical mechanics. In many cases,
however, it is possible to avoid the rather complicated formal calculations,
by making use of the symmetry properties of the bodies.
For example, if the fixed point 0 is on a symmetry axis of the homoge-
neous body, then that axis of symmetry is a principal axis for the point O.
Suppose the z-axis coincides with the symmetry axis. For the mass density
we then know that
p(x, y, z) = p( -x, -y, z).
This leads to
-J xzpdV = 0,
-J yzpdV = O.
This shows that the z-axis is principal.

Consider a homogeneous cylinder of radius R. The length of the cylinder
is L. Let the fixed point 0 be the CM of the cylinder (see the sketch below).
The symmetry axis (the length axis) is chosen as the z-axis. The other two
axes lie in a plane perpendicular to the z-axis, and pass through O. This
coordinate system is a principal system:
Ixy - J xypdV = 0,
Ixz Iyz = o.
~---:l--_y
The moment of inertia around the x-axis is obtained as follows. Consider a

thin disk as suggested on the sketch. The moment of inertia about a diameter
of this disk is (1/4)mR2, where m is the mass of the disk. The cylinder may
be perceived as a rod made up of many such disks. The total moment of
inertia about the x-axis becomes
I
xx
= ~MR2
4 + ~ML2.
12
We have used the parallel axis theorem for calculating the moment of
inertia about the x-axis. Furthermore,
Iyy
The tensor of inertia is diagonal with the three given principal moments of
inertia. Note Ixx = Iyy; we may choose the principal directions as any two
mutually perpendicular directions in the plane that is perpendicular to the
z-axis through O.
Example 13.5. Rotating Dumbbell

We return to the study of a dumbbell rotating about a fixed axis A (see Ex-
amples 13.1 and 13.4.). The system now consists oftwo homogeneous spheres,
both of mass m and radius r. This means that we have to take into account
the finite size of the spheres. The spheres are connected by a massless rod
R. The distance between the spheres is 2a. The length of the axis A is 21,
and the axis rotates without friction in the bearings Band C. The angular
velocity is w.
The motion is thus given, and the goal is to determine the torques nec-
essary to sustain the prescribed motion. In this example we shall solve the
problem using a principal coordinate system. The y-axis coincides with the
rod R and has its origin in the CM. The x-axis is perpendicular to Rand
is in the plane spanned by Rand w. In the instantaneous picture displayed
in the above figure, the z-axis is pointing out of the paper. The described
coordinate system is fixed relative to, and rotates together with, the body.
The x-axis describes a cone with opening angle 90 0 - e, the y-axis describes
a cone with opening angle e, and the z-axis describes a plane perpendicular
to w (i.e., a cone with opening angle 90 0 ).
We express the inertia tensor I by its components in the chosen principal
coordinate system:
2
2ma 2 + 2 x '5mr2
2
2 x '5mr2
2
2ma 2 + 2 x '5mr2 .
All the products of inertia are zero. The rotation vector w, has the fol-
lowing components in the principal frame: w = (wsine,wcose,Q). The com-
ponents of the inertia tensor are of course time independent in the principal
frame. In the specific case we consider here, the components of the rotation
vector ware also time independent in the principal frame.
The components in the principal frame of the angular momentum vector
become
e1 .
1[ wcose
w sin
a
,
or
Lx (2ma 2 + ~mr2) wsine,
Ly (~mr2) wcose
Lz o.
Note: L is not parallel to w. The angular momentum vector L is constant

in the principal system, i.e., L rotates with angular frequency w relative to
the laboratory frame. By employing Euler's equations we can determine the
torques necessary to maintain the prescribed motion:
Ixwx - (Iy - I z ) wyw z Nx ,
Iywy - (Iz - Ix) WzW x Ny,
Izw z - (Ix - Iy) WxWy Nz ·
Remember, the Euler equations describe how the rotation vector wanders
relative to the principal frame, which is at rest relative to the body. In our
case, W is time independent and Wz = O. Therefore, N x = Ny = 0 and
N z = -2ma 2 w 2 sinBcosB.
The torque derives from the reaction forces in the bearings. Note: N z
always points in the negative z-direction, and the torque thus turns along
with the principal coordinate system.
Note: the CM of the system (located in the point 0) remains at rest.
Describe all the external forces acting on the system. 6.
Example 13.6. Flywheel. Consider again a flywheel mounted out of align-

ment (see Example 13.2). The flywheel is assumed to be rotating without
friction and the goal is to determine the torques necessary to maintain the
motion.
We want to determine the angular momentum of the wheel relative to the

laboratory frame. We represent the angular momentum vector by its com-
ponents in a principal coordinate frame . The z-axis is always the symmetry
axis of the wheel (a disk). The origin is in the center of the wheel (see the
figure). The other two axes lie in the plane of the disk, e.g., in such a way
that the x-axis in the instant depicted is in the plane of the paper and point-
ing upwards. The x-axis is in the plane spanned by wand the z-axis. The
y-axis is then pointing out of the plane of the paper. Make sure that you
understand how this coordinate frame, at rest relative to the moving wheel,
moves relative to the laboratory frame.
The components of the tensor of inertia in the principal system are
The products of inertia are all zero in this frame.

The components of the angular velocity are w = (-wsinB,O,wcosB). We
obtain:
°
.!.MR2
1[ 1
-wsinB
4
~;;R2 wc~sB
° .
The vectors Land w do not have the same direction. The angle between L
and w is
L· w 1 + cos 2 B
cos(o:) = ILllwl = J1 +3cos2B
Question. Show by means of Euler's equations that the torque necessary to

maintain the prescribed motion of the wheel is
0,
1 .
4MR2 w 2 cosBsmB,
0.
Example 13.7. The Gyroscope. In Example 13.3 we discussed the regular

precession of a simple gyroscope: a rotating circular disk of radius r at the
end of a light rod.
p
O~-------l
x
Assume that the disk is rotating with a high angular velocity woo We
say that the disk has a high spin. Assume furthermore that the point P is
supported by a hand. By means of the hand we now lead P around in a
uniform circular motion about 0 (into the paper). The angular velocity in
this circular motion is chosen to be flo = (2gl)/(r2wO)'
If P is led with this angular velocity we may remove the supporting hand,
and the gyroscope will continue with the regular precession. The initial condi-
tions are just right for the regular precession. We have, so to speak, by means
of the hand given the gyroscope a vertical component of angular momentum.
There are no external torques in this direction, and the vertical component
of angular momentum therefore remains constant. The vertical component of
the angular momentum that we give the system matches exactly the value
required for regular precession under the influence of gravity.
If we start the gyroscope with other initial conditions the motion will be
more complicated.
Let us assume that we keep the figure axis horizontal with a high spin
(w) of the disk. Suddenly we release the figure axis, without having given it
any initial velocity in the lab frame. The eM of the system then begins to
descend, Le., dL/dt is directed downwards. The angular momentum theorem
with respect to the fixed point 0:
dL
ill = No,
must be satisfied at any moment. A downward pointing dL/dt would demand
a torque in 0 with a component pointing down. Such a torque does not exist.
The only forces acting on the gyroscope are gravity and the reaction in O. If
the suspension is really at a point (0) the reaction in 0 cannot have a torque
about O.
As soon as we release the figure axis, the gyroscope starts its "fall", with
a motion in the opposite direction of the "missing force", Le., in a direction
corresponding to the direction of the regular precession. The endpoint of the
figure axis will - as long as we may neglect friction - describe a cycloid.
The figure axis starts a complicated "nodding" motion superimposed on

a regular precession about O. The "nodding" motion is called nutation. If
the spin rotation velocity is very large, the nutation will die out quickly due
to friction in O. The regular precession remains.
When the nutation motions have been damped out, the figure axis will
not be exactly horizontal. The angular momentum from the uniform circular
motion will point vertically upwards. Therefore the angular momentum for
the spin-motion will point a little downwards, in such a way that the vertical
component of the angular momentum in 0 remains zero.
This component was zero to begin with, and it cannot change as there
are no external torques with components in the vertical direction.
We proceed to determine the general expression for the total angular
momentum of the gyroscope. The spin velocity is w. The radius and mass of
the disk, respectively, are rand M. The distance from the fixed point 0 is 1.
The point 0 is thus the fixed point around which the rigid body moves (see
the sketch below).
=~;,------;---t.. Y
x
Let the XYZ system be the laboratory frame (an inertial frame). We
might have chosen a "genuine" principal coordinate system, rotating with
the disk, in such a way that the body was at rest in that frame. We shall,
however, choose another coordinate system in which - due to the high degree
of symmetry of the body - the tensor of inertia is also diagonal with time-
independent components.
The coordinate axis xyz are chosen so that this coordinate frame follows
the gyroscope in its precession, rather than in the spin of the disk. The z-axis
lies along the figure axis of the gyroscope. The figure axis forms the angle B
with the vertical, i.e., with the Z-axis.
The x-axis lies in the XY plane (always) and the y-axis is tilted the angle
B beneath the XY-plane. The projection of the eM of the disk on the XY-
plane is called P. The angle from the X-axis to OP is called cp, the same angle
that the x-axis forms with the negative direction of the Y-axis.
The orientation of the figure axes of the gyroscope is given by the two
angles () and cpo One can determine the placement of the xyz system starting
from the XY Z system, as follows:
1: place the z-axis along the Z-axis, the x-axis along the (-Y)-axis and the
y-axis along the X-axis;
2: turn the xyz system the angle cp about the Z-axis and then an angle ()
about the new x-axis.
We shall now determine (1) the components of the total rotation vector
nand (2) the components of the tensor of inertia, with all components in
the xyz-system.
First, n, the total rotation vector. The disk rotates with an angular ro-
tation vector w about the z-axis. The total rotation vector also includes,
however, contributions from () and cpo
Let i, j, and k be unit vectors along the axes of the xyz system. A unit
vector along the Z-axis is named e z . We then have
n = -Oi + cpe z + wk .
For e z we have
e z = - sin()j + cos()k.
The rotation vector in the xyz system becomes
n = -Oi - cp sin ()j + (cp cos () + w) k,
or
n= (-O,-cpsin(),cpcos()+w) .
Note. Even if the coordinate frame used is not fixed relative to the body, the
components of the tensor of inertia are time independent. The reason for this
is the high degree of symmetry of the body. The disk is at all times parallel
to the xy-plane, and the moment of inertia about any diameter of the disk is
(l/4)Mr2.
The components of the angular momentum in the xyz coordinate system
are now determined:
1[ cp -cp-0 + w 1
IMr2
4
+Ml2 o
o o sin () ,
o cos ()
or
Lx (~Mr2+Ml2) (-0),
Ly (~Mr2+Ml2)(-cpsin()),
Lz (~Mr2) (cp cos () + w).
We cannot use the angular momentum theorem in its simple form in

the xyz frame because this frame is not inertial. We cannot use the Euler
equations directly, because the xyz frame is not fixed relative to the rotating
body, and the xyz system is thus not a "genuine" principal coordinate frame.
The problem of the general treatment of the motion of the gyroscope
is fairly complicated. For a thorough treatment of the subject you should
consult the book by Goldstein (see the literature list).
Here we shall limit ourselves to pointing out that the expressions for n
and L that we have derived contain as a special case the regular precession.
We shall study a motion where () is constant, but not necessarily equal to
1r /2.
There are no torques with components along the Z-axis; thus cj; is con-
stant. There are no torques with components along the z-axis; thus w is
constant.
Since () is assumed constant, Le., iJ = 0, there is no nutation in the motion.
Even with this assumption we see from the general expression for L that the
expression for the angular momentum is still complicated:
L= (~MR2+MI2) (-cj;sin()j)+~Mr2(cj;cos()+w)(k).
We now make the assumption that cj; « w, Le., that the spin-rotation of
the disk is very much faster that the precessional motion.
With these - very simplifying - assumptions, the angular momentum
becomes
1
L:::::: 2Mr2(wk).
The angular velocity in the precession is
Here L is a constant in the precessing coordinate frame. We evaluate the time

derivative of L, measured in the inertial frame:
dL . 1
dt
n x L = cpe z x 2Mr2wk
1
2M r 2 wcj; sin ()( -i) .
Gravity, acting on the eM, provides the torque around 0 necessary to

sustain the precessional motion. This torque is likewise pointing in the di-
rection of the negative x-axis. The equation of motion in the inertial frame
becomes
~ M r 2wcj; sin () = M gl sin () .
The angular velocity in the precession is
. 2MgI 2g1
cp=--=-
Mr 2w r 2w·
The same results are obtained in Example 13.3. The angular precession
velocity is independent of ().
The general treatment of the gyroscope is of significance both in theoreti-
cal physics and in the applied sciences. Electrons and atomic nuclei with spin
and charge perform precessional motion in an external magnetic field. This
is the basis for important techniques in the study of the structure of matter.
Inertial navigation is also based on a detailed understanding of the motion
of gyroscopes. 6
Example 13.8. Gyroscope Supported at the Center of Mass. A gy-

roscope is provided with a counterweight W such that the gyroscope is sup-
ported in its center of mass. The disk D with the mass M and radius T is
spinning around the axis A with a large angular velocity. Since the gyroscope
is supported in its center of mass there is no external torque on the system
and the gyroscope will not precess.
+
r
A eM
L
w eM
We now give the axis a tap (a light blow) with a hammer, at distance l
Jo
from the CM. The tap may be described as an impulse P = Llt F(t)dt. We
assume that the force F(t) that the hammer exerts on A during this blow is
parallel to the gravitational field.
What effect does the tap have on the gyroscope axis A? First, note the
following. If the spin of the gyroscope is very high, the axis A (the figure
axis) will hardly move during the blow and the small motion observed will be
perpendicular to the direction of the blow. On the sketch above the motion
will be into the paper.
In reality the axis A moves slightly downwards in the direction of the
blow. The axis moves on a circular cone the axis of which is nearly parallel to
the axis shown in the sketch, but points slightly into the paper (see sketch).
If no friction was present in the support (CM) the axis A would continue to
describe a cone with a small opening angle. Due to friction in the bearings,
however, the opening angle will become smaller and smaller and after a few
revolutions the figure axis A will come to rest along the symmetry axis of the
cone. The figure axis has thus experienced a slight net turn into the paper,
i.e. , in the direction perpendicular to the direction of the blow. This is the
permanent effect of a light blow on a supported gyroscope with a high value

of the spin rotation.
During the blow the gyroscope is subject to an external torque NCM =
le x F(t) with respect to the point of support. Here e is a unit vector along A.
NCM(t) is horizontal and pointing into the paper. During the blow we thus
have
dLcM
~ = NCM(t).
For the change L1LCM during the tap we find

r.6.t
L1LCM = io NcM(t)dt.
As N CM points into the paper, L1LCM is in this direction too. The value of
L1LCM is
IL1LcMI = fo.6.t llF(t) I dt = Pl.
The axis moves perpendicular to the direction of the tap.
Note. A gyroscope mounted in such a way that it can turn about its eM is
useful as a directional stabiliser. This is not because it does not turn when a
torque is applied to it, but because it stops turning when the torque ceases. If
a large nonrotating mass is mounted so that it is free to turn around its center
of mass, it would acquire and maintain a small angular velocity if subjected
to an external torque of a short duration. 6
Example 13.9. The Earth as a Gyroscope
Sun
The sketch shows the Earth at the northern summer solstice (R:: 21st of June).
In its orbit around the Sun, the Earth is - at the instant shown - on its way
out of the plane of the paper. The axis of rotation points towards the North
Star. It may be useful to emphasize that the North Star is so far away that
no change in the direction of the axis of rotation relative to the North Star
can be detected throughout a year.
As a result of the fact that the Earth is an ellipsoid of revolution, combined
with the fact that the axis of rotation is tilted relative to the plane of the
ecliptica, there is an external torque acting on the Earth. The torque is due
to the tidal field
Consider first the torque N from the Sun. N is perpendicular to the plane
of the paper and points out of the paper. Due to the fact that the Earth has
an angular momentum relative to the heliocentric reference frame (inertial
space!!), the rotation axis (Le., L) will execute a precessional motion in a
direction opposite to the direction in which the Earth moves around the Sun.
We say that the rotation axis has retrograde motion.
The torque from the Sun varies according to the position of the Earth
in its orbit. The torque is zero at each equinox (March and September) and
maximum at solstice (June, December). The precession of the rotation is then
- as mentioned - due to the tidal fields at the position of the Earth. The tides
from the Moon is about 2! times stronger than the tidal fields from the Sun;
the Moon therefore has the strongest influence on the precession of the axis
of the Earth. Almost 35", ofthe yearly precession of 50", is due to the Moon.
The axis of the Earth describes a cone with an opening angle of 23! o. A
complete revolution takes about 25800 years. Today, the axis of the Earth
points towards the North Star. In about 12 000 years the rotation axis will
point towards the star Vega in the constellation The Lyre.
The points where the instantaneous axis of rotation cuts the celestial
sphere is called the pole of the heavens. That these poles move relative to
the fixed stars was known to Hipparchus in about the year 120 BC. The
precession of the rotation axis of the Earth was first explained by Newton.
6
13.6 Problems
Problem 13.1.
"'--1_
o
tr
A
Let us assume that the support C in Example 13.3 is designed in such a way
that it has exactly the height r. Let us furthermore assume that the disk D
can roll without sliding on the horizontal surface. We still denote by Wo the
(now much smaller) angular velocity about A. Since the disk rolls without
sliding on the horizontal surface, this now determines the "angular velocity
of precession" no, in such a way that no = r(wo/l).
13.6 Problems 343
Find the force R by which the surface must act on the disk D for this
motion to be maintained.
Problem 13.2. Consider again the gyroscope supported in its CM (see Ex-
ample 13.8).
+ D • t
h
r
A eM
L
+
w eM
Let us assume that a particle with mass m falls freely through the height
h and hits the axis A in a completely inelastic collision, but in such a way
that the particle "falls off A when the collision is over" .
The particle hits the axis A at a distance l from the CM. We neglect the
effects of gravity during the collision.
(1) Find the angle .6.8 that the axis A turns (into the paper) if: M = 2.4 kg,
l = 0.30 m, Wo = 200 s-1, m = 0.10 kg, h = 0.50 m, r = 0.10 m.
(2) What condition should be satisfied by the collision time .6.t if it may be
considered a good approximation to neglect gravity during the collision?
Problem 13.3.
,gT
t
B
~~----------a----------~~
A C
¢::::=f::===:=:.===;:==o======~
5 I
D
.... r
~
A gyro-wheel D (a disk) is fastened to the end of a stiff rod S. The wheel D

is perpendicular to S. The rod S has length l and we neglect the mass of the
rod. The wheel D rotates with constant angular velocity Wo (see the figure).
The wheel D is a homogeneous thin circular disk of mass m and radius r.
The rod S goes through the center of D.
The other end of S is fastened to the mid-point 0 of a thin, rigid horizontal
axis A, which can turn without friction in two bearings, Band C, that are
fixed in the laboratory. The axle A has length a, and its mass can be ignored
in this problem.
The rod S may thus turn in a vertical plane perpendicular to the horizontal
axis A. Initially, S is held horizontal, and at rest. At a certain time the rod
is released and S begins to turn in the vertical plane through O. The initial
angular velocity of the rod is thus zero.
Neglect all frictional effects (in particular, the disk D maintains its angular
velocity wo).
(1) Find the angular velocity n of the rod S at the moment the disk D passes
through its lowest position (where S is vertical).
(2) Find the forces FB and Fe by which the axle bearing Band C respectively
act on the axle A, when S passes through its lowest position (the angular
velocity vector n of the rod is in the direction OC) .
Problem 13.4.
A wheel W is shaped as a homogeneous, fiat, circular disk. The wheel is

fastened to a thin axle A that passes through the center of the wheel. We
ignore the mass of A.
The axle A rests in two bearings in such a way that the system can rotate
without friction around a horizontal axis. Note that A is supported only from
below! The radius R of the disk is R = 50 em, and the distance between the
two bearings is 1 m. The CM of the disk is exactly at the mid-point between
the two bearings.
The wheel is fastened to the axle A in such a way that the normal to the
disk of the wheel forms an angle of e = 10 with the axle A. The acceleration
of gravity is 9 = 9.8 ms- 2 . The wheel is now set in motion, performing v
revolutions per second.
Determine the maximum value v = Vo for the rotation frequency if the
axle A must not leave the supporting bearings.
14. The Motion of the Planets
We shall now apply Newtonian mechanics to the study of the motion of

planets.
Look at the sky on a dark, clear night. Seen from the Earth, the stars and
the planets appear to be fastened inside a huge sphere, called the celestial
sphere. The celestial sphere appears to rotate, once per day, around an axis
passing through the north pole and the south pole of the Earth.
The stars are far away from the Earth. Therefore the stars seem to have
fixed positions on the celestial sphere. The stars do not change positions
relative to one another over periods of time comparable to a human life time.
The planets, on the other hand, are rather close to us. Even though from
one night to the next, a planet appears to be fastened on the celestial sphere,
over periods of weeks the line of sight, even to an outer planet, does change
noticeably relative to the stars.
The roots of modern science are found in the study of the motion of the
planets. Therefore we begin with some remarks of a historical nature.
14.1 Tycho Brahe

On the island of Hven, between Sweden and Denmark, Brahe built, in the
years 1576-1597, an astronomical research institution of historical signifi-
cance. Brahe understood the significance of precise astronomical observa-
tions, and he had the means to construct the necessary instruments. Before
the time of Brahe, the positions of the heavenly bodies were known with the
precision of 10 minutes of arc (10') . Brahe improved the precision significantly
and reached an accuracy of 1 to 2 minutes of are, which is close to the limit
of precision obtainable with the unaided eye.
The years 1576-1597, in which Brahe performed his measurements, will
be remembered as one of the most decisive periods in the history of science,
and indeed in the history of man. This is impressive, especially when we
recall that the naked-eye observational methods of Brahe were deemed to
become obsolete after half a score of years. In 1609, Galileo pointed a telescope
towards the sky, and science was changed forever.
Nevertheless Brahe's results, particularly from the observation of the
planet Mars, became of fundamental significance. As a kind of afterthought

346 14. The Motion of the Planets
one might add that if the observations of the planets made by Brahe had been
even more precise, if these observations had disclosed the "irregularities" in
for instance the motion of Mars (irregularities caused by gravitational fields
from the other planets), Kepler might not have uncovered his three simple
laws for the orbits of the planets. The gravitational law of Newton might
have been more difficult to find, and the history of man had changed. Such
speculations may be considered entertaining, but are not particularly use-
ful. Seen from the viewpoint of physics the remarks merely illustrate that
it is important to find the essential aspects of experimental or observational
material.
14.2 Kepler and the Orbit of Mars
Contrary to Brahe, Kepler initially accepted the heliocentric model of the

solar system, as described by N. Copernicus. The Sun is at rest in the center
of the system, with the planets moving in circular orbits around the Sun.
Originally Kepler looked for a connection between forces and the structure
of the solar system. He realized very soon that the periods of the planets
increase with the distance of the planet from the Sun. It was Kepler's belief
that the increase in the periods of the planets was connected to a force from
the Sun, a force that decreased with distance.
Kepler did not succeed in connecting the motions in the solar system to
the concept of force. According to historians of science, he had a quite clear
understanding of the importance of this task. The title of Kepler's principal
book (published in the year 1609) was:
A New Astronomy Based on Causation

or
A Physics of the Sky
Derived from Investigations of the Motions of the Star Mars.
Founded on Observations made by the Nobleman Tycho Brahe.
If Kepler did not succeed in explaining the dynamics of planetary mo-

tions, he did succeed in using the empirical material of Brahe in a masterly
way. Kepler condensed this material into an elegant form, a form that be-
came decisive for the work of Newton. Through a nearly superhuman effort
of calculation, Kepler succeeded in showing that planets move around the
Sun, not in circular orbits with the Sun in the center, but in elliptical orbits
with the Sun in one focus.
The difficult problem confronting Kepler was to determine the orbit of a
planet relative to the Sun, based solely on observed directions to the planet,
as seen from a rotating Earth, itself following an unknown curve on its way
around the Sun. Kepler set out to determine the orbit of Mars.
14.2 Kepler and the Orbit of Mars 347
14.2.1 The Length of a Martian Year
The time needed for a planet to complete one revolution around the Sun is
called the sidereal period. The sidereal period for a planet cannot be directly
observed, but it can be determined as shown below. Assume that the orbits
of the Earth (E) and Mars (M) are circles with the Sun (S) in the center.
The planet Mars is said to be in opposition, when the Sun, the Earth, and
Mars lie on a straight line, and Mars is closer to the Earth than to the Sun
(see the figure).
The time between two successive oppositions is called the synodic period
of the planet. The synodic period can be determined by observation. When
the synodic period is known the sidereal period may be calculated as follows.
Let the sidereal period of Mars be T and the synodic period S. The
sidereal period of the Earth - i.e., one year - is called A. The quantities T,
S, and A are the respective periods measured in days (1 day = 24 h).
Mars is an outer planet relative to Earth. In the time between two suc-
cessive oppositions, i.e., during one synodic period, Mars moves 3600 less
relative to the Sun than the Earth moves in the same time.
In one sidereal period Mars moves 360 0 relative to the Sun. In one day
Mars moves 360 0 jT relative to the Sun.
In one synodic period Mars moves (SjT)360° relative to the Sun.
During one synodic period for Mars, the Earth moves (SjA)360° relative
to the Sun.
The equation to determine T is thus:
~ 360 = ~ 360
0 0 - 3600
T A '
or
1 1 1
T A-B'
From Brahes measurements Kepler knew that the synodic period for Mars
- i.e., the time between two successive oppositions - was S = 779.8 days.
Consequently
1 1 1 -1
T= 365.24 - 779.8 = 0.0014555 days ,
T = 687 days.
To trace its orbit around the Sun, Mars thus needs 687 days = 1.88 years.
Kepler actually started by determining the orbit of the Earth. To do this
he used his knowledge about the sidereal period of Mars (687 days) to identify
the dates on which Mars was back in a given point in its orbit.
,,
,,
/42.4°
S .-------
' ......
......
......
Fig. 14.1.
We shall show only the principles used by Kepler. Consider Figure 14.1. Let
the point M mark an opposition of the planet Mars. It will take Mars 687
days to return to the point M in the orbit. During 687 days the Earth has
completed 687 /365 ~ 1.88 revolutions around the Sun. The Earth has thus
moved 1.882 x 360 0 = 677.6 0 in its orbit, or 42.4 0 less than two complete
revolutions. The Earth will then be located in the point E 1 , as shown on
Figure 14.1, i.e., 42.4 0 "behind" Mars. After an additional 687 days Mars
will again be in the point M, while the Earth will be in the point marked E2
on Figure 14.1 (assuming circular orbits).
From each successive complete revolutions of the planet Mars, Kepler was
able to find one point of the orbit of the Earth.
Brahe had observed Mars for more than 20 years. The observations in-
cluded ten oppositions of Mars.
Using the method outlined above, Kepler was able to construct the orbit
of the Earth relative to the Sun (relative to the heliocentric reference frame!).
Kepler found that, within the precision of observation, the orbit of the Earth
was a circle, but with the essential feature that the Sun was not located at
the center of the circle.
By plotting the position of the Earth at various dates, Kepler discovered
that the Earth does not move with the same speed all year round. The Earth
moves faster when it is closer to the Sun. Here is the beginning of the discovery
that later became known as Kepler's second law:
The radius vector from the Sun to the planet sweeps out
equal areas in equal amounts of time.
As we have already seen (Chapter 10), this law is a direct consequence of

the conservation of angular momentum in a central field of force.
Now Kepler knew the orbit of the Earth. The next problem was - based
on the observations of Brahe and on the knowledge of the orbit of the Earth
- to find the orbit of Mars relative to the Sun.
14.2.2 The Orbit of the Planet Mars
Kepler utilized the fact that he knew the length of a Martian year (sidereal
period, 687 days).
Fig. 14.2. Kepler's determination of the orbit of the planet Mars. (SEoMo) is one
opposition of Mars and (SE2Ml) is another
Kepler again used the oppositions of Mars. Consider the opposition

marked by the line SEoMo in Figure 14.2. The line of sight to Mars rela-
tive to the stars was known from the measurements of Brahe. When Mars,
687 days later, again is in the point marked Mo, the planet is not in op-
position, because the Earth will now be in the point marked E!, Le., 42.4°
"behind" Mars. Kepler knew the orbit of the Earth, and the position of the
Earth at any given date was now also known. That means: Kepler knew the
date on which the Earth was in the point of its orbit marked E 1 . The posi-
tion of Mars, Le. the line of sight to Mars on the day the Earth was in El
could be found from the tables of Brahe. Kepler was able to calculate the
angle SElMo. Using the distance from the Sun to the Earth as unit of length,
Kepler found one point in the orbit of Mars by triangulation.
Using the ten oppositions studied by Brahe, Kepler was able to determine
ten points in the orbit of Mars. Kepler tried to find a circle passing through
those ten points. He found for some of the points the now famous deviation
of 8' (minutes of arc) from the circle which gave the best fit (see Chapter 1) .
It is interesting to note that before Brahe the positions of the planets

were known with an accuracy of 10'. The improvement of the precision of
observation to about 2' was therefore decisive. If - before Brahe - one had
tried to fit a circle to the observation points it would have been considered
successful.
Kepler had confidence in the data of Brahe, and he took the decisive step.
The hypothesis of circular orbits must be rejected. Instead of conserving "the
ideal circular orbits" Kepler chose to believe in observations.
The author Stefan Zweig has written a book with the title Sternstunden
der Menschheit. Here is a Sternstunde in our history, far more important
than any described by Stefan Zweig. A description of Kepler's achievements
cannot be found in the book by Stefan Zweig.
Kepler found that the ten points of the orbit of Mars could be fitted to
a curve known to the mathematicians for a long time: an ellipse. Johannes
Kepler had thus reached the law that has become known as Kepler's first
law:
The orbit of a planet relative to the Sun lies in a fixed plane
containing the Sun, and each planet moves around the Sun
in an elliptical orbit with the Sun in one focus.
Through the work of Brahe and Kepler the solar system had disclosed
one of its deepest secrets.
The study of the solar system has resulted in several other decisive ad-
vances in physics: the law of gravity, the finite velocity of light (the "lingering
of light" , Ole Rpmer), and the rotation of the perihelion of the elliptical orbit
of the planet Mercury. Somewhere in the solar system - perhaps on Mars -
we may find the key to the greatest riddle of the natural sciences: the origin
of life itself.
The reason why it was not possible to fit the points of observation of Mars
into a circular orbit around the Sun is the substantial eccentricity ("flatness")
of the Martian ellipse. For the precise definition of eccentricity, see below. The
eccentricity of the elliptical orbit of Mars is e = 0.09, which is five times larger
than the eccentricity of the elliptical orbit of the Earth, and more than twelve
times the eccentricity of the orbit of Venus. It is, however, important to note
that even for Mars the deviation from the circular form is small.
Kepler's third law was published - among several more obscure results -
in the year 1619:
The square of the period of revolution of a planet is pro-

portional to the third power of the greatest semi axis of the
ellipse.
Specifically, ifT denotes the period and a the major semi axis, T2 ja 3 = C,
where C is a constant, the same for all planets.
14.2.3 Determination of Absolute Distance in the Solar System
The sidereal period of revolution Tp for a planet may be determined via the
observation of the synodic period. From Kepler's third law, the semi-major
axis a p for the planetary orbit (ellipse, see below) may then be found using
the astronomical unit as the basic measure of distance. One astronomical
unit (1 AU) is defined as the mean value of the distance of the Earth from
the Sun. Measuring Tp in years we have
T2 r,2 12
~=~=-=1
ap3 a 3E 13 .
The absolute distances in the solar system, i.e., distances measured in meters,
can be found only when one distance - for instance the distance from Earth
to Venus, or from Earth to Mars - has been determined.
The problem has not been simple to solve. Today one can measure the
distance say, from the Earth to Venus, with high precision by means of radar
signals reflected from the surface of Venus.
Historically the problem was first solved by triangulation. From two points
on the Earth, a large distance apart, the direction of the line of sight to a
planet is measured. The two directions of the line of sight will then form
a certain angle, which is larger the closer the planet is to the Earth. The
difficulties with this measurement is obviously the small value of the angle
between the lines of sight. The angle between two lines of sight from the
Earth to the Moon may be about 10 . The angle between two lines of sight
from the Earth to even the nearest planets will never be more than I' (1 arc
minute).
Mars is closest to the Earth when in opposition. In the most favorable
oppositions Mars is 0.37 AU from the Earth. Venus may come even closer
(0.26 AU). When Mars is in opposition the illuminated half sphere of Mars
is facing the Earth. Therefore Mars is easy to observe during an opposition.
The orbit of Venus lies within the orbit of the Earth. Therefore when Venus
is closest to the Earth, Venus will have its dark side facing the Earth. Venus
is therefore impossible to observe when it is closest to Earth unless the planet
passes in front of the solar disk. Venus will then be observable as a small dark
spot against the large luminous disk of the Sun. This phenomenon is called
a transit of Venus. The orbits of the Earth and the orbit of Venus lie nearly
in the same plane, but not exactly so. As a rule Venus will bypass the Sun.
A Venus transit is a rare phenomenon. They come in pairs. There were two
in the 19th century (1874 and 1882), and the next pair is 2004 and 2012.
From the first two of the mentioned Venus transits a triangulation mea-
surement was made. Based on this the AU was estimated to be between
147 x 106 km and 149 x 106 km.
Even better than Venus for the determination of the astronomical unit is
the asteroid Eros, which was discovered in 1898. The distance from Earth to
Eros may be as small as 0.15 AU. A favorable opposition of Eros took place
in 1930. At this opposition the astronomical unit was determined as 149.7 x
106 km. The present value is 1 AU = 149.598 X 106 km.
Kepler published his laws as unexplained facts. The full dynamical con-
sequences of these laws were recognized by Newton, after he had formu-
lated his general laws of motion. We shall show that gravitational attraction,
i.e., Newton's law of gravity is implied by Kepler's laws. After this we shall
demonstrate the converse: Kepler's three laws are consequences of Newtonian
mechanics and the law of gravity.
Before proceeding, we shall briefly review some results related to the ge-
ometry of conic sections.
14.3 Conic Sections

Detailed descriptions of conic sections may be found in books on geometry
or calculus. Here we give a rudimentary introduction.
The curves obtained by intersecting a cone with a plane which does not
pass through the vertex of the cone, are called conic sections. If the plane
a b c
Fig. 14.3. Conic sections obtained by intersecting a cone with a plane: (a) parabola,
(b) ellipse, (c) hyperbola
intersecting the cone is parallel to a generator of the cone (Figure 14.3a), the
conic section becomes a parabola. Otherwise the curve produced is called an
ellipse or a hyperbola, depending on whether the plane intersects one portion
of the cone (Figure 14.3b) or both portions (Figure 14.3c). A circle is a special
case of an ellipse.
The three types of nondegenerate conic sections may be characterized
within the plane, in the following manner. A conic section is the set of all
points P for which the distance from a fixed point F (the focus) and a fixed
line 1, has a constant ratio e, called the eccentricity:
14.3 Conic Sections 353
PF
PI = e, (14.1)
where PF is the distance from P to F and PI is the distance from P to 1. The

line 1, which is called the directrix, does not pass through F.
Fig. 14.4. The equation for conic sections using polar coordinates
Let 2p be the length of the chord perpendicular to the axis of the conic
section and passing through the focus F. By choosing the point P at the
endpoint of the chord, e.g., at the point A (see Figure 14.4) the defining
equation becomes
p = e(Fl). (14.2)
The quantity p is called the parameter of the conic section.
Measuring the angle from the symmetry axis as in the figure and using
(14.1) and (14.2) we find
r = FP = e(Pl) = e(Fl- r cos 8) ,
r =e (~ - r cos B) = p - er cos 8 .
Finally we get
r - p (14.3)
- 1 + ecos8 .
The expression (14.3) is the equation for a conic section for all three cases.
We get an ellipse for 0 ::; e < 1, a hyperbola for 1 < e, and a parabola for
e = 1. In Figure 14.4 only a part of the curve close to F has been shown. In
this way all three cases may be said to be included in the figure.
For the ellipse (e < 1) the angle 8 may take all values from 0 to 211". Since
e < 1 the denominator can never become zero. For the parabola (e = 1) we
have: r - t 00 for B - t 11" (or -11"). For the hyperbola one obtains all points on
the branch considered, when 8 is limited to 181 < 80 , where cos 80 == -lie;
11" 12 < 80 < 11". One finds r - t 00 for I 8 I - t 80 , which give the directions
of the asymptotes. (If 8 runs through the intervals 80 < I 8 I ::; 11" one gets
negative values of r, which might be considered to correspond to the other
branch of the hyperbola.)
Concluding: for any value of e 2: 0 and p > 0, (14.3) describes a conic

section. The result (14.3) also includes the circle, which is the special case
e = O. In general, e determines the shape of the conic section, and p determines
the size.
I
Fig. 14.5. The ellipse is the set of points where the distances to Fl and F2 have a
constant sum
The ellipse can also be defined as the set of points P where the distances
from two fixed points (the foci) have a constant sum (see Figure 14.5).
The major axis has length 2a, the minor axis 2b. The distance between
the foci is e . 2a, where - as we shall see below - e is the eccentricity. From
the definition we obtain
r + r' = 2a.
Furthermore (see Figure 14.5),
(r')2 = (2ea + r cos (})2 + (rsin())2.
Using these two equations we obtain
a(l - e2 )
r = ..,......:'-----'-:: (14.4)
1 + e cos ()
By comparing (14.4) and (14.3) we find p = a(l - e2 ).
The perihelion (smallest value of r) and the aphelion (largest value of r)
are determined by
() = 0 (perihelion) =? rmin = -p- = a(l - e),

l+e
() = 7r (aphelion) =? rmax = -p- = a(l + e).
1-e
From this
rmax l+e
(14.5)
Tmin 1- e '
or
rmax - rmin
e=
rmax + rmin
14.4 Newton's Law of Gravity Derived from Kepler's Laws 355
The connection between semi-major axis a, semi-minor axis b, and the ec-
centricity is
b=a~. (14.6)
If we - instead of the angle f) - use the angle cp == 7l' - f) as the polar angle we
will have
a(l - e 2 ) p
r- - ----'--- (14.7)
- 1 - e cos cp - 1 - e cos cp
The perihelion is at cp = 7l'.
14.4 Newton's Law of Gravity

Derived from Kepler's Laws
Newton's gravitational law is contained within Kepler's three laws. In Chap-

ter 1 we demonstrated this for the special case of uniform circular motion.
Below we present the general calculations for elliptic orbits.
Kepler's first law was formulated in two steps. Kepler first showed that
the orbit of a given planet lies in a fixed plane containing the center of the
Sun; then that the planet moves in an ellipse with the Sun at one focus.
We consider motion in a fixed plane. The problem is to deduce the 1/r 2
dependence of gravitational attraction only from the observed motion of the
planets.
Fig. 14.6. The orbit of a planet, P, around the Sun, S
We start by calculating the acceleration of a planet moving in an elliptical

orbit with the Sun in one focus. The element of area dA (see Figure 14.6) in
polar coordinates (with origin in the Sun) is
(14.8)
Kepler's second law states that the area velocity is constant and we denote
the constant by hj2. We have
. 1
dA = ~hdt, or A = '2h.
From Kepler's second law it thus follows that
2A = r 2 iJ = h. (14.9)
We seek the acceleration vector for a Kepler orbit. In polar coordinates the
two components of the acceleration are as follows (see the Appendix):
Radial component:
2
ar = ddt 2r - r (dB)2
dt = r.. - rB'2 . (14.10)
Angular component:
all = ~r [~
dt
(r2 dB)] = 2i'iJ + rB .
dt
(14.11)
It will be useful to introduce a substitution. Instead of r we shall use u == 1jr

as a new variable.
From Kepler's first law we know the shape of the curve r = r(B) or
u = u( B). The goal is to find the two components of the acceleration, a r
and all, through u = u(B), i.e., through the equation for the orbit. With this
purpose in mind we eliminate the time from (14.10) and (14.11), i.e., we
e
express i', r, iJ, and by u, (dujdB) and (d 2ujdB2).
From (14.9)
iJ =~
r2
= hu 2 (14.12)
Differentiating (14.12) with respect to time:

.. . du dB 2 3 du
B = 2huu = 2hu dB dt = 2h u dB . (14.13)
Moreover, because (dujdB) = -(ljr 2)(drjdB), we have

dr dB 1 du· du
i' = dB dt = - u2 dB B = -h dB . (14.14)
The quantities iJ, e,

and i' are now expressed by u = u(B).
We can then determine all:
or
14.4 Newton's Law of Gravity Derived from Kepler's Laws 357
a(J = o.
Based on Kepler's second law we have shown that the acceleration of the
planet is directed in a radial direction (i.e., towards the Sun). This result,
well known from circular motion, is thus also true for a general elliptical orbit.
We now seek a r . From (14.10) we see that we have to find r expressed by
u = u(B).
~: = ! (-h ~~)
d ( dU) dB 2 2d 2u
dB -h dB dt = -h u dB2 . (14.15)
By inserting (14.15) and (14.12) into (14.10) we obtain
(14.16)
From Kepler's second law we have found the radial acceleration, expressed
by the area velocity constant h, and the equation of the orbit u = u(B).
From Kepler's first law
r = p (an ellipse),
1 + ecosB
or
1
u = u( B) = - [1
p
+ e cos B] .
Thus
d2u e
dB2 = -p cosB.
Inserting (d 2ujdB2) and u into (14.16) we finally obtain
h2 1
a r =--""2. (14.17)
p r
Conclusion. The acceleration of the planet is directed towards the Sun (the
minus sign), and the acceleration is inversely proportional to the square of
the distance from the Sun.
We proceed to apply Kepler's third law, in order to demonstrate that the
constant h 2 jp can depend only on the physical nature of the Sun, Le., the
value for h 2 jp is the same for all planets.
Kepler's third law may be written
a3
T2 =C,
where C is a constant, Le., the same for all planets. We introduce the sidereal
period of revolution T into (14.9).
By integration over a complete revolution (14.9) becomes
2A = hT. (14.18)
The area A of an ellipse is A = 1rab, where a and b are the semi-major and
semi-minor axes respectively. Furthermore,
and b=a~= p .
Vf=e2
From (14.18) we therefore get, using b2 = ap,
T2 = (~A)2
h
= (21r)2 a2b2
h
= 41r2a3~2
h .
(14.19)
From (14.19) - and applying Kepler's third law, a 3 /T2 = C - we find that
h2 a 3
p = 41r2 T2 = 41r 2 C .
We may then finally write (see 14.17)
(14.20)
where C is the same for all planets, i.e., C depends (at most) on properties
of the Sun only.
From Kepler's three laws we have computed the acceleration of the planet,
and seen that it depends only on the distance of the planet from the Sun:
The acceleration is directed towards the Sun, and the acceleration is inversely
proportional to the square of the distance from the Sun.
Newton added a decisive new feature to these results in the form of a
theoretical interpretation of the derived formula. Newton introduced the Sun
as the cause of the acceleration of the planets, and this guided him to the
fundamentally new idea about universal gravitation (see Chapter 1).
This most surprising step, rightfully admired by both Newton's contem-
poraries and by later generations, was Newton's linking of the fall of bodies
towards the Earth with the motion of celestial bodies.
The interaction that makes an apple fall to the ground also holds the
Moon in its orbit around the Earth.
In Chapter 8 we proved that the Earth acts gravitationally as if all of
its mass was concentrated in the center. From (14.20) we know that the
acceleration near the surface of the Earth is (p = radius of the Earth)
41r 2C'
g=--2-'
P
The constant C' is the same for all bodies moving in the gravitational field
of the Earth. The constant may therefore be calculated from data of the
14.5 The Kepler Problem 359
orbit of the Moon: G' = r3 /T2, where r = radius of the lunar orbit and T is
the sidereal period of revolution of the Moon. Introducing numerical values,
Newton found the gravitational acceleration g near the surface of the Earth:
47r 2 r 3
g = p2T2 = 9.8 ms- 2 ,
which is in accordance with the observed value. The greatest achievement in

the history of man was completed.
14.5 The Kepler Problem
The derivation of Kepler's three laws, setting out from Newtonian mechanics
and the law of gravitational attraction, is called the Kepler problem. The
solution of this problem is one of the jewels of theoretical physics.
We start by deriving Kepler's first law. We first solve the so-called one-
s x
Fig. 14.7. A planet moving around the Sun
body problem. It is assumed that the Sun is fixed in the origin of the coordi-
nate system. We furthermore neglect the gravitational interactions between
the planets. We thus consider one planet moving in the gravitational field of
the Sun, which is at rest in the origin of an inertial system (the heliocentric
reference frame).
The angular momentum Lo of the planet around 0 is
Lo = r x mr = r x mv.
In a central force field the angular momentum is a constant of the motion.
The plane spanned by r and v is the plane of motion for the planet. The
plane of motion is fixed perpendicular to L o, and passes through the center
of the Sun.
We seek the orbit that the planet will traverse, when it is started with a
given velocity r.
We use polar coordinates. The mass of the planet is m. We write the

equation of motion of the planet:
d2r GMmr.
m -2 = - - - - - . (14.21 )
dt r2 r
The term on the right is the force on the planet. The equation of motion may
be written as:
(14.22)
where C == -GMm.
In the form (14.22) the equation is more general: for attractive forces C
is negative (gravitational forces, electron in the Coulomb field of a proton).
For repulsive forces C is positive (two electrically charged particles with the
same sign of the charge).
In polar coordinates the acceleration is
d 2r _ .. '2 1d 2'
dt 2 = a = (r - rB )er + ~ dt (r B)eo,
with e r being the unit vector along the radius vector, and eo the unit vector
perpendicular to radius vector. The equation of motion (14.22) written in
two components, one along e r and one along eo, becomes
( .. B'2) C (14.23)
mr-r =2'
r
1 d 2'
m--(r B) = O. (14.24)
r dt
From (14.24)
d 2'
dt (mr B) = 0,
or, after integration,
(14.25)
where L is the magnitude of the angular momentum of the planet relative
to O. The magnitude of the angular momentum, L = 1L I, is constant and
determined by the initial conditions. We have thus introduced a constant of
motion into the process of integration.
From (14.25) we find
. L
B=--2' (14.26)
mr
Introducing iJ into (14.23) gives
L2 C
r------
m 2r 3 - mr2 . (14.27)
The equation (14.27) is a differential equation for r as function of t. In order

to derive Kepler's first law we are interested, not in an instantaneous location
of the planet, but in the shape of the orbit. In other words: We are interested
in determining r as a function of B, not r as a function of t.
We eliminate t from (14.27) by using (14.26). First we determine
. dr
r=-
- dt '
both expressed by B instead of t.
dr dr dB dr L
(14.28)
dt dB dt dB mr2 '
or
(14.29)
We introduce a new variable u(O) = l/r(B). The reason for introducing u as

variable instead of r is that the parenthesis in (14.29) is close to being equal
to (d 2u/dB 2). We find
Using these results we obtain

d 2r L2 d 2u
- =
dt 2 -
r = - - ---
m 2r2 d0 2 . (14.30)
Introducing (14.30) into (14.27) and using l/r = u we obtain the following
differential equation for u = u( B):
d2 u Cm
d02 +u = - L2 . (14.31 )
This differential equation has the same form as the equation describing the
oscillation of a mass at the end of a spring, hanging in the gravitational field
of the Earth (see Example 2.3).
The solution of (14.31) is
Cm
u = A cos(B + <po) - £2 . (14.32)
The quantities A and <Po are constants of integration. By the substitution of

(14.32) into (14.31) it may be verified that the given expression for u is a
solution of the differential equation.
The integration constant 'Po describes the orientation of the orbit in the
plane. By choosing the polar axis in a suitable way we can obtain 'Po = 0.
The result (14.32) may thus be written as follows:
1 Cm
:;:- =AcosO- L2 . (14.33)
We proceed by introducing another constant of integration: The mechanical

energy E of the planet. The planet moves in a conservative field of force. The
energy is therefore conserved.
We express the integration constant A through E.
1 C 1 . C
E == -mv 2 + - = -m(r 2 +r 202) + -. (14.34)
2 r 2 r
Using the expressions for r (14.28) and iJ (14.26) we obtain
E = 21 m (L2)
m r
[( dOdr) 2+ r 2] + -:;:C .
2 4 (14.35)
The total mechanical energy E is - in (14.35) - expressed in terms of the

parameters of the orbit. Using (14.33) we find an equation connecting E and
A. From (14.33)
dr 2A· 0
dO = r sm.
Introducing dr IdO in (14.35) gives
1
E = -m -- (L2) (r4A 2 sin 2 0+r2) + -.
C
2 m 2r 4 r
By using (14.33) again we find
E =~ L2 A2 _ C 2m
2m 2L2 '
or
_ Cm ( 2EL2)1/2
A - L2 1 + C 2m (14.36)
Consider again (14.33). Introducing (14.36) and using C == -GMm we obtain

the expression for the orbit of the planet in polar coordinates and expressed
by two constants of the motion, Land E:
1 Gm 2M [ ( 2EL2 )1/2 ]
:;:- = L2 1 - 1 + G 2 m 3 M2 cos 0 . (14.37)
Equation (14.37) describes an ellipse with perihelion for 0 = 7r (see Section

14.3, Equations (14.7) and (14.3)). By choosing 'Po = 7r in (14.32), instead of
'Po = 0, (14.37) becomes
We prefer this choice of CPo and write our final result as
1
;: =
Gm 2M [
L2
( 2EL2 )1/2
1 + 1 + G 2m 3 M2 cosO
1 (14.38)
1 1
- = -[I+ecosOJ. (14.39)
r p
Equation (14.39) is the equation, in polar coordinates, for a conic section.
Our result, (14.38), describes a conic section with parameter
L2
and eccentricity
e=
The total energy,
E = ~mv2 _ GMm ,
2 r
may be either negative, positive, or zero.
From (14.38) we conclude:
1. For E < 0, e < 1, we have an ellipse, e = 0 corresponds to a circular
motion
2. For E > 0, e> 1, we have a hyperbola
3. For E = 0, e = 1, we have a parabola
We have shown that Kepler 's first law follows from New-
ton's second law, in combination with the law of gravita-
tional attraction.
Many other important consequenc:es concerning the motion of celestial bodies

may be read from (14.38). Let us ~onclude.
A planet, a comet, an asteroid, or any heavenly body whatsoever, gov-
erned by the gravitational field of the Sun, will traverse an orbit that is a
conic section. The form of the conic section is determined solely by the total
mechanical energy E, given by the initial conditions. If E ~ 0 the body is
not bound to the solar system. The orbit is a hyperbola, or for E = 0, a
parabola. If a body with E ~ 0 passes "close to the Sun", i.e., if such a body
appears in the solar system at all, it will happen only once. No comet with
E > 0 has been observed until now. For E < 0 we have a "bound state" of
the planet.
The angular momentum L is likewise a constant of the motion, given
by the initial conditions. The plane of the orbit is perpendicular to L. The
magnitude L of L determines the parameter p = L2 /Gm 2M.
14.5.1 Derivation of Kepler's 3rd Law

from Newton's Law of Gravity
In Chapter 10 we proved that Kepler's second law is a consequence of con-

servation of angular momentum in a central force field. We shall now show
that Kepler's third law also follows from Newtonian mechanics.
Kepler's third law is
1
T2 = C a3 == ka 3 .
The constant of proportionality k is the same for all planets.
Kepler's second law can be written as
dA L
= (14.40)
dt 2m
To prove the third law we have to introduce T, the sidereal period, into
(14.40) . By integration over a complete revolution,
(14.41 )
where A = nab (the area of the ellipse).

From (14.41)
(14.42)
From Section 14.2
and
From (14.42) we obtain
(14.43)
The constant k = 4n 2 /GM depends only on the mass of the Sun, and is thus
the same for all planets. Kepler's third law has been derived from Newtonian
mechanics.
From (14.43)
or
2
log a = 3 log T + B, where B is a constant.
Figure (14.8) illustrates Kepler's third law for the solar system. The figure
shows that log a is a linear function of log T, and the slope of the line is 2/3.
100r------.-------.------~
<fJ'
.~ 10
...o
l
.~
<fJ 1.0 1-----,fFr:-,..",+-----+------i
II
<::t
Mercury
0.1 '-----'-----'--------'
107 108 109 1010
Period, s
Fig. 14.8. Kepler's third law for the solar system, loga = ~ logT + B, B
constant. Based on Berkeley Physics Course
Kepler 's third law is a consequence of the universal law of gravity and
Newton's laws of motion. The law is valid also for elliptical orbits of moons
moving around planets. The mass M of the Sun is then replaced by the mass
of the given planet.
Newton tested the validity of Kepler's third law on the four Jupiter moons
known to him. Newton knew the periods of revolution of the moons of Jupiter
with fairly good accuracy.
The t able below shows the radius p in the orbits of the moons; p = r / R j
is measured in units of the radius R j of Jupiter. The table furthermore gives
the period of revolution T for the moons, and finally (p3/T 2). Kepler's third
law is seen to be valid to a high order of accuracy.
The radius in the orbits of the moons, as given in the table, is measured
in units of the radius of Jupiter. The knowledge of the absolute distances in
the solar system was limited at the time of Newton, and the size of Jupiter
was consequently not known.
r/R j T (s) p3/T2 (S-2)
10 5.58 1.53 x 105 7.4 X 10- 9

Europa 8.88 3.07 x 10 5 7.5 X 10- 9
Ganymede 14.16 6.19 x 10 5 7.5 X 10- 9
Callisto 24.90 l.45 x 106 7.4 X 10- 9
14.6 The Effective Potential

In this section we shall briefly describe another procedure for the integration
of the equation of motion for a planet moving in the gravitational field of the
Sun.
The angular momentum L of the planet is assumed to be different from
zero ( L == I L I = 0 corresponds to the planet moving along a radius vector,
away from or into the Sun).
From Example 10.2 it is known that the total energy of the planet may
be written as follows:
1.2 L2 GMm
E = -mr
2
+- - - -r - .
2mr 2
(14.44)
The term L 2 /2mr 2 is called the centrifugal potential energy.

We look for the differential equation of the orbit. The magnitude of the
angular momentum is
2 dB
L = mr dt .
The energy E' and the magnitude of the angular momentum L are known
constants of motion.
We transform differentiation with respect to time into differentiation with
respect to B:
d dB d L d
dt dt dB
Furthermore, we apply the variable transformation u = l/r. This means that
~: = ! (~) = - :2 ~~ = - ~ ~~ .
Equation (14.44) may thus be rewritten as
1 L2 (dU)2 1 L2
"2 m dB +"2 m u 2 - GMmu = E. (14.45)
This is a differential equation for u = u(B). The equation may be simplified

by differentiation with respect to B and division by (L2/ m )(du/dB). We find
d 2u GMm 2
dB2 +u = L2
This is the equation integrated in Section 14.5. The solution may be written
- with a suitable choice of polar axes - as u = A cos B + 1/p , where we have
introduced p == L2/GMm 2 .
Instead of the integration constant A we shall use A == e/p, where e is a
new integration constant. The possible orbits then have the form
r = p
l+ecosB'
which is the well-known equation for a conic section.
14.7 The Two-Body Problem 367
It is convenient to express the integration constant e - which obviously is

the eccentricity - by means of E. By differentiating u = (1 + e cos ())/p with
respect to () and inserting the result into equation (14.45) we find
G2M 2m 3
E = 2£2 (e 2 - 1) ,
or
2EL2
e= 1 + G2M2m 3 '
which is identical to the result found previously.
14.7 The Two-Body Problem
We have solved the so-called one-body problem: a material particle moves in

a central field of force of the type 1/r2. The model corresponds to a planet
moving in the gravitational field of the Sun, where the Sun is assumed to be
fixed at the origin.
Below we investigate the motion of two spherically symmetrical bodies
moving in astronomical space. The two bodies are assumed to move exclu-
sively under their mutual gravitational interaction. The problem is called the
two-body problem. As a model one may think of the Sun of mass M and one
of its planets, say Jupiter, of mass m . Consider Figure 14.9.
----..."'M
rl
Fig. 14.9. The two-body problem
The center of mass, CM, of the system will be either "at rest" or move
with constant velocity, because no external forces are acting on the system.
Seen from the point of view of the original Newtonian mechanics this means
that CM is either at rest in absolute space (astronomical space) or moves with
a constant velocity relative to that space. The modified form of Newtonian
mechanics are now in logical difficulties: we look for the motion of the Sun
under the influence of the gravitational field of the planet (Jupiter) . Therefore
it has no meaning to use the heliocentric reference frame as the basis for
the description. The astronomical two-body problem illustrates the profound

difficulties connected with the choice of inertial systems.
We choose eM as the origin for an inertial system (see Figure 14.9):
The radius vector to m measured from M is denoted r:
A unit vector in the direction of r is denoted e r :

r rl - r2
e r == - = - - - .
r r
The equations of motion for each of the two bodies are
d2rl GMm
m dt 2 = -~er,
M d r2 GMm
2 _
dt 2 - ~er.
If we add these two equations we reach the not surprising conclusion that the
total momentum of the system is a constant.
Our real aim is to obtain a differential equation for r = rl - r2. Transfer-
ring the masses to the right side of the equations and subtracting the second
equation from the first we get
d 2rl _ d 2r2 __ GMm [~ ~] e

dt 2 dt 2 - r2 m +M r ,
or
mM d 2 r GMm
----- - ----e
M + m dt 2 - r2 r .
The reduced mass /1 is defined as
or
1 1 1
-=-+-.
/1 m M
(14.46)
We obtain the following differential equation for r, the radius vector from M
to m:
d 2r GMm G(M + m}/1
/1 dt 2 = - ~er = - r2 er .
This differential equation has the same form as the equation we solved for
the one-body problem. We have reached a fundamental result: the motion of
m relative to M may be determined as if M were at rest at the origin of
an inertial system, if we - instead of the inertial mass m - use the reduced
14.7 The Two-Body Problem 369
mass /1 as defined in (14.46). The two-body problem is reduced to a one-body

problem for the motion of a mass /1 in the gravitational field of a mass of
magnitude M + m.
Note that m and M enter the problem in a completely symmetrical way.
We could equally well have used (-r) for a description of the motion.
We shall briefly show how the two constants of motion, the angular mo-
mentum L and the mechanical energy E may be expressed by the reduced
mass /1.
See again Figure 14.9. Consider first the total angular momentum L with
respect to CM:
LCM = r2 x M r2 + rl x mh .
Eliminate M by means of the definition of CM:
We get
LCM -r2 x mrl + rl x mrl

(rl - r2) x mrl .
The term mrl may be rewritten:
The total angular momentum for the two masses in their motion around CM
is
LCM = r x /1r.
We conclude: the angular momentum LCM may be calculated as if the mass
/1 moved around M. For this calculation the mass M may be taken to be at
rest in an inertial frame.
Next we calculate the total mechanical energy:
1 .2 1 . 2 GMm
E -mrl + - M r2 - - - -
2 2 r
1 . . 1M . . GMm
-mrl . rl + - r2· r2 - - -
2 2 r
or, using r2 = -(m/M)rl,
1 ( m2). . GMm
E =2 m+ M rl· rl - - r - .
Making use of
. M (. .) M.
m+ M m+ M
rl = rl - r2 = r ,
we find
1 .2 GMm 1.2 G(M+m)/-l
E = - /-lr - -- = - /-lr
2 r 2 r
We conclude that the total energy E may be calculated as if the mass /-l
moved around M. For this calculation the mass M may be taken to be at
rest in an inertial frame.
14.7.1 The Two-Body Problem and Kepler's 3rd Law
The differential equations describing the one-body problem and the two-
body problem are of similar form. With an easily understandable notation
the equations may be written as
GMr
one-body:
-~:;:'
G(M +m) r
two-body:
r2 r
For the one-body problem we found Kepler's third law:
T2 47r 2
~ GM·
For the two-body problem the corresponding expression becomes
T2 47r 2
a3 G(M +m)
The ratio T2 / a 3 is thus not exactly the same for all planets, due to the
fact that m varies from planet to planet. Due to the large mass of the Sun
compared to planetary masses the deviations from planet to planet are small.
14.8 Double Stars:

The Motion of the Heliocentric Reference Frame
Many stars are double stars, i.e., two neighboring stars moving under their
mutual gravitational interaction. For simplicity we assume that the two stars
move in circles around their common eM.
The distance from eM to mass M is called R, and the distance from eM
to m is denoted r (mr = MR).
We use the results from the study of the two-body problem. The motion
can be described as the motion of a single particle with the reduced mass /-l
14.8 Double Stars: The Motion of the Heliocentric Reference Frame 371
,,
m
Fig. 14.10. Double stars
moving around M. The radius in that circular motion is called p. We have:

p = r+R.
2 Mm Mm
J.Lpw = G -2- , J.L=M+m·
p
For the angular frequency we find
2 G(M +m)
w = --"--;:------'-
p3
The period of revolution is determined by wT = 21r.

Relative to what do the two stars move? It would be absurd to use the
heliocentric reference frame!
Our own solar system is "nearly a double star system". The mass of
Jupiter dominates the planetary system. In units of the mass of the Earth
the masses in the solar system are:
Sun: 332946
Jupiter: 317.9
Saturn: 95.2
The rest of the planets together: 33.7
Neglecting the mass of all the planets except the mass of Jupiter, we can
estimate the position of the eM of the solar system. The distance of Jupiter
from the Sun is
5.2 AU = 5.2 x 1.5 x 108 km,
mJ
ReM ~ Ms rJ ~ 744750 km.
The radius of the Sun is 700000 km. The eM of the solar system is thus
located about 50000 km above the surface of the Sun.
A coordinate frame with its origin in the center of the Sun, i.e., the he-
liocentric reference frame, has an acceleration relative to the eM of the solar
system. The eM of the solar system moves around the center of the galaxy.
The Sun is about 30000 light years or 3 x 10 22 em from the galactic center.
The time for one revolution around the galactic center has been estimated
as about 250 million years, or
T;:::: 8 X 10 15 s.
The acceleration of the CM relative to the galactic center is thus
v2 47r 2 r
a =- = - 2 - ;:::: 1.9 X 10- 6 cms- 2
r T
The tides in the freely falling heliocentric reference frame are so small that we
have not been able to measure them (yet). Therefore we use the heliocentric
reference frame as a local inertial reference frame.
From measurements on radioactive isotopes in minerals in meteorites (and
in rocks from the Moon and the Earth) we know that the solar system formed
4.6 x 10 9 years ago. The solar system has completed
4.6 x 109
----;o;:::: 18
2.5 x 108
revolutions around the center of the galaxy, since the system was born. What
did we meet on this long journey? Supernova explosions? Interstellar clouds?
14.10 Examples 373
Conservation laws:
1 GmM· · .
E = -mv 2 - = constant'
-' - - -
2 r
14.10 Examples
In this section we discuss a few examples of mot ion in the solar system.
Example 14.1. Planetary Orbits and Initial Conditions. We consider

a family of possible orbits for a planet around the Sun, S. The Sun has the
mass M . The planet, P, with mass m, is imagined to be started with a
velocity always perpendicular to the line SP, but with various values of the
magnitude of the initial velocity. The initial distance between S and P is ro.
a> ff
a = ff
v
We begin by determining the magnitude of the velocity necessary for a

uniform circular motion:
Vo =
ro
jGM .
The planet is then imagined to be started with an arbitrary magnitude of
the velocity Vp (still perpendicular to SP). We introduce the ratio
Vp
a=:-.
Vo
The value a = 1 thus corresponds to a circular orbit. Below we shall show
that
for a J2 the orbit is a parabola,
for a < J2 the orbit is an ellipse,
for a > J2 the orbit is a hyperbola.
Calculate first the total energy in the orbit expressed in terms of a :
1 2 GMm
E -mv - - - -
2 P ro
1 GMm
-mv~a2 - - - -
2 ro
1 1 GMm
2
-(a -1)mvo
2
+ -mvo
2
- -- .
2 2 ro
The last two terms in this expression form the energy Eo in a circular orbit.
1
E = Eo + 2(a 2 - 1)mv5·
Question: Show that the total energy in a circular orbit may be written
Eo = -! mv5·
We finally obtain
or, as Eo is negative,
E = (a 2 - 2) I Eo I
From this result we see that:
for a > J2 E>O (hyperbola),
for a J2 E=O (parabola),
for a < J2 E <0 (ellipse).
The shape of the orbit is determined not only by Newton's laws but also by
the initial conditions. This fact makes it possible to find out something about
the origin of the solar system. The fact that the planetary orbits lie nearly in
the same plane has something to do with the initial conditions of the system,
and is not dictated by the laws of force and motion. See Example 10.2. 6.
Example 14.2. Shape and Size of Planetary Orbits. Consider the figure
(see also Figure 14.4 for the definition of conic sections).
We draw the chord through the focal point F and perpendicular to the axis
of the conic section. The points where the chord intersects the conic section
are denoted A and B. All conic sections with the same value of the parameter
p pass through A and B. The shape of the conic section is determined by
the eccentricity, which again is determined by the distance d = pie to the
directrix l.
We have (for a celestial body):
p= GMm 2 '
e=
From this, the orbit for all celestial bodies, with the same magnitude of
angular momentum L , has the same value of parameter p. All such orbits
pass through the points A and B, if the conic sections have the same axis.
14.10 Examples 375
------------F·~----+_~--
p
e
The eccentricity, and consequently the shape of the conic section, is also
determined by the total energy E.
Question. Does the value of m influence the size and shape of the orbit?
Answer. No. We proceed to show that the semi-major axis of an elliptical
orbit depends only on E.
For an ellipse p = a(l - e 2 ). From p and e as given above we get (note
E < 0)
or
E=_GMm.
2a
For a circle, a = ~(r + r) = r.
Consider the following figure showing elliptical orbits with the same semi-
major axis a.
14----2 a
Planets moving in these elliptical orbits have the same energy, but not
the same angular momentum.
Of aU planetary orbits, with the same angular momentum, the circle has
the lowest energy E. This is seen from
or
(e 2 _ 1)G 2 M 2 m 3
E = 2L2
For a circular orbit e2 = o.
Example 14.3. Motion Near the Surface of the Earth
The trajectory of a cannonbaU near the surface of the Earth is - neglecting

air resistance - a parabola. The approximation made is that the acceleration
due to gravity is a constant vector.
The approximation is very good indeed, but strictly speaking we should
consider the top point in the orbit as the aphelion (apogee) in an elongated
ellipse, where the center O of the Earth is in one of the foci of the ellipse.
Assume that a cannonbaU is fired with the initial velocity Vo and from
the point P. Assume further more that the Earth is at rest in an inertial
frame. The total energy in an elliptical orbit is determined exclusively by the
semi-major axis a. The energy E is
1 2 GMm GMm
E = -mvo - - - - = - - - -
2 p 2a '
where
p radius of the Earth,

M mass of the Earth,
m mass of the canonbaU.
14.10 Examples 377
The major axis 2a is only slightly different from the radius of the Earth, p.
We write
2a = p+ 6= (1 + ~ )p
Let us estimate the magnitude of 6:
~mv5 _
2
GMm = _ GMm
P P
(1 + 6)-1
P
~mv5 _ GMm ~ _ GMm

2 P P
(1 _6)
P
We thus obtain
6~ V 02 vo2
~ 2GM/p2 2g'
where 9 is the acceleration of gravity at the surface of the Earth. We have
that 6 = h + d (see the figure).
If the start velocity of the cannonball is v =1 kms- 1 we find that
v2
6 ~ - ~51 km.
2g
The shape of the elliptic orbit depends not only on the speed Vo but also on
the firing angle.
Note. Close to the perihelion it is difficult to distinguish a "long ellipse" from
a parabola. For instance, many of the comets observed until now are in orbits
with excentricities very close to 1, i.e., many comets are in orbits that are
nearly parabolic. 6
Example 14.4. Velocities in an Elliptical Orbit
Vp
The velocity of a planet in the perihelion (perigamm) is Vp and the corre-

sponding velocity in the aphelion (apogamm) is VA. Both velocities are per-
pendicular to the axis of the ellipse, and measured relative to the heliocentric
reference frame.
From conservation of angular momentum:
mrAVA = mrpvp ,
rp 1- e
VA = -vp = --vp.
rA 1 +e
From conservation of energy:
1 2 GMm 1 2 GMm
2mVA - + e) = 2mvp
a(l - a(l - e)
By using VA = (1 - e)vP/(l + e) we obtain
Vp = ./GM1+e,
V a 1- e
VA = JGM I-e.
a 1+e
For the planet Earth:
e = 0.0167, a = 1 AU = 149.6 X 10 6 km,
JGM
-;;- = 29.78 kms-I, ~+e
- - = 1.0168,
1-e
Vp = 30.3 kms- 1 , VA = 29.3 kms- 1
For the planet Mars:
e = 0.0933, a = 1.524 AU,
JGM
-;;- = 24.24 kms -1 , ~+e
- - = 1.098,
1-e
Vp = 26.6 kms-I, VA = 22.1 kms- 1 .
Example 14.5. Hohman Orbit to Mars. When a spaceship is sent to

another planet, the ship is first placed in a so-called parking orbit around
the Earth. To enter the transfer orbit, the spaceship must leave the parking
orbit and escape the gravitational field of the Earth.
The rocket engine delivers the thrust necessary for placing the spaceship
in the interplanetary orbit. Exactly when the rocket engines should be started
depends on the relative position of the Earth and the planet of destination.
In calculating the transfer orbit from the Earth to another planet we shall
make a series of simplifying assumptions. We ignore the binding energy of
the spacecraft in the gravitational field of the Earth in the parking orbit.
14.10 Examples 379
Fig. 14.11. Hohman orbit to Mars. Launch window: El Earth at launch; E2 Earth
at arrival; Ml Mars at launch; M2 Mars at arrival
We shall briefly discuss a journey from the Earth to Mars along the so-
called Hohman orbit, named after the astronomer who first calculated this
transfer orbit.
The launch should take place as the spacecraft is on the dark side of the
Earth. The velocity of the spacecraft in the parking orbit is then in the same
direction as the velocity of the Earth in its orbit around the Sun.
Let us now assume that the spacecraft is nearly free of the gravitational
field of the Earth, i.e. , we neglect the gravitational field of the Earth. The
velocity of the spacecraft in the heliocentric reference frame is assumed to
be the same as the orbital velocity of the Earth in this frame (~ 30 km/s).
As we shall demonstrate below only a rather modest increase in the velocity
of the spacecraft relative to the heliocentric reference frame is necessary to
bring the craft into an elliptical orbit towards Mars.
The Hohman orbit is tangential to the orbit of the Earth at launch, and
tangential to the orbit of Mars at arrival. The Hohman orbit is thus a semi-
elliptical orbit, whose perihelion coincides with the orbit of the Earth and
whose aphelion coincides with the orbit of Mars.
The exact calculation of a Hohman orbit is involved, particularly due
to the fact that the plane of the orbit of Mars is slightly tilted relative to
ecliptic a (i = 1°51').
The essential aspects of the determination of the transfer orbit are nev-
ertheless present in the calculations below, in terms of our simplified model.
Assume that the orbits of the Earth and Mars are in the same plane, the
ecliptica. Furthermore, assume that the Earth and Mars perform uniform
circular motions around the Sun, with the Sun located in the common center
of the orbits.
The radius in the orbit of the Earth is 1 AU, and the radius in the orbit
of Mars is 1.52 AU. With the assumption of a circular orbit, the velocity of
the Earth is 29.8 kms- 1 and that of Mars is 24.2 kms-l.
The semi-major axis for the Hohman orbit becomes

1 + 1.52
aH = 2 = 1.26 AU.
The trip to Mars along the Hohman orbit may be determined from Kepler's
third law. The time for one complete revolution in a Hohman orbit is denoted
TH. Let TH be measured in years. The time for a complete revolution of the
Earth (one year) is called TE = 1 year, Then, from Kepler's third law:
r,2 r,2 12
-.!!=J=-=1
a3H a3E 13 '
TH = aU2 = 1.414 years.

The travel time T to Mars corresponds to one half revolution:
T = 258 days.
The sidereal time of revolution for Mars is 687 days. As the spaceship has
moved along the Hohman orbit, Mars has moved
360 0 258 S:! 1350 .

687
If Mars is 45 0 ahead of the Earth at launch, the spaceship will meet Mars at
the point where the Hohman orbit touches the orbit of Mars.
The velocity at launch. The spaceship is in the perihelion of the Hohman

orbit at launch. The initial velocity should then be
Vp = ./GM 1 + e = JGM TA ,
Y aH 1 - e aH Tp
where aH = 1.26 AU.

For the Earth we know that
JGM
- - = 29.8 kms -1 .
aE
For vp we obtain
{ 1 {i£i 1
Vp = 29.8y 1.26y 1-1- = 32.7 kms- .
When the rocket engine has released the spaceship from the gravitational
field of the Earth, the ship has the same velocity as the Earth relative to
the heliocentric frame, i.e., 29.8 km S-I. By means of the rocket engine the
spaceship should be given an increase in velocity of 6v, where
6v = 32.7 - 29.8 = 2.9 kms- I .

14.10 Examples 381
When this has taken place, the ship will "fall" along the Hohman orbit to
Mars, guided by the gravitational field of the Sun.
At the arrival to Mars, the spaceship is in the apehelion of the Hohman
orbit. The velocity of the ship is then
rp 1 1
VA = Vp- = 32.7-- ~ 21.5 kms- .
rA 1.52
If the spaceship is bound to enter an orbit around Mars the rockets must
adjust the velocity of the ship to the velocity of the planet.
Even if the Hohman orbit is inexpensive from the point of view of fuel, it
will not be used in the manned expedition to Mars, due to the long time of
~hl. 6
Example 14.6. The Face of the Moon (Spin-Orbit Locking)

Even though the Moon is spinning about its axis (which is nearly parallel
to the spin axis of the Earth), the spin period of the Moon is so close to its
orbital period that the Moon always turns the same side towards the Earth.
The back side of the Moon is thus never observed from the Earth.
Is this merely a 'cosmic coincidence' or has some physical mechanisms

driven the Earth-Moon system towards such a state? In this example we
shall see how a crude model of the influence of tidal and frictional forces can
provide the basic mechanism for the so called spin-orbit locking.
Consider the Earth (mass M) and the Moon (mass m) to be revolving in
circular orbits around their common Centre-of-Mass (CM) which is at rest
in an inertial frame (see Section 14.8).
,
,,
I
I
I
I
, I
,, :CM'\
m, :, I ,M
I
.... - .. ,'
,,
We shall consider the entire problem in a fixed plane.

Question. How far is the OM from the centre of the Earth?
Answer. Since the mass of the Moon is about 8\ of the mass of the Earth,
the OM is about II of the Earth-Moon distance from the centre of the Earth;
this is a point near the surface the Earth.
Let us examine the situation in a reference frame K with origin at the
centre of the Moon, and one axis pointed always towards the centre of the
Earth. Relative to the inertial frame in which the OM is at rest, the frame
K is rotating with constant angular velocity fl.
That the Moon is not rotating relative to the frame K means that the
Moon (relative to the inertial OM frame) is rotating with precisely angular
velocity fl. Thus, the spin-orbit locking is a mechanism that will drive, in
the frame K, the Moon towards a non-rotating state.
In the frame K, both the Earth and the Moon are at rest on the axis
connecting them. They are acted on by the centrifugal force of magnitude
Mrfl2 = mRfl 2 (driving them apart) and the gravitational for ce of magni-
tude G (r~r;)2 pulling them together. Here, Ris the distance of the center of
the Moon from OM, and r is the distance of the center of the Earth from OM.
The two forces balance when evaluated exactly at the centre of the Moon and
at the centre of the Earth.
In Section 6.5 (Tidal Fields) we saw how the gravitational field of the Sun
acted on the Earth. This is also how the Earth's gravity acts, as seen in the
frame K, at points different from the centre of the Moon. The tidal field of
the Earth acts to puU masses away from the centre of the Moon, along the
axis connecting the centre of the Moon and the centre of the Earth. The tidal
field also will push masses that are ofi' this axis back towards the axis. See
Figure 14.12.
Fig. 14.12. In the frame K, the tidal field of the Earth acting near the Moon
pulls masses along the the Moon-Earth line away from the centre of the Moon, and
pushes masses off this axis towards the centre of the Moon
The Moon is not a perfect sphere. In the calculat ion below, we shall
assume that the permanent mass distribution of the Moon is similar to that
of a short dumbbell.
••
This dumbbeU will react to the tidal field of the Earth.
14.10 Examples 383
Question. Calculate, in the frame K, the potential energy of such a dumb-

beII, placed in the tidal field of the Earth.
Fig. 14.13.
Answe1'. In the rotating frame K, there are two forces act ing on the masses of
the dumbbeII; the graviatational force and the the centrifugal force. Both of
these forces are potential forces. Suppose the dumbbell is positioned forming
an angle O with the Moon-Earth axis (Figure 14.13). Let the two masses of
the dumbbell have value m/2 each. The gravitational potential energy is then
m (1
Vg(B) = -GM- - +-
1)
2 1'1 1'2
and the centrifugal potential energy is
where 1'1 and 1'2 are the distances from the centre of the Earth to the two
masses respectivly, and P1 and P2 are the distances of the two masses from
the CM. See the figure. If each of the arms of the dumbbell are of length I,
and if the distance from the centre of the Moon to the centre of the Earth
and from the centre of the Moon to the CM are R and p respectivly, the
distances 1'1 and 1'2 are given by
1'1(0) = JI sin(0)2 + (R -I COS(0)2)

1'2(0) = JI sin(0)2 + (R + I COS(0)2)
The total potential energy as a function of O is therefore:
Plotting V as a function of O, one will get a curve with a maximum at

O = ±7l' /2 and a minimum at O = O. This means that the position of the
Moon with its off-centre masses pointed away from and towards the Earth is
a stable equilibrium position.
v
9
o
/ .,
/;'\
/
/
/
Fig. 14.14.
The Moon might still be expected to turn in harmonic oscillations with

some amplitude around the stable equilibrium position (see the next chapter,
Harmonic Oscillators), but these oscillations are damped because of the so
called tidal jriction.
Tidal friction is the result of frictional (dissipative) forces that act on
a nonrigid body rotating in a tidal field. In the frame K, tidal forces are
distorting nonrigid parts of the Moon (on the Earth, the same forces coming
from the Sun and the Moon result in a tidal bulge, see Section 6.5). If the
moon were oscillating around the stable position, each oscillation would drag
the tidal bulge away from the line connecting the Earth and the Moon. A
torque would result, act ing as a breake on the oscillation.
In the geological past, such a tidal brake has acted on the Moon, and has
brought it to complete rest at the e = O equilibrium point 1 . For a deeper
understanding of the mechanics of the Earth-Moon system, you may consult
any textbook on (theoretical) geophysics. L.
1 CurrentIy, such a breake is also acting from the Moon on the Earth. The Earth
drags the tidai bulge of the oceans aIong, but the Moon exerts a torque on
the bulge. This causes the Earth to slow down in its spin. The day therefore
increases in length. In turn, this will cause the Moon to slowly receed from its
present orbital distance, at a rate on the order of centimeters per year
14.11 Problems 385
14.11 Problems
Problem 14.1.
Assume that the Earth is at rest in an inertial frame. A rocket is started, not
along the vertical, but in a direction forming an angle e with the vertical.
(1) Calculate the magnitude of the start velocity vo, when it is assumed that
the rocket just escapes the gravitational field of the Earth.
Remark: the launch facilities of the European Space Agency are located in
South America. Why not in, say, northern Norway?
(2) This question deals with the escape velocity from the solar system from
a point in the orbit of the Earth. Assume that a rocket interacts only
with the gravitational field of the Sun. Determine the smallest velocity v
relative to the sun that a spacecraft should be given at the distance of 1
AU from the Sun, so that the spacecraft leaves the solar system.
Problem 14.2. Consider a planet moving in a uniform circular motion in

the gravitational field of the Sun. Perform an explicit calculation of the total
mechanical energy E and the magnitude L of the angular momentum for the
planet, and show explicitly that the excentricity e is zero. Use
2EL2
e2 = 1 + G 2 m 3 M2 '
M = mass of Sun,
m = mass of planet,
G = gravitational constant.
Problem 14.3. In a double star system (also called a binary star) one of the
stars has the mass m = 3 x 1030 kg and the other has the mass M = 4 x 1030
kg.
Each of the stars performs a uniform circular motion around the center
of mass (CM) of the system and relative to an inertial frame. The stars may
be considered as mass points. The distance between the stars is 10 13 m.
(1) Determine the angular velocity w of the motion of the stars.
(2) Determine the magnitude of the total inner angular momentum of the
system, i.e., determine LCM = 1LCM I, the angular momentum relative
to the CM of the system.
Problem 14.4. Consider the Earth-Moon as an isolated two body system.

The Earth and the Moon are assumed to move in circular orbits around the
center of mass (CM) of the system.
(1) Determine the position of the center of mass (CM) for the Earth-Moon
system.
(2) Determine the orbital speed and the orbital period of revolution of the
Moon. (The CM is assumed to be at rest in an inertial system.)
Problem 14.5. Comet Halley orbits the Sun in an elliptical orbit. At peri-
helion, the distance of the comet from the Sun is 87.8 x 106 km. At aphelion
the distance from the Sun is 5280 x 106 km.
(1) Calculate the period of the comet.
(2) Calculate the speed of the comet relative to the heliocentric reference
system when the comet is in the perihelion (Vp ) and when the comet is
in the aphelion (VA).
Problem 14.6. The first artificial satellite, the Sputnik 1, was launched on
October 4, 1957. Sputnik 1 had a perihelion of 227 km above the surface of
the Earth. The speed at perihelion was 8 km S-l, measured relative to the
geocentric absolute system (origin at the center of the Earth and axes that
point towards the same three fixed stars).
(1) Determine the height above the surface of the Earth that Sputnik 1 had
at aphelion.
(2) Determine the orbital period of revolution for Sputnik 1.
Problem 14.7. This problem deals with a Hohman transfer orbit to Venus.
Assume that the Earth and Venus move in the same plane (ecliptica) and
in circular orbits around the Sun. The radius of the orbit of Venus is 0.72 AU.
(1 AU = 1.5 X 108 km). Compare the present problem with Example 14.5. The
14.11 Problems 387
r1 = 0.72 AU
r2 = 1 AU
flpacecraft is in a parking orbit around the Earth. The launch into a Hohman
orbit to Venus (an inner planet) occurs when the spacecraft emerges onto
the sunlit side of the Earth. The initial velocity of the spacecraft includes
two contributions: the orbital velocity of the Earth about the Sun plus the
orbital velocity of the spacecraft around the Earth. When the spacecraft is
on the sunlit side of the Earth these contributions are in opposite directions.
A rocket thrust in the direction of the orbital motion of the craft around the
Earth will allow the spacecraft to escape from the gravitational field of the
Earth.
Once the spacecraft is essentially free of the influence of the Earth, the
spacecraft will move in an elliptical orbit around the Sun, with an initial
speed Vo relative to the heliocentric reference frame. Note: 2a = 1.72 AU.
(1) Determine Vo such that the spacecraft enters a Hohman transfer orbit to
Venus (compare with Example 14.5). Show that Vo < VE, where VE is the
orbital speed of the Earth around the Sun.
(2) Determine the travel time T to Venus;
(3) Determine the speed Vl of the spacecraft when it reaches Venus. Show
that Vl > vv, where Vv is the orbital speed of Venus in the heliocentric
reference frame.
(4) Discuss the relative positions at launch of Earth and Venus necessary for
a Hohman transfer orbit to be realized.
Problem 14.8. In this problem, the Earth is considered to be a homogenous

sphere with radius R and mass M. A coordinate system with origin at the
centre of the Earth and axes that do not rotate relative to the fixed stars is
considered to be an inertial frame.
A grenade of mass ml is fired vertically upwards from the North Pole.
The magnitude of the initial velocity of the grenade is
where G is the gravitational constant.

Ignore friction, but do consider the variation of the gravitational field
with distance from the centre of the Earth.
1. Determine the maximal height h above the North Pole reached by the
grenade. Express h as a function of R.
We now consider a different situation. See the sketch.

A satellite of mass m2 is launched from the North Pole in a direction 45°
from the vertical. The magnitude of the initial velocity of the satellite is
V2 = J3~M
The satellite will enter an orbit with the centre of the Earth as a focal
point.
2. Determine whether this orbit will be a circle, an ellipse, a parabola, or a
hyperbola. Explain your answer.
3. Determine the eccentricity of the orbit.
Problem 14.9. A planet with mass m is in circular orbit around a spher-

ically shaped star. The star, of mass M, is assumed to be at rest in an
inertial frame. We further assume that M > >m, so that we may disregrad
the motion of the star (thus, we consider a one-body problem).
The angular momentum of the planet, relative to the center of the star,
is L. The constant of gravity is G . Known quantities are thus M, m, L, and
G.
1. Determine the magnitude of the orbital velocity Va of the planet relative
to the inertial frame, and the radius R in the circular orbit.
We now assume that the velocity of the planet in some point of the orbit
is increased to a such magnitude Ve that the planet can just escape the
gravitational field of the star (ve = the escape velocity).
2. Determine the magnitude of Ve expressed by Va . Determine the shape of
the orbit of the plantet if the magnitude of the velocity of the planet is Ve.
15. Harmonic Oscillators
15.1 Small Oscillations
A particle moving in response to an elastic force proportional to the distance

of the particle from the equilibrium point is called a harmonic oscillator.
The harmonic oscillator has been treated in the Examples 2.2, 2.3, and 2.7.
In Example 2.7 we noted that harmonic motion is a good approximation to
many types of oscillations around a point of stable equilibrium.
For a one dimensional harmonic oscillator with force constant k, the po-
tential energy V is given by (motion along x-axis)
The potential energy curve for a harmonic oscillator is a parabola. In Figure

15.1 U(x) is a potential energy curve for a particle moving in a certain force
field. The particle has an equilibrium point for x = o. The parabola approx-
imating the real potential energy curve around x = 0 is shown as a dashed
line on the figure. For oscillations with small amplitudes around the equilib-
rium point, the motion of the particle may be approximated by a harmonic
oscillation; we then speak of small oscillations.
Vex)
Fig. 15.1. The potential energy function U(x) can be approximated by a quadratic
function near the equilibrium point
In Chapter 3 we saw that the mathematical pendulum, i.e., a heavy mass

fastened to a light string, will perform harmonic oscillations for small ampli-
tudes (see Section 3.4). The potential energy for the mathematical pendulum
as function of f) is:

390 15. Harmonic Oscillators
,,
,,
,
-----_JI __ -----
Fig. 15.2. Mathematical pendulum. Potential energy: U(O) = mgL(1 - cos 0)
U(8) mgL(1 - cos 8)
mgL - mgL {I _ 822 + .2..8

24 + ... }
4
1
2mgL82
For angular amplitudes, where cos 8 may be approximated by cos 8 ~ 1-

(8 2 /2) the pendulum will perform harmonic oscillations.
15.2 Energy in Harmonic Oscillators
Fig. 15.3. Harmonic oscillator
The equation of motion for the oscillator sketched in Figure 15.3 is:
mx = -kx,
or, with w5 = kim,
x + w~x = o. (15.1)
The general solution may be written as follows:
x = Xo cos(wot + cp). (15.2)
The amplitude Xo and the phase angle cp are given by the initial conditions.
The period of the oscillations is T = 21r Iwo . Let us assume we have the
following initial conditions:
15.2 Energy in Harmonic Oscillators 391
t = 0, x(O) = xo, and v(O) = 0 .

Equation (15.2) becomes
x(t) = Xo coswot .
The velocity is
v(t) = -Woxo sinwot.
We may now calculate the total mechanical energy Eo in the oscillation:
Eo = U +K,
where K is the kinetic energy and U is the potential energy.
E lk 2 1
=2 Xo cos wot + 2 mwoxo sm wot,
2 2 2 . 2
or, since k = mW5,

1 2 2
E = 2mwoxo.
The total energy is - of course - constant in time (conservative force). During
the motion the energy oscillates between potential energy and kinetic energy.
The time average of the kinetic energy K over a period is
Using ~ JoT sin 2 wotdt = ~ we obtain
For the potential energy averaged over one period we get by similar calcula-
tions
1 2 2
(U) = 4mwoxo .
The equality of the average value of the potential and the kinetic energy is
a property of harmonic oscillations. The time average of the total energy is
equal to the total energy in the system:
1
(E) = (U) + (K) = E = 2 2
2mwoxo .
15.3 Free Damped Oscillations
Consider first an oscillator moving according to the equation of motion
mx = -kx. (15.3)
This equation will have the solution given by (15.2):
x = Xo cos(wot + cp).
The oscillator would move forever, but this will never happen for any real
oscillator. Friction, in some form or another, will always occur.
We have previously considered particles moving under frictional forces
(see, for instance, Example 2.4). We now proceed to discuss an important case
of the damped oscillator. We assume that the frictional force is proportional
to the velocity of the particle. This will be the case, approximately, for a
particle moving through a fluid (oil, air). Later in your studies you will find
that a frictional force proportional to the velocity of the particle to a good
approximation describes even the damping of an oscillating electric charge
(an electron) as it emits electromagnetic waves.
There are thus several justifications for studying an oscillator that has
the following equation of motion instead of (15.3):
mx = -kx - bi; , (15.4)
where b is a constant. Rearranging the equation:

b k
x + -i;
m
+ -x = o.
m
We use the definitions
b
-m =" and
k
- =Wo,
m
2
and obtain
(15.5)
Equation (15.5) is a homogeneous, linear, second-order differential equation

with constant coefficients. The general solution of this type of differential
equations may be found in books on calculus. We shall here simply discuss
three cases.
15.3 Free Damped Oscillations 393
15.3.1 Weakly Damped Oscillations
We assume that the damping is light, so that the oscillator performs nearly
harmonic oscillations. More precisely, we assume that 'Y « Wo (weak damp-
ing). In this case the solution of (15.5) may be written as follows:
(15.6)
where
The amplitude and the phase angle e is determined by the initial condi-
tions. Note: our assumption 'Y « Wo implies that 1 - h/2wo)2 > 0, which
again means that the oscillation frequency Wd is a real number.
Question. Show by substitution that (15.6) is a solution of (15.5).
The word 'frequency' is often used for w. The correct designation of W is

angular frequency. The frequency v denotes the number of oscillations per
second.
1
v= 'T; W = 27rv.
Figure 15.4 shows a graphical representation of the solution (15.6).
x(t)
Fig. 15.4. Damped harmonic oscillations. The dashed line corresponds to e-"I t / 2 .
See (15.6)
Note. The damping makes the oscillation frequency Wd smaller than the cor-
responding frequency Wo for undamped harmonic oscillations. If 'Y « Wo we
see from (15.6) that Wd ;:::j Woo The case 'Y « Wo is by far the most important
case for harmonic oscillations.
When a hammer hits a church bell in a brief blow, the church bell starts
to ring. The surface of the bell performs harmonic oscillations with several
frequencies, and, due mainly to a coupling to the surrounding air, the oscil-
lations of the bell are damped. The energy in the oscillations of the bell is
slowly transferred to other degrees of freedom, in this case sound waves. The
bell emits sound waves corresponding to the frequency of the oscillations. The
bell will ring for some time. It is a weakly damped oscillator, I « Woo One
might say that I in this case describes the coupling between the oscillations
of the bell and the sound field.
15.3.2 Strongly Damped Oscillations
For 1/2 > Wo the oscillation frequency Wd = Wo [1 - (r/2wo) 2] 1/2 would

become imaginary. For 1/2 > Wo one obtains the following solutions of (15.5):
(15.7)
This represent a nonoscillating motion of the particle, i.e., an aperiodic mo-

tion (see Figure 15.5).
x(!)
(a)
Fig. 15.5. Aperiodic motion of a harmonic oscillator. Graph marked (a) illustrates
Equation (15.7). Graph marked (b) illustrates Equation (15.8)
Physically the solution (15.7) means, for instance, that the mass is pulled
out at a given distance from the point of equilibrium and released with zero
velocity. For 1/2 > Wo the damping is so strong that the particle cannot
oscillate at all. The particle slowly moves towards the point of equilibrium.
15.3.3 Critical Damping
For ~ = Wo the solution becomes
x(t) = Aexp (-2,t) + Btexp (-It)

-2- (15.8)
See the curve marked (b) in Figure 15.5. Equation (15.8) again describes a
nonoscillating solution. The motion for 1/2 = Wo is called critically damped
motion.
We shall not discuss the nonoscillating solutions any further. Critical
damping (r /2 = wo) is of interest in the design of some measuring instru-
ments.
15.4 Energy in Free, Weakly Damped Oscillations 395
15.4 Energy in Free, Weakly Damped Oscillations
In the case of weak damping (, « wo) the solution of the differential Equation
(15.5) is (15.6). We shall use the approximation
Wd = woVI - b/ 2wo)2 ~ Wo·

With this approximation the .solution becomes
x(t) = Xo exp ( - ~t) cos(wot + rp) . (15.9)
From (15.9) we could calculate the velocity of the mass and then the ki-
netic energy as functions of time for the weakly damped oscillator. We shall,
however, make some further simplifying assumptions.
During one period (T = 27r / wo) the amplitude will decrease by an amount
exp( -,T/2) = exp( -7r , /wo). Since ,« Wo, the amplitude will decrease very
little during one oscillation. In the following calculations we assume that
the amplitude is constant over one oscillation. The velocity then becomes
(approximately)
v~ - woxoexp (-2,t) sm(wot+rp).

. (15.10)
For the energy in the oscillations we then get
E=K+U,
1 . 1
E = 2'mw5x6 exp (-,t) sm 2(wot + rp) + 2'kx6 exp (-,t) cos 2 (wot + rp) .
Using
we get the following averages:

1 2 2
(K) = 4mwoxo exp( -,t) , (15.11)
1 2 2
(U) = 4mwoxo exp( -,t) , (15.12)
and
(15.13)
That means; the energy in the oscillations decreases exponentially. For in-
stance, pull the oscillator out to a given distance xo, and release it with the
velocity zero. The oscillator has received the (potential) energy
1 2 1 2 2
Eo = 2'kx (0) = 2' mwoxo ·
We say that the oscillator is in an excited state, with the excitation energy Eo.
In the ensuing oscillations the oscillator will loose energy due to the damping.
The decrease of the (average) energy will follow the differential equation
~~ == -,E. (15.14)
This equation has the solution
E = Eo exp( -,t). (15.15)

With T == 1/1, (15.15) becomes
E = Eo exp ( - ; ) (15.16)
The parameter T has the dimension of time and T is called the lifetime of the
excited state of the oscillator.
Consider again (15.11)-(15.13). In spite of the fact that these equations
describe the time average over one period they contain the time variable t!
These equations represents the average over one period at about the time t .
As t increases, the average values decrease exponentially. Note: the energy
decreases as exp -,t, while the amplitude decreases as exp -,t/2. The energy
is proportional to the square of the amplitude.
The frictional force, f = -bx = -m,x, is responsible for the dissipation
of energy in the oscillations. Put differently, the frictional force describes the
fact that during the oscillations part of the oscillation energy passes into
other degrees of freedom (e.g., into heat as in the damping of a pendulum or
into sound waves as in the case of the vibrating church bell).
The power dissipated is equal to the negative of the (average) rate at
which the frictional force performs work. Assuming, «wo (weak damping)
we obtain for the power dissipated:
P = (fv) ~ -(m,xv)
1 2 2
-'2mwoxo exp( -,t)
-,E,
in agreement with (15.14). We have used (15.10) and x= v.
15.5 Forced Oscillations

Consider first an oscillator without damping on which an external force F(t)
acts along the x-axis. The equation of motion is then
mx = -kx + F(t). (15.17)

15.5 Forced Oscillations 397
The external driving force may have any kind of functional dependence on
time. We shall here consider the particular case of an oscillating force F(t)
of the form
F(t) = Fa coswt. (15.18)
Note. The frequency w of the applied force is in general different from the
so-called eigenfrequency Wo = (k/m)1/2 of the undamped oscillator.
If you want a concrete situation to think about, you may assume that the
oscillating mass carries an electric charge q. If a monochromatic, electromag-
netic wave of frequency w (a light wave) passes over the mass (an electron)
the electric field vector Eo cos wt acts on the charge with a periodic force,
F(t) = qEo coswt ( Eo is in the x-direction).
The use of the expression (15.18) is not an unreasonable limitation. Any
periodic function F(t) may be approximated very accurately by a sum of sine
or cosine functions, a so-called Fourier series. A nonperiodic function may be
expressed in terms of a so-called Fourier integral.
With F(t) = Fa coswt the equation of motion may be written
..
x + W&X = -Fa
m
cos wt . (15.19)
We shall not discuss here the general solution of (15.19). We seek only one
solution, the so-called steady state solution (see below).
It is natural to assume that the oscillator starts to oscillate with the
frequency w of the applied force. Our problem may then be formulated as
follows.
Determine the value of the amplitude A for which the following expression
is a solution of (15.19):
x = Acoswt. (15.20)
Insert into the equation of motion:
_w 2A cos wt + W& A cos wt = Fa cos wt ,

m
We find that
A = Fo/m
w5 - w2
The solution of (15.19) becomes
Fo/m
x(t) = 2 2 coswt. (15.21)
Wo -w
This means that the mass oscillates with the frequency w of the applied force
and with an amplitude that depends on the eigenfrequency of the oscillator.
For w < Wo the oscillations of the mass and the applied force are in phase.
For w > Wo the oscillations of the mass and the applied force are in anti-
phase. For large values of w the amplitude of the oscillations becomes small,
w --. 00, A --. O. The oscillator cannot follow the rapidly oscillating impressed
force F(t) = Fo cos wt.
According to the result (15.21) a remarkable thing happens when w is
close to woo Actually, for w --. wo, A --. 00. The response of the oscillator to
the applied force grows without bounds. An infinite amplitude is of course
impossible in a real physical system. As we shall see below, when we include
friction, we will get a large amplitude for w = wo, but certainly not an infinite
amplitude. The phenomenon appearing for w = Wo is called resonance.
The phenomenon of resonance is well known to anyone who has pushed
a child on a swing. If we choose the right time for pushing and the right
frequency, the swing will eventually go very high, even with a slight push.
15.6 The Forced Damped Harmonic Oscillator
The solution of this general problem will contain as special cases the solution
of the unforced oscillator (with and without damping) as well as the solution
for the undamped forced oscillator.
The equation of motion is now
mx = -kx - m')'x + Fo coswt,

or
F.
x + ')'x + w6x = ~ coswt. (15.22)
m
From the theory of differential equations it is known that the solution of
(15.22) may be written as a sum of two terms: (1) the complete solution,
Xl (t), of the corresponding homogeneous equation x + ')'x + w5X = 0, and (2)
an arbitrary solution, X2(t), to the inhomogeneous equation (15.22).
We write the solution in the form
X(t) = XI(t) + X2(t) ,

x(t) = Xo exp ( - ~t) COS(Wdt + cp) + A cos(wt - B) (15.23)
The first term, Xl = Xo exp( -')'t/2) COS(Wdt + cp), is the solution obtained for
the free (no external force) damped oscillator. As before
The integration constants Xo and cp are determined by the initial conditions.

The initial conditions are, however, not interesting in the present case. The
exponential factor exp( -')'t/2) makes the solution Xl (t) a so-called transient
which dies out, as t increases, leaving the solution as
15.6 The Forced Damped Harmonic Oscillator 399
X2 = Acos(wt - e). (15.24)

This so-called steady-state solution is reached when the transient has died
out. In the steady state the mass oscillates with the frequency of the applied
force and with a constant amplitude A. With the chosen sign for the phase
angle e in the term cos(wt - e), e gives the lagging of the oscillations of the
mass behind the phase of the applied force, Fo coswt. A negative value of e
will mean that the oscillations of the mass are ahead of the applied force in
phase.
We now determine the amplitude A and the phase angle e. We rewrite
(15.24) as follows:
X2(t) = A coswt cos 0 + A sinwt sin O. (15.25)
To determine for which values of A and 0 the expression (15.25) is a solution
for the equation of motion (15.22), we differentiate twice:
X2 = Aw( - sin wt cos 0 + cos wt sin 0) ,
X2 = Aw 2( - coswt cos 0 - sinwt sin 0) .
Substituting these results into (15.22) we obtain, after some rearrangements,
[_w2 A cos 0 + l'wA sine + w5Acos 0 - : ] coswt
+ [_w 2AsinO -l'wAcosO + w5AsinO] sinwt = 0

This equation must be true for all values of t. Consequently the expressions in
the parenthesis must be zero separately. We therefore get a pair of equations
in the two unknowns, 0 and A:
(w5 - w 2) cos 0 + l'W sin 0
o.E1..} .
rnA
(15.26)
(w5 - w 2) sinO -l'wcosO
From the second equation we see that 0 can be determined (up to multiples
of 7r) as a function of w. We fix 0 as the 0 < e < 7r solution of
cotO = (15.27)
for 0 < w < 00.

Squaring the two equations (15.26), and adding, we get
from which we isolate A:
(15.28)
The expressions (15.27) for cot e and (15.28) for the amplitude A give, to-
gether with (15.23), the complete solution of the problem of the damped,
forced harmonic oscillator. Note: Fo / (mA) is always positive.
15.7 Frequency Characteristics

A harmonic oscillator moves in response to an impressed external force F .=
Fo cos wt. The transients die out exponentially with time. The steady state
response is x(t) = A cos(wt - 8), where
"/W
tan8 = 2
Wo -w
2'
Figure (15.6) shows graphs of the amplitude function A(w) and 8(w), where
8 is the phase. The joint behavior of both amplitude and phase of a (har-
monic ally) forced oscillator, often summarized graphically, is known as the
frequency characteristics of the oscillator.
(a)
wi roo
()
I (b)
I
n ---------~-----------------
I
I
I
1t
T
Fig. 15.6. (a) Amplitude, (b) phase angle as function of w/wo for a lightly damped
Cr ~ wo) forced harmonic oscillator; w is the frequency of the external force F =
Focoswt
We shall consider some limiting cases, assuming always the weakly damp-
ing case, "/ « wo, or WOT » 1, where T == 1/,,/.
15.7.1 W «wo: A Low Driving Frequency

For a very low frequency of the external driving force, i.e., for w -t 0 we
obtain
Fo
A -t - - 2 ' and 8-t0.
mwo
The amplitude is the same as the amplitude for a static force and the oscil-
lations are in phase with the impressed, low-frequency force.
15.7 Frequency Characteristics 401
15.7.2 W » Wo: A High Driving Frequency
Fo
A ---t - - 2 ' and 0 ---t 7r
mw
Note: for w = Wo, 0 = ~.
15.7.3 W c:::.< Wo: Resonance
The case where the driving frequency w is close to or equal to the eigenfre-
quency Wo is called resonance. For w = Wo we get
A =~,
m,wo an
dO
= '2'
7r
The maximum response does not occur exactly for w = woo Let us determine
the frequency Wm for which the denominator of A has a minimum:
d~ [(w5 - W2)2 + ,2W2] = 0

2(w5 - w2)( -2w) + 2,2w = 0
wm = Wo J 1- ,22 .
2wo
(15.29)
The exact oscillation frequency for the free, damped harmonic oscillator was
J
found to be [see (15.6)]
Wd = Wo 1_ ,2 .
4w5
For weak damping, i.e., for, « wo, Wm ~ Wd ~ Woo At resonance the phase
angle 0 is 7r /2. While the displacement is out of phase with the force in the
case of resonance, we can easily prove that for 0 = 7r /2 the velocity is exactly
in phase with the force:
x = Acos(wt - 0)
i: = -Awsin(wt - 0), for 0 = ~
2'
= Aw coswt .
i:
That means: in resonance, i.e., for 0 = 7r /2, the velocity of the mass is in phase
with the impressed force, F = Fo cos wt. At resonance the work performed
by the applied force on the oscillating mass reaches a maximum. Due to the
perfect phase matching the mass gets pushed at just the right times and at
the right places, and the oscillator absorbs maximum energy per unit of time
from the applied force.
In the next section we shall calculate the time average of the power ab-
sorbed as function of the impressed frequency w.
15.8 Power Absorption
The time average of the work performed by the applied force on a driven
oscillator and for a given driving frequency w is
P = (F(t)i:(t)) = (Fo coswt( -1)Aw sin(wt - B)) .

The calculation of P is somewhat tedious, but the result is interesting and
important. Using the basic identities
sin(wt - B) = sinwt cos B - coswtsinB,

and
(cos wt sin wt) = 0,
we get
P -FoAw{(coswtsinwtcosB) - (cos 2 wtsinB)}

~FowAsine . (15.30)
2
e
The phase angle is decisive for the power absorption. To find the prod-
uct A sin e, we multiply the first equation in (15.26) by "(W, and the second
equation by (w5 - w 2). Adding the two resulting equations, we find that
. Fo/m
A sm B = "(W [(
"(W
)2 + (2
Wo - w2)2]
Substituting this into (15.30), we get
1 F6 "(2w 2
P= -- .,..,.---:-;:---'-~--== (15.31)
2 "(m [("(w)2 + (w5 - w2)2]
Figure 15.7 shows the power absorption as a function of frequency for a
weakly damped oscillator.
OJ
Fig. 15.7. Absorption of power for a weakly damped, driven harmonic oscillator.
F = Focoswt
15.8 The Q- Value of a Weakly Damped Harmonic Oscillator 403
For W = Wo the absorption of power becomes
The power absorption is reduced to one half the value at resonance for
(rW)2 1
(w5 - w2)2 + "'(2w 2 = "2 '
±"'(w = w5 - w2 .
We are interested only in frequencies W close to the eigenfrequency Woo We
write
"'(W = (wo + w)(wo - w) ~ 2wo(wo - w),
or
"'(wo ~ 2wo(wo - w),
"'(
Wo -w ="2.
For frequencies W2 = Wo - ",(/2 and WI = Wo + ",(/2 the absorption has de-
creased to half the value it has at resonance. The full width at half maximum
(FWHM) of the absorption line (see Figure 15.7) is 6w = 'Y. The damping
constant "'( thus determines the width of the resonance line.
15.9 The Q- Value of a Weakly Damped

Harmonic Oscillator
The resonance properties of a given weakly damped oscillator are often char-
acterized by the so-called Q-value.
Consider first an unforced, weakly damped harmonic oscillator. The en-
ergy dissipated during one oscillation is small compared to the amount of
energy stored in the oscillations. The Q-value for such an oscillator is defined
as 27r times the mean energy stored, divided by the energy dissipated per
period. (The factor 27r could have been avoided if the work done per radian
had been used instead of the work per period.)
Q is a useful number only if Q » 1, i.e., if we consider weakly damped
oscillations ("'( « wo) . Therefore we may assume that
For Q we obtain (remember T = 27r /w )

Q = 27r~ = WoE
PT p.
Q is a dimensionless number. Previously we found (see Section 15.4)

E
P == (Iv) = -,,(E = -- .
T
From this,
Wo
Q =WOT = -.
"(
Expressed by means of Q, the energy for a weakly damped oscillator

decreases according to the equation
~E = - Wo E
dt Q'
E(t) = E(O)exp (- ~t)
The amplitude response to an applied, static force Fo is

Fo
A(w) = A(O) = -2 .
mwo
The amplitude at resonance for a driven oscillator is
F
A(w) = A(wo) = -- .
"(mwo
We thus have
.Q _ A(wo) _ Wo _
- A(O) - "( - WOT .
Sometimes Q is called "the resonance amplification" or "the Quality factor"

of an oscillator. A high value of Q signifies high amplification, i.e., small
damping .
Considered as an amplifying system the weakly damped oscillator reacts
mainly in a frequency interval of width 6w = "( around the resonance fre-
quency woo This frequency interval is often called the band width. Let the
relative band width be called W :
W == 6w =.'l.
Wo Wo
For the amplification we found
Q = Wo .
"(
The connection between the amplification Q and the relative bandwidth W

is thus
QxW=l.
15.10 The Lorentz Curve 405
This connection is fundamental for all oscillating systems. It may be a mass

on a spring or an oscillator in a radio receiver. If we want a system to react
in a broad band of frequencies we must pay the price of a decrease in ampli-
fication. The Q-value for mechanical oscillators can reach values of 10 3 . This
is obtained for, say, a vibrating violin string. A church bell will have a some-
what lower value of Q. For electric oscillators higher values are obtained. For
vibrating electrons inside atom, i.e., "antennas" emitting light waves, Q is
about 10 7 . For nuclei emitting ,),-rays, Q-values of about 10 12 may be reached.
Q is sometimes also used to describe the shape of the resonance line. The
width of the resonance line is
Wo
')'=0·
The exact value of the frequency where maximum amplitude occurs is
W = woJ1 - 2~2 .
For Q = 1000
W = Wo J 1- ~ X 10- 6 ,
W ~ Wo { 1 - ~ X 10- 6 + .. j,
where W ~ Wo with a precision of about 1 in one million.
15.10 The Lorentz Curve

Consider again the result (15.31):
p =--
1 F6 ')'2w 2
2 ')'m (w5 - w2)2 + ')'2w 2
...,.---"..--=-=---..".--..".
We assume weak damping, ')' «wo. We are then normally interested only in
frequencies close to wo0 The often used approximation
w6 - w2 = (wo + w)(wo - w) ~ 2wo(wo - w)
gives
(15.32)
This is the so-called Lorentz curve of a resonance line for a weakly damped
oscillator. You will meet this result often in deeper studies of physics, partic-
ularly in the study of emission and absorption of electromagnetic radiation.
The result (15.32) describes for instance the shape of the so-called spectral
lines emitted by atoms. The shape is similar to the one shown in Figure 15.7.
From (15.32) it is easily demonstrated that the full width at half maximum
(FWHM) of the Lorentz curve is Lw = ')'.
15.11 Complex Numbers
The solution of the equation of motion for the harmonic oscillator becomes
more elegant if we employ complex numbers in the derivation. The equation
of motion is
.. . 2 Fo
x + ,x + WOX = - cos w .
t
(15.33)
m
We represent the applied force F(t) as the real part of a complex number (we
assume F is along the x-axis) as
F(t) = Fo expiwt , (15.34)
where in principle the amplitude Fo may itself be a complex number, Fo =

a + ib. The amplitude Fo is independent of time.
Analogously, we assume a solution x(t) of the form
x(t) = Xo exp iwt, (15.35)
where Xo is the (complex) amplitude. Inserting x(t) and F(t) in the equation
of motion we obtain
[(iw)2xo + iw,xo + w6xo] exp iwt = Fo exp iwt ,

m
or
Fo
Xo = -,...-".2----=---- (15.36)
m(wo - w2 + irw) .
The new complex amplitude Xo is again proportional to the amplitude of the
force, Fo. We choose the phase angle of the force such that I Fo I = Fo is a
real number.
That means F(t) = Fo exp iwt = Fo coswt+iFo sinwt. We write Xo == CFo,
where the complex factor C is
C _ 1 (15.37)
- m(w5 - w2 + i,w) .
The factor C may be written as
C = pexpie.
15.11 Complex Numbers 407
The solution then becomes
x(t) = pFo exp i(wt + e) . (15.38)

The solution is the real part of (15.38), i.e.,
Re{x(tn = pFo cos(wt + e).

To determine p we return to (15.37):
CC* = p2
(C* == complex conjugate of C).
1 1
m(w5 - w2 + i'yw) m(w5 - w2 - if'w)
1
m 2 [(w5 - w2 )2 + f' 2 w2 ] .
The phase angle e may be found in the following way:

-1 = -1 exp - I'e = m [2
Wo - w2+ If'W
. ],
C P
f'W
tan( -e) =
w5 - w2 '
f'w
tane
Note. Here we have used a sign convention for e that is the opposite of the
one used in Section 15.6.
The solution of the equation of motion becomes

. Fo
x(t) = Re {pFo exp I(Wt + en = cos(wt + e) ,
mJ(w5 - w2 )2 + f' 2W2
the same as given in Section 15.6, except for the sign convention on e.
Clearly the use of complex numbers substantially shortens the calcula-
tions. Compare this Section with Section 15.6.
15.12 Problems
Problem 15.1. A homogeneous thin rod of length 1 m may turn without

friction around a horizontal axis through one of its endpoints. The accelera-
tion of gravity is g.
(1) Find the period T for small oscillations.
Assume now that the rod is rotating around a horizontal axis through a point
at a distance x from one of the ends of the rod.
(2) Determine x such that the period T is as small as possible, and give the
value of T.
Problem 15.2. Consider damped harmonic oscillations. Let the coefficient

of friction I be half the value of the one that just gives critical damping.
(1) How many times is the period T larger than it would be for I = O?
(2) Determine the ratio between two successive swings to the same side.
Problem 15.3. Consider an undamped harmonic oscillator (, = 0) with

eigenfrequency woo An external driving force F = Fo coswt ( where w -=I Wo )
is applied. The equation of motion is
.. Fo
x +w5x = - coswt.
m
The general solution may be written as
. Fo coswt
x(t)=Acoswot+Bsmwot+- 2 2'
mwo -w
(1) Determine the initial conditions such that the undamped oscillator will
begin steady state motion immediately.
(2) Determine the value of Xo and () in
x = Xo cos(wt + ())
for the case described in question (1).
Problem 15.4. Consider a damped, forced harmonic oscillator.
(1) Calculate the work performed by the frictional force at resonance, and
show that it is equal to the work performed by the impressed force at
resonance.
15.12 Problems 409
(2) A body is suspended from a spring. The period of oscillations is T = 0.5 s.

An impressed force F = Fo cos wt acts vertically on the body.
The amplitude Fo is Fo = 0.1 N. A frictional force f (proportional to the

velocity of the mass) acts on the body too. The amplitude at resonance, A r ,
is 5 cm. Determine the damping constant b == m"(.
Problem 15.5. A tuning fork vibrates with the frequency v = 440 cycles per
second (w = 27rV = 27r x 440 s-l). The tuning fork emits 1/10 of its stored
energy in 1 second. Determine the Q-value of the tuning fork.
16. Remarks on Nonlinearity and Chaos
16.1 Determinism vs Predictability
In Newtonian mechanics, dynamical systems are described by differential

equations which, if supplied with initial conditions, can be integrated to a
unique solution at any future time t. A unique solution means that for each set
of initial conditions there exists only one solution. A set of initial conditions
determines exactly one solution. Dynamical systems having this property are
called deterministic. See Example 12.8.
In simple systems such as occur in the examples and problems of the pre-
ceding chapters, the solution can be written explicitly in terms of elementary
functions containing the initial conditions as parameters. The correspondence
between the initial conditions and solutions is then quite transparent. One
may come to suspect that the correspondence is always this 'regular', in par-
ticular that two sets of initial conditions that are almost identical will also
correspond to two solutions that are almost identical for all times.
regularity chaos
initial
conditions
'i i
./ 1 i }l\
solutions
Fig. 16.1.
This, however, is not true in general. Many, even conceptually simple, dynam-
ical systems have the property that their solutions depend in an extremely
sensitive manner on the initial conditions, in such a way that nearly iden-
tical initial conditions quickly evolve to completely different solutions. The
number of digits needed to specify an initial condition in order to make a
meaningful prediction a time t ahead, can be an exponentially increasing

412 16. Remarks on Nonlinearity and Chaos
function of t. If, say, one additional decimal place is required in the specifi-
cation of the initial values for each second ahead that one wishes to predict,
the exact solution quickly becomes impossible to predict from an integration
of the equations of motion. Dynamical systems with this property are loosely
termed 'chaotic', and have recently (not least spurred by the evolution of
inexpensive, high-precision, high-speed computers) been the focus of muoh
attention.
A complete understanding and classification of all such systems seems be-
yond the abilities of present-day mathematics, even though much progress has
been made. The problems involved have forced mathematicians and physi-
cists to re-evaluate the very concept of what we mean by the 'solution' to a
dynamical system. In the study of such 'chaotic' systems, the emphasis has
shifted towards obtaining qualitative, global information about the system,
rather than seeking the type of detailed and local solutions discussed in the
preceding chapters.
The mathematical techniques involved in gaining even a qualitative un-
derstanding of chaotic systems are fairly sophisticated. In this chapter we
shall merely try to give the reader a glimpse of some of the aspects of the
topic.
It it essential, however, to be aware of the following fact: even the most
I chaotic' macroscopic system obeys all the fundamental laws of Newtonian
mechanics, summarized in Chapter 12. Recent advances in mathematical and

computational techniques have expanded our insight into the many complex
types of behavior that solutions to the equations of motion may have. But
the principles and conservation laws that we use to establish the differential
equations are unchanged. Mechanical systems remain governed by laws and
concepts laid down by Isaac Newton more than 300 years ago, and extended
by Albert Einstein early in the 20th century.
16.2 Linear and Nonlinear Differential Equations

The mathematical description of mechanical systems that you have seen in
the preceding chapters, has been in the form of ordinary differential equations.
In mathematics courses you will learn about the many interesting properties
of this class of differential equations. It is important to distinguish between
linear and nonlinear differential equations. An ordinary differential equation
in the unknown function x(t) is said to be linear, if the differential equation
only contains terms linear in x( t) and its derivatives, i.e., if it can be written
in the following form:
dnx dx
an(t) dtn + ... + al(t) dt + ao(t)x + f(t) = 0, (16.1)
where the functions ai(t) and the function f(t) are arbitrary. Ordinary dif-
ferential equations that are not linear, are called nonlinear. One can subse-
16.2 Linear and Nonlinear Differential Equations 413
quently distinguish all mechanical systems (or, more precisely, all models of
systems) into two classes: the linear and the nonlinear systems.
Example 16.1. Superposition. If the x(t)-independent term, J(t) is ab-

sent, so that (16.1) takes the form
dnx dx
an(t) dtn + ... + aI(t) dt + ao(t)x = 0 , (16.2)
the equation is said to be homogeneous. The single most important property

of a homogeneous linear ordinary differential equation (sometimes taken to
be the defining property) is the fact that any two solutions can be added to
form a new solution. If x(t) solves (16.2), and yet) solves (16.2), then any
linear combination z(t) = Ax(t) + By(t), where A and B are constants, will
also solve (16.2) .
An x(t)-independent term, J(t), is sometimes present on the right-hand
side (making the equation inhomogeneous), and interpreted as a 'driving'
or 'input' term. Consider now two linear equations that differ only in these
driving terms:
dnx dx
an(t) dtn + ... + aI(t) dt + ao(t)x !let) ,
dnx dx
an(t) dtn + ... + aI(t)di + ao(t)x f2(t) .
If XI(t) solves the equation with !let), and X2(t) solves the equation with
f2(t), then
will solve the equation

dnx dx
an(t)-d
n t
+ ... + aI(t)-dt + ao(t)x = !let) + f2(t) .
This additive property is called the principle of superposition. If an arbitrary
function J(t) on the right-hand side can be written as a sum
N
J(t) = L In(t) ,
n=I
where a solution of the equation with each of the Jn on the right-hand side
is known, we can simply sum these, to form the solution to the equation
with J on the right-hand side. For instance, as mentioned in Chapter 15,
every periodic driving function J can be approximated very accurately by a
linear combination of harmonic functions. Thus, if we can find the solution
to the differential equation with any harmonic function as the driver, we
can superpose these solutions to approximate the solution with the arbitrary
periodic right-hand side function f . 6.
The class of nonlinear differential equations do not in general permit su-

perposition of their solutions (indeed, a simple test for whether a given ho-
mogeneous differential equation is linear or not is to check if an arbitrary
linear combination of two solutions satisfies the equation). It is among the
class of nonlinear equations that we shall find examples of chaotic behavior.
Geometry becomes an important tool in this study.
16.3 Phase Space
The form of Newton's second law suggests that many mechanical systems can
be modeled by second order ordinary differential equations. For each coordi-
nate q used to describe the configuration of the system, one can (formally)
regard its time derivative q = w as an independent variable w. This allows
us to write, say, an equation of motion
mq = F(q)
as a system of first-order equations
dq
w ,
dt
dw 1
-F(q) . (16.3)
dt m
The right-hand side of this system describes a vector field in (q, q)-space, and
this vector field can, for each set of initial conditions, be integrated to the
curve for which the tangent vector at each point is the specified vector field.
Along such a curve within the (q, q)-space (often called the phase plane or
(qo~q)
q
Fig. 16.2. Phase space
the phase space),l one can track simultaneously both the value of the variable
1 Many texts refer to this as position-velocity space and reserve the term phase
space for a more general construction, involving the so-called canonical momen-
tum. In Selected References see, e.g. , the book by Goldstein.
16.3 Phase Space 415
q and the value of its time derivative q for each solution to the equation of
motion. The time evolution of all the phase-space points is called the phase
flow. See Figure 16.2.
When the force F does not depend on time (as in the system described
by equation 16.3), one can find the location of each phase space curve with-
out integrating the system (16.3). For, as discussed in Chapter 8, the total
mechanical energy E, which is the sum of kinetic and potential energy:
E(q,q) = ; q2 + U(q) ,
is constant along the curve in phase space. Consequently, each phase space
orbit is a collection of points where the function E(q, q) is constant, i.e., (part
of) a level curve for the function E(q, q). The particular value of the energy
on each curve is set by the initial conditions.
Example 16.2. The Simple Harmonic Oscillator. For the simple har-
monic oscillator,
mij + kq = 0,
the mechanical energy

1 1
E = _mq2 + _kq2
2 2
is constant along each solution curve. The points in (q, q)-space, for which
k 2 m' 2
2E q + 2E q = 1,
form an ellipse, with semi axis J2E/k and J2E/m respectively. For the
harmonic oscillator, each solution curve in phase space is an ellipse. D.
A solution curve can consist of a single point. Such a point is called an

equilibrium point, and is necessarily located on the q-axis (why?).
Equilibrium points correspond to stationary points of the potential energy
function U. If the potential energy function has a local maximum, the equi-
librium point is unstable (neighboring phase space points are carried away
by the phase flow), and if the potential energy function has a local minimum,
the equilibrium point is stable (neighboring phase space points are not car-
ried away by the phase flow). For an intuitive picture, consider the motion
of a ball rolling on the 'hills' of the potential energy function. The ball will
be in stable equilibrium at the bottom of an energy 'valley', and in unstable
equilibrium if resting on a 'peak'. See Figure 16.3.
Note. For the equations (16.3), two different solution curves in phase space
will never intersect or even touch. This is a simple consequence of the unique-
r U(q)
Q
I
I
I
I
I
I
I
I
I
I
I
'-'
),+.,.(
I
I
I
I
I
I
I
I
I
q
Fig. 16.3. Maxima of the potential function U give rise to unstable equilibrium
points, and minima give rise to stable equilibrium points
ness property for the solutions. If two different solution curves did have one
point in common, such a point could be seen as one initial condition evolving
to two separate solutions of the differential equation, in violation of unique-
ness.
Example 16.3. Phase Space ofthe Pendulum. Consider again the math-
ematical pendulum (see Section 3.4). We now assume that a mass m is fas-
tened to the end of a stiff rod. We neglect the mass of the rod.
Fig. 16.4.
The configuration of the system is uniquely specified by the angle () formed

with the vertical. The equation of motion,
mlB = -mg sin ()
or,
l B+ 9 sin () = 0
may be re-written as a system:
iJ = w
tV = - 9 / l sin () }. (16.4)
We first consider the so-called linear (or simple) pendulum, where we assume
that the amplitude in the oscillations are so small that sin 8 :=::: 8 (small-
oscillation approximation; see Section 3.4). With this simplification, the above
equations take the form
8=w
tV = -gil 8 } . (16.5)
If we choose 8(0) = 0 and w(O) = wo, the equations can be integrated to
8(t) woyTTgsin ((ft)
w(t) Wo cos ( (ft) .
This may be immediately recognized as a parameter representation for an

ellipse, with semi axes Wo and woffg respectively. For each set of initial
conditions, the solution traces out an ellipse in the phase plane. The period
of oscillation is
Figure 16.5 shows a few of these ellipses, corresponding to different values of

Wo = 8(0).
Fig. 16.5. Phase space for a pendulum performing small oscillations
Next we investigate what happens if we do not assume a small amplitude,

and consider the more general set of equations (16.4).
The phase-space curves of the nonlinear pendulum fall into two classes.
We first discuss the curves that resemble those of the linear pendulum and
cross the 8 axis. When a phase-space curve crosses the 8 axis, the pendulum
has no velocity at this instant; the pendulum has then swung as far as its
initial energy allows, and is turning back (for this reason such points on the
solution curves are called turning points). These curves are called libration
curves. Suppose a particular solution has turning point at 80 . The equation
for the mechanical energy along this phase-space curve can then be written
as
1
"2m(te)2 + mg(l - l cos e) = mg(l - l cos eo)
or,
9
2y(cose - cos eo)
4y9 ( sm eo
. 2 --:2 - sm
. 2
"2e)
Thus,
2 !f
Vy
dt =
. /sin 2 02
de
0 _ sin 2 rt.
V 2
Placing the origin of time where e= 0, we integrate to
2
Vy
!f t = foO -----r==d=e===
}0 V
. / sin 2 00 _ sin 2 rt.
2 2
We now perform a change of variables. A new variable angle ¢ is defined

through the equation
. A.
sin rt.
2
sIn,!, = --0- .
sin ~
Changing to the ¢ variable in the integral, we get
l' Vhi:':'¥ ,in' ¢
sn- I (sin¢,sin e;) ,

where the function sn is a Jacobi elliptic function. 2 Inverting, we have that
sin¢(t) = sn( If t,sin ~o) ,

so the e(t) curve is given through the equation:
. . eo
sme(t)=sm--:2 sn vyt,sm--:2
(fii:. eo)
The elliptic function 'sn' oscillates between +1 and -1 in a manner similar

to that of the elementary sine function, and consequently sine(t) oscillates
2 See, e.g., D.F. Lawden (1989): Elliptic Functions and Applications, Applied
Mathematical Sciences (vol. 80), Springer, Berlin.
between + sin Bo and - sin Bo. The period T is equal to four times the duration
needed to reach the first turning point (¢ = 7r/2):
4 r/ 2 d¢
T
Vi7l io VI - sin 2 e~ sin 2 ¢
4
CTi K(k) ,
V gil
with k = sin 2 e~. The function K(k) is the complete elliptic integral. As the
angle Bo ----; 0, K(sin 2en - - ; 7r/2 . Thus, for small values of Bo (i.e., small
oscillations), we recover the familiar oscillation period. For arbitrary values
of Bo, the curves qualitatively resemble the ellipses of the linear pendulum,
except that the oscillation periods are no longer independent of the ampli-
tude, rather, the period increases with increasing amplitude. For Bo ----; 7r ,
T ----; 00 (see the discussion of the separatrix, below).
The other class of curves in phase space have no turning points and con-
nect the (B = -7r) line with the (B = +7r) line. Since these two lines represent
the same configuration of the pendulum, the curves that connect them with
nonzero velocity correspond to rotational states of the pendulum, i.e., the
mass moves around in a vertical circle. Such orbits are known as rotation
orbits. In the positive 8 half plane the pendulum rotates counter-clockwise;
in the negative 8 half plane the pendulum rotates clockwise.
For these solutions, the kinetic energy never vanishes. Writing the con-
servation of mechanical energy as:
~e2
2g
- cos B = >. ,
where>. is a constant, we see (since 82 is never zero) that the value of>. must
be greater than 1 for these curves. Defining a parameter k by k = }2/(>' + 1),
we have that k must vary between 0 and 1; the greater the energy, i.e., the
larger the value of >., the closer k is to O. Introducing a new angle variable 'IjJ
by 'IjJ = B/2, we write the energy equation as
.2
'IjJ = kg2 l ( 1 - k 2sm
·
'IjJ) .
Separating variables and integrating, we get, in analogy with the libration

orbits,
or,
. B(t)
sm 2 = sn
(fg)
Vk2i t, k .
420 16. Remarks an Nonlinearity and Chaos
This leads to a period of
T = 2 k {{ K(k) ,
where again K(k) denotes the elliptic integral. For k -+ O, the period T -+ O;
these solutions have large kinetic energies and correspond to the pendulum
rotating with large angular velocities.
Separating the two classes of solutions is a pair of curves (called jointly
the sepamtrix) , which at first sight seems to violate the rule that distinct
phase-space curves never touch or intersect. The top half of the separatrix
represents the solution for which the pendulum starts out in an upright (in-
verted) position, then infinitely slowly begins to swing counter-clockwise,
completes a full circle, comes back to the upright position with zero velocity
and comes to rest there. The lower-half solution is similar, except that the
motion takes place in a clockwise direction.
Fig. 16.6. Phase space for the nonlinear pendulum
Both solutions take an infinitely long time to complete the full revolution;
most of the time being spent near the upright position where the angular
velocity drops to zero. The curves do not touch at any finite time value, but
rather originate and terminate at the unstable equilibrium point e = ±7r as
t -+ ±oo, and uniqueness is not violated.
The pendulum is nonlinear, but it is not chaotic. All solutions can be given
explicitly in terms of analytic functions . All orbits (except the two separatrix
solutions) are periodic, and initial conditions that are sufficiently 'close' will
stay 'close' for many oscillation periods.
Note also that ne ar the origin the phase-space plot for the nonlinear pen-
dulum coincides with the phase-space plot for the linear pendulum model.
This is as it should be. Nevertheless, the phase portrait for the non linear
pendulum contains, as we have seen, features that could never be guessed
from the linear model. /:::,
Example 16.4. Bifurcation in a Nonlinear Model. Consider a particle

having mass m and being able to move without friction on a circular hoop of
radius R. The hoop is rotating about a diameter, with an impressed angular
velocity w. This angular frequency is a parameter that can be changed at will.
A homogeneous gravitational field g, parallel to the axis of rotation, acts on
the system. See Figure 16.7.
I'
Fig. 16.7. Particle on a rotating hoop
The position of the particle is specified by the angle B. Three forces act on
the particle: the gravitational force, the centrifugal force, and the reaction
force from the hoop. The reaction force from the hoop has no component in
the tangential direction. Thus, the motion is governed by the competition
between the tangential component, -mg sin B, of the force of gravity (pulling
the mass towards the bottom of the hoop), and the tangential component,
(mw 2 R sin B) cos B, of the centrifugal force.
In the frame rotating with the hoop, Newton's second law for the tangen-
tial acceleration is
mRB -mgsinB + mw 2 RsinBcosB.

This is a nonlinear differential equation in B. Note, that when w = 0 , the
system becomes the familiar mathematical pendulum, the hoop being simply
another way to implement the constraint on the position of the mass.
An equilibrium position for the mass exists where B= 0, or
2
w sin B (cos B - R~2) = 0 ,
i.e., where the two force components in the B direction exactly balance, or
both vanish. The latter only occurs when sin B = 0, i.e., at the bottom or the
top of the hoop.
For sin B -# 0, the angle B = Be will be an equilibrium position if
9
cos Be = Rw 2 .
This equation has a solution for Be only if w2 is so large that

equilibrium
position
~
Fig. 16.8.
9
Rw2 :::; 1,
or,
The critical value, W = J 9 / R , of the impressed angular frequency happens

to coincide with the value of the frequency for small oscillations of the particle
at about the bottom of the hoop.
Fig. 16.9. Equilibrium positions as a function of the impressed rotation frequency

W
For values of w greater than the critical value, the mass will be in equilibrium
at the positions (symmetrical about e = 0 ):
ee = ±arccos (R~2)
Figure 16.9 shows the ee equilibrium position(s) as a function of the impressed
hoop rotation parameter w.
e
For small values of w, only the = 0 position is an equilibrium. Then, at
the critical value of w, two new equilibrium positions appear, symmetrically
around e = O. For W -> 00 , the two new equilibrium values ee approach
±7T /2. Let us examine the motion around each of the equilibrium positions.
For fixed w, the system has, in the rotating frame, a potential-energy function
Fig. 16.10. The potential energy and the phase-space flow for w less than the
critical value
Consider first the phase-space plot for the system for small values of w.
For w exactly equal to 0, the system is, from a mechanical point of view,
identical to the pendulum. For w « Vg / R, the potential energy U still has
a local minimum at B = 0, and the phase-space curves remain similar to that
of the pendulum. For higher values of w, the phase-space curves begin to
IJ
Fig. 16.11. The potential energy and the phase-space flow for w greater than the
critical value
deform, and when w passes through the value Vg / R, two new foci appear,
symmetrically placed around the origin. The potential-energy function U now
has a local maximum at B = 0, and local minima at ± arccos (ihi) .
The two new equilibrium points are consequently stable, whereas the ori-
gin has turned into an unstable equilibrium point. On Figure 16.9, this is
indicated by the line at Be = 0 being dashed to the right of the critical w
value. We say that the stable equilibrium point at the origin has undergone
a bifurcation to two new stable equilibrium points.
The analysis and classification of such bifurcations is of immense value

in the study of nonlinear differential equations. We refer the reader to the
relevant texts given in the Selected References. L:,
16.4 A Forced, Damped Nonlinear Oscillator
In Chapter 15 we saw that the equation of motion for the periodically forced,
damped harmonic oscillator is:
F.
i + ,x + w5x = -.!.l. cos wt .
m
Here, m is the mass of the particle, , is the coefficient of friction, Wo is the
oscillation frequency of the unforced and undamped oscillator, and Fo is the
amplitude of the impressed force. The solutions are well understood and can
be expressed in terms of elementary functions, for all values of the parameters
in the equation.
We shall now replace the linear restoring force F = -mw5x , with a
different type of restoring force. This is a nonlinear restoring force of the
form of F = ax - (Jx 3 , where a and (J are positive constants. A restoring
force with this dependence on x makes the origin x = 0 into an unstable
equilibrium point, and the two symmetrically placed points x = ± j al (J
into stable equilibrium points, about which the unforced mass may oscillate.
A mass moving on a rotating hoop (see Example 16.4) is subject to a force law
of similar type when the impressed rotation frequency exceededs the critical
value.
With this form for the restoring force, the forced, damped nonlinear os-
cillator has an equation of motion of the form of
F.
i + ,x - 'f/X + bx 3 = -.!.l. cos wt , (16.6)
m
where'f/ = aim and b = (JIm. A nonlinear equation of this type is known as

a Duffing equation, after the German engineer H. Duffing who early in this
century studied oscillators of this type.
We do not attempt to solve the Duffing equation. vVe give only a brief,
qualitative description of some of the solutions of the Duffing oscillator (16.6),
in order to give a feeling for the problems encountered.
Let us first consider the undamped (r = 0) and unforced (Fo = 0) case.
With no forcing or damping, the system is conservative, and the phase-space
curves of the solutions have a fairly simple appearance; see Figure 16.12.
All motions are bounded, and the motions that encircle both stable equilib-
rium points are separated from the motions that encircle only one equilibrium
point by a pair of separatrix curves, which start from and terminate on the
unstable equilibrium point at the origin.
16.4 A Forced, Damped Nonlinear Oscillator 425
Fig. 16.12. Phase-space curves of the undamped, unforced, nonlinear oscillator
If we introduce damping (friction) into the system, the phase-space por-

trait changes character. With the exception of two solution curves that termi-
nate at the unstable origin with zero energy, all phase-space curves will now
eventually lose energy and spiral in towards one or the other of the stable
equilibrium points. Two separatrix curves terminate on the unstable origin.
See Figure 16.13. There is no chaos in this system.
Fig. 16.13. Phase-space curves of the damped, unforced, nonlinear oscillator
Finally we shall briefly describe some of the solutions to the full (i.e.,
forced and damped) Duffing equation. The time-dependent forcing puts me-
chanical energy back into the system in competition with the friction force.
For some initial conditions, this competition finds a balance, leading to a
periodic orbit qualitatively like that of the unforced, undamped oscillator.
For other initial conditions, however, the solution curve winds about in a
completely nonperiodic fashion (Figure 16.14).
The fact that such a solution curve has self-intersections in (x , x)-space
is a consequence of the time dependence of the system (in (x, x, t)-space the
solution curves do not cross). Two solution curves that begin at nearly the
same point in phase space (i.e., with nearly similar initial conditions) quickly
evolve into quite different solutions. The system is deterministic, but unpre-
dictable, due to the fact that there is a limit to the precision with which the
initial values can be specified. What this limit is, may have some dependence
426 16 . Remarks on Nonlinearity and Chaos
Fig. 16.14. A single phase-space curve of the driven, damped Duffing oscillator
on the exprimental procedure used to determine the initial conditions, but

the existence of such a limit is unavoidable.
Clearly, not much can be gained from studying particular non periodic so-
lution curves. Instead, we shall turn to a technique known as a Poincare sec-
tion, after the French mathematician H. Poincare (1854- 1912) . In the present
case, this amounts to marking, instead of the complete solution curve, only
those points for which the time variable is an integer number of, say, the
period 27l' /w of the forcing (this is a natural time unit in the system).
Instead of a continuous curve, we get a sequence of dots in phase space.
After many periods, one finds that the dots settle into a fairly narrow region
of phase space (see Figure 16.15) and wander about only within this region.
.. , -,- ..
r!J ., ;;:--;:"i;
~-,
1; x
~: .'
....
-',-,
'.
..' ~
\."~~~~.~-".o:' "
Fig. 16.15. A 'strange attractor' in the Duffing system
Remarkably, many different initial conditions eventually display the same

Poincare section, i.e., the dots an the solution curve are eventually attracted
into the same region in phase space. For this reason, such a region is called an
attractor. Moreover, computer studies indicate that this attracting region has
unusual geometric properties; we shall not go into these here, but merely note
that they have given rise ta the term strange attractors being used about this
and similar point sets. Ta get qualitative information about the long-term
behavior of the system, one must study the motion an the attractor. Often,
16.5 Liapunov Exponents 427
the map from One dot of the Poincare section to the next is studied in its
own right.
For many initial values, the final state of this forced, damped, nonlinear
oscillator is chaotic: a nonperiodic motion on an attracting set, with sensitive
dependence on initial conditions.
16.5 Liapunov Exponents

In the phase space for a system, we pick two points close to each other and let
each of these points be the initial conditions for a solution curve. See Figure
16.16. In the figure, phase-space curves evolving from such a pair of initial
conditions are drawn.
.
q
Fig. 16.16. Phase-space curves evolving from two initial conditions
The distance d between the phase-space points at time t tends to increases

as time passes. If this distance d increases exponentially with time, i.e., if
d = do exp (At) ,
where do and A > 0 are constants, the system will evidently have a sensitive
dependence on the initial conditions. The positive constant A gives a quan-
titative measure of the rate at which solutions corresponding to two nearly
identical initial conditions are separated by the phase-space flow. This is the
idea behind the so-called Liapunov exponents.
Consider in phase space a flow that is generated by the set of equations:

q h(q,w) ,
w fz(q, w) .
Let (qo(t), wo(t)) be a particular solution to this system. A measure of how
rapidly nearby orbits separate from (qo(t), wo(t)) can be extracted from the
corresponding linearized system (the linear approximation to the above flow,
around (qo (t), Wo (t) ). This linearized system is given by:
[ !] [
!liJ.
8w
f!h.
8w
where the elements of the matrix are to be evaluated along the solution
(qo(t),wo(t)). Next, let e be a unit vector in phase space at (qo(O),wo(O)).
See Figure 16.17.
e(t)
10
Fig. 16.17. The linearized flow evolves the vector e to the vector e(t)
The linearized system will evolve the vector e forward to a new vector e(t).
The Liapunov exponent is designed to extract the average exponential growth
rate (if any) of the size of the vector e(t) , as it evolves with the phase-space
flow.
We therefore define the Liapunov exponent A(qO ,wo)(e) , for the orbit
(qo(t), wo(t)) and the direction e, to be:
. 1
A(qO , wo)(e) = hm -In 1 e(t)
t--+ oo t 1 ,
where 1 e(t) 1 denotes the magnitude of the vector e(t).

Consider as an example the following (linear) system
(16.7)
As this system is linear, we can immediately write down the solution 1 e(t) 1
for any initial vector e = (el' e2) :
[ el(t) ] = [ exp(at) 0 ] [e 1 ] = [ el exp(at) ]

e2(t) 0 exp(bt) e2 e2 exp(bt)
The magnitude of the vector e(t) is consequently:
1e(t) 1= Jei exp(2at) + e~exp(2bt).

If, say, a > b, and el of. 0, we have that
A(e) lim
t--+oo
~t In (exp(at) Vlei + e~exp(2(b - a)t))
. In (d + e~exp(2(b - a)t))
a+ hm =a.
t--+oo 2t
16.6 Chaos in the Solar System 429
Thus, for the flow (16.7), the above limit will extract the dominant expo-
nential rate a among the directions for e. This is true in general. For an
n-dimensional system, one can find, by varying the vector e, up to n distinct
exponents.
For nonlinear systems, Liapunov exponents are mainly accessible through
numerical studies, 3 and can be defined in a manner intrinsic to Poincare
maps. The existence of a positive Liapunov exponent is generally taken to be
one of the indicators for chaos.
16.6 Chaos in the Solar System
With the realization that all the planets in the solar system interact with
each other, slightly perturbing each other's orbits, the question arises whether
or not these interactions might over time lead to significant changes in the
orbital parameters for a given planet. One might imagine that many small
perturbations might accumulate and cause a planet to fall into the Sun, or,
in the other extreme, to escape from the solar system, drawing its energy to
do so from all the other planets combined.
Between the orbits of Mars and Jupiter (see Figure 16.18), a vast number
of masses ranging in size from large rocks to small planets orbit the Sun.
Though some of the largest are known by individual names, they are all
collectively called asteroids, and the most densely occupied region between
Mars and Jupiter is known as the asteroid belt. Each of these small masses
orbits the Sun in accordance with Kepler's laws. But each small mass is
also subjected to a time-varying gravitational field, particularly from Jupiter
(which is by far the mo'st massive of the planets; see Section 14.8) .
Mercury Venus Earth Mars Asteroids Jupiter

Fig. 16.18. Most asteroids have orbits between the orbit of Mars and the orbit of
Jupiter
Responding to the effective potential (see Section 14.6), the radial motion
of a planet or an asteroid is, disregarding the influence from other planets, an
oscillation around a mean radius. This oscillation is practically undamped.
Consider now the time-varying gravitational tug on an asteroid from
Jupiter. The radial equation of motion for the asteroid is now roughly like
that of the weakly damped, forced oscillator (see Section 15.5). The magni-
tude of the gravitational disturbance from Jupiter is quite small compared
to that from the Sun, but magnitude alone is not decisive. In Section 15.5 we
3 see, e.g ., Bennetin et ai, Meccanica, 15 (1980).
saw that a weakly damped oscillator reacts very strongly to a time-varying

force if that force oscillates with a frequency near the natural frequency of
the oscillator. This is the phenomenon of resonance.
How can the asteroid be in resonance with Jupiter, when, as we know from
Kepler's third law, they have different semi-major axes and subsequently dif-
ferent periods? The answer is that a rational relation between the orbital
periods of the two bodies will lead to a similar phenomenon. If, say, Jupiter's
period is 5/3 that of the asteroid, the two bodies will return to exactly the
same relative position for every 5 orbital revolutions of the asteroid. This
type of resonant interaction can, over many orbital periods, lead to a signifi-
cant change of, e.g., the excentricity of the orbit of the asteroid, causing the
asteroid to depart from its nearly circular orbit in the asteroid belt.
a (AU)
2 2.5 3 3.5
t
1/3
tt
2/5 3/7
t
1/2
Fig. 16.19. The number n of asteroids as a function of the semi-major axis a of
their orbits. The most important resonances (period of asteroid/period of Jupiter)
are indicated
The so-called Kirkwood gaps (named after the astronomer who discov-
ered them) are regions in the asteroid belt where the density of asteroids is
remarkably low compared to other regions.
A plot of the number of asteroids as a function of the radius (Figure
16.19) clearly shows the Kirkwood gaps. On the figure some of the values of
the semi-major axis that (through Kepler's 3rd law) correspond to a rational
ratio between the period for an asteroid at that location and the period of
Jupiter are indicated.
How will the planets affect each other? About 100 years after Newton,
the French mathematician Lagrange (1736-1813) demonstrated that it would
take at least several millions of years for a dramatic perturbation of, e.g.,
the Earth's orbit to occur. Thus, from the time perspective of the human
civilization, the solar system may be considered as being stable.
Nevertheless, the mathematical question of determining the stability of
the planetary orbits for all time, still remained open . In 1885, the Swedish
16.7 Problems 431
King Oscar II announced a prize contest for a mathematical proof that

the planetary orbits would remain stable forever. The prize was awarded
to Poincare who submitted, not the requested proof, but a remarkable in-
vestigation into the properties of the simplest nontrivial problem, namely,
the problem of three masses interacting gravitationally. Poincare's pioneer-
ing work laid the basis for most of the later studies into what is now known
as the theory of deterministic chaos.
Even today, the issue about the long-term stability of the solar system
is not settled permanently. Recent extensive numerical calculations 4 strongly
suggest that the orbital parameters of most planets vary slowly, but chaot-
ically. The axes of rotation for some planets (e.g., Mars) also seem to be
undergoing chaotic fluctuations in their directions.
16.7 Problems
Problem 16.1. Which of the following (systems of) differential equations

are linear?
(a) exp(t)x - x + sint x = 0,
(b) x-2xx+tx=0,
(c) X = w, W = cos wt x + 2w ,
(d) x = w, W = sin x ,
(e) d:~~X) + (E - V(x)) 1jJ(x) = 0 .
Problem 16.2. Sketch qualitatively the phase-space curves corresponding

to the potential functions below:
(a) U(x) = 5Jx 2 +5- 0.5x 2 ,
(b) U(r) = -~ + r12' r >0,
(c) U(r) = --it for 0 < r < Rand U(r) = -~ for 0 < R < r < 00 .
Problem 16.3. For the nonlinear pendulum (Example 16.3), show that, if
e
we choose = 0 for t = 0, motion on the separatrix is given by
4 see, e.g., Laskar and Robutel, Nature 361, 608, (1993), or J. Wisdom, p. 109 in
Dynamical Chaos, M. Berry et al. (eds.) Royal Society of London (1987).
432 Remarks on Nonlinearity and Chaos
Bsep(t) = 4 arctan [exp (If t)] - 7r for - CX) t

< < CX).
Problem 16.4. Show that the Liapunov exponents for the harmonic oscil-
lator
[ ~] [
all vanish.
Problem 16.5. How many days would the Earth year be shorter, if the
Earth were to be in exact 1/12 resonance with Jupiter?
Appendix. Vectors and Vector Calculus
Addition and Subtraction of Vectors

Physical quantities possessing both magnitude and direction can be described
mathematically as vectors. A vector can be visualized as an oriented line
segment (an arrow). Vectors can be added (and subtracted) to form other
vectors, and they can be multiplied by real numbers (see the figure below).
-0
b 3b
The Dot Product, a· b

Let a be a vector with length a, and b be a vector with length b. The dot
product a· b, also known as the scalar product, is defined by
a· b == abcos(a, b),
where cos(a, b) denotes cosine of the angle between the two vectors. Note
that no coordinate system is involved in the definition of the scalar product.
The value of the scalar product does not depend on the orientation of the
angle (a, b) (why?). From this, we see that
a·b=b·a.
The scalar product is commutative. If a . b = 0, we say that the vectors are
orthogonal. And, since cos 0 = 1, we have that a· a = a 2 ; we often use the
434 Appendix. Vectors and Vector Calculus
notation a· a = 1a 12. The vector e a == a/a is a unit vector, Le., a vector of

length 1, in the direction of a. We can divide vectors with a number, but not
vice versa - the scalar product has no inverse. The equation a . x = c thus
has no unique solution for the vector x.
We often employ the set (i,j, k) of three unit vectors. These are unit
vectors, i· i = j .j = k· k = 1, directed along the axis of a cartesian coordinate
system, and mutually orthogonal: i . j = j . k = k . i = O.
An arbitrary vector a can be written as a combination of the three unit
vectors:
The three numbers, ax, ay and a z are called respectively the x, y, and the
z coordinates of the vector a. We have that ax = a . i, ay = a . j and a z =
a· k. We identify a vector with its coordinates, writing a = (ax, ay, a z ).
Expressed in terms of coordinates, a± b = (ax±b x , ay±by, az±b z ), and ka =
(ka x , kayk, ka z ), where k is a constant. Expressed in terms of its coordinates,
the length of a vector a is given by
a = ra.-a = va; + a~ + a~ ,
and the unit vector in the direction of the vector a thus has coordinates
Appendix. Vectors and Vector Calculus 435
,
........ I ~~
.............~'
The Cross Product, a X b
Let a be a vector with length a, and b be a vector with length b. The cross
product, also known as the vector product, is defined as
a x b == absin(a, b)n(a, b)
where n( a, b), is a unit vector giving the direction of a x b according to the

following right-hand rule:
To find the direction of the cross product, place your right hand such that the
fingers are in the direction of the first vector a and the palm of your hand is
towards the second vector b. The direction of a x b is then perpendicular to
the plane spanned by a and b, and in the direction of your thumb.
axb
Note that the result of the multiplication this time is a new vector, not
a number, and the result does depend on the relative orientation between
the vectors a and b. We employ no coordinate system, but an orientation
convention (right handed in the present case). The angle between the vectors
a and b is now the smallest angle through which one can turn the vector a
until it is aligned with b. That angle is always between 0 and 180 0 • If we
interchange a and b, the angle is unchanged, but the right hand rule makes
the vector n(b, a) point opposite the vector n(a, b). Consequently,
b x a = -a x b,
so the vector product is anti-commutative. If b is along the direction of a,

sin(b, a) = 0, so that, in particular, a x a = 0 .
In a coordinate system with orthonormal vectors (i,j, k), we can find the
coordinates for a x b in terms of the coordinates of a and the coordinates of
b.
ixj k
~
1 J
jxk =
kxi j
Note: the vectors i, j, and k are oriented in such a way that they form the
basis of a right-handed coordinate system.
Using the above formulae, and the anti-commutative law, one obtains the
cartesian coordinate expression for the cross product:
axb (axi + ayj + azk) x (bxi + byj + bzk)
(ayb z - byaz)i + (azbx - bzax)j + (axby - bxay)k.
Briefly,
An important formula which we shall use several times, and which can be
verified by direct (and repeated) use of the coordinate expression for the
vector product, is the so called "bac-cab" formula:
a x (b xc) = b( a . c) - c( a . b)
The scalar and vector products give rise to a number of similar identities that
can be verified by checking the coordinate expressions of both sides of the
equation. Below we list just a couple of these:
a .(b xc) = b . (c x a) = c .(a x b) ,
(a x b) . (c x d) = (a· c)(b· d) - (a· d)(b· c).
Vector Calculus
Consider a vector r which is the position vector for a particle P (see the figure
below). Suppose that r has coordinates (x, y, z) in some cartesian coordinate
system:
r = xi + yj + zk .
If the particle P is in motion relative to the coordinate system, the vector r
varies with time in such a way that the coordinate values (x(t), yet), z(t)) are
Appendix. Vectors and Vector Calculus 437
differentiable functions of time. We define the time derivative of the vector r

to be
The vector v == dr / dt is called the velocity of the particle P relative to

the chosen coordinate frame. The vector a == dv / dt = d 2 r / dt 2 is called the
acceleration of the particle P relative to the chosen coordinate frame.
dv d2r
a=-=-=
- dt dt 2
dr
dt
We now consider the particular case of motion confined to a plane.

Suppose the vector r is the position vector for a particle moving within the
plane. The set (x, y) of cartesian coordinates specify the vector r in the sense
that r = xi + yj. We shall now introduce another set of mutually orthogonal
unit vectors, (en eo).
The vector e r is the unit vector in the direction of r, and eo is perpen-

dicular to e r and in the direction of increasing e (see the figure). The two
direction vectors are known when e is known.
If the position of the particle changes with time, r = ((dx/dt), (dy/dt))

2 2), 2 2))
and r = ((d x/dt (d y/dt are the cartesian coordinates of the velocity
vector v and acceleration vector a, respectively.
What are the coordinates of the vectors v and a expressed by e r and eo?
We need to know how the vectors e r and eo change with time. In cartesian
coordinates these unit vectors are given by
er ( cos 8, sin 8) ,
eo (- sin 8, cos 8) .
By differentiation with respect to time we get

de r .
- = er B( - sin 8, cos 8) ,
dt
deo .
-=eo B( - cos 8, - sin 8) .
dt
The direction vectors thus change with time according to
er Bee,
eo -Be r .
This result can also be seen geometrically by inspection of the figure. The
unit vectors change in direction, but not in magnitude.
For the position, velocity and acceleration we obtain by differentiation:
r r(t)e r ,
r r(t)e r + r(t)er ,
r r(t)e r + 2r(t)er + r(t)e~ .
By using the formulae for er and ee, and rearranging, we finally get
The polar coordinates of the velocity and the acceleration vectors are conse-
quently
(r,rB) = (~:,r~~)
(r-rB2,2rB+re) = (~:~ -r (~~y ~! (r2~~)) . ,
Selected References
[1.) Born M. (1962): Einstein's Theory of Relativity, Dover Publications, New York
[2.) Born M. (1949): Natural Philosophy of Cause and Chance, The Clarendon
Press, Oxford
[3.) Feynman R. P., Leighton R. B. and Sands M. (1963): The Feynman Lectures
on Physics (vol.1), Addison-Wesley, Massachusetts
[4.) French A. P. (1971): Newtonian Mechanics, W. W. Norton, New York
[5.) Goldstein H. (1980): Classical Mechanics (2nd ed.), Addison-Wesley, Mas-
sachusetts
[6.) Kittel C., Knight W. D. and Ruderman M. A. (1973): Berkeley Physics Course
(vol. 1), McGraw-Hill, San Francisco
[7.) Newton 1. (1686): Philosophire Naturalis Principia Mathematica, (English
translation by Motte A. 1729), University of California Press, Berkeley and
Los Angeles (1962)
[8.) Guckenheimer J. and Holmes P. (1986): Nonlinear Oscillations, Dynamical Sys-
tems and Bifurcations of Vector Fields, Springer, Berlin
[9 .) Ott E . (1993): Chaos in Dynamical Systems, Cambridge University Press, Cam-
bridge
[10.) Wiggins S. (1990): Introduction to Applied Nonlinear Dynamical Systems and
Chaos, Springer, Berlin
Answers to Problems
Chapter 1
1.1 (1) w = 33.02 s-l. Between the hand and the first mass
(2) No
1.2 (1) M = 6.48 X 10 23 kg
(2) T = 684 days
1.3 8=Mg
1.4 (1) 8 1 = m(lw 2:,j29) , 82 = m(lw 2 ,;,j2g)
(2) w 2 :::- ,;2g/l

1.5 v = y'Mgr/m
1.6 (1) 8 = mlw 2
(2) cos()= g/lw 2
(3) sin () = 0
1.7 (1) 8 = ~ e~2 + V39), R = ~ (g - ,?lw2)
(2) w2 = 2g/V3l .
1.8 (1) v = (J6, 2,,;2)
(2) v = 2V3
(3) j3 = 54.7°
(4) 'Y = 65.9°
1.9 70.5°
1.10 t=~s, d=(~,-~,O)m
1.12 (1) aT = 4.9 ms- 2 , aN = 0
e
(2) w = = 0, jj = w = 7 s-2
(3) 8 = 4.24 N
Chapter 2
2.1 Ve = 5.03 X 10 3 ms- 1 , Ve = 2.38 X 103 ms- 1

2.2 (1) h = 22.5 m
(2) 42.9 m
2.3 a=g/9,8=2.18N
2.4 f = m~MF
442 Answers to Problems
2.5 tanB> JL
2.6 (1) dry ~ 174.6 kmh- 1 iey ~ 87.1 kmh- 1
(2) rmax = 4.4 em
2.7 (1) JL = tan CPo
(2) F=~
V1+J.t 2
2.8 (1) d max = 4.9 em
(2) Yo = 4.9 em
(3) a = g, upwards
2.9 Hint: K(.:::h1 + .:1x2) = k2.:1X2; k1.:1x1 = k2.:1x2
T -- 21r M(kl +k2)
klk2
2.10 (1) F = ~z
(2) F(r) = ~~2 (L2 _ r2)
2.11 (1) VMoon/Vesc = Jl - 1/60 = 0.9916
(2) Vmax ~ 42 kms- 1 + 30 kms- 1 = 72 kms- 1
2.12 F = (m2/m1)(M + m1 + m2)g
2.13 (1) a = (1/3)g
(2) S = 70 kg = 686 N
2.14 v = 331.9 ms- 1
2.15 B = 45°
2.16 (1) J=3(Mg/L)S
(2) J = 3Mg
2.17 (1) u = M"'.;.m Vo
(2) T = J.t9(1~m/M)
_ 1 v6
(3) D - "2 IJ-g(1+m/M)
2.18 (1) F = pau 2
(2) F = 9000 N
(3) 27 m
(4) F = pa(u - v)2
2.20 (1) U1 = ::;:~v
(2) U2 = m2.;'M V
(3) J = (~~ki)2' r = l+~/M
(4) J = ;2
2.21 v = ;;~m Vo
2.22 (1) JL = 2mg/V re l ~ 26 kgs- 1
(2) b = g/Vrel ~ 3.3 X 10- 3 s-1
2.23 F = JLV
Chapter 3
3.1 (1) 11.4 kg* = 111.7 N

Answers to Problems 443
(2) 26.4 kg* = 258.7 N

(3) 84.3 kg*
(4) 55.7 kg*
Chapter 4
4.1 x'(t') = 5t,2 + 4

v' x(t') = lOt' m s-1
a'x = ax = 10 ms- 2
Chapter 5
5.1 (1) v = w x r = 211"(2, -4.;2,2.;2) ms- 1

(2) a = w x (w x r) = 411"2(12.;2, -4, -20) ms- 2
5.2 a = R.w 2 = 3.3 X 10- 2 m S-2
Chapter 6
6.1 Cork (floats) to the left, lead (sinks) to the right.

6.2 a= g/p,
6.3 (2) T = 211"/f
6.4 T = 84.4 min
6.5 T = 84.4 min
6.6 d = 2.66 mm, east
6.7 (1) W = ..jg/p,R
(2) v = 8.6 ms- 1
6.8 (1) Centrifugal force: mw 2 p. Coriolis force: 2m (v x w)
(2) Xo = ~..j p,2g2 - 4V 2W2
6.9 () = 30°
6.10 (1) Wo = y'f7"(g-:"/r--'-)-ta-n-"()
( 2) W = 211"v w 2 > !l. sin O+y c,:,s 0
, - r CosO-J.LsmO
(3) p,-l = tan ()max
6.11 (1) w = (O,wcosP,-wsinp)
(2) mx = -2mwysinp
my = 2mwxsinp
(3) y = 10.4 m
6.13 Llt = 2(M~:)J.Lg
6.14 T = ws7nq,
Chapter 8
8.1 31.9 m
8.2 59.6 x 103 ms- 1 :::::: 60 kms- 1
8.3 cosO=~
8.4 = 7.9 kms- 1
(1) Vo
(2) to = 7rJR/g = 42.2 min
8.5 (1) R = 7mv 2
- mg(2 - 3cosO)
(2) 0 = 60°
(3) F = 4mg
(4) 0 = 45°
8.6 (1) h = R/2
(2) F = 4mg, horizontal
8.7 (1) To = (1/2)mgL
(2) U = n!l..;gr, v = ~~~..;gr, q = (n!1)2
(3) n --+ 00, q --+ 0, reflection from infinite mass.
(4) cos 00 = 1-!
(~~~)2
(5) n --+ 00 cos 00 --+ 1/2
8.8 h max = a + l2/(27r 2a)
Chapter 9
vVmfk
vv
9.1 (1) ill =
(2) ill = (M~'::,)k
9.2 x = D MAf.-m
9.3 (1) d = ILV Vi--'(m""-+-p.-)-(M-M-+-m-+-P.)-k
(2) T = T/4 = (7r/2) In/k , n = ;:++:1~

(3) interchange M and m.
9.4 (1) S = 2~9 (m';'M) 2 v 2
(2) ilp:::::: 1 kgms- 1 , ilP:::::: 720 kgms- 1
9.5 Highest when the people jump off one by one, as described in (b)
9.6 d = 18.1 m
9.7 (1) U = M"':-mv:::::: :v
(2) P = mv
(3) V eM -- 2M+m m V ~ m V
~ 2M
(4) K = !(2M + m)v6M

(5) E -- 2M+m
1 m 2v2 ~ 1 m 2 2
~ 2M v
(6) ill = "'';kvM
(7) T = 27rJM/2k
9.8 v = J2ghM;;.m
9.9 (1) YCM = ~gt2
(2) a(l) = 2g
a(2) = 0
(3) z = (L/2) + (mg/2k) cos )2k/m t
Chapter 11
11.1 T = 27r/fi :=:::! 1.1 s

11.2 (1) T = 27rJR2/g2iL2
(2) T = 27r J R2 /g~L2
11.3 (1) TI = ~(2M + m)l2wr
(2) W2 = 2AftrwI, T2 = TI (1 + 2~)
(3) W3 = WI, T3 = TI
(4) v = lWI)l + 2~
11. 4 (1) CM falls along the vertical.
(2) T = ~ML2e2 (sin 2 0 + t)
u = -Mg~ (1 - cosO)
(3) Y. -- sm
. 0 3gL(I-cos 8)
4-3cos28
11.5 (1) VCM = ~

(2) T = t~ (1
+ 12b)
(3) OC = L 2 /12d; below 0
d ~ L/6
m,fiih
11.6 (1) U = 2M+m+I/r2
(2) Q 2M+I/r2
= mg h 2M+m+I/r2
11.7 (1) v = 3:::!~
(2) S = (1/2)l, d = (3/2)l
Chapter 12
12.1 (1)
(2)
VCM
Vrn =
=
.;r¥ 4Mm;3m
2 '-(M....:+:.c:m....:)..::.c
ga"":(-ct,;:=-2;--I-)(-4M-+-
3m-)
12.2 (1) v = 5.4 ms- I

(2) v = 4.4 ms- I
12.3 a = (mg/Mr) sin 00
12.4 h = (3/2)r
12.5 to = 7r fIT : : 50 min.

12.6 (1) t l' = ~7/"9
Vo
( 2) D = 49/"9
12~
(3) W = ~MV02
12.7 J1min = 2a/7g
12 . 8 mvo
W = MR+2mR
12.9 F = lMg, upwards.
12.10 (1) Wo = ~ ;:;r
(2) v = (2M/m)JgL/3
(3) F1 = ~Mg
(4) F2 = ! 'it fii
(5) ..::1t« T, T = 27r I¥i
12.11 (1) VCM = v/2
(2) w = 6v/5L
(3) ..::1T = (1/10)Mv 2
(4) d = (2/3)L
12.12 (1) (p = (39/ L)(sinOo - sinO)
(2) sin 0 1 = (2/3) sin 00
(3) rCM = (1/3)y'gL
12.13 (1) F1 = (1/28)Mg
(2) F2 = (41/14)Mg
(3) h = (7;22 - L !)
Chapter 13
M1'3 W 2
13.1 R=~+Mg
13.2 (1) ..::10 = 2r;:;;0f : : : 0.04 rad :::::: 2°

(2) ..::1t«.J2h19:::::: 1/3 s
13.3 (1) D= JI21;~/4
(2) FB !m [g + W
= 2 r:
+ won]
Fe = !m [g + w 2 - 1': won]
13.4 l/ = wo/27r = 1O.8s- 1
Chapter 14
14.1 (1) Vo = 11.2 kms- 1 (independent of 0)

(2) v ~ 42 kms- 1
14.3 (1) w = 6.83 X 1O- lO s- 1
(2) LCM = 11+":nR2w ~ 1.17 X 10 47 kgm 2 s- 1
14.4 rCM = 4660 km
T = 2.35 X 10 6 s, V ~ 1 kms- 1
14.5 (1) T = 75.6 year ~ 76 year
(2) Vp = C:[P!)e) ~ 5.5 x 10 4 ms- 1 ~ 55 kms- 1
VA = CM(I-e)
a(l+e)
~
~
0 . 09 x 10 4 ms -1 ~
~.
0 9 k ms -1
14.6 (1) ha = 1047 km
(2) T = 1.62 ~ 1 h 37 min
14.7 (1) Vo = 27.2 kms- 1
VE = 29.8 kms- 1 , 8v = -2.6 kms- 1
(2) T = t = 145.6 days
(3) VI = 37.8 kms- 1 Vv = 35.0, kms- 1
(4) e = 54°
Chapter 15
15.1 (1) T = 1.64 s

(2) Tmin = 1.52 s
15.2 (1) Td/Tf = 2/../3
(2) X2/Xl = exp (-27r/../3)
15.3 (1) x(O) = ~ w2~w2
a
v(O) = 0
(2) Xo = Fo~
m wo-w
e=o
15.4 (1) W = !2 fl.
ml'
(2) b = 1/27r Nsm- 1
15.5 Q ~ 26240
Chapter 16
16.1 (a) , (c), (e)

16.5 About 4 days
Index
Absolute space 30, 155 Centrifugal force 116, 160

Absolute time 30 Centrifugal potential energy 232, 366
Acceleration 29 Centripetal acceleration 107
- absolute 103 Chaotic oscillator 427
- centripetal 107 Circular motion 12
- comoving 104 Closed system 56, 195
- Coriolis 104 CM of solar system 371
- of gravity 11 Collision approximation 204, 206, 280
-- varition with latitude 118 Communication satellite 13
Action-reaction 29 Comoving velocity 103
Andromeda 83 Conic section 352, 374
Angular momentum Conservation law
- around CM 228 - for angular momentum 221
- conservation of 253 - for mechanical energy 188
- definition 219 - for momentum 54
- for a system of particles 224 Conservative force 52, 169
- referred to various points 256 Copernicus 1, 346
- Theorem 219 Coriolis force 128, 138
Angular velocity vector 95 Coupled differential equations 46
Aphelion 232 Critical damping 394
Aristotle 1 Cyclotron radius 45
Asteroids 429
Astronomical unit (AU) 53, 351 Damped oscillations 392
Asymmetry 159 Deterministic equntions 299, 411
Asymmetry of nature 186 Dimensional analysis 20
Attractor 426 Dot product 433
- strange 426 Double stars 370
Atwood machine 41, 262 Duffing equation 424
Ballistic pendulum 206 Earth

Bay of Fundy 125 - bulging of 120
Bifurcation 421, 423 - rotation of 98
Binding energy 183 Eccentricity 352
Brahe 1, 345 Effective potential 232, 366
Eigenvalues 327
Celestial sphere 345 Einstein 110, 155, 162
Center of gravity 227 - box 111
Center-of-mass (CM) theorem 193, Elastic collision 56
195 Elastic force 34
Center-of-mass frame 197 Ellipse 352
Central force 221 Elliptic function 418
Central force field 170 Energy 167
450 Index
- conservation of 172 - in rotational motion 313

- of Kepler orbits 363 -law of 27
Epicycles 3 - moment of 240
Equation of motion 31 - tensor 321
Equilibrium point 415 Inertial frame 31
Equivalence principle of mechanics - global 137
109 - local 136, 137
Escape velocity 54 Inertial mass 74
Euler equations 328, 330
Invariance 158
Exact differential 174
Fictitious forces 108, 164 Konig's theorem 198

Field strength 180 Kepler 1, 346
Force 29 -laws 4
- constant 34 Kinetic energy 51, 167
Forced oscillations 396 - of a rotating rigid body 330
Forces Kirkwood gaps 430
- external and infernal 194
- tidal 121 Lagrange 430
Foucault pendulum 137 Liapunov exponents 427
Frequency 36 Libration curves 417
Frequency characteristics 400 Lorentz 44
Friction 38, 40 - curve 405
- rolling 288 - force 44
Frictionless curve 181
Mach 120, 162
Galilei transformation 84
Magnetic field 44
Galileo 1, 90
Map 427
-laws 2
Geocentric reference frame 98, 122 Mathematical pendulum 115
Gradient 174 Mechanical energy 52, 172
Gravitational constant 10, 17 Milky Way 83
Gravitational field 9 Moment of inertia 240
- around a homogeneous sphere 175, - calculation of 245
179 - circular disk 246
Gravitational mass 73 - homogeneous rod 246
Gravitational potential 180 - rectangular plate 249
Gravity 9 11 - solid sphere 248
- spherical shell 247
Harmonic oscillator 33, 385 Monkey 226
- energy in 386
Heliocentric frame 85, 98, 101 Newton 1,120
Hohman orbit 378 - bucket 142
Homogeneous gravitational fields 114 - experiment 76
Hooke's law 34
-laws 27-31
Hoop 421
North Star 342
Horizontal throw 14
Hyperbola 352 Nutation 336
Impetus 1 Opposition 347

Impulse 55, 168 Ordinary differential equation
Inelastic collision 56, 202 -linear 412
Inertia -- homogeneous 413
- arm of 291 - nonlinear 412
Index 451
Parabola 352 - around a fixed axis 243

Parallel axis theorem 241 - of the Earth 97
Particle 31 - vector 238
Pendulum 115, 182
Perihelion 232 Satellite 78
Period of oscillation 75 Saturn 127
Perpendicular axis theorem 242 Scalar product 433
Phase space 414 Second law of thermodynamics 187
Physical pendulum 263 Semi-major axis 355
Physical reality 159 Semi-minor axis 355
Pitch of a helix 47 Separatrix 420
Poincare 426 Sidereal period 347
Polar coordinates 355 Small oscillations 385
Potential energy 52, 167, 171 Spherical shell 176
- calculation of 174 Spring force 175
- due to electric forces 183 Strange attractor 426
- effective 232 Superposition 413
Potter's wheel 226 Symmetry 156
Power 168, 266 Synodic period 347
- in rotation of a rigid body 254
Precession 107 Tension 19
- of the Earth 342
Thomson J. J. 48
- regular 319
Tidal forces 121
Principal coordinate system 327
Torque 220
Principia 1
- of gravity 227
Q-value 403 - with respect to an axis 223
Torsion pendulum 255
R!ilmer 350 Turning point 417
Reaction 29 Two-body problem 367
Reduced mass 368
Relativity principle of mechanics 87, Vectors 433
159
Resonance 398 Weightless 79
Retrograde motion 342 Work 51,167
Right-hand rule 435 - and energy 51
Right-handed coordinate system 436 - done by a force 167
Rigid body 193, 237 - in rotation of a rigid body 254
Roche limit 126
Rocket propulsion 57 Yo-Yo 295
Rolling 289
Rotation Stefan Zweig 350

Elements of Newtonian Mechanics

Uploaded by

Copyright:

Available Formats

Elements of Newtonian Mechanics

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Elements of Newtonian Mechanics

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

Who are the authors of the book?

Who are the authors of the book?

Elements of Newtonian Mechanics

Advanced Texts in Physics

Third Revised and Enlarged Edition

With 341 Figures and 131 Problems with Solutions

Library of Congress Cataloging-in -Publication Data

Third Edition 2000. Corrected 2nd Print ing 2002

A Note for the Instructor

JMK & PGH

Copenhagen JMK €9 PGH

Preface to the Second Edition

Copenhagen JMK & PGH

This book is intended as a textbook for an entry-level university course in

1. The Foundation of Classical Mechanics .. . . . . . . . . . . . . . . 1. . .

3. Gravitational and Inertial Mass. . . . . . . . . . . . . . . . . . . . . . . . . .. 73

4. The Galilei Transformation .... . ......................... 83

5. The Motion of the Earth ................................. 95

6. Motion in Accelerated Reference Frames ................. 101

7. The Problem of Motion ................................... 153

8. Energy ................................................... 167

9. The Center-of-Mass Theorem ............................. 193

10. The Angular Momentum Theorem ........................ 219

11. Rotation of a Rigid Body ....................... . ......... 237

12. The Laws of Motion .................... .. . ........... . .... 279

13. The General Motion of a Rigid Body .. . .......... . . .. ..... 313

14.6 The Effective Potential .............................. 366

15. Harmonic Oscillators . ................................... 389

16. Remarks on Nonlinearity and Chaos . .................... 411

16.4 A Forced, Damped Nonlinear Oscillator ................ 424

Appendix. Vectors and Vector Calculus ...................... 433

Selected References ..................................... . ... 439

Answers to Problems ... .. .... . ................... . ......... 441

Index ......... . ............................................ 449

1.2 Prerequisites for Newton

J. M. Knudsen et al., Elements of Newtonian Mechanics

where C has the same value for all the planets.

1.3 The Masterpiece

By eliminating v, we get from (1.3) and (1.4) that

From Kepler's third law we have:

Fig. 1.1. The gravitational interaction between a particle A and a spherically

However, in Principia, Newton showed the important theorem that the

Inserting this into (1.10) we get

T 27 days 7 hours 43 minutes

Inserting these values into (1.12) we find:

gcalculated = 9.8 m s-2 .

This means that,

gmeasured = 9.8 m s-2 .

This beautiful agreement is rightfully considered a high point in the history

where C depends on the mass of the Sun. If a law of general gravitational

p' = M47r 2 e (1.14)

47r 2 e = GM, (1.18)

where G is a universal constant. The numerical value of G depends only on

G = 6.668 X 10- 11 N m 2 kg- 2 .

law of gravity was modified in the beginning of this century, by Einstein's

Example 1.1. The Acceleration of Gravity. Calculate the acceleration

As is well known, g is sometimes called the gravitational field strength at the