NCRP REPORT No.

108

CONCEPTUAL BASIS FOR

CALCULATIONS OF
ABSORBED-DOSE
DISTRIBUTIONS

Recommendations of the
NATIONAL COUNCIL ON RADIATION
PROTECTION AND MEASUREMENTS

Issued March 31, 1991

Sexond Reprinting February 1, 1995

National Council on Radiation Protection and Measurements

7910 WOODMONT AVENUE / Bethesda, MD 20814
 LEGAL NOTICE
This report was prepared by the National Council on Radiation Protection and Mea-
surements (NCRP). The Council strives to provide accurate, complete and useful
information in i t . reports. However, neither the NCRP, the members of NCRP,other
persons contributing to or assisting in the preparation of this report, nor any person
acting on the behalf of any of these parties: (a) makes any warranty or representation,
express or implied, with respect to the accuracy, completeness o r usefulness of the
information contained in this report, or that the use of any information, method or
process disclosed in this report may not infringe on privately owned rights; or (b)
assumes any liability with respect to the use of, or for damages resulting from the use
of any information, method or process disclosed in this report, under the Civil Rights
Act of 1964. Section 701 et seq. as amended 42 U.S.C.Section 2000e et seq. (Titk VII1
or any other strrtutory or common law theory gooerning liability.

Library of Congress Cataloging-in-PublicationData

Conceptual basis for calculations of absorbed-dose distributions.

p. cm.-(NCRP report; no. 108)
Includes bibliographical references and index.
ISBN 0-929600-16-9
1. Radiation dosimetry. 2. Ionizing radiation-Measurement.
I. National Council on Radiation Protection and Measurements.
11. Series.
QC795.32.RX66 1991 91-9135
539.7'22'0287-dc20 CIF'

Copyright 8 National Council on Radiation

Protection and Measurements 1991
All rights reserved. This publication is protected by copyright. No part of this publica-
tion may be reproduced in any form or by any means, including photocopying, or
utilized by any information storage and retrieval system without written permission
from the copyright owner, except for brief quotation in critical articles or reviews.
Preface

The idea for this report emerged, in the early 1970's, from the need
of a n NCRP Scientific Committee to characterize the beta-ray depth-
dose distribution in connection with immersion doses. It was realized,
however, that the calculation of such a distribution was only a small
part of the very much larger task concerned with the theoretical,
mathematical and computational concepts involved in the develop-
ment of absorbed-dose distributions in general. To address this issue
in an allencompassing manner, the NCRP formed Scientific Commit-
tee 52 on the Conceptual Basis of Calculations of Dose Distribution.
In either external or internal irradiation, the absorbed dose is
usually non-uniform in any structure and, in particular, in the
human body. This non-uniformity is to be distinguished from the
stochastic variations that exist even in regions where the dose is
uniform and that are the subject of microdosimetry and not this
report. Many illustrations of absorbed dose non-uniformity come to
mind: for example, the absorbed-dose distributions from hot parti-
cles, from internal emitters, from radiation therapy, from radiation
accidents and from environmental radiation. There can even be addi-
tional non-uniformity with respect to time of the non-uniform distri-
bution, for example, in the redistributions of administered radioac-
tivity in the body.
For all absorbed-dose calculations, there is a source (or sources) of
radiation and a receptor (or receptors) of some of the energy of this
radiation, with or without intervening material between the source
and receptor. The calculation of absorbed-dose distributions requires
specification of the sources and receptors, characterization of their
geometrical relationships and consideration of the physical interac-
tions of the radiations involving attenuation, scattering and the
production of secondary radiations. All these processes are consid-
ered in the basic transport equation, the general theorems and prop-
erties and the methods of solution of which are described in the
transport theory.
The report is a systematic presentation, discussion and compila-
tion of all the concepts involved. It contains some complicated mathe-
matics that will be of interest to the mathematically knowledgeable,
but that should not discourage those not mathematically inclined.
iv / PREFACE
The text of the report contains detailed explanations of all the con-
cepts and of the consequences of the equations so that, even omitting
the mathematics, a broad and comprehensive understanding can be
obtained of what is entailed in the calculation of an absorbed-dose
distribution.
The cutoff date for the report is about two years ago and, hence,
the report is lacking in the most current references. However, this
field does not evolve at a rapid pace and the current literature is,
therefore, not abundant and can be reviewed easily.
In accord with the recommendations of NCRP Report No. 82, SI
Units in Radiation Protection and Measurements, as of January 1990,
only SI units are used in the text. Readers needing factors for conver-
sion of SI to conventional units are encouraged to consult Report No.
82.
This report was prepared by NCRP Scientific Committee 52 on
Conceptual Basis of Calculations of Dose Distributions. Serving on
the Committee during the preparation of this report were:
Harald H. Rossi, Chairman
105 Larchdale Avenue
Upper Nyack, New York
R. G. Alsmiller, Jr. William C. Roesch
Engineering Physics and 1646 Butternut
Mathematics Division Richland, Washingtan
Oak Ridge National Laboratory
Oak Ridge, Bnnessee
Martin J. Berger Lewis V. Spencer
5011 Elm Street Post Office Box 87
Bethesda, Maryland Hopkinsville, Kentucky
Albrecht M. Kellerer Marco A. Zaider
GSF Institut fiir Strahlungbiologie Radiological Research
Neuherberg, Germany Laboratory
Columbia University
New York, New York
NCRP Secretariat: Thomas Fearon (1976-80)
J a m e s A. Spahn, Jr. (1981)
J a m e s T. Walker (1982-84)
Constantine J. Maletskos (1985-91)
The Council wishes to express its appreciation to the Committee
members for the time and effort devoted to the preparation of this
report.
 PREFACE, 1 V

Especial thanks are due to Marco Zaider for his contribution to

the editing of scientific aspects of this report.
Warren K. Sinclair
President
Bethesda, Maryland
5 March 1991
Contents

.
1 Introduction .................................... 1
1.1 The Concept of Absorbed Dose . . . . . . . . . . . . . . . . . . 1
1.2 Dose Measurement and Dose Calculation ......... 3
1.3 Elements of Dose Calculations . . . . . . . . . . . . . . . . . . 4
.
2 Transport Formalisms . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.1 Concepts in Dose Calculations . . . . . . . . . . . . . . . . . . 6
2.2 Transport Equation ........................... 8
.
3 Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3.1 Specification of Sources ......................... 14
3.2 Simplified Representations of Sources . . . . . . . . . . . . 15
.
4 Receptors ..................................... 17
.
5 Cross Sections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
5.1 Schematization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
5.2 General Aspects of Required Cross Sections . . . . . . . 22
.
6 Transport Theory-General Theorems and
Properties ..................................... 26
6.1 Integral Form of the Transport Equation . . . . . . . . . . 26
6.2 Iterative Solutions (Orders of Scattering) . . . . . . . . . 27
6.3 Density Scaling Theorem . . . . . . . . . . . . . . . . . . . . . . 28
6.4 Fano's Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
6.5 Energy Conservation . . . . . . . . . . . . . . . . . . . . . . . . . 29
6.6 Superposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
6.7 Adjoint Transport Equation .................... 31
6.8 Reciprocity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
6.9 Transport Equations in Commonly Used Coordinate
Systems .................................... 34
.
7 Transport Theory-Methods of Solution .......... 36
7.1 Introduction ................................. 36
7.2 Radiation Equilibrium and Space-Integrated
Radiation Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
7.3 Continuous Slowing-Down Approximation (CSDA) . . 40
7.4 Numerical Integration Over Energy ............. 43
7.5 Elementary Problems Involving Particle Direction . . 45 '

7.5.1 Thin-Foil Charged Particle Problems . . . . . . . . 45

7.6 Penetration Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
7.6.1 The Moment Method . . . . . . . . . . . . . . . . . . . . . 47
7.6.2 Discrete-Ordinates Transport Codes ......... 49
 CONTENTS 1 vii
7.6.2.1 Neutron-Photon Transport . . . . . . . . . . 49
7.6.2.2 Dosimetry Calculations By the Method
of Discrete Ordinates . . . . . . . . . . . . . .
7.7 Spectral Equilibrium and Related Concepts . . . . . . .
7.7.1 Aspects Applicable to All Radiations . . . . . . . .
7.7.2 Electrons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7.7.3 Photons and Neutrons . . . . . . . . . . . . . . . . . . . .
7.8 Radiation Quasi-equilibrium ...................
7.8.1 Transient Equilibrium . . . . . . . . . . . . . . . . . . .
7.8.2 Non-uniform Sources . . . . . . . . . . . . . . . . . . . . .
7.8.3 Non-uniformity in the Internal Dosimetry of
Radionuclides . . . . . . . . . . . . . . . . . . . . . . . . . . .
7.8.4 Non-uniform Media . . . . . . . . . . . . . . . . . . . . . .
.
8 Monte-CarloMethods . . . . . . . . . . . . . . . . . . . . . . . . . . .
8.1 Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8.2 Analog Monte-Carlo and Variance-Reduction
Bchniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8.3 Transport Codes . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8.3.1 Neutron-Photon Transport a t Energies 520
MeV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8.3.2 Electron-Photon Cascades . . . . . . . . . . . . . . . . .
8.3.3 Nucleon-Meson Transport a t Energies >20
MeV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8.3.4 Dosimetric Calculations . . . . . . . . . . . . . . . . . .
9. Geometric Considerations . . . . . . . . . . . . . . . . . . . . . . .
9.1 Absorbed Dose in Receptor Regions . . . . . . . . . . . . . .
9.2 Reciprocity Theorem . . . . . . . . . . . . . . . . . . . . . . . . . .
9.3 Isotropic Point-Source Kernels . . . . . . . . . . . . . . . . . .
9.4 Point-Pair Distance Distributions and Geometric
Reduction Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
10 Calculation of the Dose Equivalent . . . . . . . . . . . . . . .
List of Symbols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
Appendix A Information about Cross Sections for
Transport Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A.l Photon Cross Sections . . . . . . . . . . . . . . . . . . . . . . . .
A.l.l Photoelectric Effect . . . . . . . . . . . . . . . . . . . . . .
A.1.2 Fluorescence Radiation and Auger Electrons .
.
A 1.3 Incoherent (Compton) Scattering . . . . . . . . . . .
A.1.4 Pair Production . . . . . . . . . . . . . . . . . . . . . . . .
A.1.5 Coherent (Rayleigh) Scattering . . . . . . . . . . . .
A.1.6 Photonuclear Effect . . . . . . . . . . . . . . . . . . . . .
A.1.7 Attenuation Coefficient . . . . . . . . . . . . . . . . . .
A.1.8 Energy-Absorption Coefficient . . . . . . . . . . . . .
A.1.9 Photon Cross-Section Compilations . . . . . . . . .
viii / CONTENTS
A.2 Cross Sections for Charged Particles . . . . . . . . . . . . . 113
A.2.1 Elastic Scattering of Electrons by Atoms . . . . 113
A.2.2 Elastic Scattering of Protons by Atoms . . . . . . 117
A.2.3 Scattering of Electrons by Atomic Electrons . . 119
A.2.4 Scattering of Protons by Atomic Electrons ... 120
A.2.5 Electron Bremsstrahlung . . . . . . . . . . . . . . . . . 122
A.2.6 Continuous Slowing-Down Approximation . . . 126
A.2.7 Stopping Power . . . . . . . . . . . . . . . . . . . . . . . . . 128
A.3 Neutron Cross Sections . . . . . . . . . . . . . . . . . . . . . . . 139
A.3.1 Classification of Interactions .............. 139
A.3.2 Data Compilations . . . . . . . . . . . . . . . . . . . . . . 143
A.3.3 Kerma Factors . . . . . . . . . . . . . . . . . . . . . . . . . 145
~ . Nuclear
4 Cross Sections for Charged Particles a t
High Energies .............................. 147
A.4.1 Interactions of Pions below 100 MeV . . . . . . . 147
A.4.2 Nuclear Interactions of Hadrons above 100
MeV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
Appendix B. Examples of Absorbed-Dose and Dose-
Equivalent Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
B.1 Absorbed Dose from Neutrons in Tissue-Equivalent
Material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
B.2 Shielding of Manned Space Vehicles Against
Galactic Cosmic-Ray Protons and Alpha Particles . . 172
B.3 Skyshine for Neutron Energies 5400 MeV . . . . . . . . 178
.
Appendix C A Compilation of Geometric Reduction
Factors for Standard Geometries . . . . . . . . . . . . . . . . . . . . 185
C.l The Autologous Case (A = B) . . . . . . . . . . . . . . . . . . 185
C.2 The Heterologous Case (A # B) . . . . . . . . . . . . . . . . 188
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
1. Introduction

1.1 The Concept of Absorbed Dose

The effects of radiation on matter are initiated by processes in

which atoms and molecules of the medium are ionized or excited.
Over a wide range of conditions, it is a n excellent approximation to
assume that the average number of ionizations and excitations is
proportional to the amount of energy imparted to the medium by
ionizing radiation1 in the volume of interest. The absorbed dose, that
is, the average amount of energy imparted to the medium per unit
mass, is therefore of central importance for the production of radia-
tion effects, and the calculation of absorbed-dose distributions in
irradiated media is the focus of interest of the present report. It
should be pointed out, however, that even though absorbed dose is
useful as an index relating absorbed energy to radiation effects, it is
almost never sufficient; it may have to be supplemented by other
information, such as the distributions of the amounts of energy
imparted to small sites, the correlation of the amounts of energy
imparted to adjacent sites, and so on. Such quantities are termed
stochastic quantities. Unless otherwise stated, all quantities consid-
ered in this report are non-stochastic. A discussion concerning sto-
chastic quantities is given in ICRU Report 33 (ICRU, 1980).
The absorbed dose, D, is defined (ICRU, Report 33) as the quotient
of d by dm:

where d2 is the mean energy imparted by ionizing radiation to matter

of mass dm.
The energy, E,imparted to the volume containing dm is defined as

'Ionizing radiation consists of directly ionizing and indirectly ionizing radiation.

Directly ionizing radiations are charged particles (electrons,positrons, protons, alpha
particles, heavy ions) with sufficient kinetic energy to ionize or excite atoms or
molecules. Indirectly ionizing radiations are uncharged particles (photons, neutrons)
that set in motion directly ionizing radiation (charged particles) or that can initiate
nuclear transformations.
2 / 1. INTRODUCTION

where Ei, (E,,,) is the sum of energies of all the charged and
uncharged ionizing particles that enter (leave)the volume, excluding
rest mass energies, and XQ is the algebraic sum of all changes
(decreases: positive sign; increases: negative sign) of rest-mass
energy in mass-energy transformations occumng in the volume.
A few clarifications of Equation (1.1) are necessary a t this point.
The energy imparted results from random discrete energy deposition
events by individual ionizing particles andlor their secondaries. The
quantity E is therefore stochastic in nature and governed by a (nor-
malized) probability distribution function fv(d where V is the volume
containing m. The mean value of e,

is the quantity referred to in the definition, Equation (1.1).In addi-

tion, this equation implies a limiting process2such that V +. 0. The
absorbed dose, Dfi), is thus defined a t a given position ? in the
irradiated object (see footnote 2), and is a non-stochastic quantity.
Furthermore, in general D c ) changes with and this variation is
termed the "dose distribution."
A second important aspect refers to the temporal pattern of dose
accumulation. Let

be the dose increment at ? during the time interval CtJt+dtl. This

equation defines the absorbed-dose rate, D@,t). This aspect is impor-
tant in understanding the relation between dose and biological effect.
A final remark concerns the relation between.dose and its stochas-
tic counterpart, the specific energy, z, defined as
z = dm. (1.5)
The specific energy is always measured in a non-zero volume, V, and
its mean value, 2, has the value of the average absorbed dose, Dv, in
that volume of

2Becauseof the discrete manner in which energy is imparted, the limiting process
V- 0 is obviouely an idealization. At all times the volume V should contain a large
number of atoms and molecules.
 1.2 DOSE MEASUREMENT AND DOSE CALCULATION / 3
It follows (ICRU, Report 33) that
DP) = lim Z
v+ 0
FEV,
where the limiting process is such that ? is contained in the volume
v.
The stochastic variations of specific energy that occur even in a
medium that is uniformly irradiated must not be confused with
variations of absorbed dose in a medium that is not uniformly irradi-
ated. The latter often occurs with internally deposited radionuclides,
especially with alpha particles. For example, if 2S9pUparticulates,
having a diameter of the order of a micrometer, are lodged in tissue,
substantial variations in absorbed dose occur over comparable
dimensions.
The theoretical and experimental study of the distributions in z is
the objective of microdosimetry (see ICRU, 1983) and is beyond the
scope of the present report.
No formal definitions of other radiation quantities are given here
because they can be found in many publications and because their
meaning is usually obvious from the context. The units employed
are those of the International System (SI) except that, in accord with
common practice, the electron-volt (eV) is frequently used as a unit
of energy.

1.2 Dose Measurement and Dose Calculation

The absorbed dose of ionizing radiation may be determined by two

entirely different methods: measurement and calculation.
Measurement has the decided advantage in that it requires-at
least in principle-only one auxiliary quantity, the calibration factor
of the instrument employed. The construction of the instrument from
materials of suitable atomic composition can largely obviate even
the need to identify the incident radiation. The experimental deter-
mination is also often simpler and more rapid.
Calculations, on the other hand, usually require extensive subsid-
iary information, including configuration of sources, disposition and
atomic composition of matter intervening between them and the
points of interest, and corresponding data for objects that are merely
in the vicinity of source and receptor if they produce significant
scattered or secondary radiation. Information is also needed on the
nature and energy distributions of the primary radiations, on the
4 1 1. INTRODUCTION
cross sections for the production of secondary radiation, and on the
magnitude of material constants such as attenuation coefficients and
stopping powers. These must again be known for each type and as a
function of energy for all significant primary and secondary radia-
tions. Errors or uncertainties in any of the variables in this lengthy
catalog can result in incorrect dose assessment and further inaccura-
cies may also be due to simplifications that must often be made. Even
with such simplifications, the calculations tend to be complex and
expensive.
Theoretical assessments are nevertheless essential in a number of
cases. Measurements can be made only when dose is actually
received. Retrospective determinations of doses that have been
received, or evaluations of the doses to be received, are generally
done by calculations. Calculations are also necessary when it is
impractical to place a dosimeter at the point of interest as in most
instances of radiotherapy or the dosimetry of internal emitters. Cal-
culations can be performed when the required spatial resolution
cannot be attained with a physical device or when radiation intensit-
ies are too high or too low to permit measurement with available
equipment.
Calculations can also play a significant role in the interpretation
of dose measurements. For example, the quantity actually measured
is often not the absorbed dose itself, but the ionization in an air-filled
cavity chamber. Theory, in the form of the Bragg-Gray principle and
its elaborations, can then be used to estimate what the absorbed dose
would be in the medium, a t the point of measurement, in the absence
of the dosimeter. Furthermore, the presence of the dosimeter usually
perturbs the radiation flux. A theoretical model of radiation trans-
port in and near the dosimeter may be needed for the appropriate
perturbation corrections.
Apart from any immediate utility, theoretical dosimetry often
results in a deeper understanding of basic aspects of the propagation
and absorption of radiant energy. This has led to the development of
concepts and theorems that are of value in their own right and whose
application may also lead to simplifications of measurements or
reductions in their required number.

1.3 Elements of Dose Calculations

This report is designed to present an outline of the methodology

of theoretical dosimetry. From the definitions above, it is clear that
the calculation of dose requires a description of the radiation fields
 1.3 ELEMENTS OF DOSE CALCULATIONS / 5
in terms of sources of particles, of the physics of their interactions,
and of receptors. With this description, one should be able, in princi-
ple a t least, to predict the flow of energy in and out of volumes of
interest and to calculate the dose.
Two general methods, not always equivalent, are used in these
calculations. In the so-called deterministic approach, the radiation
field is characterized by functions describing the phase-space density
of particles a t a given point and its distributions with respect to
energy, angle and particle type. The term "deterministic" refers to
the fact that these density functions (or their distributions) are sub-
ject to mathematical laws termed transport equations, and, there-
fore, given the sources, boundary conditions and interaction coeffi-
cients, one can calculate their values a t each point and time.
In the second approach, the particle trajectories are simulated
individually according to the stochastics of the physical interactions.
The treatment is in terms of probability distributions, for instance,
the probability that a given number of particles (in a n energy and
solid angle interval) traverse an element of area. The stochastic
approach (e.g., a Monte Carlo method) allows the calculation of any
deterministic quantity as the average of the corresponding stochastic
one over its probability distribution, although, of course, a large body
of additional information is available (e.g., the moments of these
distributions). The simulation of particle trajectories is performed
with Monte-Carlo techniques.
Because, in the calculation of dose (a non-stochastic quantity),
the additional information provided by the stochastic method is not
needed, one prefers to use the transport equations. For many practi-
cal situations, however, the complexity of the source-receptor sys-
tems makes a deterministic approach impractical and then the Mon-
te-Carlo method remains the only suitable one.
The basics ofthe deterministic approach to dose calculation (trans-
port formalisms) are presented in Section 2 together with several
simple examples. Specific elements of radiation transport are covered
in succeeding sections. They are: sources (Section 3), receptors (Sec-
tion 4), and interaction coefficients (Section 5).
Dose calculations are sometimes based on general theorems (Sec-
tion 6),or on the use of transport methods (Section 7), or they may
utilize the stochastic approach (Section 8). Section 9 presents some
geometric considerations relevant to this subject.
Section 10 deals with calculations of the dose equivalent, which is
the absorbed dose weighted by the biological effectiveness of the
charged particles producing it. These calculations are performed for
use in radiation-protection evaluations.
2. Transport Formalisms

2.1 Concepts in Dose Calculations

The primary objective of dose calculations is to determine the

absorbed dose a t a given location (the point of interest) in a finite
mass (the receptor) due to ionizing radiation originating from a
region (the source) that is outside the receptor for external irradia-
tion and within the receptor for internal irradiation. A further objec-
tive of dose calculations can be the determination of the dose equiva-
lent a t the point of interest. From Equation (1.21, it is obvious that
the physical information on which a dose calculation is based
includes the distribution in number, energy and direction of all
particles entering or leaving a differential volume surrounding the
point of interest or originating within it (sources).This type of infor-
mation, in a non-stochastic formulation, is the object of transport
theory. In the following, the basic concepts and quantities of trans-
port theory are introduced and the relation between these quantities
and the calculation of dose is established in a formal manner.
Consider a medium. Define
f i ,E, 2, t)d3rdEdu = the expected (mean) number of
radiation particles or photons (hereafter referred
to as "particles") in a volume element d3r about 3, (2.1)
with energy E in dE and moving with directions in
the solid angle du about ii (unit vector) a t time t.
If 8 and a are the spherical coordinates of Ti, then du=sin8 d8 da,
and n is the particle density distribution with respect to E and ti. The
particle density, NG, t) is then

Let 3 be the velocity of a particle of energy E. Then

where v is the magnitude of .;

 2.1 CONCEFTS IN DOSE CALCULATIONS / 7
A closely associated concept is the particle current-densitydistribu-
tion with respect to E and 2:
7 G,E,Z, t)-dSd.Edu = expected number of particles
that cross an area dS per unit of time with energy
E in dE, moving in directions within du about Z at (2.4)
time t (by definition, & is perpendicular to the
surface element dS).
The particle current density is:

A simple relation holds between n and 7:

The quantity in square brackets in Equation (2.6) is the flux-density

distribution, 4, in E and 2,

which, when integrated, yields the flux density (also called fluence
mte)
6,t) = J#bu &i,.E,2,t). (2.8)

It is important to note that, although @andz4re measured in the

same units, they have very different meanings: J - & is the net flow
rate of particles through dS (i.e., outgoing minus ingoing particles)
while specifies the total rate of particles going through dS.
The rate of change in the energy absorbed in a volume can be
related directly to jG,E,ii,t). If sources are present in the volume,
one can define, in analogy with Equations (2.1) and (2.4),
sG, E,ii,t)d3rdEdu = the expected number of particles
produced per unit of time in d3r about 7, etc. (2.9)

One has then:3

3Qis the change in r e ~ maas

t in interactions within the volume.
8 / 2. TRANSPORTFORMALISMS

With the use of Green's theorem (conversion of a surface integral to

a volume integral), the first term on the right-hand side of Equation
(2.10) becomes:

(The minus sign above results fkom the convention of dS pointing

outward, i.e., in the direction of particles leaving the volume).
By considering an infinitesimal mass dm of density p,

one arrives at the final expression:

or, by integrating over the time, t, to an equivalent relation for D.

This is the desired (albeit formal) relation between the dose, D G , ~ )
a t a point 7, over time t, and the field quantities 4 and s (or 3 and
s). The calculation is thus reduced to the question of calculating $J
or y, which constitutes the objective of transport theory. 4, s and Q
in Equation (2.13) include all types of particles present in the field.
The flux density distribution, +(T, E,;,t), satisfies the transport
equation. Because of its fundamental importance in dose calcula-
tions, this equation is derived in full detail in the next section.

2.2 The Transport Equation

Consider a volume V with surface area S. The net change per unit
time in the total number of particles of a given type in V within the
phase-space element dEdu is given by4:

'All the quantities used in the discussion of the transport equation refer to a single
type of particle as opposed to the more general definitions given in the previous
section. For simplicity the same notation is used.
 2.2 THE TRANSPORT EQUATION 1 9
One should note that s includes here not only "true" sources, but
also contributions from other types of radiation in the field undergo-
ing transformation to the particle type to which Equation (2.14)

(3mll
refers. This is discussed later in more detail. The term - repre-
sents changes in n (per unit time) due to collisions in the medium.
Equation (2.14) is simply the balance of the particles entering or
leaving the phase-space element dEdu throughout the volume V, per
unit time.
With:

Equation (2.14) becomes:

Because this equality is valid for any volume V, it follows that

which is the transport equation for nG, E,z,t).

Consider now the collision term. Let
p&, E) = probability per unit pathlength of a particle (2.18)
a t ;' and with energy E to have an interaction.
If the interactions are uncorrelated, then f i is proportional to the
density of scattering centers in the medium, n,:
CLT.G,
EE) = nMG)u(E) (2.19)
The interaction coefficients and a are also called macroscopic and
microscopic cross sections, respectively. Also:
Vp&, E ) = number of interactions per unit time expe-
(2.20)
rienced a t 7 by a particle with energy E.
One can defme various distributions oft.+ with respect to its vari-
ables. For example:
cL,G,E+E$+zr)dE'du' = the probability per unit dis-
tance traveled that a particle with E and will <
produce, as a result of an interaction at 7, a particle (2.21)
(including the primary one itself, i.e., a scattering)
with energy E' and direction ii',
10 / 2. TRANSPORT FORMALISMS
and

where pa corresponds to particle absorption. The absorption term,

however, will be formally omitted here with the convention that such
interactions are represented by transitions E + E' = 0. All the
quantities defined above, Equations (2.18 to 2-22], correspond to
mechanisms contributing to the rate of change in n due to interac-
tions. Thus, particles are removed from the phase-space element
dEdu by collisions ( k )while other particles scatter in from other
elements [p,(7,E'+E,ii1-Z)]:

Introducing Equation (2.23) into the transport equation, Equation

(2.171, and changing from n to 4 [Equation (2.711 gives:

This is the classical form of the linear transport equation with the
interaction probabilities explicitly shown.
It is important to remark also that Equation (2.24)is linear because
it involves C#I only to the first power (for instance, the other well
known transport equation, the Boltzmann equation for gases,
belongs to a nonlinear transport formalism; see also Subsection 6.1).

Examples
In order to further elucidate the concepts introduced above, several
examples of practical application of the transport equation, Equation
(2.24), are given here. These examples also allow the introduction of
a number of additional concepts widely used in dose calculations.
The presentation here follows closely that of Rossi and Roesch (1962)
and Roesch (1968).
Possibly the most typical dose calculation involves the situation
where a constant radiation source is used to irradiate over a given
time interval (O,t,). One is interested in the total dose delivered.
 2.2 THE TRANSPORT EQUATION 1 11
The transport equation, Equation (2.241, can be simplified by
observing that

and by defining:

w,E,<) = r 4G, E$,t) dt,

where t is the fluence distribution in E and 2, and
(2.26)

sOG,E,G) dEdu = dEdu

6sG, E,Z,t) dt, (2.27)

where Sois the expected number of particles produced per unit vol-
ume during [O,t,] in dEdu about E and ii. With these simplifications
and definitions, Equation (2.24), on integration over time, becomes

For a field with only one type of charged particle and Q = O (no
secondary particle transport), one obtains the dose delivered during
[O,tol by inserting Equation (2.28) in the time integral of Equation
(2.13):
12 / 2. TRANSPORT FORMALISMS

Because F~is defined per unit length, x , one recognizes in Equation

(2.29) the stopping power of the medium

The dose, then, is

The quantity in square brackets in Equation (2.31) is called the

mass stopping power of the medium for the charged particles under
consideration. Equation (2.31) equates the energy lost by the parti-
cles (- dEldx) and the energy absorbed in the volume (- D),an
equality that holds only because transport by secondary particles
was neglected.
Consider now a slightly more complicated situation, namely a field
with two kinds of particles, charged and uncharged. With subscripts
c and u indicating the particle types, one has now a system of coupled
transport equations similar to Equation (2.28):

The dose is then given by

The first integral in Equation (2.33) can be transformed in a man-

ner similar to that of Equation (2.29). Assume further that S, = 0
(no charged particle sources). Then:
 2.2 THE TRANSPORT EQUATION / 13
Now, let q, (Eu+E:) be the change in rest mass energy following a
transition Eu+E: and in which a secondary charged particle of
energy E, - E: - q, is produced. The energy transfer coefficient of
the uncharged particles, pk(Eu),defined as the fraction of E, trans-
ferred to kinetic energy of the charged particles per unit distance
traversed, is

pk(Eu)=
'I
- dE: p, (E,+E:) [E, -E: - Q,(E,+E:)I.
Eu
(2.35)

The first integral in Equation (2.34), with the correction for rest
mass energy changes, q,, represents, then, the sum of the initial
kinetic energies of all the charged particles created by the uncharged
particles per unit mass:
1
KG) = IdE,E u ' P u ~E,)
, - pk(Eu). (2.36)
P
K is called the kermu of the uncharged particles.
Frequently the second integral in Equation (2.34) is negligibly
small compared to K (the same holds, then, for dQJdm). The condition
when ly,(?, Ec)does not depend on F' (i.e., is constant over the field)
is called charged particle equilibrium. Then, because $'PC= 0, one
has

a relation useful'in photon or neutron dosimetry.

3. Sources

Ionizing radiation is produced by accelerators that impart kinetic

energy to charged particles such as electrons, protons or other atomic
nuclei. It is also emitted in radioactive decay and in both fission and
fusion, and it is incident upon the earth from outer space (cosmic
rays). The radiation field at a point consists not only of the particles
and photons from these original sources, but of secondary radiations
produced in intervening or proximal matter: deceleration of charged
particles produces x rays, nuclear reactions produce neutrons and
other particles and quanta, and absorption and scattering usually
produce secondary radiations as well as changes in the energy and
direction of the primary. Any one of these original or secondary
processes can be designated as the "source" for the purposes of dose
calculations, as is most convenient in the circumstances.

3.1 Specification of Sources

A calculation of the absorbed dose starts with the solution of the

transport equation, Equation (2.28)above. The sources can be repre-
sented in this equation in two ways: The quantity So,the number of
particles or quanta with a given energy emitted per unit volume of
the medium, is specified. For example, for radionuclides distributed
in a medium, the emission can be obtained from the activity per
unit volume of the radioactive material. Alternatively, the sources
themselves are not specified; instead, the number of particles or
quanta per unit area and their energy incident on the surface of a
medium are given. The specification of such a boundary condition is
a convenient way of dealing with a source whose structure plays no
significant role in determining the radiation field in the material.
Examples are beams of radiation from accelerators and cosmic rays.
Both methods of specifying sources can be used in a single calcula-
tion, but joint use is avoided here. Instead the calculation is broken
into parts. For example, in the calculation of secondary radiation
fields, the primary radiation may be given as a boundary condition
and the distribution in energy, direction, and position of the radia-
tions it produces is then calculated. These results are then used as
the source for calculation of the remainder of the radiation field.
 3.2 SLMPLIFIED REPRESENTATIONS OF SOURCES 1 15
Instead of using boundary conditions, one can devise functions for
fictitious sources on the surface of the medium that will produce the
same radiation field. The radiation emerging from these fictitious
sources must have the same distribution in energy and direction a t
the surface of the medium as the radiations from the actual sour-
ces. Whenever appropriate, this replacement will be assumed in
the rest of this report because it simplifies discussion of the tran-
sport equation and its solution by using only one method of source
representation.

3.2 Simplified Representations of Sources

The analytical methods for solving the transport equation are

difficult to apply to complex configurations of sources. The Monte-
Carlo method is better in this regard, but increasing complexity may
make its cost prohibitive. A common practice, therefore, is to replace
the complicated source structure with a simpler structure whenever
doing so will retain the desired accuracy. For example, a real 60Co
gamma-ray source is a small pellet of material containing the cobalt
inside a sealed capsule to prevent loss of radioactive material. This
arrangement can often be simulated by a point source because the
radiation field of the point source differs significantly from that of
the real source only a t small distances. Furthermore, the pellet
source can usually be assumed to emit photons isotropically.
Even complicated sources can often be simplified. Sources, such as
nuclear reactors, massive accelerator targets, etc., have structures
of such complexity that calculating the radiation emerging from
them is in itself a major task apart from the computations of the
subsequent absorption and transmission of that radiation. In such
cases, the emerging radiation and the subsequent transport are fre-
quently calculated separately. For example, people in the vicinity of
a nuclear reactor, but not employed there, are apt to be a t distances
large enough that the fission source can be simulated by a point
source emitting the spectrum calculated to emerge from the real
source. On the other hand, workers may be close to the reactor;
therefore, for them, its geometrical structure may have to be taken
into account.
Another way of simplifying complex sources is to consider them
divided into very small pieces, each of which can be treated as a point
source. This method is particularly valuable if the radiation field a t
a given distance from each of the points is the same except for a
factor (proportional to how much radiation is emitted a t that point)
16 / 3. SOURCES
independent of that distance. Then, the field need be calculated for
a point source only once and the total field obtained by combining
this point-source field for all the point sources that make up the
complex source. For this method to work, the source and surrounding
media must be reasonably homogeneous out to distances from the
source a t which no one point source makes a significant contribution.
Sources other than the point source can sometimes be treated in
the same way. If the sources and media are such as to make the
radiation field one- dimensional, "plane" sources that emit uniformly
over an infinite plane may be useful. For example, in calculations of
doses from radioactivity, the emission from each element. ofthe plane
is isotropic and the plane source is called a "plane-isotropic source".
Plane sources are also useful in simulating the boundary conditions
for radiation incident from outside the me&um. Then the emission
from the plane should not be isotropic; rather the emission should
be only in the direction or directions of the incident radiation. These
are "plane-directional sources".
If a source emits radiation with a spectrum of energies, calcula-
tions can be made for sources that emit monoenergetic particles or
quanta ("monoenergetic sources") and the solutions combined in
proportion to the abundance of each type in the emitted spectrum.
Further simplification results if the spatial conditions, as described
above, permit straightforward superposition of point or plane
sources; then one need only calculate monoenergetic point or plane
sources with suitable directions of emission.
Superposition is similarly used for fields containing different kinds
of particles. Provided that the different kinds of radiation are not
coupled (i.e., one does not give rise to the other), calculations are
made for quanta and for one kind of particle a t a time and the
solutions added. Actually, most mixed fields are coupled (e.g., elec-
trons and photons produce each other) but often to a small enough
degree that separate calculations for each radiation are adequate,
with coupling taken care of by iterative corrections. In high-energy
phenomena such a s cosmic-ray showers, however, the coupling is so
large that it must be included ab initio.
4. Receptors

The aim of absorbed-dose calculations is usually to determine the

dose in tissue. But doses in other media also are required sometimes,
for example, in the parts of an instrument for measuring radiation.
Because of this diversity, the non-specific term receptor is used for
the object in which the dose is to be determined.
For the same reasons that lead one to simplify the structures of
the sources of radiation in transport calculations, one also simplifies
the structures of receptors. For example, early calculations aimed a t
estimating neutron doses for purposes of radiation protection used a
receptor consisting of a 30-cm thick slab of tissue-equivalent mate-
rial that was infinite in lateral extent (NCRP, 1957). Later calcula-
tions used increasingly more realistic receptors. The next one used
was a tissue-equivalent cylinder 30 cm in diameter and 60 cm high
(Auxier et al., 1968). More recently, a n anthropomorphic receptor
was used in which body surfaces, organ boundaries, and bone vol-
umes were defined by simple mathematical equations (Snyder et al.,
1969). Figure 4.1 illustrates this receptor. Further refinements of
this receptor have been made to simulate the two sexes and to simu-
late children (Cristy and Eckerman, 1985).
In complex irradiation conditions, it may be economical to calcu-
late first the radiation field with no receptor present and then the
field in the receptor, using the receptor-free field for the boundary
conditions on or near the surface of the receptor. Clearly, such a two-
step calculation is possible only if introduction of the receptor does
not materially alter the receptor-free field outside the receptor. For
example, the transmission through thick shields is ordinarily evalu-
ated without considering a receptor; subsequently, one starts with
the energies and directions of the radiation emerging from the shield
to calculate the dose distribution in the receptor. An extension of
this step-wise procedure is followed, for example, in calculating the
doses from nuclear weapons in the following sequence: the radiation
emerging from the weapon; the radiation from the rising fireball;
the transport of these radiations to points on the bare ground; the
effects of houses and buildings; and the penetration in a receptor
simulating the human body. In the first two steps, the sources are
fissions, nuclear reactions, and radioactive emissions; in the follow-
18 / 4. RECEPTORS

Fig. 4.1. An anthropomorphic model receptor. Dimensions and coordinate system

of adult human phantom. (After Snyder et a[.,1969.)
ing steps, the input conditions are the fluences calculated in the
preceding step.
Apart from simplifying the calculation, the step-by-step approach
has the advantage that calculation of one receptor-free field suffices
for calculations in a number of different receptors or in any one
receptor by a number of different methods.
The calculation of the dose in the receptor essentially follows the
propagation of radiation from the receptor-free field after it crosses
the receptor boundary. One can do just the reverse: start with radia-
tion a t a point in the receptor and follow it backwards in time till
it crosses the receptor boundary. Then one must superpose these
backward solutions in proportion to the receptor-free fluences a t the
boundary. These backward calculations are done by what are called
adjoint methods. In principle, the radiations could be followed back-
ward to the points a t which they emerged from the original source.
The advantage of the adjoint calculation is that one such calculation
suffices for any number of different receptor-free field calculations.
Receptor-free calculations and measurements in nearly receptor-
free conditions are frequently employed in radiation protection. The
results can, for convenience, be characterized by the kerma in some
material at a given point. Because of differences in how they attenu-
ate radiation, different receptors placed a t that same point will pro-
duce different depth-dose distributions.
Calculations relating to the measurement of radiation present a
somewhat different aspect of the receptor concept: here one is con-
cerned with the receptor as a n instrument rather than as a biological
target. Occasionally the receptor-free condition is employed as a n
intermediate step. Even when the instrument is calibrated to read
some quantity that will exist in the biological target, design compro-
mises, differences in calibration conditions, and other conditions,
almost always exist to make the reading and the quantity differ. The
art of making measurements is to be aware of these differences and
to avoid (or minimize) them. Calculations of radiation fields are
often used to understand and correct for the differences between the
instmment-receptor and body-receptor conditions.
5. Cross Sections

5.1 Schematization

The penetration, diffusion, and slowing down of ionizing radiation

in an extended medium, and the associated transfer of energy to the
medium, can be treated as resulting from a large number of binary
interactions between radiation particles and certain constituents of
matter. The radiations of interest in radiation dosimetry are photons
(X rays, gamma rays, continuous bremsstrahlung, characteristic x
rays), neutrons, and charged particles (electrons, mesons, protons,
alpha particles, heavy ions). The interactions of interest are either
electromagnetic or nuclear, and take place with various scattering
centers such as atoms, molecules, atomic electrons, nuclei, or individ-
ual nucleons in nuclei. The choice of the appropriate scattering ten-
ters in a transport calculation depends on the nature of the interac-
tions, and also on the approximation in which they are treated. In
any case, a dosimetry calculation usually involves the solution of a
multiple scattering problem, starting with an assumed knowledge
of the cross sections for single interactions.
In order to have a framework for the discussion of cross sections
needed for dosimetry calculations, it is useful to discuss the schemati-
zation in which the cross sections are used. The schematization,
which is assumed in the derivation of the linear transport equation,
Eq. (2.241, and which also underlies the formulation of the Monte-
Carlo method, is based on the following assumptions and approxima-
tions:
A. The various radiations are treated as particles with well-defined
trajectories, and the effects of successive collisions with scattering
centers are treated in terms of probabilities rather than in terms of
probability amplitudes, that is, quantum-mechanical interference
effects are disregarded. This approach precludes the treatment of
phenomena of minor importance for dosimetry such as the diffraction
of photons, electrons or neutrons a t low energies, where the particle
wavelengths are of the same order- of magnitude as the spacing
between atoms. This schematization assumes that the radiation par-
ticle wavelengths are small compared to the sizes of collision targets
and also small compared to the average distances between collisions.
 5.1 SCHEMATIZATION 1 21
Finally, it should be noted that quantum-mechanical interference
effects are neglected only in the treatment of multiple collisions,
whereas the underlying single-scattering cross sections are derived
on a quantum-mechanical basis.
B. The particle trajectories in the medium are assumed to be zig-
zag paths consisting of free flights interrupted by sudden localized
interactions with scattering centers. The particle momentum and
energy are assumed to be constant during each free flight, changing
suddenly a t the time of the interaction which terminates the free
flight.
C . The scattering centers are assumed to be distributed throughout
the medium with a specified average density distribution, but other-
wise randomly. This assumed randomness precludes the treatment of
channeling phenomena, that is, the modification of charged-particle
penetration in crystalline materials that occurs when the direction
of the motion of the particle coincides, within certain critical angles,
with one of the symmetry axes or planes of the crystalline medium.
D. The probability that an interaction of type j will occur in a small
segment A t of the flight path is assumed to be equal to yAt where
p, is the interaction coefficient [see Equation (2.18)l. It is assumed
further that, in the limit of vanishingly small At, the probability of
more than one interaction of type j in A t is negligibly small. The
probability that no interaction of type j will occur in A4 is therefore
equal to 1 - p,Al. The interaction coefficient can be expressed as the
product of two factors [see Equation (2.19)l

where n is the density of scattering centers, and oj is the cross section

for the interaction. Note that p, is a probability per unit pathlength,
n is the number of scattering centers per unit volume, and the cross
section ujhas the dimensions of an area. The total probability of an
interaction per unit pathlength, taking into account all modes of
interaction, is given by
= Zp,. (5.2)
j

The probability that a free flight will be terminated when its length
is between C and t + d t is given by

and the mean free path between collisions is

E. The irradiation intensity is assumed to be low enough to satisfy

22 1 5. CROSSSECTIONS
two conditions: (i) The interactions between the radiation particles
are negligible compared to their interactions with the medium. This
condition does not hold for certain electron accelerators with intense
beam currents of lo4 to lo6 A used in radiation effects studies. (ii)
The properties of the medium are not changed significantly as the
result of the irradiation, i.e., it is extremely unlikely that the radia-
tion particles have interactions with scattering centers whose prop-
erties or distribution have been altered as the results of earlier
interactions with other particles.

5.2 General Aspects of Required Cross Sections

Within the framework of the schematization indicated above, the
following information is needed in terms of cross sections, interaction
coefficientsor related quantities:
A. Location-of next collision. Suppose the particle has emerged
from the previous interaction (or has been injected into the medium
by the radiation source) at the point 7, with energy E and in direction
--,
u. The next interaction will occur a t the point
-+
r =3, + e5, (5.5)
where e is the length of the free flight. The probability distribution
of the distance t o the next interaction is thus determined by the
distribution of lengths of free flights, p(e), as given by Equation (5.31,
and thus depends only on the total interaction coefficient C(T.
B. Type of collision. With probability M h , the collision is of type j.
C. Outcome of collision. Detailed differential cross sections are
required that provide the following information, in terms of probabil-
ities for various contingencies:
(i) What happens to the particle making the collision? Is it scat-
tered or absorbed? If scattered, through what angle is it
deflected, and how much energy does it lose?
(ii) What happens to the struck collision target? How much recoil
energy does it receive, and in what direction is it set in
motion? Is the collision elastic, with the target left in its
ground state, or is it inelastic, with the target left in an excited
state? Is there any long-lasting modification of the target,
e.g., as the result of an ionization or fragmentation? What is
the expected chain of events when the target, or some of its
fragments, are eventually de-excited?
(iii) What kinds, and how many, of secondary particles are emitted
by the struck target, either promptly or after some delay?
What are their energies and directions of motion?
 5.2 GENERAL ASPECTS OF REQUIRED CROSS SECTIONS 1 23
In order to specify the various outcomes of radiation interactions,
one needs a vast set of electromagnetic and nuclear cross sections.
Fortunately, the complete array of cross sections may be required
only in the most detailed types of calculations. Many other transport
methods are based on approximations that reduce the cross-section
requirements. For example, in the treatment of neutron diffusion,
the simplifying assumption is often made that elastic scattering
is isotropic in the center-of-mass system. In the treatment of the
penetration and diffusion of charged particles, the continuous-slow-
ing-down approximation is often used, in which the kinetic energy
of the charged particles is assumed to decrease continuously along
the particle tracks a t a rate given by the stopping power (mean
energy loss per unit pathlength). The applicability of this approxima-
tion derives from the circumstance that the transfer of energy from
charged particles to the medium takes place via many thousands of
transfers, most of which are small (less than - 30 eV), with only
rare collisions involving large energy transfers. In the continuous-
slowing-down approximation, only stopping-power data are needed
as input, but detailed inelastic-scattering cross sections are needed
when energy-loss straggling is taken into account.
Other reductions of the cross-section requirements can result from
the nature of the problem under consideration. For example, in a
calculation of the intensity and energy spectrum of the gamma rays
emitted from a large encapsulated 60Coteletherapy source, one can
concentrate almost entirely on the penetration, diffusion and slowing
down of the gamma rays in the source and treatment head, and can
disregard the penetration of the secondary electrons also produced.
Similarly, when calculating the spatial distribution of absorbed dose
from x rays, it may not be necessary to consider the transport of
energy by photoelectrons and Compton-recoil electrons. Because the
ranges of these secondary electrons are small compared to the mean
free paths of the x rays, it is a good approximation to compute the
absorbed dose in terms of the integral in Equations (2.36) and (2.37)
where their transport is neglected.
Neither experimental data nor theoretical results by themselves
are, in general, adequate to provide the comprehensive and complete
cross-section sets required for transport calculations. It is often nec-
essary to use a n amalgam of experimental and theoretical cross
sections derived from a critical data analysis. Such an analysis
involves the reconciliation of different experimental and theoretical
results, the interpolation and extrapolation of cross sections, and the
application of consistency checks.
The critical evaluation of cross sections is particularly difficult for
interactions in which the energy of the incident particle is of the
24 1 5. CROSSSECTIONS

same order of magnitude as the differences between internal energy

levels of the scattering center. The cross sections, then, are character-
ized by thresholds and resonances, and vary irregularly as functions
of particle energy and atomic number. These conditions make the
theory, and even interpolation with respect to energy and atomic
number, quite difficult. Such difficulties are particularly severe for
neutron cross sections, and also arise for photoelectric cross sections
at low energies near the K-, L-, . . . , absorption edges.
It is often necessary to adapt the format of the cross-section input
to the requirements of the method for a dosimetry calculation. For
example, when the moment method is used (Fano et al., 1959), the
angular dependence of the scattering cross section must be available
in the form of expansion coefficients that allow the cross section to
be written as a sum of Legendre polynomials. For neutron transport
calculations in the multi-group approximation, the various cross
sections must be available in the form of averages over specified
energy bins. The application of the Monte-Carlo method may require
cross sections converted to cumulative probability distributions.
Thus, there is a need for specialized data libraries adapted to the
needs of particular transport computer programs. Many such librar-
ies have been collected and are available from the Oak Ridge Radia-
tion Shielding Information Center (RSIC, 1981).
To give a complete and detailed account of the cross sections needed
for dosimetry calculations is beyond the scope of this report. Table
5.1 lists a number of textbooks and monographs that contain authori-
tative discussions of the various cross sections. In Appendix A, the
characteristics of some of the more important photon, neutron and
charged-particle cross sections are briefly described and illustrated,
and references are given to review articles and data compilations of
up-to-date cross-section information.
 5.2 GENERAL ASPECTS OF REQUIRED CROSS SECTIONS 1 25
TABLE5.1-Textbooks and monographs dealing with cross sections
Reference Particles
Bethe (1933) electrons, protons
Bohr (1948) charged particles
Rossi (1952) photons, electrons
Bethe and Ashkin (1953) photons, charged particles
Heitler (1954) photons, electrons
Evans (1955) photons, charged particles, neutrons
Price et al. (1957) photons, neutrons
Fano et al. (1958) photons
Amaldi (1958) neutrons
Birkhoff (1958) electrons
Evans (1958) photons
Weinberg and Wigner (1958) neutrons
Fano (1963) protons, alpha particles, mesons
Northcliffe (1963) heavy ions
Beckurts and Wirtz (1964) neutrons
Mott and Massey (1965) electrons
Auxier et al. (1968) neutrons
Bichsel (1968) charged particles
Evans (1968) photons
Hubbell and Berger (1968) photons
Goldstein (1971) photons, neutrons
Bichsel (1972) charged particles
Jackson (1975) charged particles
Ziegler (1980) heavy ions
Starace (1982) photons
ICRU (1984) electrons, positrons
6. Transport Theory-
General Theorems
and Properties

6.1 Integral Form of the Transport Equation

A number of useful properties and theorems concerning the solu-

tions of the transport equation can be obtained if the differential
representation of Equation (2.24),

is changed to an integral one. In Equation (6.1):

4 = 447,E,Z,t),
& =- kF, EE),
and

For simplicity, only the variables and twill be shown explicitly in

the following.
Consider a change of variables:

One can rewrite Equation (6.1) in these new variables as

In this representation, the gradient term in Equation (6.1) has been

eliminated, and Equation (6.4) can be solved formally to yield:

x exp [ - J dpl&G -plZ Ell.

0
P
 6.2 ITERATIVE SOLUTIONS (ORDERS OF SCATTERING) 1 27
This is the integral form of the transport equation. Equation (6.5) is
not a n actual solution, because S contains 4 in its definition (Equa-
tion (6.2)) which is of course the unknown function. The interpreta-
tion of Equation (6.5) is straight forward: contributions to the flux
density distribution, 4, come from particles generated by the general-
ized source S a t any position (7-pi& O s p ~ m and time (t-plv) and
are attenuated exponentially according to the coefficient h , inte-
grated between their point of origin and ;.' When the medium is
homogeneous (i.e., C(T does not depend on ?) the exponential factor
in Equation (6.5) simplifies to exp [ - h(E)pl.

6.2 Iterative Solutions (Orders of Scattering)

Consider for simplicity a homogeneous mediuin and define two

operators T and K as follows:

where f is an arbitrary function. The transport equations, Equations

(6.2) and (6.5), can be written as
4 = TS,
and
S = K4 + s.
where K plays the role of a "scattering" operator, which generates a
source term from a flux density, and T performs rather the alternate
operation ("translation" operator).
Equations (6.7) can be combined to yield

where I is the unit operator and, for reasons apparent below, the
notation So= s is introduced.
By introducing the operator (I - Tk)-I one can rearrange Equation
(6.8') as
28 1 6. TRANSPORT THEORY-THEOREMS AND PROPERTIES
Then,

with the obvious relations

etc.
The formalism described by Equations (6.9) and (6.10), known as
the Neumann expansion, has a simple interpretation in terms of
sequences of collision events. Thus, the first term, do, represents
particles arriving a t ;:directly from the source [see Equation (6.6)l
and after attenuation. The second term, dl, represents particles that
had a collision and scattered once to produce a new "source", S1=
Kdo, whose output was transported, TS,, to the point ,; and so on,
hence the name "orders of scattering." It is clear from Equations
(6.10) that the Neumann expansion, Equation (6.91, expresses an
iterative solution to the transport equation.

6.3 Density Scaling Theorem

A simple but important theorem holds that, if two configurations

differ in physical dimensions by only a scaling factor 5 4 , then one
can multiply the densities everywhere in the smaller by the inverse
scaling factor to bring about precise agreement between radiation
fields in these two configuations a t all corresponding points (Spencer
et al., 1980). This theorem requires, as a precondition, a similar
scaling relationship between the radiation source functions in the
two configurations and the density-independent interaction coeffi-
cients.
This scaling theorem can be understood from Equation (6.8) or
(6.8'1, where the translation operator, T, and the scattering operator,
K, operate sequentially; any factor that cancels when both are
applied is essentially irrelevant. Thus, if distances are decreased by
a factor t< 1,then scattering probabilities are increased by the same
factor. But the last term in Equation (6.8) remains unchanged only
if the source is also intensified by the factor 5-l.
 6.4 FANO'STHEOREM 1 29
6.4 Fano's Theorem

This well-known and important theorem (Fano, 1954) states that,

in any medium with only density variations from one point to
another, if there is also a radiation source function that differs from
one point to another only in proportion to the density ratio, the
resulting radiation field is everywhere the same, independent of local
densities. Two cases with different density functions can then be
equated by equating source strengths per unit density, i.e., using a
single characteristic number.
To prove this theorem, note that in the combined operator (TK) of
Equation (6.81, the integrals involve only the differential dp multi-
plied by the local density because both and p, are proportional to
the local density. Thus, if distances are expressed by a variable 77
defined as

~ ( p=
) 1
0
P
dp1p(7-prz), (6.11)

then the transport Equations (6.8 and 6.8') will depend only on
local density in the final (source) term. But if So is also everywhere
proportional to the local density, Equations (6.8 and 6.8') can be
written entirely in terms of position-independent functions, and, as
a consequence, must be independent of position.
This theorem can be applied in plane symmetry to uniform radia-
tion sources confined to a plane. There results a plane-scaling theo-
rem which applies to configurations with differing plane-density
variations (Spencer et al., 1980).

6.5 Energy Conservation

While energy conservation in radiation transport is a basic expec-

tation, it is instructive to formulate an expression of it. One formula-
tion of energy conservation is that for particles of a given type, in a
given band of kinetic energies, the input of kinetic energy into this
band through radiation sources exactly matches the losses through
the various interactions by the particles.
To obtain a n expression of energy conservation, it is easiest to
utilize the differential form, Equation (2.28), for the transport equa-
tion. There is no interest in time dependence; hence, one may ignore,
or integrate out, this variable. Each term is multiplied by the energy
E, and integrated over energy above an arbitrary cutoff designated
by A. With the use of a n expression for the source term that sepa-
30 1 6. TRANSPORT THEORY-THEOREMS AND PROPERTIES
rately identifies scattering interactions, i.e., Equation (6.21, the
energy input into the particle field above A is expressed by the term
Es.
IAmdE
The main complication, then, concerns the scattering interactions
that may regenerate a particle still in the beam, albeit of lower
energy. By changing the order of integrations in the scattering term
this possibility can be formulated precisely. The resulting expression
is shown in Equation (6.12), with explicit indication of most variables
omitted to clarify the meaning.

One needs mainly an understanding of the expression in the curly

brackets. The quantity, p&, clearly refers to the total energy of
particles that interact. The integral term represents the energy
remaining in the beam, i.e., the interactions that leave particles with
energy above A. Because this subtractive term represents the beam
energy that remains, the full curly-bracket factor weights in accor-
dance with the part of the interaction energy that leaves the beam.
The irrelevant variables, position and direction, can be removed
by integrations that also remove the gradient term. The remaining
terms equate total energy removed to total energy injected, i.e.,
energy is consenred.

6.6 Superposition

The transport equation is a linear equation, which means that any

linear representation of a given source function, s, can be translated
into a solution using the same superposition for solutions for the
component sources. Thus, if

sG, E,;) = I~x,

r E,ZA g(z), (6.13)

then the solution is

where +f is the solution when f is used as the source function in the

same configuration. Note that the variable z may be discrete, and
 6.7 ADJOINT TRANSPORT EQUATION 1 31
the integral, then, is replaced by a sum. The individual source terms
f may be products of functions of the different variables in arbitrary
combinations. The main requirement is that the configuration, the
interaction coefficients, and the boundary conditions remain the
same for different 4, and for +8.
This theorem is essential for procedures based on Fourier trans-
form methods, Legendre transform methods, etc. It will be utilized
in Section 7 of this report, which deals with methods of solution of
the transport equation.
One of the more interesting aspects of superposition in the space
variables is the full generality of sources uniformly distributed on
infinite plane surfaces. It can be shown that point-delta functions
6G -7,) can be represented by superpositions of functions
6 6 . C;:-;,)I (Morse and Feshbach, 1953; Berger and Spencer, 1959)
corresponding to arbitrary planes passing through 3,. This represen-
tation constitutes a proof of the completeness of such systems ofplane
sources.

6.7 Adjoint Transport Equation

The adjoint operator Ht corresponding to an operator H is defined

by:

where f and g are arbitrary functions and the inner product may be
defined as

where g* is the complex conjugate of g.

Consider the transport equation, Equation (2.24), and define a n
operator L such that

With this definition Equation (2.24) becomes:

Without significant loss of generality, one can assume th2t the

boundary condition on 4 (and therefore on L) is such that +(R.,E,i)
32 1 6. TRANSPORT THEORY-THEOREMS AND PROPERTIES
is zero at a surface& for directions pointing into the medium. Then,
the adjoint of L can be obtained from Equations (6.15) and (6.17):

~t = -
-.
.v + CI.r
+

II
- d ~ '~U~~,(E-+E:Z+S~~),
for E'<E.
(6.19)

Equation (6.19) is valid when the functions on which Lt operates are

such that they are identically zero on the surface R, discussed above
for any directions pointing out of the medium. As will become clear
shortly, for dose calculations this condition describes a natural
restriction.
In analogy with Equation (6.18), one can write, now, an adjoint
transport equation for the adjoint flux density 4t a s

where the adjoint source, st, is arbitrary.

The physical significance of the adjoint equation can be illustrated
as follows. Consider the time-independent versions of Equations
(6.18) and (6.201, as in Equation (2.28).
L+ = Soand Lt$t = g. (6.21)
Because L and Lt are adjoints, it follows from Equation (621) that
(#,So> = (SL,$)- (6.22)
Now let So represent a monoenergetic, monodirectional unit point
source of charged particles,

and let Si be the mass stopping power of a receptor at a point 7'

where p is the local density,

Then Equation (6.22) yields

where the notation shows that 1+9is a solution of the transport equa-
tion corresponding to the point source, Equation (6.23). But $' of
Equation (6.25) is the dose at the point ;f' in the receptor due to the
point source in Equation (2.31).
 6.8 RECIPROCITY 1 33
It can be shown, as might be intuitively obvious, that for a single
radiation with an extended source, s,(;f,
E,;), the dose is given by

The adjoint solution, I)+,provides, therefore, a direct relation between

the dose in a receptor and a distribution of sources.

6.8 Reciprocity

Reciprocity is the elementary idea that source and detector "can

be interchanged" without changing the response of the detector.
This concept is particularly easy to visualize when both source and
detector are idealized points, and when emission and response func-
tions are isotropic. But it is not hard to prove that, even then, this
concept can hold only under fairly rigid limitations. Thus, consider
two semi-infinite media in juxtaposition, e.g., lead and water. If a
detector and a source of gamma rays are confined to idealized points
on opposite sides of the interface, interchange will surely change the
detector response because of differences in the scattered radiation
components in the vicinity of the detector in the two media.
Complicated barrier shapes, such as in ducts that require one or
more changes of direction in any trajectory to the detector, are likely
to have a similar lack of reciprocity unless the configuration as a
whole is symmetric with respect to interchange of source and detec-
tor. This condition is a general requirement, and it is not obeyed by
the previous case of adjoining, but different, semi-infinite media.
If either source or detector is not isotropic, then, even if both are
points in a medium which is infinite and homogeneous, one cannot
interchange only their positions. Consider a point isotropic detector
and a point monodirectional source. To expect reciprocity, one must
accompany their interchange with a reversal, or reflection, of the
source angular distribution. Thus, a second requirement for reciproc-
ity is that an interchange of position must be accompanied by a
reversal of both source and detector angular distributions.
But sources and detectors can be complicated, and their inter-
change is not always a simple matter. Consider three points in an
infinite homogeneous medium. At the f i s t point is a source, while
detectors with equal response functions are a t the other two points.
This case is to be compared with a single detector at the first point
and two equal sources a t the other two points. Source emission and
detector response functions are unchanged between the two cases.
34 1 6. TRANSPORT THEORY-THEOREMS AND PROPERTIES
Then it can be seen that addition of the two detector responses in the
first case gives exactly the same result as the two source contribu-
tions to the detector in the second case.
Note that in the first case just discussed, two quite different detec-
tors a t the two points would destroy the simple interchange concept,
as would two quite different sources in the second case. But, if a t
different source points the emission is identical, and at different
detector points the response is likewise identical, the interchange of
source and detector can apply to very complex sources and equally
complex detectors. The number of points in the simple illustration
above can be increased and the assignment of any point to source or
detector can be arbitrary. Also, either source or detector can have
the form of a surface or volume configuration provided there is unifor-
mity of source emission or detector response per unit surface area or
per unit volume. Note that this argument for reciprocity, unlike that
of Section 9.2, does not depend on isotropic emission and detection
properties, and, for infinite homogeneous meda, it applies to scat-
tered components as well.
Introduction of, say, different detector or source types into a single
configuration destroys the concept of simple interchange; but it is
possible, though perhaps seldom useful, to appeal to superposition
and apply reciprocity to detector and source pairs if the configuration
otherwise permits. For the special cases of isotropic sources and
detectors, see Section 9.2.

6.9 Transport Equations in Commonly

Used Coordinate Systems

The simplest and probably most studied form of the transport

equation is the plane symmetry case. Here the single space variable
z = ? .'FI is measured along the basic reference axis 'FI, which is fixed
in space. Either integration over lateral planes or uniformity on
lateral planes is assumed for the source and, consequently, for the
flux density 4. Directions are measured relative to 'ii also, and we
have 4=4(z,uZ, E,t) where u, = ;I- R and
- arp
u . Vrp = u,-.
_t

az
In expressions such as Equation (6.5), the generalized variable can
be written as follows: ?+z, &uZ, 7 -pii+z-pu,.
Almost as important is the transport equation in spherical symme-
try. The natural reference axis for this case is the radial unit vector,
- 6.9 TRANSPORT EQUATIONS 1 35
n. For spherical sources, 4 varies only along a radial line: All varia-
tions of 4 from one point to another on any of the concentric spheres
vanish. Hence 4=4(r, y,E,t), where r = 3 .Z and u, = u - n = cos
+ +

Or, and the convection term takes the form

+ -
u.V4 =
a4
ur-
ar
a4
+ --.sinr 8, ae,
The simplest problems involving more than one space variable
may often be studied in cylindrical coordinates (z,p,P) with corres-
ponding unit vectors (7E,7i,z). Again, the most common case utilizes
a source geometry for which there are no variations on circles cen-
tered on the axial line and lying in one of the lateral planes (fixed
value of 2). Then 4 = +(z,p; 8,,f?; E,t), where 7 = 7ip + xz,
- a a l a
u = up I + uzX + up;, and u .V = uz - + updp + pug q.
+ +

az
84 vanishes
If there is no dependence of the source on f?,the term with -
af?
and

The direction cosines are, then, u, = cos 8, and up = sin B,cos+.

These three cases cover the requirements of much of the literature,
and demonstrate the pattern for other types of symmetry that may
arise.
7. Transport Theory-
Methods of Solution

7.1 Introduction

With very few exceptions, the transport equation, Equation (2.281,

can be solved numerically only with the use of computers. To set up
the solution of the transport problem on a computer requires, for
instance, the replacement of a function by sets of values of this
function (e.g., a t discrete points of the variables).
There are two general ways in which the transport equations
are expressed as discrete values, namely the discrete ordinate (DO)
method and the method of expansion with known functions. In the
DO method, a continuous function, Nx), is replaced by a set of num-
bers, {4i),defined by

Similarly, derivatives and integrals are replaced, for example, with:

and (7.2)

where the ai are known and depend on the algorithm used for evalu-
ating the integral. In this way, the transport equation is replaced
with a set of algebraic equations for the unknowns di.
In the second method, one approximates the function &x) using
a finite set of known functions f,(x):
 7.1 INTRODUCTION / 37
where the unknowns now are the expansion coefficients 4i. As an
example, it is customary to represent the angular dependence of the
flux density using spherical harmonics, Ye,(;):

or, for one-dimensional problems, to use the Legendre polynomials,

P,(cosfI):

The solution obtained by truncating the sum a t t' = N is known as

the P, approximation scheme.
A full transport problem can have up to seven independent vari-
ables,;( E,$) and a complete representation in terms of discrete
values will result in a system of equations with a large number of
unknowns. This complexity may force one to rely on Monte-Carlo
procedures, to be discussed in Section 8. But successful solutions to
transport equations have resulted from judicious choice of problems
with a minimum set of variables and with well-chosen representa-
tions for the required cross sections. Thus, questions of importance
are how to minimize the number of variables required, how to
describe usefully the most important features of required cross sec-
tions, how to keep numerical integration manageable, and what to
do about boundaries and boundary conditions.
It has turned out that problems with few variables, especially
when the dominant cross-section features are well-described, have
been instructive regarding the nature of the solutions as well as the
most appropriate methods for general use. Later methods, designed
to attack less specialized problems, have utilized this information
and in some cases have incorporated the elementary procedures and
results into basic routines for wide use. Thus thin-foil studies of
electron penetration and energy loss are applied directly when Mon-
te-Carlo programs describe penetration as a succession of thin-foil-
type segments. Additionally, there are cases in which a basic one- or
two-variable problem gives the essential information required by
important applications.
The discussion of such approaches is the objective of this section.
38 1 7. TRANSPORT THEORY-METHODS OF SOLUTION
7.2 Radiation Equilibrium and Space-Integrated
Radiation Fields

An important simplification occurs in the transport equation,

Equation (2.28), if the medium is effectively unbounded and homoge-
neous and the sources are uniformly distributed. Then the fluence
distribution, $, will not depend on position and, because the gradient
term is zero, Equation (2.28) can be integrated over the angular
variable to yield:

This is the "radiation equilibrium" form of the transport equation.

The energy, E, is the only variable, i.e., one deals with an energy
loss problem.
Several modifications of this equilibrium equation are useful.
First, substituting Equation (2.22) with omission of the p, term,
integrating over the energy from E to infinity and combining inte-
grals (with reversal of order of integration) gives:

The physical meaning of this equation is that the number of particles

of energy greater than E that acquire an energy below E (left-hand
side) equals the number of particles emitted with energy greater
than E (right-hand side). Next, if the equilibrium Equation (7.6) is
multiplied by the energy E before the integration and if the energy
integration is extended to zero, the result is

This equation is simply an expression of the law of conservation of

energy: the energy lost by all particles (left-hand side) equals the
total energy emitted from the source (right-hand side). On substitut-
ing the definition of stopping power, Equation (2.301, the conserva-
tion equation for charged particles becomes

On substituting the definition of the energy-transfer coefficient,

Equation (2.351, the conservation equation for photons becomes
 7.2 RADIATION EQUILIBRIUM 1 39

1 d(hv) #hv) hhv p, = (7.10)

Finally, the conservation equation for neutrons can be handled

the same way as for photons. But, when neutron scattering is sym-
metrical in the center-of-mass system, the ratio of the neutron energ-
ies before and after the scattering is independent of the initial
energy. This fact can be exploited by multiplying Equation (7.6) by
ln(E) rather than E and integrating. The result is

Now, introducing 6, the average logarithmic decrement (average of

the natural logarithm of the ratio of the energies before and after a
collision) (Marshak, 1947), Equation 7.11 becomes

Because all radiations generate secondary radiation of other types

(which may, in turn, produce the original type again), a multi-compo-
nent transport equation should have been used to obtain a multi-
component equivalent of Equation (7.6). This has been done in the
study of the slowing down of electrons, because their secondary parti-
cles are also electrons, indistinguishable from the primaries. Spencer
and Fano (1954) made the necessary change to the integral equation;
Equation (7.7), by adding to its right-hand side another source term
that represents the generation of secondary electrons by other elec-
trons in the medium, as shown in Equation. 7.13.

An equation identical in form with the equilibrium equation can

be formulated for many radiation fields that are not themselves
equilibrium fields. If the transport equation, Equation (2.28), is inte-
grated over a volume, the first term can be transformed by Gauss's
theorem:

where dS is a vector element of area of the surface bounding the

40 1 7. TRANSPORT THEORY-METHODS OF SOLUTION
volume of integration. The right-hand side is the net number of
particles leaving the volume. This number is zero in at least two
cases: a) radiation equilibrium, and b) if the radiation field is limited
in extent and the volume is taken large enough to include all of the
field. Then in terms of a new quantity,

the volume integration gives

Equation (7.16) is identical in form to Equation (7.6). The physical

meaning of the identity can be seen by considering, as an example,
the special case of a point source. The volume integral of Equation
(7.15) then takes a sample of the radiation everywhere around the
source. The equilibrium field samples the same way, because any
given location in the field is in every possible position relative to a
point in the infinite, uniform distribution of point sources.
The preceding mathematical operations correspond physically to
the familiar total-absorption experiment. In determining W values
or G values, for example, the entire tracks of charged particles are
contained in the material of interest. This material defines the vol-
ume of integration above.
The quantity y is called the distribution ofpath length with respect
to energy and is also called the degradation spectrum (Fano, 1953).
If d3r = dS dL (with dL along the direction c),then $dS is the
number of particles traveling in that direction that enter d3r,and dL
is the length of their path while in d3r.Thus, on integrating over the
volume, y(E) is seen to be the total path length traveled at energy E
by all the radiation in the volume of integration.
Because the path-length distribution applies to more than just the
radiation-equilibrium field, many authors prefer to write about it
rather than about equilibrium. But it makes no essential difference
which they choose, because the equations are identical.

7.3 Continuous Slowing-Down Approximation (CSDA)

One of the obvious complications in solving the transport equation,

Equation (2.28), is the coupling in the macroscopic cr_osssection, p,,
of energy transfer (E1+E) and angular scattering (ul+u). For the
transport of charged particles, this situation can be alleviated based
 7.3 CONTINUOUS SLOWING-DOWN APPROXIMATION (CSDA) 1 41
on the observation that their interactions with the medium usually
can be divided, within a good approximation, into two classes: a)
very frequent collisions that result in small energy transfers, but
no appreciable angular deflection, and b) infrequent collisions that
result in angular scattering but negligible energy transfer.
Formally, this amounts to rewriting p, in the form
(E1+E,
p*., ;I+;) = p,(Ef+E) ti(;' -); (7.17)
+ p,(G1+G) ti(Er-E).
It is customary to make the further assumption that the type-a
collisions can be described by a continuous energy-loss process in
terms of the stopping power dEldx. Thus,

This equation is known as the continuous slowing-down approxima-

tion (CSDA). With this equation and Equation (7.17), the first term
on the right hand side of Equation (2.28) becomes

Similarly, the term pp) in Equation (2.28) becomes

The quantities u and 6' can be interchanged in Equation (7.20)

because p, depends only on 2 . I'.
Finally, one can replace the variable E by R, the residual CSDA
range of a particle of energy E. Thus,
42 1 7. TRANSPORT THEORY-METHODS OF SOLUTION

noting that

By introducing Equations (7.19 through 7.22) in Equation (2.28), one

obtains the charged particle transport equation:

Under equilibrium conditions, the divergence term in this equa-

tion is zero and the equation can be integrated over direction and
solid angle to give

Integrating from R to infinity and then changing back to the energy

variable gives

If the source is monoenergetic, the approximation becomes

By substituting the CSDA of Equation (7.25) into Equation (7.9)and

changing the order of integration, one can show that CSDA satisfies
the energy conservation law. The form of the conservation law sug-
gests that the equilibrium field must depend on the inverse of the
stopping power; hence, it should be noted that the CSDA satisfies the
law-it is not derived from it. Any other approximation should sat-
isfy the conservation law too; a rigorous solution would, of course,
satisfy it.
As mentioned above, the CSDA was devised for charged particles
because their numerous small energy losses often can be adequately
approximated by a continuous slowing-down process. For photons
and neutrons, this is certainly not an adequate approximation. Nev-
ertheless, the CSDA can be applied and the results used as a rough
guide. For monoenergetic sources of photons of energy hv,

This result satisfies the conservation-of-energyequation, Equation

 7.4 NUMERICAL INTEGRATION OVER ENERGY 1 43
(7.10). With E substituted for hv, this equation also applies to neu-
trons. But, particularly below about 1MeV, neutrons tend to lose a
constant average fraction of their energy per collision, regardless of
their energy; as a consequence, a better solution is obtained using
the average logarithmic decrement. The CSDA for a monoenergetic
source is then (for no neutron capture)

which satisfies the energy-conservation law in the form of Equation

(7.12).

7.4 Numerical Integration Over Energy

The solution of the radiation equilibrium equation, Equation (7.61,
requires different methods for the different types of particles.
In the case of photons, direct numerical integration of Equation
(7.6) is fully adequate to evaluate this elementary integral equation,
using standard procedures. The scattering function, p8, is propor-
tional to the Klein-Nishina cross section which varies smoothly
(except a t cut-off)and rather slowly with both energy and energy loss.
The integrations can be performed in energy, energy loss (E-E'), or
photon wavelength.
The neutron cross sections are highly variable (see Appendix A),
with smooth functions f i and psfor hydrogen, but regions of irregu-
larity for most other nuclei. When the detailed structure of the
resonance regions is not of importance, the energy scale can be
divided into regions Ei+,<E<Ei. A useful form of Equation (7.6)
results from integration over each of a set of energy intervals
En+,< E < E n :

where &(E) = 1 En

E,+1
dEfpS(E'+E),with the integration over E' start-
ing only a t E for E > En+,.
The scattering integral is replaced by a
sum, so that Equation (7.29) takes the simple and soluble form
(E, is the maximum energy of neutrons emitted):
n-1
4-4 / 7. TRANSPORT THEORY-METHODS OF SOLUTION

If the function IME) were known numerically or analytically, Equa-

tion (7.30) would be correctly given by choosing functions W(E) satis-
fying W(E) = &El to insert into the following expressions for K, and
Kni:

and, for i < n,

Because such knowledge of 4 is ordinarily absent, one may choose

the next alternative, i.e., the evaluation of Equations (7.31) and
(7.32) with a weight function W(E) resembling &El so far as it is
known. Then, with values so obtained for K, and K,~,one can solve
the triangular set (Equation 7.30) of linear equations for values
dE I,NE)to give approximate solutions in terms of energy groups.
IE: 1

Electron slowingdown spectra have been determined by using

a n equation obtained by integration of Equation (7.6) to obtain a n
equation much like Equation (7.29),

The various terms in this equation give additions or removals of

particles from the energy region above E; Equation (7.33) is fully
equivalent to Equation (7.6). Various methods have been used to
solve this type of equation, whose scattering kernel function is less
singular and, thus, more manageable than that of Equation (7.6).A
straightforward method that works well is to use direct numerical
integration except for the last energy interval (for E' just above
E ) where the scattering function is strongest; then use a special
quadrature formula for that interval5.
The curly bracket term in Equation (7.33) is strongly peaked a t
E' = E , even though a major cancellation occurs between the two
terms in the brackets. Note that, if one approximates this "scatter-

'Numerical integration is essentially a replacement ofsegmentsofa weight function

( o h n unity) by one to several Dirac delta-function terms with strength constants
chosen to give accurate values for one to several elementary integrals over the weight
function for the segment. The main factor multiplying qHE) can be replaced by the
sum of two Dirac delta functions at the ends of the segment between Enand E n + ,in
a manner that gives accurate results if $ does not vary strongly in that segment.
 7.5 ELEMENTARY PROBLEMS 1 45
dE
out" function with - -6(E1-E), the result is a "continuous slow-
dx
ing-down" type of solution, which was discussed in the previous
section.

7.5 Elementary Problems Involving Particle Direction

Several important problems can be investigated in terms of energy
or distance. As a first example, consider photon directional distribu-
tions as functions of the energy, a problem studied very early in the
development of modern transport theory. This problem requires only
the evaluation of Equation (2.28) without the differential terms,

+ So(E,G . R). (7.34)

The scattering kernel is based on the Klein-Nishina cross section.
Application of the Legendre transformation gives
m

and

Each coefficient, rCI,,is determined by evaluation of a directly soluble

integral equation

For the photon case, ordinary numerical techniques s f i c e , except

that, for a peaked source function such as So= 6(E -E,) 6 6 . $ - 11,
the solution is singular in the direct beam, and discontinuous in the
first scattered beam. These discontinuous features are computed
separately (i.e., orders of scattering) (Spencer and Fano, 1951; Fano
et al., 1959).

7.5.1 Thin-Foil Charged Particle Problems

To treat the problem of electron angular distributions a h r pene-
tration of thin foils, the transport equation can again be taken as
46 / 7. TRANSPORT THEORY-METHODS OF SOLUTION

Equation (2.28). The scattering function is basically given by the

interaction of electrons with the screened Coulomb field, which does
not involve significant energy losses, and with the atomic electrons,
which involves significant direction changes only rarely in the
knock-on (large-loss) cases. One proceeds by assuming that the elec-
tronic interactions lead to a continuous energy loss (CSDA) and that
the interactions with the nuclear field should be corrected to account
for knock-on deflections. The resulting equation can be put in the
simple form

for a source function for monoenergetic electrons emitted in a single

direction, g and where z is the perpendicular distance from the plane.
This equation is directly soluble using the Legendre transformation.
Then, if it is noted that tc, = 2.rr du ps(u)Po(u),one can solve for

the coefficients $e(z) = 2n duJr(z,u)Pt,giving

with

where u = ; - u'. Various summation procedures have been sug-

gested. Of these, the most flexible and best known is that due to
Moliere (1948). This procedure utilizes a well-known "small-angle
approximation," in which the sum in Equation (7.37)is transformed
to an integral, using
(e + :I+ - a , and Pe+Jo(crf3),so that

and
 7.6 PENETRATION STUDIES 1 47
MoliBre's procedure involves writing the exponent in the scaled form
1
zAu21n(u2/b)= - x2 + -x21nx2.
B (7.39)

The two terms can be roughly identified as the Gaussian peak and
the largedeflection tail. One immediately finds that x2 = zABu2,
where B is the solution of
B - lnB = ln(bzA) (7.40)
The term exp [ + B-'x21nx21is expanded into powers of the exponent
which converge rapidly when B >> 1.There results a series of which
the first term is a Gaussian in s = e l m , and the next two terms,
which can be expressed a s functions fT1)(s),F2'(s) (tabulated in
Molihre, 1948), give additive corrections that evaluate the large-
deflection tail accurately.

7.6 Penetration Studies

The following is a brief introduction to several of the best-known

procedures. Because each has a large literature with many refine-
ments, only essential concepts are presented.

7.6.1 The Moment Method

Consider the expansion, Equation (7.3), with known functions,
fi(z):
+XE,z) = C$i(E)fi(z)
I

The n-th moment of + is, then,

If the moments of $ are known and the moments of fib) can be

calculated, Equation (7.42) is a system of equations for ei(E). Once
the $JE) are calculated, one can reconstruct +from Equation (7.41).
To include the angular distribution of the radiation, +,, is further
expanded in spherical harmonics (Fano et al., 1959; Krumbein and
O'Reilly, 1968).
Application of both Legendre and moments transformations yields
an inter-linked system of integral equations that is directly soluble
if one proceeds from high values of the Legendre index 42 and low
48 1 7. TRANSPORT THEORY-METHODS OF SOLUTION

values of the moment index n (i.e., n = O), in the direction of low 4

and high n. For each 4, there results a series of energy-dependent
moments. Whether one is concerned with particular energies or spec-
tral integrals of interest, the primary problem is then the construc-
tion of spatial distributions from the moments (Fano et al., 1959;
Krumbein and O'Reilly, 1968).
In the case of the construction of spatial distributions for
uncharged particles, the penetration process is dominated by expo-
nential trends. One may, in principle, then choose to represent the
distributions by an exponential, exp( - a r ) , multiplied by polynomi-
als of r, with the known moments representing conditions that fix
the coefficients multiplying the polynomials (Spencer and Fano,
1951; Fano et al., 1959; Spencer 1971). Such a procedure will be
accurate only when a value for the coefficient c* can be found such
that exp( -ar) provides a good first approximation of the spatial
distributions a t great depth (many mean free paths); otherwise, the
procedure may give a poor result (Spencer et al., 1980).
A competitive, and in many ways more flexible, method for con-
structing spatial distributions uses a sum of exponentials, exp( - a,r),
with multiplicative coefficients and with, a t least, some of the cc,
values determined such that the representation agrees with the
known moment values from the numerical solution of the transport
equations for the moments. This method has worked well, even when
polynomial approaches have failed (Fano et al., 1959).
In both of the approaches just referred to, numerical convergence
is tested with use of a n increasing number of moment values in the
representation, and accuracy estimates are based on the rate of
convergence. A third method combines exponentials with polynomial
forms, and yields a somewhat more satisfactory error estimate (in
this case a n estimate of an error bound) (Spencer, 1972; Spencer and
Simmons, 1973).
The calculation of angular distributions requires evaluation of the
Legendre sum from space distributions for a few of the lower Leg-
endre coefficients. The complexity of these distributions has led to
procedures incorporating the lower-order-scattering components in
some detail (Fano, et al., 1959; Spencer et al., 1980). Even so, such
methods for computing angular distributions have not been entirely
satisfactory because of sharp features requiring many harmonics for
a description.
In the case of charged particles, particularly electrons, deep pene-
tration trends are not well known, and further, they do not dominate
the spatial distributions to a similar extent. Moment-method calcula-
tions have nevertheless been successfully performed (Spencer 1955,
1959) in the continuous slowing-down approximation. These data
 7.6 PENETRATION STUDTES 1 49
have been useful, but the approach has not been vigorously pursued,
in part because of the greater flexibility of the stochastic approaches.

7.6.2 Discrete-Ordinates Transport Codes

The discrete-ordinates method is perhaps the most powerful
method available for obtaining a deterministic solution to coupled
neutron-photon transport equations. The method is not often used
for charged particles, and applications are restricted mainly to those
problems where charged particle transport may be neglected or
treated very approximately. Furthermore, a t the present time the
available discrete-ordinates codes are applicable only to one- and
two-dimensional problems; i.e., to transport problems in which the
particle fluxes are dependent on no more than two space dimensions.
Only a few of the discrete-ordinates codes are discussed below, but a
comprehensive list of the available codes may be obtained fkom the
Radiation Shielding Information Center (RSIC) of the Oak Ridge
National Laboratory.

7.6.2.1 Neutron-Photon Tmnsport

The approximation of neglecting charged-particle transport effec:

tively limits the application of discrete-ordinates codes to the study
of neutron-photon transport. However, because of fission and fusion
reactor studies and military applications, neutron-photon transport
calculations a t energies 5 2 0 MeV are of considerable practical
importance and much attention has been directed to using discrete-
ordinates codes to carry out such calculations.
The accuracy of transport calculations depends on the differential
cross-section data that are used. Because of the practical importance
of low-energy neutron-photon transport, the cross-section data avail-
able are quite good and have been systematized as the Evaluated
Nuclear Data Files (ENDFIB) that are available from the National
Nuclear Data Section Center at the Brookhaven National Labora-
tory (Garber, 1975). These cross-section data are continually
updated, contain all of the relevant nuclear interactions, and in
general represent the best available information a t any given time.
For use in discrete-ordinatescodes, the differential cross-section data
in the ENDF files may not be used directly, but must be put in
multigroup form; i.e., must be averaged over specified energy inter-
vals, and the angular distributions must be expanded in Legendre
polynomials. In general, the energy-group structure and the required
50 1 7. TRANSPORT THEORY-METHODS OF SOLUTION
number of Legendre polynomials is problem-dependent, i.e. for best
accuracy, they should be carefully designed for a particular problem,
but this is time-consuming and requires considerable experience.
There are available from the Radiation Shielding Information Cen-
ter a t the Oak Ridge National Laboratory several multigroup cross-
section sets (Roussin et al., 1980; Ford et al., 1976; Santoro et al.,
1981) that are suitable for use in a variety of problems. Multigroup
cross-section data for carrying out neutron-photon transport calcula-
tions for energies up to 400 MeV are also available (Alsrniller et al.,
1986).
The most widely used one-dimensional discrete-ordinates code is
ANISN (Engle, 1967) and there are two two-dimensional codes, DOT
(Rhoadesand Mynatt, 1973)and TWOTRAN (Lathrop and Brinkley,
1973) that are often used, which are, for most purposes, quite mmpa-
rable. There is available, also, a version of TWOTRAN, called TRI-
DENT (Seed et al., 1977), that allows a triangular rather than a
rectangular spatial mesh, and this feature can be useful in some
problems. All of these codes can be used to carry out adjoint as well
as forward calculations. Appendix B.3 shows an example of results
obtained using the discrete-ordinates code DOT in the adjoint mode.
The various codes were written for a wide group of users and are
well documented. Copies of the codes and documentation are avail-
able from the Radiation Shielding Information Center.

7.6.2.2 Dosimetry Calculations By the Method of Discrete

Ordinates.

In general, discrete-ordinates codes provide the fluence of neutrons

and photons as a function of energy and position, but give no informa-
tion about charged particles. To obtain the absorbed dose, the particle
fluences are multiplied by suitable conversion factors as a hnction
of energy and then integrated. The kerma factor is often used as the
appropriate conversion factor (Appendix A). Kerma factors in tissue
have been calculated by Caswell et al. (1980) at energies I30 MeV
and by Dimbylow (1980,1981,1982) at energies up to 50 MeV, and
have been measured by Brady and Romero (1979). For radiation
protection purposes, the particle fluences are often converted to
absorbed dose using the fluence-to-absorbed dose conversion factors
recommended by the National Council on Radiation Protection and
Measurements (NCRP, 1971), by the International Commission on
Radiological Protection (ICRP, 1973) and by the American Nuclear
Society (ANS, 1977).
 7.7 SPECTRAL EQUILIBRIUM AND RELATED CONCEFTS / 51
7.7 .Spectral Equilibrium and Related Concepts

7.7.1 Aspects Applicable to All Radiations

Equilibrium spectra show certain common feature-s, although they
are expressed somewhat differently for the different radiations.
Immediately below the source energy of a monoenergetic source, the
spectrum has a noticeable fine structure. In this region, the spectrum
is dominated by particles that have suffered only one or a few scatter-
ings. The first feature of the fine structure is, roughly, the spectrum
of once-scattered particles; it is modified as the energy decreases by
contributions from multiply scattered particles. In the spectrum of
the oncescattered electrons, Figure 7.1, peaks due to multiply-scat-
tered particles are also visible. The once-scattered spectrum of pho-
tons and neutrons, Figures 7.5 and 7.6, is very broad, and the princi-

ELECTRON KINETIC ENERGY, keV

Fig. 7.1. Ratio of the calculated equilibrium spectrum of 1.99-keV electrons in

helium to that of the CSDA as a function of electron kinetic energy. The use of the
ratio reduces the total variations with energy to allow the small oscillatory component
to stand out. (After Douthat, 1975.)
52 1 7. TRANSPORT THEORY-METHODS OF SOLUTION

pal evidence of its presence is the discontinuity a t the minimum

energy that a particle can have after a single collision. The multiply
scattered contributions are so smeared that the individual compo-
nents are not distinguishable.
Fano (1953) called the high-energy part of the equilibrium spec-
trum, including the fine structure, the "transient" phase; he called
the low-energy part the "steady state." For energies less than about
an order of magnitude below the source energy, the equilibrium
spectra for different source energies are similar and approach a
spectrum characteristic of the medium. After a sufficient number of
scatterings, the degraded spectra of particles in a given medium
emitted a t various energies can be indistinguishable if properly nor-
malized (see Figure 7.2).

7.7.2 Electrons
Curve 1 in Figure 7.3 is an equilibrium spectrum for electrons
calculated by Spencer and Fano (1954); curve 2 is based on the

1 10 lo2 lo3 1o4 1o5 lo6

ELECTRON COLLISION ENERGY, eV

Fig. 7.2. Electron energy degradation spectra in water. Scaled collision density,
nc(T,TJ (TIT,), as a function of the electron energy, T , for various initial electron
energies, To.The usual track length distribution, y(T,TJ, is equal to A(n nc(T,TJ,
where A (T) is the mean free path between inelastic collisions. The collision density
should also include a contribution from the first collisions made by previously unscat-
tered electrons; this contribution, represented by a delta function 6(T- T J , is not
shown. Also not shown is the fine structure of n,(T,TJ near T,. (After Berger, 1984.)
 7.7 SPECTRAL EQUILIBRIUM AND RELATED CONCEPTS 1 53

2.04 MeV
I

0.01 0.1 1.0

ELECTRON KINETIC ENERGY, MeV
Fig. 7.3. Curve 1: Charged-particle equilibrium spectrum for electrons and their
delta rays from 2.04-MeVsources in aluminum. Curve 2: Same spectrum with delta
rays omitted, approximately the same as the CSDA. (AfkrSpencer and Fano, 1954.)

Continuous SlowingDown Approximation. The increasing difference

between them with decreasing energy is due to secondary electrons.
The two curves coincide above one-half the source energy because of
the rule in electron physics that the higher-energy electron emerging
from a collision is identified with the incident' electron-i.e., there
are no secondaries with energy above one-half the source energy.
Figure 7.4 shows Spencer-Fano calculations for different media
(McGinnies, 1959).
Figure 7.2 shows electron spectra in water for a number of initial
energies and extending down to 10 eV (Berger, 1984).These curves
are normalized for equal total energies emitted from the sources and
show the approach to indistinguishable steady-state spectra. The
figure also illustrates the large fluences of low-energy particles that
are generated in a medium.
Figure 7.3 does not show the fine structure near the source energy
because the method of calculation did not employ a fine enough
energy mesh. Figure 7.1 does show the fine structure, for 1.99-keV
54 7. TRANSPORT THEORY-METHODS OF SOLUTION

1o4 10" 1cis 10"

ELECTRON KINETIC ENERGY, MeV
Fig. 7.4. Charged-particle equilibrium spectra for electrons and their delta rays
from 1.038-MeV sources in different materials. (After McGinnies, 1959.)

electrons in helium (Douthat, 1975). The energy-loss spectrum for

helium is partly resolved and exhibits peaks separated by about 25
eV for once-scattered, twice-scattered, etc. electrons. The oscillations
in the equilibrium spectrum are damped rapidly because the multi-
ply scattered spectra broaden a s the number of scatterings increases
and because the equilibrium spectrum is a sum over all orders of
scattering.

7.7.3 Photons and Neutrons

Equilibrium spectra for photons and neutrons have less immediate
applicability than those for electrons. However, spectra for photons
and neutrons are useful as illustrations of the kinds of spectra that
photon and neutron sources would generate and useful as approxima-
tions of spectra arising from natural background radiation.
Figure 7.5 shows photon spectra calculated by Karr and Lamkin
(1949). The discontinuity due to once-scattered photons a t the mini-
mum energy loss in Compton scattering is visible.
Figure 7.6 shows the ratio of an equilibrium neutron spectrum in
graphite calculated by Placzek (1946) to the spectrum resulting from
 FLUENCE PER ENERGY INTERVAL, crn" (M~v)" A
A
A o 8
56 / 7. TRANSPORT THEORY-METHODS OF SOLUTION

RELATIVE NEUTRON ENERGY

Fig. 7.6. Rstio of the equilibrium neutron spectrum in graphite to that of the
CSDA as a function of neutron energy expressed as a fraction of the initial neutron
energy, To.(After Placzek, 1946.)

the CSDA. It demonstrates that the CSDA is a good approximation

below the transient stage. The discontinuity resulting from the maxi-
mum energy loss in neutron elastic scattering is quite large. For
graphite, both p and [ in the CSDA, Equation (7.28),are roughly
constant at low energies; thus the steady.state is a 11T spectrum.

7.8 Radiation Quasi-equilibrium

Several topics involving quasi-equilibrium will be treated in this

section because they are closely connected with the equilibrium con-
cept and because the relative simplicity of equilibrium fields makes
their treatment easier.
 7.8 RADIATION QUASI-EQUILIBRIUM 1 57
7.8.1 Transient Equilibrium
For up to tens of MeV for photons and hundreds of MeV for neu-
trons, their charged-particle secondaries have ranges that are much
smaller than the distance over which they are attenuated apprecia-
bly. Thus, in the interior of an irradiated medium, the "source" of
the charged particles is nearly uniform over distances comparable
to their range and the conditions for an equilibrium charged-particle
field are nearly met. Because of the relatively slow attenuation of
the primaries over small distances in the interior, both 9 and So
in transport Equation (2.28) vary approximately in proportion to
(1 - k), where x is distance measured along the direction in which
u . 99 is a maximum (e.g.,along a primary beam) and k is an attenua-
tion coefficient characteristic of the primaries and is much smaller
than p. Substitution in the transport equation leads to the equilib-
rium equation, Equation (7.6), except that p is replaced by (p - k).
Because of the relative sizes of p and k, the effect of this replacement
is small. It tends to represent more higher-energy particles than
would be present in a true equilibrium field, but the increase is
generally small.
In transient equilibrium, the charged particle fluence is smaller
than that in the case of true equilibrium. Roesch (1958) showed that,
for photon beams, the dose a t a point in the medium is (1 - kX)
times that in the corresponding equilibrium field. Here, k is the same
as above and the distance X is a particular function of the dose due
to the charged particles emitted in a small volume element. This
distance is always less than the maximum range of the charged
particles and is typically one-third of the maximum electron range
for photons in the MeV range; for these photons, the kX correction
is a few percent.
The term "transient" is applied to these near-equilibrium fields
because of the similarity-in space-to transient equilibrium in
time of radioactive decay chains.

7.8.2 Non-uniform Sources

Transient equilibrium is an example of a nonuniform source distri-
bution that results in small departure from equilibrium. How non-
uniform can the sources be before the departure is appreciable? This
question has no general answer because it depends too much on the
specifics of the source distribution and the radiation considered and
on what is meant by "appreciable." But, in general, some degree
of departure from uniformity can often be tolerated because the
58 1 7. TRANSPORT THEORY-METHODS OF SOLUTION
penetration of the radiation tends to damp it out. The field a t a
given point is the result of radiations emitted out to sometimes
considerable distances from that point and, therefore, represents an
average effect of the nearby sources. The averaging tends to reduce
the effect of source nonuniformity.
A simple example will show mathematically how this damping
effect works. The equilibrium field, $-(E), due to isotropic sources
can be calculated from the field, $pi(E,x),a t a distancex from a plane-
isotropic source by

Fields due to one-dimensional source distributions can also be calcu-

lated in like manner. For example, consider the distribution
S = So[l+ sin(wx)]illustrated in Figure (7.7). It oscillates between 0
and 2S0;but it has the same average, So,as the uniform distribution
that gave the equilibrium field of Equation (7.43). The field at a point
xo is given by

and equals the equilibrium field plus an oscillatory term whose

coefficient is the Fourier transform of the plane-isotropic field. If the
source distribution oscillates rapidly compared to the predominant
frequencies in this transform, then the oscillatory term in the field
is negligible; if it oscillates slowly, then the oscillatory term is
approximately i),(E)sin(wxo). To illustrate, suppose $pi(E;x)is an
exponential function, exp( -x/xl); then

If the penetration of the radiation is large compared to the distance

between oscillations of the source (wx' >>I),then the oscillatory term
is negligible and the field is essentially the equilibrium field; if it is
small (w xl<<l), then the multiplier in Equation (7.45) is nearly 1
+sin (wxJ and the field closely follows the source distribution as
shown in Figure (7.7).

7.8.3 Nonunifonnity in the Internal Dosimetry of Radionuclides

Radionuclides acquired physiologically are seldom, if ever, laid
down uniformly in tissues. Nevertheless, because of difficulties in
 7.8 RADIATION QUASI-EQUILIBRIUM 1 59

DISTANCE, x, (arbitrary)
Fig. 7.7. The field due to a one-dimensional oscillatory source-strength distribu-
tion as a function of distance (Equation 7.45). (---)the resulting fluence density when
the factor of the oscillatory term of Equation (7.45) is 0.9; (. . . .) when it is 0.1.

determining the actual distribution in an organism, dosimetrists

commonly assume that the distribution is uniform. This is equivalent
to the assumption that radiation equilibrium exists. The question
naturally arises: how good is this assumption?
The example just treated can serve as a prototype for dealing with
this question. The absorbed dose varies with position in the same
way as the fluence:
D(x,) = D, [ l + sin (wx,)]/[l + (wx')~I. (7.46)
Here D, is the absorbed dose if the activity is uniformly distributed.
Suppose we ask what fraction of the tissue receives a dose within
some fraction ofD,, say (1- a)D, to (1+ a)D,. From Equation (7.461,
all of the tissue does if w x 1 > v m . For smaller wx: a fraction
(211~)arcsin {a[l+(wxl)q) does. This fraction approaches a limit,
(211~)arcsin(a), for radiations for low penetration.
Even for a = 0.25 (i.e., doses within 25% of D,), all the tissue is
within the specified dose range if wx' is greater than about 1.7. If
the activity peaks are a few millimeters apart, x' must be greater
60 / 7. TRANSPORT THEORY-METHODS OF SOLUTION
than about one millimeter-a limit satisfied by beta particles in
the MeV range, but not satisfied by any alpha particles. For alpha
particles with x' about 10 micrometers (about one-third of their
range) the activity peaks would have to be just a few micrometers
apart for all doses to be within 25% ofD,. For peaks a few millimeters
apart, alpha particle doses would be at the limit (21.rr) arcsin (0.25),
or 16% of the tissue would be in the specified dose range.

7.8.4 Non-uniform Media

The same damping by particle penetration that reduces the effect
of source nonuniformity reduces the effect of medium nonuniformity
also. But mathematical analysis with, for example, plane-isotropic
source functions would be too laborious, because the functions would
now also depend on the spatial structure of the medium.
In the cavity theory of radiation dosimetry, the nonuniformity
represented by the cavity is commonly assumed not to perturb appre-
ciably the equilibrium field there in its absence (c-f.Burlin, 1968).
Some ad hoc attempts have been made to correct for nonuniformity in
the case of solid-state detectors, which constitute similarly localized
mass anomalies, sometimes considered as thick "cavities" (c.f. Bur-
lin, 1968).Zheng-Ming (1980)has applied rigorous transport theory
to cavities.
8. Monte-Carlo Methods

8.1 Principles

The Monte-Carlo method provides a computer-based technique to

simulate the interactions of particles in a medium. This section
serves as an introduction to this technique.
Before Monte-Carlo transport calculations are described, it is use-
ful to consider briefly the main mathematical tool utilized, namely
the simulation (or generation) of random variables distributed
according to a given probability function utilizing various computer
algorithms. A good introduction to Monte-Carlo methods can be
found in Messel and Crawford (1970).
Consider first a continuous random variable, X, the realizations
(numerical values) of which are distributed according to a density
function f(x). In other words f(x)dx represents the (normalized) proba-
bility that X takes a value between x and x + dx:

and

Perhaps the easiest way to generate values, x, distributed accord-

ing to Equation (8.1) is based on the fact that for any density flx),
the random variable y:

is uniformly distributed between 0 and 1:

One recognizes in y the probability distribution F(x), that is

The statement, Equation (8.4), follows then by observing that F(x)

is a monotonically increasing function of x (from 0 to 1)and therefore:
P[F(x) s F(xl)] = P(x Ix') = F(xr) (8.6)
62 1 8. MONTE-CARLOMETHODS
which is equivalent to Equation (8.4) integrated from 0to some value
y' = F(x1).
Any modern computer has available software for generating pseu-
dorandom numbers, 5, uniformly distributed between 0 and
The procedure to generate values of x is then accomplished by the
following algorithm:
a) select a random number, 5, between 0 and 1
b) from F(x) = 6, calculate x = F-'((). (8.7)
In order to apply the procedurejust described,one needs an analytical
representation of flx) that can be integrated [Equation (8.3)l and
inverted [Equation (8.7)].These conditions are not always met and
alternative procedures will be presented later in this section.
There is a corresponding basic selection process employed for a
discrete random variable X, the realizations of which, xi, occur with
probabilities proportional to the numbers mi. The normalized prbba-
bilities are then ail 1 N

j=l
o,. Imagine on a straight line, adjacent
segments of length u,lCuj, odCuj, etc., forming a total segment of
length 1. A random number, 5, uniformly distributed on (0,l)will
then be contained in one of these segments and thus select a value
xi. This may be done by using an algorithm in which

and selecting the value xi corresponding to the first yi which is

negative.

These numbers are generated by deterministic algorithms but are nevertheless

sutliciently uncorrelated. The most widely used algorithm is the congruential method,
in which a sequence of integers xi is generated according to the rule

The pseudorandom numbers themeelves are obtained by forming the aequence d m .

A particularly thorough analysis of the congruential method has been made by Knuth
(1981) who recommenda appropriate values of the parametera a, c and m. The pseudo-
random numbers xJm, on the basis of theoretical analyses and empirical tests, are
expected to be satisfactorily uncorrelated and suitable for moat Monte-Carlo calcula-
tion. Residual correlations can be decreased by using several sequences of pseudoran-
dom numbers.
 8.1 PRINCIPLES 1 63
As an example, consider an angular distribution density fl6)dO. An
analytical expression might be available for f(6) and the procedure,
Equation (8.71, can be applied. Quite frequently, though, the theoret-
ical prediction, f(O), might provide only a poor representation of the
data and the researcher will decide to use measured values, fl6,), to
describe Re). Then the realizations of e are de facto discrete and equal
to the set {Oil and the procedure, Equation (8.8), can be applied; it
amounts to representing a continuous curve flO) by (f(6,)).
Other selection procedures have been developed. For instance,
consider the density fix) where xE[a,b] and assume that the method,
Equation (8.7), is not feasible. One might generate values of x as
follows: First, change the variable from x to y, where y&[0,11,and
then

y is distributed according to:

Next, let

A plot of g*@) will be contained in a square of side = 1. Select, now,

two random numbers 5, and 5,. If

the value of y [and therefore x, Equation (8.9)] is accepted; if not

repeat the procedure.
In the procedure following Equation (8.12), one encounters a typi-
cal problem in computer simulation, namely the efficiency for the
selection method. In this case, the acceptance rate is proportional to
the area under the curve, g*(y). It will be very inefficient, in terms
of computing time, to use this procedure if, for instance, g*(y) i s
sharply peaked. The efficiency in selecting values of X is usually
optimized with the following very powerful method applicable in
almost any situation:
Assume flx) can be written:

where
64 1 8. MONTE-CARLOMETHODS
A value x can be chosen as follows:
a) Select i with probability proportional to cri [Equation (8.811;
b) Select x from ((x) [Equation (8.7)l;
C) Accept x with probability gi(x) [Equation (8.811;
d) If x is rejected go back to a).
It can be shown (Messel and Crawford, 1970) that the acceptance
probability for this method is I/&;. Very ingenious decompositions
for particular functions f(x) are available, such that the maximum
sampling efficiency is attained.
Consider, as an illustration of this procedure, the Klein-Nishina
cross section for the Compton scattering of a photon with energy k
to energy k' (the energies are in units of the electron rest mass).
From Equations (A.3) and (A.6) of Appendix A one has

where
c = k'lk
and

A useful decomposition of du is (Messel and Crawford, 1970):

where

and

The Monte-Carlo transport of particles is now illustrated with the

simple example of low energy electrons (e.g.,a few keV). Typically,
an electron of energy E is started at the origin of the coordinate
system in the z direction. Three types of interactions are usually
considered: elastic scattering, impact ionization and impact excita-
tion. Let the (angle-integrated) interaction coefficients for these pro-
cesses be p,,(E), pion(E),h X ( E )respectively,
, and let:
 8.1 PRINCIPLES 1 65
P(E) = + pion@') + p,(E) (8.15)
be the total macroscopic cross section.
The first decision to be taken is how far the electron travels before
an interaction occurs. The probability that a free flight will be termi-
nated when its length is between s and s+ds is [Equation (5.3)l
f(s)ds = @xp( - ps) ds. (8.16)
From Equation (8.71, given a random number, 5, one has:
F(s) = 1 - exp( -ps) = 5,
so that (8.17)

(5 and 1 - 5 have exactly the same distribution). The electron is

moved to position s and a decision is made [Equation (8.811 as to
which of the three possible interactions takes place. If elastic scatter-
ing occurs, no energy is locally deposited and a scattering angle has
to be sampled [for instance from the Rutherford expression, Equation
(A.151, with perhaps Equation (8.7)]. The electron is transported in
the new direction to the next interaction selected as described above
[Equation (8.17)]. The situation is similar for impact excitation with
the exception that now energy is deposited locally and the energy of
the electron decreased by the corresponding amount. Should the
outcome be an ionization process, one ends up with two electrons
(primary and secondary)and an energy and angle have to be sampled
for each one. Both electrons have to be transported now in the code.
The simulation continues until some end-of-calculation criterion is
satisfied, for instance E < Em,, or the electrons are outside some
receptor volume. The calculation must be repeated for a sufficiently
large number of-primary particles to obtain the desired statistical
accuracy for the parameters of interest,
A close examination of the sampling procedures just described
reveals a very important feature, namely that, with the exception of
the macroscopic cross sections of Equation (8.15), one needs in the
calculation only relative values of the distributions of with respect
to any other variable (i.e., the shape of the distributions).
In such an "event by event" simulation, all the 'information con-
cerning the particle track is available. Unfortunately, in most practi-
cal cases such a complete mode of simulation is prohibitive in terms
of computer resources, and various simplifying schemes must be
applied. The calculations are then simplified by considering collec-
tively certain kinds of interactions, for instance, treating the energy
66 1 8. MONTE-CARLOMETHODS

loss with CSDA or the angular deflections with multiple-scattering

formulae. Typically the transport medium is divided into segments
and the quantities calculated are: a) the energy deposited in the
segment, and b) the angular deflection following penetration through
the segment. The choice of the segment thicknesses is usually a
compromise between two requirements: 1)the total number of seg-
ments should be sufficiently small to make the calculation time
reasonable, and 2) the segment thickness should be appropriate for
the approximations used (CSDA, multiple scattering) to be appli-
cable.

8.2 Analog Monte-Carlo and Variance-Reduction

Techniques

As described above, the transport of particles through complex

geometries involves various kinds of interactions that, in general,
are governed by known probability laws. In practice, it is possible to
incorporate these laws into Monte-Carlo computation schemes so
that the processes that occur during the transport are simulated in
the calculations and, thus, the computations are an analog of physi-
cal reality. When this is the case, the calculations are said to be
analog Monte Carlo calculations. Such calculations are simple and
easy to understand, but they are, in some instances, not very efficient.
If, in a particular transport problem, the events of interest occur
only infrequently in nature, they will occur only infrequently in the
analog simulation, and large amounts of computing time will be
consumed in carrying out calculations that are of little interest for
the problem under study.
A variety of techniques, usually called variance-reduction tech-
niques, are available for treating transport problems by other than
analog methods (Spanier and Gelbard, 1969). The variance of a result
obtained by Monte-Carlo calculation is a n estimate of the statistical
confidence one can have in the result. In simple terms, variance-
reduction techniques amount to "biasing" the Monte-Carlo selection
procedures in such a way that rare events of interest occur more
frequently than in nature and compensating for this bias by weight
factors. A detailed discussion of the various types of variance-reduc-
tion techniques can be found in standard texts such as those of
Goertzel and Kalos (1958) and Spanier and Gelbard (1969). Unfortu-
nately, i t is difficult to establish general rules concerning which of
the various techniques are most appropriate for specific problems.
However, the power of Monte-Carlo methods is greatly enhanced by
 8.3 TRANSPORTCODES 1 67
the use of such techniques and, therefore, extensive Monte-Carlo
calculations should not be undertaken without some familiarity with
the available methods of variance reduction.

8.3 Transport Codes

A variety of computer codes have been written to carry out the

transport of various types of particles with Monte-Carlo methods,
only a few of which are discussed below. A comprehensive list of
available codes may be obtained from the Radiation Shielding Infor-
mation Center (RSIC) of the Oak Ridge National Laboratory. The
codes discussed here are, for the most part, those that authors of this
report have found useful in dosimetry studies. They are reasonably
well documented, and are available to users either through the code
authors or through the RSIC.

8.3.1 Neutron-Photon Transport a t Energies 5 2 0 MeV

Because of fission and fusion reactor studies and some military
applications, the transport of neutrons and photons with energies
5 2 0 MeV is of considerable practical importance. Consequently, this
energy region deserves special consideration, and much effort has
been expended on developing methods and cross-section data for
carrying out such transport calculations.
At neutron energies of 5 2 0 MeV, any charged particles that are
produced by nuclear interactions have very short ranges and usually
need not be transported. They are, however, important for some
dosimetry purposes and, thus, the type and energy of such particles
may be needed. The photons produced by the nuclear interaction of
these low-energy neutrons are relatively low in energy (510 MeV)
and may be transported by the methods described below under elec-
tron-photon cascades. At these low photon energies, charged-particle
secondaries, i.e., electrons and positrons, usually need not be trans-
ported. The low-energy photons also produce some neutrons (photo-
neutron production), but the number is small and for most practical
purposes can be neglected.
Of the many computer codes available for carrying out low-energy
neutron-photon transport calculations by Monte-Carlo methods, only
two, namely, MCNP (Thompson, 1979) and MORSE (Emmett, 1975)
will be discussed.
MCNP and MORSE are general purpose Monte-Carlo codes that
were written for a wide variety of users, and both are frequently
68 8. MONTE-CARLOMETHODS

used. Both contain routines that allow calculations to be carried out

in complex geometries, and both contain a wide array of variance-
reduction techniques that can be selected by the user.
MCNP and MORSE differ in one important respect; namely,
MCNP is a continuous-energy code and MORSE is a multigroup
code. A continuous-energy code uses interpolation in the available
cross section data to obtain cross section information at all energies
that are needed during the transport. For a multigroup code, on
the other hand, the cross-section data are preprocessed into energy
groups, and these preprocessed data are used in the transport calcula-
tion. Thus, the particles are not given a precise energy but belong to
an energy group, and, when interactions occur, the results used are
those obtained by averaging over the group energies. Multigroup
Monte-Carlo calculations are inherently less accurate than continu-
ous-energy Monte-Carlo calculations, but nevertheless, when the
group structure is chosen appropriately (for a given problem) the
two methods give results of comparable accuracy in most cases.
Multigroup Monte-Carlo methods tend to be slightly faster than
continuous-energy Monte-Carlo methods, but the actual savings in
computing time is very problem-dependent and is usually not large.
The chief advantage of the multigroup method over the continuous-
energy method is that adjoint calculations (see Section 6.7) can be
carried out readily.

8.3.2 Electron-Photon Cascades

Because there are three species of particles in electron-photon
cascades and many physical processes must be considered, they are
usually treated theoretically by the use of Monte-Carlo techniques.
A general discussion of the use of Monte-Carlo methods to treat such
cascades was given by Berger (1963), and the reader is referred to
his paper and references given therein for details. (See also, Berger
and Seltzer, 1968, 1982).
In practice, electrons, positrons, and photons with energies less
than some specified value, E,, are not of interest, and particles below
the cutoff energy are no longer considered to be part of the cascade.
The actual value of E, will, of course, depend on the particular
application being considered. If the particles produced by photon
processes are of sufficiently high energy, i.e., above the cutoff energy,
E,, they become part of the cascade.
Positrons lose energy and change directions by much the same
process as electrons, but positrons will also undergo annihilation and
produce two photons.
 8.3 TRANSPORTCODES 1 69
Several computer codes are available for carrying out calculations
of electron-photon cascades by the Monte-Carlo method. The
required cross-section data base and schematizations are incorpo-
rated in the ETRAN code (Berger and Seltzer, 1968), and in the
independently developed EGS code (Ford and Nelson, 1978). The
ETRAN code combines a treatment of electron transport by the
path-segment model (Berger, 1963) with the simulation of photon
transport by conventional Monte-Carlo techniques, taking into
account all orders of the electron-photon cascade. In the path-seg-
ment model, the exceedingly numerous Coulomb collisions between
electrons and atoms are not sampled individually. Instead, each
electron track is divided into a moderate number of segments (less
-
than loo), and the energy loss in each segment is sampled from a
theoretical energy-loss straggling distribution that takes into
account the combined result of many inelastic collisions. Each seg-
ment is further divided into two or more subsegments, and the net
angular deflection in each subsegment is sampled from a theoretical
multiple-scattering distribution that takes into account the com-
bined effect of many elastic collisions. In EGS, the angular deflections
are treated by a path-segment model similar to that used in ETRAN;
small energy losses from inelastic collisions are treated in the contin-
uous-slowing-down approximation (with use of a restricted stopping
power), and large energy losses are sampled individually.
The schematizations employed in the ETRAN and EGS codes,
particularly the path segment model, give rise to errors whose mag-
nitude is difficult to estimate from first principles. However, numer-
ous comparisons have been made that indicate that the ETRAN code
gives reliable results (see, e.g. Berger and Seltzer, 1968,1970, 1972;
Berger et al, 1970; Seltzer and Berger, 1974; Seltzer and Berger,
1978). Experimental verification of the validity of the EGS code has
been provided by Ford and Nelson (1978) who also show that the
results from ETRAN and EGS are in satisfactory agreement.
The ETRAN code treats homogeneous media only, and is limited
to slab geometry in its one-dimensional version, and to cylindrical
or spherical symmetry in its three-dimensional version. The EGS
code permits the treatment of multi-material media; however, the
main code is limited to the generation of electron and photon tracks,
and the user is expected to develop specific geometrical routines
for extracting information from the sampled particle histories. The
ETRAN code, with its schematization and cross-section data base
left unchanged, has been generalized by various authors to allow the
treatment of transport in multi-material media. The generalizations
include the TIGER code of Halbleib and Vandevender (1975) for slab
geometry, the CYLTRAN code of Halbleib and Vandevender (1976)
70 1 8. MONTE-CARLOMETHODS
for cylindrical geometry, and the SPHERE code of Halbleib (1978)
for spherical geometry. The treatment of even more complex multi-
material configurations has been incorporated into the SANDYL
code of Colbert (1973) and the ACCEPT code of Halbleib (1980). A
modern reference for electron-photon cascade calculations, including
discussions of all the codes mentioned above, can be found in the
book edited by Jenkins et al. (1988).

8.3.3 Nucleon-Meson Transport at Energies >20 MeV

At nucleon energies >20 MeV, the experimental information on
differential particle-production cross sections from nucleon-nucleus
collisions is limited and the nuclear-physics data needed to carry out
transport calculations must, for the most part, be obtained from
nuclear-physics models. The model that has been used most exten-
sively t o obtain the needed data is the intranuclear-cascade-evapora-
tion model (Bertini, l963,1967,1969a, 1969b,1972a, and the refer-
ences given in these articles), but more recently the pre-equilibrium
model (Kalbach, 1978a, 197813)has also been used. The intranuclear-
cascade-evaporation model as implemented by Bertini is used in
the High-Energy Transport Code, HETC (Chandler and Armstrong,
1972), that is described below.
HETC is a transport code that utilizes Monte-Carlo techniques to
simulate the nucleon-meson cascades produced by incident neutrons,
protons, charged pions and muons. HETC was designed for all energ-
ies less than or of the order of many hundreds of GeV and for essen-
tially arbitrary material compositions and source-geometryconfigu-
rations. The code takes into account charged-particle energy loss due
to atomic ionization and excitation, multiple Coulomb scattering of
incident charged particles, elastic acd nonelastic neutron-nucleus
collisions, nonelastic collisions of protons and charged pions with
nuclei other than hydrogen, elastic and nonelastic collisions of nucle-
ons and charged pions with hydrogen, pion and muon decay in flight
and a t rest, and particle production from negatively charged pion
capture a t rest. A detailed discussion of the manner in which these
interactions are taken into account is beyond the scope of this report.
Some discussion of the manner in which nucleon-nucleus nonelastic
collisions are treated is given below, but more detailed information
on all of the interactions will be found in the work of Alsmiller et al.
(1970) and Amstrong et al. (1972).
In HETC, nucleon-nucleus nonelastic collisions in the energy
range of approximately 15MeV to 3.5 GeV and pion-nucleus inelastic
collisions a t energies C2.5 GeV,including negatively charged pion
 8.3 TRANSPORT CODES 1 71
capture a t rest, are treated by means of the intranuclear-cascade-
evaporation model of nuclear reactions as implemented by Bertini et
al. (1972a). When a nucleon-nucleus or pion-nucleus collision occurs,
this model gives the correlated energy-angle distribution of all neu-
trons, protons, charged and neutral pions as well as the atomic
number, mass and kinetic energy of the residual nucleus. The model
also gives the total energy emitted from the collision in the form of
photons, but not the photon spectrum. A large number ofcomparisons
between calculated results obtained with the model and experimen-
taldata will be found in the work ofBertini (1963,1967,1969,1972a).
The model is primarily intended to describe inelastic collisions a t
incident energies >50 MeV, but, because of the lack of other sources
of data, it is used in HETC a t lower energies down to -15 to 20 MeV.
Comparisons between calculated results and experimental data a t
these lower energies have been given by Alsmiller and Hermann
(1970) and Bertini (1972b). A unique feature of HETC is that it
contains the intranuclear-cascade-evaporation code as a subroutine,
and, thus, uses the code itself rather than data in the code to deter-
mine the collision products when a nonelastic collision occurs. This
use of the intranuclear-cascade-evaporation code as a subroutine
substantially reduces the data handling problems in HETC, but it
also makes the core requirements for HETC very large.
The intranuclear-cascade-evaporation model described above
becomes inadequate a t nucleon energies -2.5 GeV because the model
accounts only for double pion production from nucleon-nucleon colli-
sions and only for single pion production from pion-nucleus collisions.
At nucleon energies >3.5 GeV and pion energies >2.5 GeV, particle-
production data for nonelastic collisions are obtained in HETC from
an approximate extrapolation model due to Gabriel et al. (1970).
Basically, the model uses particle-production data obtained from the
intranuclear-cascade-evaporation model a t energies - 3 GeV and
uses energy- and angle-scaling relations to estimate particle-produc-
tion data for higher-energy nucleon-nucleus-or pion-nucleus colli-
sions. The extrapolation model is approximate, but it provides a
detailed description of each high-energy collision, e.g., it gives the
mass, charge, and kinetic energy of the residual nucleus from the
collision, and for this reason it is well suited for use in transport
calculations.
In HETC, neutrons with energy 220 MeV, protons, I T +IT-, , p+
and p- are followed until they eventually disappear by escaping
from the geometric boundaries of the system, undergoing nuclear
absorption, coming to rest, or in the case of pions and muons, decay-
ing. Neutrons with energies <20 MeV are not transported with
HETC but are treated with codes such as MCNP (Thompson, 1979)
72 1 8. MONTE-CARLOMETHODS
and MORSE (Emmett, 1975) because these codes are designed for
the transport of neutrons of such energies and are capable of using
the best available cross section data. Electrons and positrons from
muon decay and photons from IT" decay and inelastic nuclear colli-
sions are not transported with HETC, but information concerning
these particles is provided by the code so they may be transported
with other codes such as EGS (Ford and Nelson, 1978) if this is an
essential part of the problem. HETC also provides the mass number,
charge and location of all of the residual nuclei produced by inelastic
collisions. This information allows a determination of the induced
activity in the transport medium as a result of the cascade. Finally,
HETC provides the energy distribution of all of the residual nuclei
produced by nuclear collisions and the energy distribution of the
low-energy heavy particle, i.e., deuterons, tritons, 3He's, and alpha
particles produced by nuclear reactions. These energy distributions
are of considerable importance in various kinds of dosimetry calcula-
tions. HETC does not contain any variance-reduction techniques; if
such techniques are employed they must be introduced by the user.
Calculated results obtained with HETC have been compared with
a large variety of experimental data and, in general, very good
agreement has been obtained (Alsmiller, 1976;Armstrong and Chan-
dler, 1972 and the references given therein). Examples of calcula-
tions carried out with HETC will also be found in Appendix A.

8.3.4 Dosimetric Calculations

In general, Monte-Carlo transport codes provide information on
particle "histories", and the user of the code may utilize this informa-
tion to provide any specific result of interest. A particle history
contains sufficient information to describe in detail all the changes
the particle undergoes as it travels through the medium. A particle
history will include such information as the energy, direction of
travel and statistical weight of the particle as it crosses the internal
boundaries in the geometric system. A particle history will also
include information on the position a t which reaction products are
formed and the changes in these reaction products as they travel
through the geometric system. In principle, the history information
may be combined to answer a variety of questions that are of interest
in dosimetry. For example, the energy deposition in some specific
volume may be obtained by taking the difference between the energy
of a particle as it leaves the volume and as it enters the volume and
summing over all such particles. Much more complex quantities,
e.g., dose equivalent, linear-energy-transfer (LET) spectra, lineal-
 8.3 TRANSPORTCODES 1 73
energy (y) distributions, etc. can be obtained equally well by the
appropriate combinationsof the history information.As in all Monte-
Carlo calculations, only a statistical estimate of these quantities can
be obtained, and, if very complex quantities are considered, the
number of histories that are needed to obtain statistically significant
results-as indicated, for example, by the variance of the r e s u l t
may be prohibitive. This difficulty may, to some extent at least, be
ameliorated by using the variance-reduction techniques mentioned
earlier, but, even with such techniques, the computing time required
to obtain satisfactory statistical results for complex dosimetry ques-
tions may be quite large.
9. Geometric Considerations

The multiplicity of factors entering dosimetric calculations will

rarely permit rigorous analytical solutions. As described in Sections
6 to 8, complex numerical integrations or Monte-Carlo computations
a r e frequently required. Geometrical considerations are, then,
mainly required for the utilization of whatever spatial symmetries
may be inherent in a problem.
However, a common class of problems allows a rigorous treatment
of the geometry of source and receptor regions and is, therefore, to
be discussed first. Such problems are the determination of mean
absorbed doses or absorbed-dose distributions in receptors (target
regions) that are exposed to ionizing radiation from one or more
source regions that contain uniformly distributed radioactive
nuclides. The special condition for this class of problems requires
that source and target regions are part of a uniform extended medium
with boundaries sufficiently far away as to be disregardable. Often,
this requirement is adequately met in dose calculations for internal
emitters for which the approach considered in this section is espe-
cially designed. For unbounded uniform media, the computations
are greatly simplified because multidimensional integrals over the
volumes of interest can be reduced to one-dimensional integrals that
contain only two functions. One of the functions, the isotropic point-
source kernel, characterizes the radiation; the other function, the
point-pair distance distribution, characterizes the geometry of the
source and receptor regions and their relative positions.
In the following subsection, the essential definitions and general
formulae will be given, then the point-source kernels will be
described, and, finally, the point-pair distance distributions and
related functions will be considered and will be illustrated by numer-
ical examples. [For specific applications in the medical use of radio-
nuclides, see MIRD (1988).]
9.1 Absorbed Dose in Receptor Regions

Assume a source region, A, that contains a uniformly distributed

radionuclide with specific activity (activity per unit mass) a and
consider h r t h e r a receptor region, B, that is exposed to the radiation
from A. The regions A and B may consist of several parts. They may
 9.1 ABSORBED DOSE IN RECEPTOR REGIONS / 75
also partly or fully overlap. As stated above, regions A and B are
treatecas part of an infinite uniform medium of density p.
Let E(x) be the average energy imparted per transition to the
medium a t a distance up to x from a point source; this is call* the
dE(x) is
integral isotropic point-source kernel. The derivative e(x) = -
dx
termed the differential point-source kernel. Isotropic point-source
kernels have been computed for a variety of radiations and will be
discussed in some detail in Subsection 9.3.
The absorbed-dose rate a t a distance x from a volume element dV
with specific activity (activity per unit mass) a is, then,

The absorbed-dose rate a t a point due to the source region A is,

accordingly,

where x is the distance from the point to the volume element dVA.
The average dose rate in the receptor region B due to radiation
from the source region A is, therefore,

where VB is the volume of the receptor region B.

This integral, or similar integrals over two volumes with argu-
ments that depend only on the distance between the two differential
volume elements, are common in dosimetric calculations. It is impor-
tant, therefore, that the 6-dimensional integral can be replaced by a
one-dimensional integral. Berger (1970) has shown that the solution
can be expressed in terms of the isotropic point-source kernel and
the point-pair distance distribution.
The point-pair distance distribution and the related quantity, geo-
metric reduction factor, will be considered in Section 9.4 for specified
geometries. At present, it is sufficient to consider the general formu-
lae. The point-pair distance distribution, pa(x), is defined as the
probability density of distances between random points uniformly
distributed in A and random points uniformly distributed in B. Thus,
p,(x) dx is the probability that a random point selected in A and a
random point selected in B are separated by a distance between x
and x+dx.
76 1 9. GEOMETRIC CONSIDERATIONS

Let h(x) be a function of the distance x; then, one can reduce the
6-dimensional integral over the regions A and B to a one-dimensional
integral

where x is the distance between the two differential volume elements

dVAand dVB, and VAand VB are the volumes of the regions A and
B. Here, and in the following, the limits of integration in the one-
dimensional integral are 0 and w. It is understood that p,(x) may
be zero in certain ranges of x, so that, effectively, the integral ranges
from xminto x-.
Combining Equation (9.3) and (9.4), one obtains

where a is the total activity in A, and the right hand expression must
be used if A has zero volume.
A point-pair distance distribution usually cannot be determined if
one deals with unbounded regions. It is then practical to use a modi-
fied quantity, called the geometric reduction factor (Berger, 197017,
U&) = VBpm(x)/4~x2, VB 0. + (9.6)
If B has zero volume, U,(x) is not defined; but one can then utilize

One can show that the geometric reduction factor, UAB(x),is equal
to the probability that a random point of A ends up in B, if shifted
in a random direction by distance x; a more detailed discussion of
the point-pair distance distribution and the geometrical reduction
factor is given in Section 9.4.
With the geometrical reduction factor one can rewrite Equation
(9.5) as

(-
- -
a
vB /e(x) u-(x) b, for VB f 0,VA finite

1
-
(9.8)
Dm = a I d x ) UBA(x)&, for other situations,
but with VA > 0.

'In the case of an extended region A, the term average geometric reduction factor
has formerly been employed. The proximity function, s,(x) = VB . pAB(x),
can also be
used, and is closely related to U,(x).
 9.2 RECIPROCITY THEOREM 1 77
If the receptor region B is unbounded, the mean dose rate may not
be meaningful; one may then consider the rate of energy imparted
to B as

1m ( = a 1e(x)um(x)61,

= a VB le(x) UB*(x)&,
for V,>O, VAfinite

for other situations.

(9.9)

As before, a and a are total activity and specific activity in A, respec-

tively
These equations permit a variety of dosimetric computations, pro-
vided the receptor and target regions are part of an infinite or suffi-
ciently extended uniform medium. The derivation of the quantities
e(x) and U,(x) is usually a far simpler problem than the direct
solution of Equation (9.3). More importantly, the formalism just
described permits a convenient separation of radiation-transport cal-
culations into two independent parts: the point-source kernel e(x)
and the geometrical factor Um(x), which is independent of radiation
properties and depends merely on the configuration of the source and
receptor regions.
Examples for the two functions are given in Appendix C; specific
problems may require separate numerical computations or Monte-
Carlo calculation4 of the two functions. Also, it should be noted that
much of the formalism described in this subsection is equivalent to
the method of absorbed fractions described in ICRU Report 32 (ICRU,
1979).

9.2 Reciprocity Theorem

The definition of the point-pair distance distribution is symmetri-
cal in A and B, i.e., pm(x) = pBA(x).It follows, from Equation (9.5),
that the mean dose rate in B from the activity a in A is equal to the
mean dose rate in A from the activity a in B. This is the reciprocity
theorem.
On the other hand, if the specific activity a is kept the same, the
energy imparted to B from A is equal to the energy imparted to A
from B. The reciprocity theorem applies, also, if one of the regions is
a point; the mean dose rate in B from a point source A of activity a
is equal to the absorbed dose rate a t A, if the activity a is uniformly
distributed in B.
It must be noted that the reciprocity theorem holds under the
condition of A and B being part of a sufficiently extended uniform
78 1 9. GEOMETRIC CONSIDERATIONS
medium. Approximate validity can still hold for non-uniform media.
For the consideration of non-uniform media the density f i )will
depend on position. If the specific activity (activity per unit mass) in
A is constant, Equation (9.3)is to be replaced by the relation

where MAand MBare the mass of A and B, and where the dose rate,
DGA,2,), from a unit activity a t point 2, to point 2, depends on the
composition and density of the medium throughout the vicinity of 2,
and f B that is relevant to the radiation transport Gom f Ato 2,.
For the reciprocity theorem to hold, DGA, must equal DGB,
zA),i.e., an activity a t point 2, must produce the same dose at point
zB,as the same activity a t 2, produces a t 2,. For scattered radiation,
this identity will commonly not apply. However, for the unscattered
primary radiation, it can usually be assumed. It applies strictly for
the unscattered radiation if all interaction coefficients are propor-
tional to density; in this latter case DG,, 2,) is equal to a function
h(s)/x2,where h(s) depends only on the type of radiation and on the
integral, s, of the density along the straight line between 2, and zB.
9.3 Isotropic Point-Source Kernels

Point-source kernels have been calculated for a variety of radia-

tions; examples of the results are given below. First, two particularly
simple examples will be treated, although these are of more didactic
than pragmatic interest.
For photons of sufficiently low energy, E,, only the photoelectric
effect plays a role and, if energy transport by the photoelectrons,
Auger electrons and fluorescent radiation is disregarded, one obtains
E(x) = Eo(l- exp( - @)I,
and (9.11)
e(x> = pE, exp( - F ) .
In a first approximation, the tracks of a-particles can be treated
a s straight lines with continuous energy loss equal to the stopping
power. With these assumptions, and if energy transport by h a y s is
disregarded, one has
E(x)= E(RJ -E(R, -x)
and (9.12)
e(x) = S(R, - x ) ,
 9.3 ISOTROPIC POLNT-SOURCE KERNELS / 79
where E(R) is the energy of the particle with residual range R and
S(R)is the stopping power at the same energy and remaining range.
Figure 9.1 gives the resulting integral, isotropic point-source ker-
nels normalized to the initial energy E, for different energies, E,, of
the a-particles. The differential kernels are frequently called Bragg-
curves.
The derivation of isotropic point-source kernels for electrons must
account for various factors such as curvature of the particle track,
energy-loss straggling and energy transport by secondary electrons.
Such computations have been performed by Berger (1971,1973) and
by a number of other authors who have used Monte-Carlo simula-
tions of charged particle tracks.

I I
I

10 MeV

0 20 40 60 80 100 120
DISTANCE, pm
Fig. 9.1. Integral, isotropic point-source kernels, normalized to the initial energy,
E,, for cr particles in water. The ordinate is the fraction of the emitted energy that is
deposited along the particle track, while the abscissa is the distance that the particle
has traveled. The electronic stopping power is taken from Ziegler (1980).
80 1 9. GEOMETRIC CONSLDERATIONS
Figure 9.2 gives the resulting integral, isotropic point-source ker-
nels normalized to the mean initial electron energy for beta rays and
monoenergetic electrons of various energies. The point kernels are
given in a scaled form, i.e., as a function of the variable XI%, where
x is the distance from the source and xS0 is the radius of the sphere
about the source within which 90 per cent of the emitted energy is
deposited. One notices that the scaled, integral point kernels are not
very sensitive to the energies of the beta rays and electrons.
High energy photons require a consideration of the scattered pho-
ton contribution and, depending on the nature of the problem, also
a consideration of energy transport by charged secondaries. Figure

0.8- 'H 0.000219

'lP 0.351

0.7- 1 MeV 0.328

0.05 MeV 0.00318

0.6-

o d.1 d.2 013 014 015 0:s d.7 0:s 019 1.0
RELATIVE DISTANCE, xlx,,

Fig. 9.2. integral, isotropic point-source kernels, normalized to the mean initial
electron energy for beta rays and monoenergetic electrons. The ordinate is the fraction
of the emitted energy that is deposited within a sphere of radius x about the point
source. The abscissa is the relative distance, d%, where % is the radius of the sphere
within which 90 per cent of the energy is deposited. The resulta are for interactions
with water, and x, distances are shown in the tabulation i n the figure. The curves
were developed from data in Berger (1971).
 9.4 POINT-PAIR DISTANCE DISTRIBUTIONS 1 81
9.3 gives a numerical example for a photon, isotropic point-source
kernel. Neutrons present a similar problem where the consideration
of the scattered radiation is essential.

9.4 Point-Pair Distance Distributions and Geometliic

Reduction Factors

The point-pair distance distribution and the related quantity, geo-

metric reduction factor, can, in principle, be computed for any con-
figuration of interest. However, the presence of complicated multiple
integrals may make the numerical work difficult. Monte-Carlo meth-
ods have been developed that simplify calculation of multiple inte-

50-

0.01
PHOTON ENERGY, MeV
Fig. 9.3. Energy deposition around monoenergetic, point-isotropic sources of pho-
tons with energies up to 2 MeV in an unbounded water medium. The figure shows the
radii of spheres around the point source within which the indicated fractions of the
emitted photon energies are absorbed. Several different plots are possible; in this case,
the radii of the spheres are plotted as function8 of the source photon energy. The
curves were developed from data in Berger (1968).
82 9. GEOMETRIC CONSIDERATIONS

grals. Analytical solutions exist, however, for a variety of simple

geometries.
Two important cases can be distinguished and will be identified
by the terms autologous and heterologous. The term autologous refers
to the condition that source region and target coincide, i.e., A=B.
The term heterologous refers to cases where source region and target
differ, although they may partly overlap.
A special case is the condition that either the source region or the
target region is a point while the other region is an extended domain.
In certain cases, this situation can be treated rigorously, and, then,
it is possible to obtain relatively simple integrals in the calculation
of the spatial distribution of absorbed dose around extended source
regions.
In the autologous case, the point-pair distance distributions are a
familiar concept of geometrical probability, and they have-
although for different applications, or from a purely mathematical
standpoint-been treated in the literature [see, for example, Kendall
and Moran (1963), Weil (1983), Stoyan et al. (1987)l. There is also a
relation between the geometric reduction factor, U(x), for a convex
body and the chord-length distributions, f(x), that result when the
body is randomly traversed by straight lines that are uniformly and
isotropically distributed (Kellerer, 1971, 1984):

where V and S are volume and surface of the body, and 4VlS is its
mean chord length. Furthermore, for a convex body, U(x) is the
sum distribution of the lengths of rays in random directions from a
random point in A to the surface of A.
For both the autologous and the heterologous cases, the two quanti-
ties p,(x) and U,(x) are equivalent. The definition of the point-pair
distance distribution pAB(x)as the probability density of distances
between pairs of random points leads to solutions for the spheroid
and for the cylinder [Equations (C.7 and C.lO) in Appendix Cl. The
separate definition of UAB(x)in Subsection 9.1 can be formulated in
a way that is also suitable for computations: If A(x,8) is the translate
of A by a distance x in the direction 6, then U,(x) is numerically
equal to the volume of the intersection A(x,B)nB uniformly averaged
over all directions and divided by VA(for related considerations see
Enns and Ehlers, 1978). This form of the definition leads to the
solution for the sphere or for two concentric spheres A and B [see
Equations (C.3 and C.15) in Appendix Cl. The relation
US(x) = (VAUA(x)+ VB UB(x) - 2 VAUAB(x))1 (VA-VB) (9.14)
 9.4 POINT-PAIR DISTANCE DISTRIBUTIONS 1 83
is then used to obtain the geometric reduction factor for the spherical
shell S = A - B [see Equation (C.4)I.
Finally, it may be noted that the geometric reduction factor can
also be defined in two-dimensional space. In this case, it can be
calculated from the point-pair distance distribution by the relation

where SBis the area of B.

 Calculation of the Dose
Equivalent

The objective of the preceding sections was to describe in detail,

with both text and equations, the theoretical intricacies of the con-
cepts involved in the determination of absorbed dose under various
conditions and by various methods. Because of the eventual interest
in health effects and risk, attention turns to the dose equivalent, the
quantity that places on a common basis the effects on tissues of
absorbed doses from various radiations. For completeness, therefore,
this short section is intended only to show how the calculation of
the dose equivalent follows from the concepts for the calculation of
absorbed-dose distributions.
The biological effectiveness of a given absorbed dose of ionizing
radiation depends, among other things, on the type and energy of
the charged particles that produce it. In particular, the biological
effectiveness generally increases with increasing linear energy
transfer (LET) up to approximately 125 keV/p (ICRU, 1970). Conse-
quently, the absorbed dose of a high-LET radiation that causes a
given effect must be multiplied by a factor, termed the RBE (relative
biological effectiveness), in order to obtain the absorbed dose of low-
LET radiation that causes the same effect.
Experimental radiobiology has demonstrated that the RBE
depends on many factors including absorbed dose, absorbed-dose
rate, type of effect, and environmental factors. The role of these
variables is-with some theoretical justification-ignored a t the low
doses of interest to radiation protection, and a quality factor has
been defined that weights the absorbed dose for the biological effec-
tiveness of the charged particles producing the absorbed dose (ICRU
1970, 1986). This quality factor, Q, is chosen to be a continuous
function of the unrestricted LET, i.e., L, or stopping power, of the
charged particles producing the absorbed dose. IfD(L,,;) is the distri-
bution of absorbed dose in L, a t a point of interest, i,in the receptor,
then the dose equivalent, H, is given by

-:1
Hf?) = D(L,,F)Q(LJdL, (10.1)

where L,, and L,, are the minimum and maximum values of L,
 l o . CALCULATION OF THE DOSE EQUIVALENT / 85
for which D(L,,7) has a nonzero value. Equal values ofH are intended
to correspond to equal probabilities of deleterious effects regardless
of the type of charged particles that are producing the absorbed doses.
The recommended quality factors as a hnction of L, are given in
Table (10.1) (ICRU, 1980). In the last several years, changes in the
value of the neutron quality factor have been recommended (ICRP,
1985; ICRU, 1986; NCRP, 1987) and simplified relationships are
being considered between Q and LET (ICRP, 1991). However, the
values used here serve to illustrate the principles of the calculations
and of the examples given later.
TABLE10.1-Quality factor, Q as a fumtwn of Lma

keV/prn
3.5 or less 1
7.0 2
23. 5
53. 10
175. or greater 20
"ICRU (1970);ICRP (1977).

To calculate the dose equivalent from Equation (10.1), the

absorbed-dose distribution in L, must be known, and this dose distri-
bution can be calculated if the type and energy distributions of the
charged particles producing the absorbed dose are known. If the
receptor is irradiated by uncharged particles, e.g., neutrons, then
in general, the production and transport of the secondary charged
particles produced by the uncharged particles must be carried out
before the dose equivalent can be calculated. If, however, the range
of the charged particles is small compared to the attenuation length
of the uncharged particles, then the dose equivalent may be calcu-
lated in a n approximation that is similar to that used in calculating
the absorbed dose using kerma factors (ICRU, 1980). In this approxi-
mation, the dose equivalent, HP), may be obtained from the equa-
tions

where
4fE,F) = the fluence per unit energy of uncharged particles a t
energy E and position F,
Em,,, Emin = the maximum and minimum energy a t which &E,?)
has a nonzero value,
c = summation over all charged particle species,

n,(E,E1) =the number of charged particles of type i per unit

mass produced in the energy interval dE' by the
collision in the receptor of an uncharged particle
with energy E,
= the maximum and minimum energy a t which
Eli,-, E'i,min
ni(E,E1)has a non-zero value,
R i(E ') = the range in the receptor of a particle of type i with
initial energy E',
L,i = the unrestricted LET of charged particles of type i in
the receptor,
and dl is a differential path length. Note that if Q in Equation (10.3)
is numerically set equal to unity, then the quantity Z(E) is the kerma
(ICRU, 1980) and Equation (10.2)yields the absorbed dose as well
as the dose equivalent a t point 7.
 LISTOFSYMBOLS / 87

List of Symbols

In choosing the various symbols, certain general principles have

been followed whenever possible. Vectors have been denoted by an
arrow above the symbol (e.g., F). A dot above a letter denotes differen-
tiation with respect to time. Primes are frequently used to denote
quantities which have been subjected to a transformation of some
kind. Subscripts 0 frequently denote initial or equilibrium values.
As is customary, complex conjugates are denoted by an asterisk.
This list of symbols is not intended to be complete; it lists only
the important symbols, and those that may possibly give rise to
ambiguity.However, all symbols are defined in the text as they occur.

V Vector operator del

a Total activity in a source region of interest
A Source region
B Binding energy
B Receptor region
c Velocity of light in a vacuum
c Charged particle designation
-
C Shell correction factor
L
Generalized inverse cosine
Auxiliary function
width
Vector element of area
Vector element of path length
Element of volume
Volume element of a source region A
Volume element of a receptor region B
Absorbed dose
Absorbed dose in tissue for uniform activity distribu-
tions
Absorbed dose at a radial distance r
Absorbed-dose rate
88 / LIST OF SYMBOLS
D(L,,r) Distribution of absorbed dose in L, at a point r
DAB Mean absorbed dose rate in receptor region B from the
activity in source region A
Dv Average absorbed dose in volume V
e(x) Differential point-source kernel
eV Electron volt
E Particle energy
E* Excitation energy
Ec Charged particle energy
Ei, Sum of energies of ionizing particles that enters the
volume of interest
Et Sum of energies of ionizing particles that leaves the
volume of interest
En Neutron kinetic energy
El, Uncharged particle kinetic energy
E(R) Energy of a-particles with remaining range R
E(x) Average energy imparted within a distance x from a
point source
VqJ) Atomic form factor
f(E) Scaled nuclear-recoil energy loss function
fdE) Energy probability function
G(gJ) Nuclear charge form factor
h Height
hv Energy of photons
H' Adjoint operator
Hfi) Pose equivalent a t point F
I Mean excitation energy
I Unit operator
j Type of collision
r u t Particle current distribution
J(r,t) Particle current
&$,Z Unit vectors in cylindical coordinates
k Energy of incident photons
K Atomic shell designation
K Scattering operator
k' Energy of scattered photons
K(r) Kerma of uncharged particles
kkEJ Kerma factor
 LIST OF SYMBOLS / 89
Relativistic and spin correction factor
Screening correction factor
Stopping number
Atomic shell designation
Atomic shell designation
Electron mass
Particle density distribution
Mass
Electron rest energy
Mass of incident ion
Mass of the target atom
Pion mass
Proton mass
Density of scattering centers
Screening parameter
Number of protons or neutrons
Radial unit vector
Number of atoms of type j
Particle density of the medium of interest
Avogadro's number
Particle density
Momentum transfer
Distance
Residual range
Probability that a fkee flight will be terminated when
its length is between s and s + cls
Legendre polynomial
Point pair distance distribution
Momentum transfer
Sum of kinetic energies of all charged particles created
by uncharged particles per unit mass
Rest mass energies
Quality factor
Position vector
Sphere radius
Classical electron radius
Original particle position
Initial range of a-particles
LIST OF SYMBOLS
Surface area of the volume of interest
Adjoint source operator
Mass stopping power of a receptor
Source density
Stopping power of an a-particle with remaining range R
Stopping power
Restricted stopping power
Proximity function
Collision stopping power
Nuclear recoil stopping power
Incoherent scattering function
Time
Translation operator
Particle kinetic energy
Positron kinetic energy
Electron kinetic energy
subscript to indicate uncharged particles
Direction of solid angle
Geometric-reduction factor (heterologous case)
Geometric reduction factor when source region and
receptor region coincide (autologous case)
Two dimensional geometric-reduction factor of cross
section c (autologous case)
Two dimensional geometric reduction factor of cross
section C. (heterologous case)
Volume
Nuclear volume
Volume of source region A
Volume of receptor region B
Energy transferred to recoiling electrons
Energy transferred from neutron to a recoiling nucleus
Maximum energy transfer
Average energy given to charged particles in a single
neutron-charged particle interaction
Energy weight function
Distance between source and receptor regions
Total path length traveled at energy E by radiation in
the volume of interest
 LIST OFSYMBOLS / 91
Spherical harmonic
Axial distance
Specific energy
Mean specific energy
Atomic number
Specific activity
Ratio of particle velocity to velocity of light
Density-effect correction fador
Infinitesimal thickness
Dirac delta function
Cutoff energy
Energy imparted
Mean energy imparted
Energy imparted to a receptor region
Eccentricity
Average logarithmic decrement
Center of mass deflection angle
Nominal charge
Effective charge
Deflection angle of faster electrons
Deflection angle of slower electrons
Deflection angle of secondary electron
Compton wavelength
Attenuation coefficient
Interaction coefficient of type j
Macroscopic cross section
Particle scattering cross section
Total probability of interaction per unit length
Total neutron mass interaction coefficient
Energy dependent nuclear interaction coefficient
Total uncharged particle interaction coefficient
Particle absorption interaction coefficient
Energy absorption coefficient of photons of energy k
Uncharged particle energy transfer coefficient
Total cross section
Radial distance in cylindical coordinates
Density
Nuclear charge density
92 1 LIST OFSYMBOIS
Microscopic cross section
Cross section for the jth type of interaction
Transport cross section
Rutherford scattering coefficient
Coherent (Rayleigh) scattering cross section
Incoherent scattering cross section
Scattering cross section that includes binding effects
Klein-Nishina scattering cross section
M+ller scattering cross section
Azimuth angle in cylindrical and polar coordinates
Adjoint flux density
Flux density distribution
Flux density
Spacedependent photon fluence of energy k
Scaled energy brernsstrahlung loss function
Fluence per unit energy of neutral particles a t energy
E and position F
Neutron fluence
Fluence
Equilibrium radiation field due to isotropic sources
Fluence a t a distance x from a plane-isotropic source
Deflection of charged particle
Deflection in the axial direction
Deflection in the radial direction
Appendix A
Information about Cross
Sections for Transport
Calculations

Some general aspects of the cross sections needed for dosimetry

calculations were discussed in Section 5. The purpose of this Appen-
dix is to describe and'illustrate some of the more important cross -

sections for photons, neutrons and charged particles, and to provide

information about up-to-date critically evaluated cross-section data.
Attention is focused on the cross sections needed for conventional
(macroscopic) dosimetry. Little is said about the cross sections at
very low energies (i.e., below 1keV) for charged particles needed for
microdosimetry or track-structure calculations.

A.l Photon Cross Sections

The various known interactions between photons and constituents

of matter are listed in Table A.1. The three most important interac-
tions, which are often the only ones included in transport calcula-
tions, are photoelectric absorption by atoms, incoherent (Compton)
scattering by electrons, and pair production in the electric field of
atomic nuclei and electrons. Figure A.1 indicates the values of the
photon energy hv, and the atomic number Z for which each of these
three interactions is predominant. Next in importance are coherent
(Rayleigh) scattering by atoms, and photonuclear interactions. Fig-
ure A.2 is a plot of the various cross sections as a function of photon
energy for carbon and lead.

A. 1.1 Photoelectric Effect

In this process, a photon is absorbed by an atom, and an electron
is ejected with kinetic energy, T, from the atom:
T=k-B, (A.1)
 Scattering
Interaction with: Absorption Elastic (coherent) Inelastic (incoherent) Multi-photon effects
Atomic electrons
-Z4 (low energy)
T, -Z6 (high energy)
p-------------- -
I
-7
Nucleons Photonuclear reactions ; Elastic nuclear Inelastic
I (y, n), (y, P), I scattering, Nuclear scattering,
;photdsslon, etc.
- : ( ~ , d - s (7, 7')

- - - - - - - - - . I

Electric field (1) Electron-positron

surrounding pair production in
charged particles field of nucleus,
( E r 1.02 MeV)
(2) Electron-positron
pair production in
electron field

(3) Nucleon-antinucleon
pair ~roduction

Mesons Photomeson production Modified

(E L 150 MeV) (nV)
.Solid-line boxes indicate the major effects contributing to photon attenuation in matter. Dotted-line boxes indicate additional effects
that contribute more than 1%over particular energy ranges.
bFromHubbell and Berger (1968).
 A . l PHOTON CROSS SECTIONS 1 95

PHOTON ENERGY, hv,, MeV

Fig. A.1. Locus of equal atomic cross section for Compton and photoelectric inter-
actions (u = T ) , and for Compton and pair-production interactions (u = K ) . The
incident photon energy is hv,, and Z is the atomic number of the atoms in the absorber.
Compton collisions have larger cross sections than any other mode of interaction in
the entire domain of medium-energy photons marked "Compton effect dominant"
(From Evans, 1958.)

where k is the energy of the incident photon and B the binding energy
of the atomic shell from which the electron is ejected. At a given
energy, the stronger the binding, the larger is the photoelectric cross
section. At photon energies greater than the K-shell binding energy,
approximately 80 percent of the photoelectrons are ejected from the
K shell. The photoelectric cross section (per atom) is proportional to
- -
Zmlkn,where m 4 and n 3. At low photon energies the dependence
of the photoelectric cross section on the photon energy is not mono-
tonic but exhibits sharp discontinuities a t the K -, L - . . .absorption
edges (see Figure A.2), that is, a t energies a t which the photon has
just enough energy to eject a n electron from a particular shell.
At low photon energies, the photoelectrons are ejected mainly a t
right angles with respect to the direction of the incident photon. At
high energies, the peak in the angular distribution is in the forward
direction; half of the photoelectrons are emitted within a cone whose
half-angle (with respect to the photon direction) is approximately
equal to 20" a t 1 MeV, 30" a t 0.5 MeV, and 60" a t 100 keV.
One of the most comprehensive sets of accurate photoelectric cross
sections is that calculated by Scofield (1973) based on the numerical
solution of the Dirac equation and the numerical evaluation of the
pertinent matrix elements. Excellent reviews of the theoretical and
experimental aspects of the photoelectric process have been given by
 CARBON 12 ' 6 )
1

PHOTON ENERGY, MeV

I I I 1 I I I I I
>\?< -
LEAD lZ*821
?.Ft0 -
-

U,,, .EXPERIMENT lmq

-
-

cn
cn
0
-
IT:

1DJ 1@' 100 1w

PHOTON ENERGY, MeV

Fig. A.2. Photon cross sections for carbon and lead. (From Hubbell et al.,
 A.l PHOTON CROSS SECTIONS 1 97
Fano and Cooper (1968) and Starace (1982) with emphasis on the
low-energy region, and by Pratt, et al. (1973) covering the energy
region above 10 keV.

A.1.2 Fluorescence Radiation and Auger Electrons

When an electron is ejected from the K shell, the vacancy can be
filled by an electron from the L or M shell jumping into the K shell.
The energy thereby made available is either converted into an x ray,
or is expended in the ejection of an Auger electron from the L or M
shell. Thus, either one or two additional vacancies in the outer L or
M shell are produced, and these in turn are filled by electrons from
outer shells jumping into the L or M shell. In the course of the de-
excitation of the atom, a cascade of many x-ray photons and Auger
electrons is emitted, and the atom is left in a multiply-ionized state.
Similar, but smaller cascades, are initiated when a photoelectron is
ejected from L or higher shells.
The probability that a vacancy is filled by emission of an x-ray
photon rather than an Auger electron is called the fluorescence yield.
Surveys of the experimental data on fluorescence yields and the
Auger effect have been made by Bambynek et al. (1972) and Burhop
and Assad (1972). The fluorescence yield, %, for the K shell is shown
in Figure A.3. It is almost zero for low-Z and almost unity for high-
Z atoms.

A.1.3 Incoherent (Compton) Scattering

At energies a t which incoherent scattering is the dominant mode
of photon interaction with matter, it is usually a good approximation
to consider the process to be a n elastic collision of a photon with a
free electron that is initially a t rest. The application of the laws of
conservation of momentum and energy to such a collision leads to
the Compton relation between the increase, AA, in the wavelength
of the scattered photon and the angle 0 through which it is deflected:

where A, = hlmc = 2.42 x 10-lo cm is a constant known as the

Compton wavelength (h is Planck's constant, m is the electron mass
and c the velocity of light). Expressed in terms of the photon energy,
this relation becomes
98 1 APPENDIXA

0 20 40 60 80 100
ATOMIC NUMBER, Z
Fig. A.3. Fluorescence yield for the K shell as a function of atomic number. ( A h
Bambynek et al., 1972.)

where k is the energy of the incident photon, ktthat of the scattered

photon, and mc2 = 0.511 MeV is the electron rest energy.
The electron recoil angle 4 is given by
cot 4 = (1 + k l d ) tan (812) (A.4)
and lies between 0"and 90". The energy transferred to the recoiling
electron is

and lies between 0 and 2k2/(mc2+ 2k).

 A.l PHOTON CROSS SECTIONS 1 99
For unpolarized photons, the cross section for Compton scattering,
differential in the deflection angle 8, is given by the Klein-Nishina
formula:

x {l +
y2(1 - ~ 0 ~ 0 ) ~
(1 + cos2e) [ l + y (1- c0sO)l
sine do,]
where y = klmc2 and re = e2/mc2 = 2.82 x 10-lScm is the classical
electron radius. Graphs of the Compton energy-angle relation and of
the differential and integrated Klein-Nishina cross section can be
found in a report by Nelms (1953). A thorough review of the theoreti-
cal and experimental aspects of Compton scattering was given by
Evans (1958).Figures A.4, A.5 and A.6, taken from Evans, illustrate
the number us angle distributions of scattered photons and recoil
electrons, and the number us energy distribution of the recoil elec-
trons.
Because the target electrons are neither free nor at rest, the Comp-
ton relation between energy loss and deflection angle does not hold
precisely. Photons incident with energy k and deflected through an

CROSS SECTION, 102' cmzlradian electron

Fig. A.4. Collision differential cross section as number-us-angle dietribution of
scattered photons, dGu)/dO. The direction of the incident photon is along the abscieea
a t the angle of 0". Numerically, dcu)ld8 is the probability per incident photon and per
electronlcm2of scattering material that the scattered photon will be directed into the
angular interval which lies between two cones whose half angles are 0 and 0 + dB.
(From Evans, 1958.)
100 1 APPENDIXA

CROSS SECTION, lo2' cm2/radian electron

Fig. A.5. Collision differential cross section as number-vs-angle distribution of
recoil electrons, d(,u)/d4, projected into the angular interval between4 and 4 + d4,
for the primary photon energies shown in the curves. The incident photon is from the
left along the abscissa. (From Evans, 1958.)

angle can have various energies k" after the scattering that are
distributed around the value k' given by the Compton relation. This
is illustrated in Figure A.7 with experimental and theoretical results
of Schumacher (1971) for 662-keV gamma rays scattered through
135" in lead.
The effect of the binding of the atomic electrons is to reduce the
scattering cross sections, particularly at small angles. This reduction
can be calculated, in good approximation, with the use of an incoher-
ent scattering function S(q,Z). The scattering cross section with
binding effect included, duBD,is calculated from the expression
dam = duKN S(q,Z), (A.7)
 A.1 PHOTON CROSS SECTIONS 1 101

RECOIL ELECTRON ENERGY, MeV

Fig. A.6. Number-vs-energy distribution of Compton recoil electrons, for seven
values of the incident photon energy hv,, in cm2(millibarn)/keV,per free elec-
tron. The energy spectrum of scattered photons is obtained by transformingthe energy
scale from T to hv, - T for each curve (From Evans, 1958.)

where

is the momentum transfer. The incoherent scattering function,

S(q,Z), represents the probability that the atom is excited or ionized
102 / APPENDIXA

PHOTON ENERGY, keV

Fig. A.7. Compton crow sectionsversus secondary gamma energy k',for a primary
energy k = 661.6 keV and for a scattering angle of 0 = 135".Circles: experimental
resulta for lead. Curve a: calculated for the K shell of lead. Curve b: calculated for
lead taking into account the K, L, M and N shells. Curve c: experimental results for
scattering from free electrons. (After Schumacher, 1971).
 A. 1 PHOTON CROSS SECTIONS 1 103
when the photon imparts a momentum q to any of the atomic elec-
trons. Incoherent scattering functions, calculated with the use of
Hartree-Fock wave functions, have been tabulated for all elements
by Hubbell et al. (1975). Figure A.8 compares the cross sections duKN
and duBDfor 84.4 keV gamma rays in tungsten and shows that the
cross section calculated with the incoherent form factor is in good
agreement with measurements of Rao (1981); see also Rao and Rao,
1981.
I I I I I I I I I
804
TUNGSTEN
- Klein-Nishina cross section -

---- Calculated with the use of

the form factor from
Hubbell et al. (1975) -

,- 1 Experiment of D. V. Rao
-

SCATTERING ANGLE, degrees

Fig. A.8. Differential, incoherent scattering cross sections as a function of scatter-
ing angle for 84.4-keV gamma rays. (From Rao, 1981.)

Binding effects lower the total scattering cross section compared

to the value predicted by the Klein-Nishina formula. However, a t
low energies where this reduction is significant, scattering is unim-
portant compared to photoelectric absorption. As indicated in Table
A.2, the inclusion of binding effects lowers the total photon-interac-
tion cross section by a t most a few percent. Therefore, the binding
effects have ofZen been disregarded in transport calculations.

A.1.4 Pair Production

In this process, the photon disappears and an electron-positron
pair is created. The process takes place in the electric field of the
104 / APPENDIX A
TABLEA.2-Percent decrease of the total photon attenuation coefficientdue to
binding correction for incoherent scattering
Photon Percent decrease
energy C A1 Cu Ag Pb
keV
10 2.5 0.3 0.0 0.1 0.1
20 6.2 1.2 0.2 0.4 0.1
50 3.2 3.7 0.9 0.3 1.6
100 1.3 2.4 2.2 0.9 0.4
200 0.0 0.8 1.9 1.3 0.6
500 0.0 0.2 0.5 0.5 0.8

atomic nucleus, with a cross section k, proportional to 22.Some recoil

momentum (but practically no energy) is transferred to the atom as
a whole. The energy k of the photon is given to the electron-positron
pair, i.e.,

where T-and T+are the kinetic energies of the electron and positron
and 2mc2 is the combined rest energy of the pair. The threshold
energy for the process is k = 2mc2 = 1.022 MeV. The energy is
shared approximately randomly between the electron and positron.
Figure A.9 shows the scaled cross section as a function of the fraction
of the total kinetic energy of the pair electrons that is acquired by
the positron, calculated according to the theory of Bethe and Heitler
(1934). For high incident photon energies, the pair emerges predomi-
nantly in the forward direction. The average angle between the
direction of the photon and the directions of the electron and positron
has the order of magnitude mc2/T'.
Pair production also takes place in the electric field of the bound
atomic electrons. Recoil momentum and energy are transferred to a
single electron which is thereby ejected from the atom. In a track
visualization device (such as a photographic emulsion or a bubble
chamber), three charged particles are observed: the electron-positron
pair and the ejected atomic electron. For this reason the process is
also called triplet production. The threshold for triplet production is
k = 4 d = 2.044 MeV. The cross section for pair production in the
field of the Z atomic electrons, K,, is equal to (qJZ)K, where K , is the
cross section for pair production in the field of the atomic nuclaus.
The proportionality constant % is a quantity of order unity. As shown
in Table A.3, q,/Z actually goes to zero near the threshold and reaches
values of 1.1 to 1.2 a t very high energies.
A positron created in a pair-production event eventually recorn-
bines with an electron. The energy made available by the disappear-
 A.l PHOTON CROSS SECTIONS 1 105

Fig. A.9. Differential pair-productioncross section expressed as a function of the

dimensionless quantity P = (137/9J2) (k-2rnc2) (d~,/dT+). The dashed curves take
into account the screening of the nuclear charge by the atomic electrons. (From Evans,
1955.)

ance of the eledron-positron pair is converted into annihilation radi-

ation. The most likely annihilation process is two-photon annihila-
tion, resulting from the col.lision of a positron with a free electron.
Most collisions with free electrons (80 percent in the case of lead)
occur when the positron has slowed down nearly to rest in the
medium. The two annihilation photons, going off in essentially oppo-
site directions, then each have an energy mc2 = 0.511 MeV. Two-
photon annihilation can also occur when the positron is still in flight.
The probability for this to occur is greatest when the kinetic energy
-
of the positron has the rather low value of 0.5 MeV. A rarer process,
106 1 APPENDIXA
TABLEA.3-Contribution of pair production in the field of atomic electrons to the
total pair production cross section. The total cross section is proportional to
22 (1 + ~),,/a
Photon energy
Carbon Lead
MeV

the emission of one annihilation photon, occurs when the positron

collides with a strongly bound atomic electron. The ratio of one-
photon to two-photon annihilation is largest for positron kinetic
- -
energies of 10 nu?, where it amounts to 20 per cent for lead.
The theory of Bethe and Heitler, as summarized, for example, in
the book by Heitler (1954), provides an explanation of all the essen-
tial features of pair production. However, the Bethe-Heitler theory,
derived in first-order Born approximation, needs corrections a t all
energies if accurate cross sections are to be obtained. A review of the
complicated analytical cross-section formulas going beyond the first-
order Born approximation is given by Motz et al. (1969).A combina-
tion of theoretical formulas and numerical factors for the accurate
evaluation of the total pair-production cross section is given by Hub-
bell et d.(1980).

A.1.5 Coherent (Rayleigh) Scattering

This process involves the coherent scattering of photons by bound
atomic electrons such that the atom absorbs no energy. The photon
is deflected but its energy is left unchanged. In good approximation,
the Rayleigh cross section can be calculated with the use of an atomic
form factor F(q,Z), from the expression

where
 A.l PHOTON CROSS SECTIONS 1 107
is the momentum transfer. F(q,Z) represents the probability that
the recoil momentum q is transferred from the photon to the Z
electrons of the atom without any resulting excitation or ionization.
Tables of form factors derived from Hartree-Fock wave functions
have been given by Hubbell et al. (1975).
A review of the theory of Rayleigh scattering, including a treat-
ment more rigorous than the use of form factors, can be found in a
paper by Kissel et al. (1980) which also contains tabular results
for elastic scattering by lead. The elastic scattering cross section
(differential in deflection angle) for atomic numbers Z = 2 to 26 and
for photon energies from 0.5 to 200 keV, calculated with the use of
form factors, has been presented in graphical form by Brown (1975).
Angular distributions for scattering by carbon, takenfrom that refer-
ence, are shown in Figure A.lO.
The Rayleigh cross section peaks in the forward direction and is
large only a t energies where the photoelectric cross section is even
larger. Therefore, elastic scattering (involving no energy loss and
only a small deflection, and dominated by photoelectric absorption)
is often omitted from transport calculations.

A.1.6 Photonuclear Effect

This effect consists of the absorption of a photon by a nucleus

and the subsequent emission of nuclear reaction products. The most
likely outcome is the emission of a single neutron. Other possibilities
are the emission of two or more neutrons and the emission of charged
particles and gamma rays. The energy dependence of the photo-
-
nuclear cross section is characterized by a peak 5 MeV wide at an
energy located between 30 MeV and the threshold energy (- 5 to 15
MeV). Experimental photonuclear cross sections for nitrogen and
oxygen are shown in Figure A.11. (The broad maximum, as illus-
trated in the figure, has been referred to traditionally as the "giant
dipole resonance.")
Photonuclear cross sections are relatively small, amounting at
most to 5 to 10 percent of the total electromagnetic cross section.
Relative magnitudes of the photonuclear cross section are indicated
in Table A.4 for a few elements.
The photonuclear cross sections vary irregularly from one nucleus
to another, and must be obtained experimentally. A guide to the
literature on photonuclear reactions is available in the form of a
bibliographic index (Fuller et al., 1973; Fuller, 1985). Extensive
collections of photonuclear cross-section data, in graphical form, can
108 / APPENDIX A

SCATTERING ANGLE, degrees

SCATTERING ANGLE, 'degrees

Fig. A.lO. Calculated, normalized differential cross sections for coherent (Rayle-
igh) scattering of photons for various photon energies. Upper figure: energy range 0.5
to 10 keV. Lower figure:energy range 15 to 200 keV. To obtain absolute cross sectiona
in units of cmYsr, multiply values from figuresby Z2r,2, where rz = 7.94 x m2.
(After Brown, 1975.)
 A . l PHOTON CROSS SECTIONS 1 109

PHOTON ENERGY, MeV

Fig. A.11. Phobnuclear cross sections for nitrogen and oxygen. (After Berman,
1975.)
110 / APPENDIX A
 A.l PHOTON CROSS SECTIONS 1 111
be found in papers by Biilow and Forkman (1974), Berman (1975),
and Berman and Fultz (1975).

A.1.7 Attenuation Coefficient

The photon attenuation coefficientrepresents the total probability,
per unit pathlength, for the photon to have an interaction of any
kind. It is convenient to reduce the explicit dependence on the density
of the medium by using the mass attenuation coefficient p/p which
is given by

Here p is the density of the medium, in g ~ m -N,

~ ,= 6.022045 x
is Avogadro's number in mol-I, M A is the molar mass in g mol-',
and CJp is in units of cm2g-'. The component cross sections (cm2)in
the bracket in Equation (A.12) are, respectively, for photoelectric
absorption, incoherent (Compton) scattering, pair production in the
field of the atomic nucleus and the atomic electrons, and coherent
(Rayleigh scattering). In data tables, the photonuclear cross section
is usually excluded from the sum. Attenuation coefficients are some-
times given with the coherent scattering cross section omitted, for
users who disregard this effect from their transport calculations.

A.1.8 Energy -Absorption Coefficient

The various channels of energy transfer from photons to electrons
and from electrons to photons are indicated in Figure A.12.
The energy transfer can be calculated in terms of an energy-
absorption coefficient p,,(k). In terms of the cross sections for various
interactions, the energy-absorption coefficient can be expressed by
a n equation similar to Equation (A.12):
N
Pen - a [T f, + Oincfme + (Kn + fKl.
~ e )
P MA
The factors f,, qnc,and f, represent the average fraction of the photon
energy that is converted into kinetic energy of secondary electrons
in the respective interactions. These factors do not include the energy
re-radiated in the form of fluorescence, bremsstrahlung, and annihi-
lation radiation (as indicated in Figure A.12). Mass attenuation-
coefficients and energy-absorption coefficients for selected materials
are shown in Figure (A.13).
112 / APPENDIX A

I
r h h 7 energy S&nirg
f i r pducf i n \
\ \
~lod/&c A Annihi/ofin

+
o&olpIion
ff-ce
~1ofion

I I
t 1 1

1 - 1
Ionizafion and &itofion
Riohe/ecfhs,Auger e l e c f m ~ ~ ~

h okqh
Nuclear

Fig. A.12. Flow diagram of conversion of energy from one form to another in the
course of photon energy absorption in a medium. (From Hubbell and Berger, 1968.)

A.1.9 Photon Cross-SectionCompilations

Many critical data analyses have been carried out leading to the
compilation of photon cross sections, attenuation coefficients, and
energy-absorption coefficients. Except at energies below 10 keV, -
there has been an increasing tendency to base such tabulations
largely on theoretical cross sections. Experimental data (most o h n
in the form of attenuation coefficients) then serve mainly as a check
on the calculated cross sections. Thirteen photon cross-section compi-
lations that have appeared in the last thirty years are listed in Table
A.5. The relation between these compilations is complex. Usually,
each compilation includes some data from older compilations. Thus,
 A.2 CROSS SECTIONS FOR CHARGED PARTICLES / 113

/---

/' 10-2
10-2 lo-! I 10 102

PHOTON ENERGY, MeV

Fig. A.13. Photon mass-attenuation coefficient, LJp (solid line), and mass energy-
absorption coefficient, pJp (dashed line), as a function of photon energy. (After
Hubbell, 1982.)

the older compilations are usually not completely obsolete, and they
also retain explanatory text of lasting value. The most up-to-date
photon cross-section data above 10 keV for dosimetry calculations
can be obtained by combining the data given in Hubbell (19771,
Hubbell et al. (1980), and Hubbell (1982). A complete compilation of
photoabsorption and coherent scattering data a t low energies (30 eV
to 10 keV), based on experimental data, has been given by Henke et
al. (1982).

A.2 Cross Sections for Charged Particles

A.2.1 Elastic Scattering of Electrons by Atoms

In this process, the electron is deflected but loses practically no
energy. The target atom receives some recoil momentum but is nei-
 TABLEA.5--Compilations of photon cross sections 2
Authors Region wvered Substances included
SK
Davisson and Evans (1952) 100 keV to 6.5 MeV 24 elements +
White-Grodstein (1957) 10 keV to 100 MeV 24 elements and 5 compounds
Hubbell and Berger (1968) 10 keV to 100 MeV 23 elements and 4 compounds
McMaster et al. (1969) 1 keV to 1MeV elements Z = 1 to 94
Hubbell (1969) 10 keV to 100 GeV 23 elements and 13 compounds
Storm and Israel (1970) 1keV to 100 MeV elements Z = 1 to 100
Bigga and Lighthill (1971, 1972) 10 eV to 10 GeV elements Z = 1to 100
Veigele et al. (1973) 100 eV to 1 MeV elements Z = 1to 94
Hubbell (1977) 100 eV to 20 MeV H, C, N, 0, Ar and 7 compounds
Plechaty et al. (1981) 100 eV to 100 MeV elements Z = 1to 98
Hubbell et al. (1980) 1 MeV to 100 GeV elements Z = 1to 100
Hubbell (1982) 1 keV to 20 MeV 40 elements and 45 compounds
Henke et al.(1982) 30 eV to 10 keV elements Z = 1 to 94
 A.2 CROSS SECTIONS FOR CHARGED PARTICLES 1 115
ther excited nor ionized. The peak, elastic-scattering, cross-section
differential in deflection angle occurs a t very small angles but deflec-
tions up to 180" are possible.
There is a large body of theory dealing with the elastic-scattering
cross sections. The numerous theoretical expressions, developed in
various approximations, have been reviewed by Motz et al. (1964).
The most accurate cross sections are obtained by a phase-shift analy-
sis involving the numerical solution of the Dirac equation, for a
screening taking into account the presence of the electron cloud
around the atomic nucleus. The most extensive calculations of this
kind a t low energies have been carried out by Riley et al. (1975) for
selected elements with atomic numbers between 2 and 92 and for
electrons with energies from 1keV to 256 keV.
At higher energies, it is a good approximation to express the differ-
ential cross section as the product of three factors that can be evalu-
ated independently:
da(8) = d~E(8)K,(B) K,1(6), (A.14)
where duR is the Rutherford cross section for scattering by a bare
point charge, K,, is a screening correction, and &,
is a correction
that takes into account spin and relativistic effects. The Rutherford
cross section is given by

where re = 2.817938 x 10-lS cm is the classical electron radius, Z

is the atomic number, mc2 = 0.5110034 MeV is the electron rest
mass, T is the kinetic energy of the incident electron, and 8 is the
angle through which the electron is deflected.
The screening correction takes into account the reduction of the
probability of small deflections due to the screening of the nuclear
charge by the atomic electrons, and can be computed in various
approximations. In the impulse approximation,
K,(6) = [l - F(q,Z)12, (A.16)
where F(q,Z) is the atomic form factor, which, in turn, is a function
of the momentum transfer
q = (2/mc2)[T(T + 2mc2)]" sin (612). (A.17)
The atomic form factor used here is exactly the same quantity as
that used in Section A. 1.5 for the calculation of coherent (Rayleigh)
scattering of photons, and can be obtained from the tables of Hubbell
et al. (1975).
116 / APPENDIX A
An alternative screening correction is
K,(o) = (1 - c o ~ e ) ~ /-( iC O S ~+ 2q12, (A.I~)
where

is a screening parameter derived by Moliere (1947) who used a

screened Coulomb-potential, Thomas-Fermi model of the atom, and
where p is the velocity of the electron relative to the velocity of light.
Actually, Equation (A.19) is an approximation to a more rigorous,
and much more complicated, expression given by Molibre whose
derivation and numerical evaluation has been discussed by Zeitler
and Olsen (1964).
The screening factor, Equation (A.181, is only a crude approxima-
tion to the angular cross section at small angles, but provides an
accurate value of the transport cross section, ul,which is the key
parameter for the calculation of multiple elastic scattering. This
transport cross section is defined as
o1 = 2n d0 (1 - cos6) sin 6 d0 = u (1 - a) ,
(A.20)
- v
dR -
where a is the total elastic scattering cross section and cos0 is the
mean cosine of the deflection angle.
The correction factor, L,, is given by the ratio uMot,,/uR of the
Mott to the Rutherford cross section, where urn, is the cross section
obtained by Mott (1929) from an analytical solution of the Dirac
equation for scattering by a static unscreened Coulomb potential.
Mott's cross section is expressed in terms of a slowly converging
series with complicated coefficients; it has been evaluated numeri-
cally and tabulated by Doggett and Spencer (1956). Illustrative
results are given in Table A.6.
TABLEA.GRatio of Mott to Rutherford cross section, u,,,&~~for electron9
Scattering Ratio
-
Annle
13 Carbon Lead
Kinetic energy of electrons
degrees 10 MeV 1 MeV 0.1 MeV 10 MeV 1 MeV 0.1 MeV
0 1.000 1.ooo 1.000 1.000 1.000 1.000
15 0.999 1.000 1.004 1.127 1.108 1.024
30 0.961 0.967 0.996 1.358 1.315 1.040
60 0.788 0.814 0.946 1.918 1.857 1.336
90 0.532 0.586 0.868 1.981 1.974 1.672
120 0.269 0.352 0.786 1.324 1.444 1.866
150 0.074 0.179 0.725 0.422 0.683 1.956
180 0.002 0.115 0.702 0.007 0.328 1.978
"From Doggett and Spencer (1956)
 A.2 CROSS SECTIONS FOR CHARGED PARTICLES 1 117
The factorization in Equation (A.14)gives accurate results only
when there is no overlap between the two correction factors, i.e.,
under conditions such that, at a given angle, only I(, or only &,
departs significantly from unity. This condition is not satisfied for
Iow-energy electrons, The region of validity of the factorization is
indicated in Figure A.14 in terms of the transport cross section ul.
In this figure, comparisons are shown between values of ulobtained
with the factorization and the more accurate results of Riley et al.
(1975).It can be seen that the factorization is adequate for electron
-- -
energies down to 128 keV in gold, 64 keV in copper, 16 keV -
in aluminum, and 4 keV in beryllium. For the differential cross
section itself, the factorization is expected to break down a t some-
what higher energies.
When the electron wavelength is sufficiently small, the nucleus
can no longer be treated as a point charge, and there occurs a great
reduction in the probability of scattering through angles greater
than Om,, - AIR, where A is the electron wavelength divided by
2m,and R is the nuclear radius. The effect of finite nuclear size is
important for electrons with energies greater than - 50 MeV. For
review of this topic, see, e.g., Hofstadter (1956)or Friar and Negele
(1975).

A.22 Elmtic Scattering of Protons by Atoms

The Rutherford cross section for this process is given by

where 8 is the deflection angle in the laboratory system. The factor

H (8, MdM) arises in the conversion from the center-of-mass to the
laboratory system, depends on the ratio of MdM of the proton mass
to the mass of the target atom, and is given by

H is smaller than unity, by an amount that is significant only at

Iarge angles. For collisions with hydrogen atoms, H is 0.9948a t 30
degrees, and 0.8889 a t 60 degrees. For collisions with carbon atoms,
H is 0.9993a t 30 degrees and 0.9895 at 60 degrees.
The eIastic scattering cross section given by Equations (A.21)and
(A.22)should be reduced at very small angles to take into account
 ELECTRON ENERGY, MeV
Fig. A.14. Scaled transport cross section for elastic scattering of electrons by
atoms as a function of electron energy, for various atomic numbers. The ordinate is
B2ul,where a,is the transport cross section [Equation (A.2011 and B is the velocity of
the electron relative to the velocity of light. The circles represent the results of
an exact phase-shift calculation by Riley et al. (1975).The curves are approximate
calculations based on the factorization indicated in Equation (A.14).The screening
correction, &,was evaluated according to Equation (A.16)for the dashed curves,
and according to Equations (A.18)and (A.19) for the solid curves. ( f i r Seltzer and
Berger, 1982.)
 A.2 CROSS SECTIONS FOR CHARGED PARTICLES / 119
the screening of the nuclear charge by the atomic electrons and, also,
at large angles to take into account the effect of the finite size of the
nucleus. This can be done by applying to Equation (A.21) a combined
screening factor

where F(q,Z) is the atomic form factor, and where

G(q,Z) = (4dq) (A.24)

is a nuclear-charge form factor that depends on the charge density

mu of the nucleus. For information about my, see, e.g., Friar and
Negele (1975). Protons of interest in dosimetry are hardly ever ener-
getic enough so that a relativistic correction factor Km1for Equation
(A.21) would depart significantly from unity.

A.2.3 Scattering of Electrons by Atomic Electrons

Such collisions result in an angular deflection of the incident elec-
tron and a transfer of energy to the target electron that is ejected
from the atom. When the energy transfer is large compared to the
binding energy of the target electron (a condition usually well satis-
fied for valence electrons), it is a good approximation to treat the
collision as obeying the Mgller (1932) cross section, derived on the
assumption that the target electron is free and initially at rest. This
cross section is given by

where a = F / ( T + mc2)2,T is the kinetic energy of the incident

electron, p is its velocity divided by the velocity of light, mc2 is
the electron rest energy, and W is the energy transferred from the
incident to the target electron.
The factor in curly brackets in Equation (A.25) converts the Ruth-
erford cross section to the Mgfller cross section, and takes into account
exchange effects (due to the indistinguishability of the incident and
target electrons). The value of W in an electron-electron collision
could range from 0 to T (in the absence of binding effects). However,
by convention, W is interpreted as the difference between T and the
kinetic energy of the faster of the two electrons after the collision.
120 / APPENDIXA
Therefore, the largest possible value of W is actually Tl2, and the
cross section is normalized accordingly.
It follows from the conservation of energy and momentum in the
collision, that the deflection angle 8, of the faster of the two electrons
is given by
sin2& = (2WIT)l[(Tld) (1 - WIT) + 21, (A.26)
and that the deflection angle 8, of the slower electron is given by
sin28, = 2(1 - W/T)I[(TId)(WIT) + 21. (A.27)
Taking into account all possible values of the energy transfer .W, one
finds that Of lies between 0 and 8*, and 8, between 8* and I,where
2
sin2@ = 2 / [ ( T l d ) + 41. (A.28)
The relations in Equations (A.26) and (A.27) between energy loss
and deflection angle do not hold rigorously because the target elec-
trons are bound and are in motion when struck by the incident
electron. By way of example, this is indicated in Figure A.15, which
gives the cross-section differential in energy loss as measured by
Missoni et al. (1970) for 400-keV electrons scattered through an
angle of 20". It can be seen that the energy spectra of the scattered
electrons have maxima at, but are distributed around, the energy
predicted by Equation (A.26). The higher the atomic number, the
greater is the spread of the spectrum around the peak value. There
are as yet not enough experimental data or theoretical results for a
systematic and comprehensive inclusion of binding effects in the
cross section. However, binding effects are included correctly in stop-
ping power theory (see Section A.2.7), a t least under conditions when
the incident electron has a velocity larger than those of the atomic
electrons.

A.2.4 Scattering of Protons by Atomic Electrons

The cross section for the collision of a proton with a free electron
is given by

where W is the energy transferred to the electron, and

 A.2 CROSS SECTIONS FOR CHARGED PARTICLES 1 121

N
E
0
0.8

s
N

2-
0
F
Z:
0.4
(I)
(I)
0
u
0

SCATTERED ELECTRON ENERGY, keV

Fig. A.15. Inelastic scattering cross sections per atomic electron, (l/ZMZuldMT,
as a function of the energy of the scattered electron for electron-atomic collisions with
an initial electron kinetic energy of 400 keV and a scattering angle of 20". Results
are given for thin targets of carbon (Z = 6 , 6 1 ~ g l c m ~copper
), (Z = 29,64 wglcm2),
and gold (Z = 79, 50 Bg/cm2).For electron-electron collisions (M#ller scattering), the
kinetic energy T', of the scattered electron in the laboratory system is indicated by
the arrow on the abscissa energy scale. (After Missoni et al. 1970.)

is the maximum possible energy transfer. The angular deflection of

the proton is quite small. Of chief interest is the emission angle 8, of
the secondary electron (delta ray) which is approximately equal to

Because the binding and non-zero momentum distribution of the

atomic electrons are neglected, Equations (A.29) and (A.30) are only
approximate, and an exact theory is not readily available. Figure
A.16, taken from a review by Wilson and Toburen (1980), shows the
distributions in energy and angle of the delta rays set in motion
in molecular gases of low average atomic number by protons with
energies from 0.3 to 2 MeV. As in the case of electrons, proton
stopping power theory (to be discussed in A.2.7) includes an accurate
treatment of the average energy loss of protons in collisions with
atomic electrons.

A.2.5 Electron Bremsstrahlung

In this process, the electron is accelerated in the field of the
screened atomic nucleus or in the field of the atomic electrons, and,
as the result of this acceleration, emits a bremsstrahlung photon.
When a particle of charge e and mass M is accelerated in the field of a
charge Ze, its acceleration is proportional toZe2/Mand the radiation
probability or intensity is proportional toZ2e4/iW.Due to the appear-
ance of M2 in the denominator, bremsstrahlung by heavy particles
is negligible compared to brernsstrahlung by electrons, and can be
disregarded in the present context. The emission of bremsstrahlung
involves the coupling of the electron with the electromagnetic field
of the emitted photon. Therefore the cross section for the emission of
-
bremsstrahlung is 11137th the cross section for the elastic scatter-
ing of electrons by atoms.
Bremsstrahlung in the electric field of the atomic nucleus is dis-
cussed first. The quantum-mechanical theory for this process, in
the first Born approximation, was initially developed by Bethe and
Heitler (1934). A theory not depending on the first Born approxima-
tion, but making use of a high-energy approximation, was later
developed by Davies, Bethe, and Maximon (1954) and Olsen (1955).
This DBMO theory assumes that the electron energies before and.
after the bremsstrahlung event are large compared to the electron
rest energy, and gives accurate results for electron energies greater
than - 50 MeV. Both the Bethe-Heitler and the DBMO theories take
into account the effect of the screening of the nuclear charge by the
atomic electrons. The cross section formulas are complicated; they
have been systematically reviewed, together with similar results of
other authors, by Koch and Motz (1959).
More recently, a phase-shift analysis method has been developed
by Tseng and Pratt (1971)for calculating the bremsstrahlung cross
section without use of the first Born approximation, through the
numerical solution of the Dirac equation for the case of a screened
Coulomb potential. Pratt et al. (1977) have presented extensive
tables of bremsstrahlung cross sections, as functions of the energy
 A.2 CROSS SECTIONS FOR CHARGED PARTICLES 1 123

1.0 MoV

. -
.r

6 1.I
MeV 2 MeV

EMISSION ANGLE, cos 8

Fig.A.16. Doubly differential m s s sections for delta-ray ejection from molecules

by protons as a function of emission angle. Cross sections are given per valence
electron, i.e., K-shell electrons are not counted. The energy above each figure is the
proton energy. The energies of the delta-ray electrons are given at the right of each
figure. Dashed lines in figuresC and D are drawn merely to aid the eye. (After Wilson
and Toburen, 1980.)
124 / APPENDIXA
of the emitted photon, for incident electrons with energies from 1
keV to 2 MeV in elemental substances with atomic numbers Z from
2 to 92. These results were based on elaborate numerical computa-
tions of the cross sections for key elements and energies, and exten-
sive interpolation.
The explicit dependence of the bremsstrahlung cross section on
atomic number, electron energy and photon energy can be greatly
k du
reduced by introducing the scaled cross section pL, where P is
Z2dk
the velocity of the incident electron (in units of the light velocity), k
is the energy of the emitted photon and duldk is the cross section
differential with respect to k. The scaled cross section is shown in
Figure A.17 for 50-keV and 50-MeV electrons in carbon and gold, as
a function of the ratio, klT, of the photon energy to the incident
electron kinetic energy. The results a t 50 keV are from Pratt et al.
(1977), and the results a t 50 MeV are from DBMO theory. It can be
seen that the residual dependence of the scaled cross section on T
and Z is indeed small.
With use of the accurate cross sections available below 2 MeV
and above 50 MeV as anchor points, it is possible to make reliable
interpolations to obtain cross sections a t intermediate energies (Selt-
zer and Berger, 1985). The reliability of such interpolations can be
confirmed by comparisons with directly calculated cross sections
given by Tseng et al. (1979)for 5- and 10-MeVelectrons in aluminum
and uranium.
Of particular interest for dosimetry calculations is the scaled ener-
gy-loss function

which is proportional to the radiative stopping power discussed in

the next Section. The subscript n is a reminder that Equation (A.32)
pertains to bremsstrahlung in the field of the atomicnucleus. Figure
A.18 shows the scaled energy-loss function in six materials, includ-
ing the directly calculated values below 2 MeV and above 50 MeV
and the interpolated curves a t intermediate energies.
The bremsstrahlung cross section, - d2u differential in the energy
dkd0'
and direction of the emitted photon, was calculated by Tseng et al.
(1979), and these results were used by Kissel et al. (1983) to prepare
a tabulation of the doubly-differential cross section for selected val-
ues of atomic number between 2 and 92, and selected energies of the
incident electron between 1 and 500 keV.
 A.2 CROSS SECTIONS FOR CHARGED PARTICLES 1 125

0 0.2 0.4 0.6 0.8 1.O

RATIO, k/T

Fig. A.17. Scaled differential cross section for bremsstrahlung production in the
field of the atomic nucleus as a function of the ratio of the photon energy, k, to the
kinetic energy of the incident electron, T, in carbon and gold. The scaling factor is
described in the text. The curves for 50 keV have been plotted from the tables of Pratt
et al. (1977). The result. for 50 MeV have been calculated according to the theory of
Davies et al. (1954) and Olsen (1955).

The total cross section, including bremsstrahlung not only in the

field of the atomic nucleus but also in the field of the Z atomic
electrons, would be proportional to Z2 + Z if the screening of the
atomic electrons were the same as the screening of the nuclear
charge. Actually the cross section is proportional to 22 + q&, where
 2'47

T.MeV T,MeV
Fig. A.18. Scaled radiative energy-low cross section, q5&,, for bremsstrahlung in
the field of the atomic nucleus as a function of electron energy, T. Points below 2 MeV
are from the calculationsof Ratt et al. (1977), and pointa above 50 MeV are from the
highenergy theories of Davies et al. (1954)and Oleen (1955).Curves are from a least-
squares fit to the theoretical points. ( f i r ICRU. 1984.)

q~ is a variable quantity of order unity (analogous to the quantity

qp introduced earlier for pair production by photons). Values of q~
applicable to the scaled energy-loss function are shown in Figure
A.19.

A.2.6 Continuous Slowing-Down Approzimution

When charged particles travel through an extended medium, they
gradually lose their energy as the result of many thousands or tens
 A.2 CROSS SECTIONS FOR CHARGED PARTICLES 1 127

Fig. A.19. Ratio of the scaled cross sections 4 & , and that represent the
radiative energy losses resulting from the emission of bremsstrahlung in the field of
the atomic electrons and in the field of the atomic nucleus, respectively, as a function
of electron energy, T. The total radiative stopping power is proportional to p k , +
24+
,. (ARer ICRU, 1984.)

of thousands collisions with atomic electrons. In most of the colli-

sions, the energy loss is small (typically less than 100 eV). As dis-
cussed earlier, a proton can lose up to 22110,000, and an electron up
to 1/22, of its kinetic energy in a single collision.
Some transport phenomena depend critically on the stochastic
fluctuations between successive energy losses. There are many prob-
lems in dosimetry, however, for which the continuous slowingdown
approximation can provide s d c i e n t l y accurate answers a t greatly
reduced computational cost (see Section 7.3). In this approximation,
128 / APPENDIXA
the charged particles are assumed to lose their energy continuously
along their tracks, at a rate given by the stopping power S(T). The
stopping power is defined as the mean energy loss of a charged
particle of energy T per unit pathlength. If energy-loss fluctuations
are disregarded, the particle kinetic energy is then a non-stochastic
function of the pathlength traveled.
Suppose that the interaction coefficient (probability per unit path-
length) for a particular process j (e.g., elastic scattering) is &. In
the continuous slowing-down approximation, the probability of an
interaction per unit energy interval is

where S(T) = -dT/ds is the energy loss per unit path length (stop-
ping power). Suppose that the particle had an interaction of type j
when it had energy To.The probability pj(T)dTthat the next interac-
tion of this type will occur when the particle energy is between T
and T + dT is then given by
dT'
pj(T)dT = exp dT (A.34)

The error introduced by the continuous slowing-down approxima-

tion can be reduced by limiting the approximation to energy losses
smaller than a preassigned value, say, W,.In other words, all energy
losses smaller than W,are lumped together into a restricted stopping
power S,(T),and energy losses greater than W,are treatedstochastic-
ally, taking into account energy-loss fluctuations.

A.2.7 Stopping Power

The stopping power of a medium for a charged particle denotes the
average energy loss per unit pathlength that the particle suffers
while traveling through the medium. There are actually several
kinds of stopping powers corresponding to major modes of energy
loss. The collision stopping power takes into account energy losses
associated with the excitation and ionization of atoms or molecules.
The radiative stopping power takes into account energy losses associ-
ated with the emission of bremsstrahlungphotons. The nuclear recoil
stopping power (often referred to as nuclear stopping power) takes
into account energy losses associated with the transfer of kinetic
energy to recoiling atomic nuclei. The collision stopping power is
important for electrons as well as for heavy charged particles. The
radiative stopping power is important for electrons only, and is large
 A.2 CROSS SECTIONS FOR CHARGED PARTICLES 1 129
a t relativistic energies and for media of high atomic number. The
nuclear recoil stopping power is important mainly for alpha particles
and heavy ions a t low energies. A list of stopping-power compilations
is given in Table A.7.
-
At energies above 10 keV for electrons and 1-2 MeV per atomic
mass unit for heavy particles, the collision stopping power can be
evaluated from Bethe's stopping-power theory (with some correc-
tions), and experimental data are needed only to specify certain
parameters in the stopping-power formula. At lower energies, the
role of theory is reduced, and one must rely on experimental stopping-
power information.
Let S,,(T) denote the collision stopping power as a function of the
1
kinetic energy T of the charged particle, and let -S,,,(T) denote the
0
mass collision stopping power, where p is the density of the medium.
In terms of the cross section duldW for energy transfer to atomic
electrons in inelastic collisions, the mass collision stopping power is
given by

Here, W- is the maximum possible energy transfer [equal to TI2

for electrons, specified by Equation (A.30) for protons]. The factor
Na(Z/MA) represents the number of electrons per gram of the
medium. Na = 6.0221367 x loz3mol-' is Avodagro's number, Z is
the atomic number and M A is the atomic mass of the target atom.
Summing over the probabilities for all possible atomic excitations
and ionizations, Bethe (1930, 1932) evaluated Equation (A.35) in a
quantum-mechanical calculation carried out in the first Born approx-
imation. In this calculation, the binding of the atomic electrons was
taken into account for &stant collisions (involving small energy
transfers). Close collisions (involving energy transfers large com-
pared with atomic binding energies) were treated as if the atomic
electrons were free. The resulting collision stopping-power formula
for electrons is

The corresponding formula for heavy particles with charge ze is

130 1 APPENDIX A
TABLEA.7-Stopping-power a d mnge tables
A. Electrons
Berger and Seltzer (1964) 10 keV to 1000 MeV; 21 elements and 20
compounds; based on Bethe theory; radiative
stopping power also given
Pages et al. (1972) 10 keV to 100 MeV; 59 elements and 57
compounds; based on Bethe theory; radiative
stopping power also given
ICRU (1984) 10 keV to 1000 MeV; 25 elements and 46
compounds; based on Bethe theory; radiative
etopping power also given
B. Protons
Barkas and Berger (1964) 2 to 5000 MeV; universal tables of stopping
power and range as functions of mean
excitation energy; also data for 15 elements
and 22 compounds; data for muons and pions
also given; based on Bethe theory
Janni (1966) 100 keV to 1000 MeV; 46 elements'and 28
compounds; based on Bethe theory
Anderson and Ziegler (1977) 1 keV to 100 MeV; all elements, Z = 1to 92;
based on experimental data
C. A l p h Particles
Ziegler (1977) 1 keV to 100 MeV; all elements, Z = 1to 92;
based on experimental data; theoretical nuclear
stopping power also given
D. Heay Ions
Northcliffe and Schilling (1969) 0.0125 to 12 MeVIamu; representative ions,
1 103; 21 elements and 3 compounds; b e d
on experimental data
Oldenbwg and Booz (1972) 100 eV to 6 MeV; 10 elements and 5
wmpounds; protons, deuterons, alpha particles,
12C, 14N, 160,19F and 32Srecoil nuclei; baaed on
theories of Bethe and of Lindhard, Scharff and
Schiott; theoretical nuclear stopping power also
given
Ziegler (1980) 0.2 to 2000 MeVIamu; all elements, 47
representative ions; based on Ziegler's version
of Lindhard theory; theoretical nuclear
stopping power also given
Hubert et al. (1980) 2.5 to 100 MeVIamu; all ions, Z = 1 to 103; 19
elements (solids); based on experimental data
 A.2 CROSS SECTIONS FOR CHARGED PARTICLES 1 131
where p is the particle velocity in units of the light velocity, I is the
mean excitation energy of the medium, 6 represents the density-
effect correction, CIZ the shell correction, andzL, andz2L, are correc-
tions that take into account departures from the first Born approxi-
mation. The four,correction terms represent refinements of the origi-
nal Bethe theory. The quantity in the curly brackets on the right-
hand side of Equation A.37 is called the stopping number, L. Note
that

2nN,<rnc2 = 0.153536 MeV cm2mol-'. (A.38)

The mean excitation energy, I , is a material constant and represents

the stopping characteristics of the medium. It depends on the elec-
tronic structure of the medium, especially the arrangement of the
valence electrons. Therefore, the value of Z is sensitive to chemical
binding effects and to t h e physical s t a t e of aggregation of the
medium. Accurate ab-initio calculations of Z have been made so far
only for simple atomic gases. Generally, values of I having high
accuracy can be obtained only with use of experimental information.
The most commonly used method is to extract values of Z from stop-
ping-power and range measurements (usually for protons or alpha
particles), using the stopping-power Equation (A.37), which requires
the best possible estimate of the various correction terms in that
equation. An alternative method coming into increasing use is to
calculate I from dipole oscillator-strength distributions for gases or
from dielectric response functions for solids. This method involves
the use of photon-interaction cross sections to determine the mean
excitation energy, because the dipole oscillator-strength distribution
is proportional to the photoelectric cross section and the dielectric
response function can be obtained in terms of the complex index of
refraction.
A recent analysis of experimental information on values of I ,
including stopping-power, range, oscillator-strength and dielectric-
response data, can be found in a report of the ICRU (19841, which
also presents recommended values of Z for all the elements and for
many compounds. The estimated uncertainties of the recommended
values of I vary from 2 to 15 per cent, depending on the available
experimental information. The associated uncertainties of the colli-
sion stopping powers are several times smaller. The recommended
values of Z for elements from ICRU (1984) are shown in Figure A.20
in the form of a plot of the ratio ZIZ as a function of atomic number
Z. It can be seen, that for Z z 20, this ratio is approximately 10 eV
but varies irregularly with Z. This irregular Z-dependence, which is
also predicted theoretically, is due to the effects of atomic shell
132 / APPENDIXA

22;;

20 -
t @
4
4
GASES.
GASES,
SOLIDS.
EXPERIMENTAL
INTERPOLATED
EXPERIMENTAL

P -SOLIDS. INTERPOLATED

18-

0 10 20 30 40 50 60 70 80 90 00
ATOMIC NUMBER, Z

Fig. A.20. Dependence of the mean excitation energy, I, for elemental substances
on the atomic number 2.I is roughly proportional to 2.To indicate the residual, non-
smooth, dependence on Z in greater detail, the ratio IlZ is plotted as a function of Z.
(From ICRU, 1984.)

structure. For Z < 20, the irregularity of the 2-dependence is even

more marked.
The density-effect correction, 6 in Equations (A.36) and (A.37),
takes into account the reduction of the collision stopping power due
to the polarization of the medium by the incident particle. The correc-
tion depends on the dielectric properties of the medium. In the stan-
dard treatment of Sternheimer (19521, these properties are repre-
sented by the simple binding energies of the medium as input param-
eters. Values of the density-effect correction evaluated with up-to-
date parameters can be found in ICRU (1984) and in Sternheimer
et al. (1982). The collision stopping power of water for electrons,
calculated according to Equation (A.36), is shown in Figure A.21.
The density-effect correction is appreciable in this case, and reduces
the collision stopping power by 1.2 per cent at 1MeV,11.5per cent
a t 10 MeV, 21.5 per cent a t 100 MeV, and 29.3 per cent a t 1000MeV.
 A.2 CROSS SECTIONS FOR CHARGED PARTICLES 1 133

ELECTRON KINETIC ENERGY, MeV

Fig. A.21. Electron stopping power in water. Dashed curve: mass collision stop-
ping power; dotted curve: mass radiative stopping power; solid curve: total mass
stopping power (collision plus radiative).

Figure (A.21) also gives the mass radiative stopping power ofwater
for electrons, calculated f r ~ m
the expression

where 4,,, is the scaled energy-loss function from bremsstrahlung

in the field of the atomic nucleus given by Equation (A.32), and the
term I~$Zrepresents the contribution of bremsstrahlung in the field
of the atomic electrons (see Figure A.19). Table A.8, calculated from
Equations (A.36) and (A.39) gives-for water and selected other
materials-the electron kinetic energies T, and T,,a t which the
contribution of Srad to the total stopping power, S,, + Smdramounts
to 5 per cent and 50 per cent, respectively.
134 / APPENDIXA
TABLEA.8--Energies T,and T, at which the mdiutive stopping power is equal to
5 percent and 50 percent, mspectively, of the total stopping power.
Element Z TK Tso
MeV MeV
H 1 23.8 382
C 6 6.44 95.5
A1 13 3.40 50.8
cu 29 1.48 24.3
Ag 47 0.789 16.1
W 73 0.352 10.6
U 92 0.233 8.95

Bethe's derivation of the stopping power formula was based on

the assumption that the velocity of the incident particle is large
compared to the velocities of the atomic electrons, an assumption
which is often not satisfied, particularly for K-shell electrons. The
term CIZ in Equation (A.37) is a correction that compensates for this
approximation. It is called a shell correction because it was first
calculated separately for collision with K-shell electrons (Livingston
and Bethe, 1937; Walske, 1952)and with Lshell electrons (Walske,
1956). Another theoretical treatment of shell corrections, based on
Lindhard's statistical model, was developed by Bonderup (1967).
The theoretical shell corrections are only approximate and are often
supplemented by semi-empirical shell corrections derived from the
analysis of stopping-power data (Andersen and Ziegler, 1977; ICRU,
1984). The terms zL,, and z2L, in Equation (A.37) are small correc-
tions, important mainly at low energies, that represent departures
from first Born approximation theory. The correction zL,,has been
calculated by Ashley, Ritchie, and Brandt (1972,1973) who provide
tables of their numerical results. It should be noted that the appear-
ance of odd powers of z in this correction will result in stopping-
power data differences between particles that have the same mass
and charges equal in absolute value but opposite in sign. The correc-
tion z2L, is contained in a stopping-power formula of Bloch (1933)
derived by an impact-parameter method and without use of the first
Born approximation.
The derivation of the shell corrections, and the zL1and 2L2correc-
tions, have been based on the assumption that the incident particle
can be considered equivalent to a perturbing potential whose center
moves with constant velocity. This assumption, while satisfied for
protons down to rather low velocities, is much less applicable to
electrons, so that there appears to be no sound basis for extending
 A.2 CROSS SECTIONS FOR CHARGED PARTICLES 1 135
the corrections to electrons. By way of example, the relative im-
portance of the various corrections for the proton stopping power in
gold is indicated in Figure A.22, which shows a plot of the ratios
(CIZ)IL, (zL,IL), (z2L,)IL and (6/2)lL, where L is the stopping number
(the quantity in curly brackets on the right-hand side of Equation
(A.37).
Electron collision stopping powers a t energies below 10 keV can
be obtained through the use of experimental values of the energy-
transfer cross section in Equation (A.35), or through the use of semi-
empirical dielectric response functions (see, e.g., Ashley, 1982). A
survey of low-energy electron stopping powers obtained by various
authors for selected materials is given in ICRU (1984).
Collision stopping powers for protons in all elements have been
tabulated by Anderson and Ziegler (1977), mainly on an experimen-
tal basis. These authors took into account all available proton stop-
ping-power measurements, smoothed the data as functions of proton
energy and atomic number, and fitted them by simple formulas. In
the energy region 1keV to 1MeV, these formulas are strictly empiri-

PROTON ENERGY, MeV

Fig. A92. Relative corrections to the stopping power, L [curly brackets in Equa-
tion (k37)], for gold as a function of proton energy. The quantity CIZ represents the
shell correction and &'2, the density-effect correction. The Barkas correction,zL,, and
the Bloch correction, SL,, account for departures from the f i s t Born approximation.
(After ICRU, 1984.)
136 / APPENDIX A
cal. Above 1 MeV, they have the form prescribed by the Bethe theory,
and contain empirical shell corrections. Similar tabulations of colli-
sion stopping powers for alpha particles in all elements have been
published by Ziegler (1977).
Proton and alpha-particle stopping powers have been measured
for many compounds. Nevertheless, there are many materials of
interest in dosimetry for which no experimental data are available.
For these, it is necessary to invoke Bragg's additivity rule. According
to this rule, the mass collision stopping power of a compound can be
approximated by the weighted sum of the mass collision stopping
powers of the atomic constituents; the weighting factors are the
fractions by weight of the respective constituents. Due to the neglect
of chemical binding effects, the use of the additivity rule introduces
some error, which typically may amount to several percent a t low
energies. The extensive, and somewhat contradictory, literature on
the uncertainties associated with the use of the Bragg additivity rule
has recently been surveyed by Thwaites (1983).It is often possible
to improve the accuracy of the additivity rule by applying it to atomic
constituents in specific chemical environments, e.g., by assigning
different values to the collision stopping power for the atomic constit-
uent oxygen, depending on whether the oxygen in the compound is
singly or doubly bonded. Such distinctions can be developed on the
basis of the available experimental information for compounds.
Thwaites also discusses the sensitivity of the collision stopping power
to the physical state of aggregation of the material. In general, the
collision stopping power of gases is slightly higher than that of
liquids or solids of the same composition, the differences amounting
to a few percent.
To illustrate the energy-dependenceof the collision stopping power
at low energies, Figure A.23 gives results for electrons, protons
and alpha particles in water vapor. The quantity plotted is the
mass collision stopping power divided by the square of the pro-
jectile charge, z2, as a function of the particle velocity v in units of
v, = ct137. The stopping-powercurve for electrons is similar to that
for protons but lies below it a t all velocities, mainly due to differences
in the kinematics of close collisions with atomic electrons. The stop-
ping-power curves for protons and alpha particles coincide for veloci-
ties greater than 2v,, but, a t lower velocities, the alpha-particle curve
lies below the proton curve, due to a difference in effective charge.
When an alpha particle of low velocity travels through a medium, it
does not remain fully ionized but occasionally picks up and then
again loses an atomic electron. Therefore, the nominal charge ze in
the stopping-power formula must be replaced by an effective charge
ze* < ze. For protons in gases, there is a similar, but much smaller,
difference between nominal and effective charge.
 A.2 CROSS SECTIONS FOR CHARGED PARTICLES 1 137

PROTONS
a d r a , a'
a I Lev

ELECTRONS

Fig. A.23. Mass collision stopping power divided by z2 for electrons, protons, and
alpha particles in water vapor as functions of particle velocity expressed as the ratio
of the particle velocity, o, to the Bohr velocity, v, = ~1137.Particle-energy scales are
shown below the figure. ( f i r Berger, 1978.)

The heavier the ion, the greater is the reduction of the stopping
power due to the replacement of the nominal by the effective charge.
This is illustrated in Figure A.24 which gives the collision stopping
powers (divided by z2) of various ions in aluminum, plotted as func-
tions of the kinetic energy per atomic mass unit. The determination
or prediction of the effective charge is a difficult task. The extensive
literature on this subject has been reviewed by Betz (1972), who
suggests an empirical formula
z*
-= 1 - C exp [ - u/(u,,z~)l, (A.40)
Z
138 / APPENDIXA

,(jS 104 ,,53 10-z

I K) 10' 10' D~ lo5
RELATIVE PARTICLE ENERGY, MeV/amu
Fig. A24 Mass collision stopping powers divided by z2for heavy ions in aluminum
as a function of the relative particle energy in atomic mass units. The solid curves
represent the sum of the electronic and nuclear stoppingpowers and the dashed curvea
represent only the electronic stopping powers. The curves are based partly on theory
and partly on experiment. (After Northcliffe and Schilling, 1970.)

with two parameters, C and y to be determined empirically for each

ion-target atom combination. For gaseous targets, the values C = 1
and y = 213 are reasonable estimates in many cases.
The d i d curves in Figure A.24 are fits to experimental data, and
the dashed curves are estimated values of the collision stopping
power evaluated according to the theory of Lindhard, Scharff, and
Schipt (1963) that predicts a Pa-dependence of the collision stop-
ping power a t low energies. The difference between the solid and
dashed curves represents the nuclear recoil stopping power S,,,(T).
The nuclear recoil stopping power for heavy charged particles can
be calculated by a classical treatment of the collision between an
incident ion and the target atom, assuming an appropriate screened
Coulomb interaction. The results of such calculations (averaged over
many ion-atom combinations) can be expressed in scaled form as,
 A.3 NEUTRON CROSS SECTIONS 1 139
in terms of a scaled energy variable

and a scaled nuclear-recoil energy-loss function f(4. In these equa-

tions, ze and Ze are the charges, and M and M,the masses, of the
incident ion and the target atom, respectively. Ziegler (1980), in his
evaluation of heavy-ion stopping powers, has adopted the following
scaled energy-loss function,

which has been given by Wilson et al. (1977). This formula is a fit to
numerical results obtained with a krypton-carbon potential, and is
stated to give results in good agreement with experimental stopping-
power data for many other ion-atom combinations.

A.3 Neutron Cross Sections

A.3.1 Classification of Interactions

The various interactions of interest for transport calculations can
be classified as follows:
1) Elastic scattering, (n,n);
2) Inelastic scattering, (n,nl);
3) Inelastic scattering accompanied by the emission of charged
particles, (n,nlp), (n,nla);
4) Inelastic scattering resulting in fragmentation of the target
nucleus, e.g., the reactions 7Li(n,n')at, 'jLi(n,nl)ad,
12C(n,n')3a;
5) (n,2n) reaction;
6 ) Reactions from which no neutron emerges, (n,p), (n,d), (n,t),
(n,a), (n,2p), (n,2a), (n,pa);
7) Radiative neutron capture, (n,y).
Elastic scattering is the result of two processes. The first of these
is direct (or potential) scattering due to the reflection of the neutron
wave from the sudden change of potential a t the nuclear surface.
This process predominates a t intermediate and low energies, and its
cross section is approximately four times the geometric cross section.
The second process is resonance (capture) scattering, and involves
the capture of the neutron by the target nucleus, the formation of a
140 1 APPENDIXA
compound nucleus, and the re-emission of the neutron. This is a
resonance process that is highly favored when the sum of the binding
and kinetic energies of the neutron is equal to the energy difference
between an excited and the ground state of the nucleus. The cross
section for elastic scattering has anenergy dependence that is smooth
when potential scattering predominates, and has an irregular char-
acter when resonance scattering predominates.
In an elastic-scattering event, the nucleus is left in the ground
state, and an amount of energy W is transferred from the neutron to
the recoiling nucleus. The relation between the energy transfer and
the center-of-mass deflection angle OCM of the neutron is

where Enis the neutron kinetic energy, and A is the ratio of the mass
of the target nucleus to the mass of the neutron. The largest possible
energy transfer is

In a collision with hydrogen (A = I), a neutron can lose its entire

kinetic energy in one elastic collision. The deflection angle B in the
laboratory system is given by

a s ecM
cos 8 =
(1 + 2A cos OCM +A2)liz'
and is always smaller than the center-of-mass system deflection
angle Om. For hydrogen, 6 = ocM/2
The angular deflection in the center-of-mass system, at energies
away from resonances, is almost isotropic until the neutron energy
is greater than -10 A-213MeV. For isotropic scattering,
- 2
cos eCM= 0, and cos 0 = -
3A '
(A.47)

Experimental cross sections for elastic neutron scattering from 1 6 0

at various energies, as summarized by Cullen and Hlavac (1972),
are shown in Figure A.25.
Inelastic scattering is similar to capture elastic scattering, except
that the nucleus is left in an excited state. The neutron loses an
amount of energy W + E*, where E* is the excitation energy and
where W is the kinetic energy given to the recoiling nucleus. The
 A.3 NEUTRON CROSS SECTIONS 1 141

COSINE SCATTERING ANGLE, cos OCM

0
18Z
r
,a'
. . .
ea
. #

ia'
d .

id
. t

10'
. 1
e4 > eS ,a7
I
ENERGY, eV

Fig. A.25. Angular dependence of the cross section for the elastic scattering of
neutrons from 160.Figure A: angular distribution (normalized to unity) as a function
of the cosine of the scattering angle in the center-of-mass system for various neutron
energies. Figure B: average cosine of the scattering a n d e in the center-of-masssystem
as a function of the neutron energy. (After Cullen and Hlavac, 1972.)

relation between the energy transfer and the center-of-mass deflec-

tion angle is

-- E*
1 + A,'
where A, is the ratio of the mass of the recoiling (excited) nucleus to
the neutron mass. The de-excitation of the excited nucleus takes
place mainly via the emission of gamma rays, and the knowledge of
this gamma-ray spectrum is an essential part of the neutron cross-
section information for transport calculations.
The neutron interactions, 2) to 6), listed above are characterized
by sharp thresholds and resonances. By way of example, experimen-
tal cross sections for some of these interactions, with 160as target,
are shown in Figure A.26 as functions of the neutron energy. Figure
A.27 shows the total cross sections for scattering from hydrogen,
carbon, nitrogen and oxygen and from a 4-component tissue-equiva-
142 / APPENDIXA

I 1
," 80. -

i -
!2
ELASTIC

--
--_- - - --__
-
YI. '
L . .A
ma ID, D. 10' 1C 10' 10'

ENERGY. eV

I@' THRESMOLO
2.W MeV

10.

-.X
THF€SHOLD
8 lo' MH

@
,.
E

,,
Z Y
B8 10'

I@' - THREWOLO
I O M MeV

10' -

5 8 7 8 9 1 0

ENERGY, MeV

Fig. A26. Illustration of cross sections for various neutron interactions with '
'
0
as a function of neutron energy. Note change in energy scales between upper and
lower figures. (ARer Cullen and Hlavac, 1972.)
 A.3 NEUTRON CROSS SECTIONS 1 143

NEUTRON ENERGY, MeV

Fig. A.27. Total neutron macroscopic cross sections for a specified tissue and its
constituents as a function of neutron energy. The assumed percentage composition by
weight is 10.2 H, 12.3 C, 3.5 N and 74.0 0. The macroscopic cross section is the
equivalent of a linear attenuation coefficient for gamma rays (see text). (After Chen
and Chilton, 1979.)

lent material containing these elements. All of these figures show

an irregular energy-dependence due to resonances.
The radiative capture of neutrons is the predominant process for
energy transfer to the medium a t low energies (a few eV and lower),
resulting in the subsequent emission of energetic gamma rays from
the target nucleus (e.g., the 2.225-MeV gamma ray emitted after
neutron capture in hydrogen). However, for tissue or tissue-like
materials, the 14N(n,p)reaction transfers more energy than the
gamma rays.

A.3.2 Data Compilations

As indicated in the preceding discussion, the cross sections for
neutron interactions are more complex than those for photons or
charged particles. The number of processes that must be considered
is greater than for the other radiations. Whereas photon or charged-
particle cross sections are, with some exceptions, smooth functions
of the particle energy, the neutron cross sections tend to be irregular
and characterized by resonances and thresholds. These features
depend sensitively on the structure of the target nucleus, and inter-
polation of cross sections with respect to atomic number is often
difficult.
For photons and charged particles, the theoretical evaluation of
cross sections plays an important and often dominant role, and exper-
imental data are often used mainly to provide confirmation of theo-
retical results. For neutrons, the situation is reversed. Measure-
ments constitute the principal source of cross-section data, and the
theory mainly plays the role of providing sophisticated tools for the
interpolation, extrapolation, and checking of experimental data. One
of the important theoretical tools is the optical model in which the
scattering of neutrons is calculated with a complex potential. The
real part of the potential takes into account elastic scattering, and
the imaginary part inelastic scattering. By parameterizing the scat-
tering process in terms of the real and imaginary parts of the poten-
tial, one can more easily interpolate the cross sections with respect
to neutron energy. An important class of theoretical tools consists of
statistical models that describe the formation of a compound nucleus
and the subsequent emission of particles. Another part of cross-
section theory deals with direct interactions in which particles are
emitted from the nucleus prior to the formation of the compound
nucleus.
In response to the requirements of nuclear technology, there has
been, during the past forty years, and continues to be broad activity
in the area of neutron cross-section evaluation. Extensive biblio-
graphic references to measurements, calculations, reviews and eval-
uations of neutron cross sections can be found in CINDA (1986a,
1986b), a computer index of neutron data prepared as a cooperative
effort of the U.S. National Nuclear Data Center (NNDC) a t the
Brookhaven National Laboratory, the USSR Nuclear Data Center
(Obninsk), the NEA Data Bank (Saclay, France), and the IAEA
Nuclear Data Section (Vienna). A source list of nuclear data bibliog-
raphies, compilations and evaluations can be found in Burrows and
Holden (1978).
The U.S. National Nuclear Data Center (NNDC) maintains an
Evaluated Nuclear Data File (ENDF) that contains information on
neutron cross sections up to 20 MeV, on related nuclear constants,
and on photon-production and -interaction cross sections. The file
ENDF consists of two parts. ENDFIA contains evaluated data sets
(often more than one per material) that are still in preliminary form.
ENDFB contains evaluated data sets (only one per material) that
are in final form and have been recommended by the Cross Section
Evaluation Working Group (CSEWG), a group of scientists from
 A.3 N E m O N CROSS SECTIONS 1 145
many laboratories whose work is coordinated by NNDC. As new data
and evaluations become available, new versions of ENDFIA and
ENDFA3 are issued. The latest version released is ENDF/B V, which
is documented in a report by Kinsey (1979). Work on ENDF/B VI is
in progress. Evaluation methods and procedures currently used a t
NNDC are described in the proceedings of a recent conference
(Magurno and Pearlstein, 1981). Various publications of NNDC pro-
vide information on neutron cross sections in tabular or graphical
form. Particularly important are successive editions of Brookhaven
National Laboratory Report BNL 325 (see, e.g., Garber and Kinsey,
1976; Mughabghab, Divadeenam, and Holden, 1981). The complete
body of data in ENDFA3 is contained in a magnetic-tape library for
computer use.
ENDFB neutron cross section data are widely used as input for
transport calculations, both in the U.S.and elsewhere. In many
cases, the data are first used in processing codes (see, e.g., RSIC,
1981) that put the data in the form required by various transport
computer programs. The ENDFB cross sections are given for a grid
of discrete energies. Many of the processing codes convert such data
into cross sections averaged over specified energy bins.
For neutron dosimetry in biology and medicine, the coverage in
energy above 20 MeV provided by ENDFA3 is not sufficient, and
workers in this field must make their own data analyses a t energies
that are of interest in radiation therapy and protection. The efforts
of many groups to extend the evaluation of neutron cross sections up
to 50 MeV are described in a symposium sponsored by NNDC (Bhat
and Pearlstein, 1980).

A.33 Kerma Factors

The transfer of energy from neutrons to secondary charged parti-
cles can be calculated as a n integral over the product of the neutron
fluence and a kerma factor. Let K denote the kerma, i.e., the kinetic
energy released to matter per unit mass. The kerma a t position r in
the medium is given by

where K r J , ) is the neutron fluence (differential in neutron energy

En)a t Ti, and where kXEn) is the kerma factor. The latter can be
evaluated from the expression
146 / APPENDIX A
where p is the total mass interaction coefficients for neutrons, and
where WEJ is the average amount of energy given to charged parti-
cles in a single interaction, the average being taken over all possible
modes on interaction.
One of the most extensive evaluations of kerma factors is that of
Caswell et al. (1980), which is given in tabular form for 19 elements
and 15 compounds, from thermal energies up to 30 MeV. The total
neutron-interaction cross sections for the various interactions were
taken from the Evaluated Nuclear Data File (ENDF/B IV up to 20
MeV), and horn new data analyses by Caswell et al. a t higher energ-
ies. Even below 20 MeV, the required input data were not always
complete. When the h a 1 state of excitation of the nucleus was not
known, all levels were assumed to be equally likely to be excited. For
interactions other than elastic and inelastic scattering, the angular
distribution of reaction products was, in general, not known, and
the assumption was made that the reaction products are emitted
isotropically in the center-of-mass system.
Kerma factors for carbon, nitrogen and oxygen are shown in Figure
A.28 as functions of the neutron energy. Because of cross-section
resonances and thresholds, kkE,) is a highly irregular function of
En.To minimize these irregularities, Caswell et al. tabulated the
kerma factors as averages over 116contiguous energy bins extending
from 0.026 eV to 30 MeV. The kerma factors for hydrogen are esti-
mated to be accurate to 1 to 2 percent, and those for the other
elements, a t energies below 20 MeV, to about 20 percent. Above 20
MeV, the kerma factors are stated to be only rough estimates.
In Table A.9, further details of the work of Caswell et al. are given
pertaining to kerma factors for a four-element tissue-equivalent
material. This table shows, for various neutron energies, the percent-
age of the kinetic energy that is transferred to photons, alpha parti-
cles, recoiling carbon, nitrogen or oxygen nuclei, and electron-posi-
tron pairs. Also given are the percentage contributions to the energy
transfer caused by neutron interactions. The table clearly demon-
strates the dominant role played by elastic scattering from hydrogen.
Further work on neutron cross sections in the energy region 20 to
50 MeV has been done by Dimbylow (1980,1981,1982) for elements
that are constituents of tissue. Dimbylow used optical-model fits of
experimental total cross sections to estimate elastic and non-elastic
cross sections, and used the latter to normalize the individual reac-
tion cross sections estimated according to a statistical evaporation
model. He also used corrections to include the emission of particles
in the pre-equilibrium phase of the compound nucleus, and took into
account the possibility of direct excitation, in (n,n') scattering, of
 A.4 NUCLEAR CROSS SECTIONS / 147

NEUTRON ENERGY, MeV

Fig. A.28. Kerma fadors as a function of neutron energy for carbon, nitrogen, and
oxygen. The factors have not been averaged over energy bins. (From ICRU, 1977.)

low-lying levels in C, N, and 0 nuclei before formation of a compound

nucleus.
Figure A.29 shows kerma factors calculated by Dimbylow (1982) for
C, N, 0, and tissueequivalent material, and compares these fadors
with results of his earlier calculations and those of other authors.

A.4 Nuclear Cross Sections for Charged Particles at High

Energies
A.4.1 Interactions of Pions below 100 MeV
When pions travel through an extended medium, they undergo
numerous Coulomb collisions with atoms that result in a gradual
 TABLEA.9-Energy transfer for ICRU tissue" by neutronsb.
\
Percentage of Kerrna
Particle Fleaction Neutron energy
MeV %
Cd
0.5 5.0 9.0 14.5 19.5 5
Protons

a particles

Carbon recoils 12C(n,n)'2C

160(n,a)L3C
160(n,n'a)12C
Nitrogen recoils 160(n,n'p)l6N 0 0 0 0.1 1.5
Oxygen recoils 160(n,n')160
160(n,n')w0
e;e+ 'W(n,n')160
Miscellaneous 0.4 0.9 1.6 2.4 2.4
aAssumedtissue composition: 10.2%H, 12.396 C, 3.5% N and 74.0%0 by weight.
bFromCaswell et al. (1980).
 A.4 NUCLEAR CROSS SECTIONS / 149
--
10
--
8-
-
1
' 2~
I I I 1 I

ALSM lLLER '

6 -

a
LL

-
Y
BEHROOZ

I I I I I I
10 20 30 40 50 60 70
NEUTRON ENERGY. MeV

10
- I
16
I ! I I I --
-
8- 0 -
i
6 -
DIMBYLOW
-
-
-

I I I
10 20 30 40 50 60 70
NEUTRON ENERGY, MeV

Fig. A B . Kerma factors as a function of neutron energy for carbon and oxygen.
In comparison to Figure A.28, this figure extends to higher neutron energies and
includes the results of several investigators. ( k o m Cross, 1986.)
150 / APPENDIX A
energy loss, and also undergo occasional nuclear interactions that
can result in scattering or absorption. In the continuous slowing-
down approximation, the probability, p(T)dT, that a pion with initial
kinetic energy Towill undergo its first nuclear interaction when its
energy is between T and T + dT is given by

where ~(2') is the energy-dependent nuclear interaction coefficient

and S(2') is the pion stopping power.
When negative pions are at the end of their range and have slowed
down to rest, they are captured into the Bohrorbits ofatoms and react
with the nucleus. The time required for this capture and reaction to
occur is several orders of magnitude shorter than the natural decay
time of the pion. When the negative pion is absorbed by the nucleus,
its rest energy (139.6 MeV) is transferred to the constituents of the
nucleus, and the nucleus explodesinto a "star" of charged and neutral
secondary particles. The high local density of energy deposition by
secondary particles from these stars is considered potentially useful
in radiation therapy.
Cross sections for pion-induced nuclear interactions have been
described in papers dealing with the calculation of the spatial pattern
of absorbed dose in water (Turner et al. 1972; Armstrong and Chan-
dler, 1972; Wright et al. 1979) and in tissue-equivalent material
(Buche and Przybilla, 1981) for pions with energies below 100 MeV
(typical of therapeutic pion beams). The cross sections given by Buche
and Przybilla are largely based on experimental data, and are partic-
ularly complete for interactions with 12C. For pions in flight, the
following interactions are considered:
1) Elastic scattering, 12C(m-,T-) 12C;
2) Inelastic scattering, l2C(=-,~-)12C*;
3) Quasielastic proton scattering, 12C(lr-,n-p) "B;
4) Quasielastic neutron scattering, 12C(n-,a-n) llC;
5) Charge exchange scattering, 12C(~-,a"n) "B;
6) Nuclear absorption, T- + 12C+ "star'' of secondary particles.
Illustrative results for total and partial cross sections for pion inter-
actions with 12Cand 160are shown in Figures A.30 and A.31.
The spectra of the secondary particles emitted subsequent to the
capture and absorption of negative pions at rest have been measured
by Mechtershimer et al. (1979)for interactions with 12C,and by Klein
et al. (1979) for interactions with 12C,14N,and 160The . results for
12C,as compiled by Biiche and Przybilla (1981), are given in Table
 A.4 NUCLEAR CROSS SECTIONS 1 151

I I
lb 2b 3b LO Sb 60 70 80 90 100 110 120
ENERGY, MeV

Fig. A.30. Total and partial m a s sections for the interaction of negative pions
with 12C and 160as a function of the pion kinetic energy. ( f i r Biiche and F'rzybilla,
1981.)
152 / APPENDIX A

rc' . "c, '60

ENERGY, MeV
. 1 Cross sections for elastic scattering of pions from I2C and 160aa a
function of pion kinetic energy, evaluated from experiments (data points) and deduced
for "j0by partial wave analysis (curves). (After Biiche and Pnybilla, 1981.)

A.10 and are compared with corresponding calculations of Guthrie

et al. (1968) according to the intranuclear cascade model (to be
described in the next subsection). In a paper by Miinchmeyer et al.
(1982), extensive experimental data can be found for the energy
spectra of protons, deuterons, alpha particles, and other heavy ions
emitted following the absorption of stopped negative pions in oxygen-
containing organic compounds.

TABLEA.lO-Particle emission after a- capture in 12C.

Calculationd Experimentb
Particle Number per Kinetic energy Number per Kinetic energy
type n" capture per ~ rcnpture
- n- capture per n- capture
MeV MeV
neutron 2.79 64.16 2.50 76
proton 1.03 20.14 0.486 10.4
deuteron 0.258 2.82 0.356 6.3
triton 0.080 1.03 0.250 3.0
aHe 0.029 0.383 0.064 0.6
alpha 1.10 16.84 0.767 5.6
Li 0.208 1.22 0.239 1.6
Be, B 0.125 0.30 0.273 0.3
Total 106.9 103.8
.From Guthrie et al. (1968).
bFrorncompilation of experimental data in BDche and Pnybilla (1981).
 A.4 NUCLEAR CROSS SECTIONS 1 153
When pions slow down to rest in a compound rather than in an
elemental substance, one must estimate the relative probabilities of
capture by the various atomic constituents. According to Fermi and
Teller (1947), the probability of capture by the jth constituent is
proportional to n?,, where n, is the number of atoms of type j in the
compound and Zjis the atomic number. Actually, the relative capture
probabilities are influenced by the arrangement of the valence elec-
trons of the compound (Ponomarev, 1973), and the probability of
capture by the constituent hydrogen in a n organic compound is
greatly suppressed. Biiche and Przybilla have adopted the Fermi and
Teller approach described above, but have assumed that there is no
capture by hydrogen.

A.4.2 Nuclear Interactions of Hadrons above 100 MeV

Cross sections for hadron-hadron and hadron-nucleus interactions,
a t energies up to 1000 GeV, are required for dosimetry calculations
pertaining to radiations from high-energy accelerators, such as those
a t Fermilab, SLAC or CERN. An excellent review of these cross
sections and their use in transport calculations can be found in a
report of the ICRU (1978).
In hadron-hadron or hadron-nucleus collisions, a great variety of
secondary particles are produced. However, for dosimetry calcula-
tions it is usually sufficient to consider only protons, neutrons, and
pions because other types of particles are produced in relatively small
numbers. For example, the rate of K-meson production is about ten
times smaller than the rate of pion production.
Total and elastic hadron-hadron collision cross sections, from a
recent compilation by the Particle Data Group (1986), are shown in
Figure A.32 as functions of the momentume,p, of the incident hadron
in the laboratory system. The dependence on s, the square of the
total energy in the center-of-mass system, is also indicated. It can be
seen that the total cross sections are approximately constant for pc
greater than a few GeV, and the elastic cross section for pc greater
than - 100 GeV. Hadron-nucleus cross sections are similarly con-
stant a t high energies. Such high-energy total and elastic cross sec-
tions, given by the Particle Data Group (1986) are shown in Table
A . l l for various materials. These cross sections are proportional to
-
Am, where A is the atomic weight, with m 314 for the total cross
section.

BThe relation between the kinetic energy, T, the momentum,p, and the rest mass,
MC2, is T = (p2C2 + Wc4)* - M S .
154 1 APPENDIX A

k
5
i
0
5 lo'
LU

Fig. A.32. Illustrations of hadronic total and elastic cross sections as a function
of momentum, p,,-, in the laboratory system and of energy squared, s, in the center-
of-mass system. (ARer Particle Data Group, 1986.)
 TABLEA.ll--Crosssections for the interactions of hadrons with nuclei at high energies"
Nucled Nuclear'
total inelastic Nucled Nucled
Atomic Atomic c m cross collision interaction
number weight section ~ection length length
Material Z A UT UI AT A1 -
cm2 x loz4 cm2 x 10% g~ m - ~ gc ~ - ~
Hz 1 1.01 0.0387 0.033 43.3 50.8
D2 1 2.01 0.073 0.061 45.7 54.7
He 2 4.00 0.133 0.102 49.9 65.1
Li 3 6.94 0.211 0.157 54.6 73.4 $
Be 4 9.01 0.268 0.199 55.8 75.2
z
C
Nz
6
7
12.01
14.01
0.331
0.379
0.231
0.265
60.2
61.4
86.3
87.8 s
oz
Ne
A1
8
10
13
16.00
20.18
26.98
0.420
0.507
0.634
0.292
0.347
0.421
63.2
66.1
70.6
91.0
96.6
106.4
E
0
Si
Ar
14
18
28.09
39.95
0.660
0.868
0.440
0.566
70.6
76.4
106.0
117.2 g
U)

V)
Fe 26 55.85 1.120 0.703 82.8 131.9 M
Cu
Sn
29
50
63.54
118.69
1.232
1.967
0.782
1.21
85.6
100.2
134.9
163 8
Xe 54 131.3 2.120 1.29 102.8 169 U)
W 74 183.85 2.767 1.65 110.3 185
Pb 82 207.19 2.960 i .77 116.2 194 b

U 92 238.03 3.378 1.98 117.0 199

 TABLE
A.11-Cross sections for the interactions of hadrons with nuclei at high energies. (continued)
w
Nucled Nucleaf U1
total inelastic Nucled Nucled Q,
Atomic Atomic rr088 cross collision interaction
Material
number
Z
weight
A
section
UT
section
UI
length
h
length
AI
--.
-
Air. 2VC, 1 atm. 62.0 90.0
Hz0 60.1 84.9 %
Shielding concrete 67.4 99.9 2
SiO, (quartz) 67.0 99.2 Z
u
H, (bubble chamber 26°K) 43.3 50.8 r?
D, (bubble chamber 31°K) 45.7 54.7 +
H-Ne mixture (50 mole percent) 65.0 94.5
Ilford emulsion G5 82.0 134
NaI 94.8 152
Bfi2 92.1 146
BGO (Bi,Ge,0,3 97.4 156
Polystyrene, scintillator (CHI 58.4 82.0
Lucite, Plexiglas (C,H,O,) 59.2 83.6
Polyethylene (C&) 56.9 78.8
Mylar (C5H102) 60.2 85.7
Borosilicate glass (F'yrex) 66.2 97.6
co2 62.4 90.5
Methane CH, 54.7 74.0
Isobutane C,H,, 56.3 77.4
Freon 12 ( C C l p 3 gas, 26"C, 1atm. 70.6 106
Silica Aerogel 65.5 95.7
GI0 plate 62.6 90.2
Qrom Particle Data Group (1986). (Information from the Particle Data Group is updaied periodically.)
a t 804240 GeV for neutrons (==a for protons).
Cainal.ade
-
- awl - uslh - aPwi =; for neutmns at 60-375 GeV.Note that a,(p) ukn). -
*Me= Free path between collision (AT) or inelastic interaction (AS,calculated A = A/(Nm), where N is Avogadro's number.
 A.4 NUCLEAR CROSS SECTIONS 1 157
Inelastic interactions of hadrona with nuclei give rise to spallation
reactions involving the emission of a considerable number of nucle-
ons and pions from the target nucleus. Especially for target nuclei
of intermediate and high atomic number, a variety of product nuclei
are produced and many of these are radioactive. Various empirical
formulas have been developed to describe the reaction products from
spallation processes. The best known of these are the formulas of
Rudstam (1966), which have been further refined by Silverberg and
Tsao (1973 a, b). Computer programs for the evaluation of these
formulas have been published by Routti and Sandberg (1981).
Information about the products of spallation reactions, including
the distribution in energy and angle of secondary nucleons and pions,
can be obtained from calculations based on the intranuclear cascade
model which was f i s t used by Goldberger (1948) and further devel-
oped and elaborated at the Oak Ridge National Laboratory (Bertini,
1963,1969; Bertini et al. 1978), at the Brookhaven National Labora-
tory and Columbia University (Chen et al. 1971) and at the Joint
Institute of Nuclear Research a t Dubna (Barashenkov et al. 1969).
In Bertini's calculations, initially limited to energies up to 350
MeV, and later extended to 2.9 GeV and 1000 GeV,the nucleus is
assumed to be spherically symmetric, with a dense core and outer
regions of lesser density. The distribution of the bound nucleons
in the model nucleus is assumed to be that of a mixture of zero-
temperature Fermi gases for protons and neutrons. The maximum
kinetic energies of the nucleons are assumed to be (E, + 7) MeV,
where 7 MeV is an approximate value of the separation energy and
E, is the Fermi energy, given by

where N is the number of protons or neutrons, V is the nuclear

volume, and M is the nucleon mass.
The successive interactions of the incident hadron with the constit-
uent nucleons of the model nucleus (involving the transfer of kinetic
energy to the target nucleon and the production of additional parti-
cles) are simulated by the Monte-Carlo method. The interactions of
the recoiling nucleons and of the additional secondary particles with
the constituent nucleons are also simulated. An intranuclear hadron
cascade is thus generated which is followed until the participating
particles either escape from the nucleus or reach an energy lower
than a pre-assigned cut-off value. The Pauli exclusion principle is
approximately satisfied by not allowing, in the Monte-Carlo simula-
tion, the occurrence of collisions in which a particle emerges from a
collision with an energy lower than the Fermi energy. When the
cascade process has come to an end, holes are left, and particles have
158 / APPENDIXA
been captured in the nucleus, which is, therefore,' lefi in an excited
state. The de-excitation of this excited state is calculated according
to a statistical evaporation model, and results in the emission of low-
energy protons, neutrons, deuterons, tritons, etc.
Bertini's model was designed to be applicable to intranuclear cas-
cades initiated by particles with energies above 50 MeV, but appears
to give reasonable results also a t lower energies. For example, the
calculation of neutron interaction cross sections according to the
intranuclear-cascade models leads to kerma factors (labelled Alsmil-
ler in Figure A.29) that are in fair agreement with the results of
other calculations. The intranuclear-cascade model has. also been
applied by Guthrie et al. (1968) to obtain the prediction of the spec-
trum of particles emitted as the result of the capture of negative
pions. As shown in Table A.lO, the results are in fairly good agree-
ment with experimental results.
The results of the Monte-Carlo simulations of intranuclear cas-
cades with computer programs developed at Oak Ridge, Brookhaven,
and Dubna have been compared by Barashenkw et al. (1972) for the
case of 150- and 300-MeV protons incident on 27Aland 181Tanuclei.
There is good agreement in regard to the calculated energy spectra
of protons emitted at various angles and the curves of yield relative
to mass of reaction products, but there are considerable discrepancies
in regard to the distribution of nuclear excitation energies. Some of
these comparisons are shown in Figure A.33.
The results of Bertini et al. (19781, at energies fr6m several GeV
up to several hundred GeV, are in good agreement with experimental
data in regard to the total hadron-nucleus cross section, the spectral
shapes for protons and pions emitted from target nuclei, and the
production of various nuclides in spallation reactions. Serious dis-
crepancies are found, however, in regard to the number of shower
particles per incident hadron. These discrepancies are due to the
use of inadequate semi-empirical formulas for secondary particle
production by Ranft (1967), and can be expected to be removed when
more realistic production cross sections are used.
More realistic secondary-particle-production cross sections have
been developed for use in transport calculations by RanR (19721,
Ranft and Routti (1972,1974),and Chirikovet al. (1977).An informa-
tive discussion of these cross sections can be found in Van Ginneken
and Awschalom (1974). RanR and co-workers developed semi-empiri-
cal formulas with the qualitative behavior and functional form
expected theoretically. The approximation formulas satisfy Feyn-
man's scaling law (Feynman, 1969; Bali et al. 1970), according to
which the particle-production cross section, in the limit of very high
energies, depends only on the variable 2pJE,, and h,where p~ and
 EXCITATION ENERGY OF PRODUCT NUCLEUS. MeV
CHANGE IN MASS NUMBER. AA IU

0
Fig. A.33. Calculated differential cross sections as a function of emitted proton energies a t a particular angle, of residual mass yields and 5
of residual excitation energies for protons incident on aluminum and tantalum. The figures show comparisons between three intranuclear-
cascade models. Solid curves and dots: JINR model (Barashenkov et al., 1969);dashed curves and squares: BNLColumbia model (Chen et d.. __
1971);dotted curves and crosses: ORNL model (Bertini, 1969).(Excerpted h m Barashenkov et d., 1972.)
160 / APPENDIXA
J+ are the longitudinal and transverse momentum of the emitted
secondary particle in the center-of-mass system, and where Em is
the total center-of-mass energy for the incident and emitted particle.
In order to simplify the sampling in Monte-Carlosimulations, Ranft
furthermore made the approximation of factorizing the production
cross section into a function of 2pJE, times a function of h.
The functional form of the approximation formulas is taken from
the Hagedorn-Ranft thermodynamic model (Hagedorn and Ranft,
1968;Ranft, 1970).In this theory ofhadronic matter at high tempera-
ture, a hadron-hadron collision gives rise to a system with a tempera-
ture corresponding to 160 MeV that emits secondary particles in a
process similar to boiling. The thermodynamic model was developed
for hadron-hadron collisions, but Ranft later adapted it to fit the
experimental proton-nucleus cross sections obtained by Allaby et al.
(1970) at 19.2 GeVlc. For secondary proton and neutron production
in nucleon-nucleus collisions, RanfYs formulas have the form

and, for the production of nucleons or pions by incident pions, they

have the form

where M,,is the pion mass. Numerical values of the parameters a,,
and b, are given by Ranft (1972) and up-dated by Chirikov et al.
(1977) for interactions with H, Be, Al, Cu, and Pb targets.
Equations (A.53) and (A.54) take into account mainly very ener-
getic secondary particles that are emitted with an angular distribu-
tion strongly peaked in the forward direction. They do not take into
account the nucleons with energies up to a few hundred MeV and
the somewhat more diffuse angular distributions that are ejected
from the nucleus as the result of the intra-nuclear-cascade process.
Ranft and Routti (1972), therefore, used a particle-production spec-
trum that consists of two parts: (a)the high-energy spectrum given by
Equations (A.53) or (A.54); and (b) a low-energy spectrum calculated
from a parameterized fit to the results of the intranuclear-cascade
calculations of Bertini (1963) and Barashenkov et d.(1%9,1972).
Illustrative results pertaining to the number of particles produced
 A.4 NUCLEAR CROSS SECTIONS / 161
per incident hadron and to the energy used for their production, are
shown in Table A.12 for part (a) and in Figure A.34 for part (b).
Neutral pions decay very close to the point where they are produced
because of their short mean life (0.83 x lo-"% in their rest system).
With a probability of 98.8 per cent they decay into two gamma rays,
and with probability of 1.2 per cent they d a y into a single gamma
ray and an electron-positron pair. The gamma rays, which have a
combined energy equal to the sum of the TO rest energy (134.96
MeV) and the kinetic energy, initiate a photon-electron cascade. The
combined cross sections for hadron and electromagnetic cascades and
a Monte-Carlo calculation of the production of photons, electrons,
and positrons from primary proton beams have been described by
Ranft and Nelson (1979).
The cross sections needed for the calculation of the electromagnetic
cascade have been discussed in Sections A.l and A.2, with the excep-
tion of photon-nucleon and photon-nucleus interactions a t very high
energies. Up to 25 MeV, the photoabsorption process is dominated
by the giant dipole resonance already mentioned in Section A.1. At
energies from 50 to 200 MeV, photons mainly interact with proton-
neutron clusters inside the nucleus. The interaction cross section is
proportional, then, to the cross section for the photodisintegration of
deuterium, and the cross section is denoted, therefore, as a quasi-
deuteron cross section. At energies above the threshold for pion
-
production (- 135 MeV for a0and 140 MeV for a'), the photons
interact with the constituent nucleons of the nucleus, which are
raised to a n excited state and then decay into a stable nucleon and
a pion. The reaction mechanisms in the three regimes are indicated
in Figure A.35 for the case of photon interactions with lg7Au.Photon
cross sections for interactions with 160below the pion threshold are
shown in Figure A.36, and the cross sections for aoand a +production
from protons are shown in Figure A.37.
The mean life of negative and positive pions is rather long (2.60
x lo-' s in their rest frame), so that most of them are slowed down
to rest before decaying. Depending on the initial pion energy, from
0.1 to 1per cent of the pions decay into a muon and a neutrino while in
flight. These muons are quite important for shielding considerations
because they are highly penetrating. They do not interact strongly
with nuclei and are slowed down mainly by electromagnetic interac-
tions with atoms. Joseph (1969) has calculated muon range-energy
tables according to Bethe's stopping power theory by applying a
density-effect correction, and a small correction for muon brems-
strahlung, pair production and nuclear excitation.
 TABLEA.12-Multiplicities N iand inelasticities Ki of proton and pion production in proton-nucleus collisions calculated by Ranft and 1

Routti (1972) from Equations (A.53 and A.54). Ni represents the average number of emitted particles per incident proton. K, is the fraction
of the available energy (incident proton energy minus residual excitation energy of the target nucleus) that is used to produce

Momentum (lab)
the various p r t i c l e ~ . ~
10 GeVlc 20 GeV/c 100 GeV/c
5z
M
Ni K Ni K Ni Ki
H+ 3e
H2
Be
1.28
1.16
0.28
0.25
1.32
1.21
0.30 1.36 0.31 *
0.27 1.24 0.28
A1 1.27 0.26 1.35 0.30 1.40 0.32
Cu 1.21 0.22 1.37 0.26 1.46 0.28
Pb 0.69 0.10 0.69 0.09 0.67 0.08
71 +

Hz 1.46 0.19 3.00 0.19 4.17 0.19

Be 1.62 0.13 2.05 0.13 3.14 0.13
Al 1.79 0.14 2.27 0.14 3.51 0.14
Cu 1.46 0.11 1.87 0.11 2.89 0.11
Pb 1.23 0.09 1.57 0.09 2.45 0.09

Hz 1.67 0.13 2.15 0.13 3.41 0.13

Be 1.01 0.075 1.29 0.075 2.03 0.076
A1 0.97 0.073 1.24 0.073 1.98 0.073
Cu 0.94 0.073 1.24 0.073 2.01 0.073
Pb 0.75 0.062 1.07 0.062 1.70 0.062
"Ranft and Routti (1972).
 A.4 NUCLEAR CROSS SECTIONS 1 163

0 1 2 3 4 5
KINETIC ENERGY, GeV

0 1 2 3 4 5 6
KINETIC ENERGY, GeV

Fig. A.34. Results of intranuclear-cascade calculations as fitted to formulas by

Ranft and Routti (1972). A. Average number of protons and neutrons emitted as a
function of incident proton energy for various target atomic weights. B. Amounts of
energy spent for the production of these particles. (ARer Ranft and Routti. 1972.)
164 / APPENDIXA

I I I I I
10 100 1000
PHOTON ENERGY, MeV
Fig. A.35. Cross section for photonuclear interaction with '("Au ae a function of
photon energy. (After Forkman and Schroder, 1972.)
 A.4 NUCLEAR CROSS SECTIONS / 165

(y.p)+ly.n)
from p-shell \ I

Fig. A S . Photonuclear cross sections as a function of incident photon energy for

'9.Curve A: calculated total cross section for single-nucleon emission (y,p) + (y,n)
from the p-shell. Curve B: calculated total cross section for (y,pn). Curve C:sum of A
and B, i.e., sum of the cross sections for (y,pn), (y,n) and (y,p). Curve D: sum of the
cross sections for (y,pn) and (y,n). The experimental results shown are (i) the cross
seetion for total nuclear photon absorption taken from Ahrens et d.(1975) (Mainz)
($1; (ii) the cross section for total photon absorption with the exception of those
channels in which no neutron is emitted, according to Berghre and Carlos (private
communication, Saclay) ($4 (from Gari and Hebach, 1981).(Note: The (*) and the (4)
indicate the data points in the figure.)
166 / APPENDIXA

0
100 200 300 400 500 600 700 800 900 1000
ENERGY, MeV
Fig. A.37. Cross eections for photomeson production as a function of photon
energy. C w e a: y + p -,+,curve b: y + p -, n + n+.(After ICRU,1978.)
Appendix B
Examples of Absorbed-Dose
and Dose-Equivalent
Calculations

In this appendix, several examples of absorbed-dose and dose-

equivalent calculations are presented. In Section B.l, results of the
transport of neutrons with energies 5 4 0 MeV through tissue equiva-
lent material are presented and compared with experimental data.
In Section B.2, the problem of shielding manned spacecraft against
galactic cosmic-ray protons and alpha particles is considered. The
calculated results in both Sections B.l and B.2 were carried out using
Monte-Carlo methods, but the two applications are quite different
because of the different energy ranges involved and because only in
the second application is the transport of charged particles signifi-
cant.
In Section B.3, the problem of neutron "skyshine," i.e., the trans-
port of neutrons through the atmosphere is considered for neutron
energies 5400 MeV. The calculations in this section are different
from those presented earlier in that they are carried out using deter-
ministic-discrete ordinates-rather than stochastic methods. Also,
in this section, adjoint techniques are used to illustrate their useful-
ness.

B.l Absorbed Dose from Neutrons in Tissue-Equivalent

Material

Theus (1973) and Rogers (1973) measured the absorbed dose as a

function of depth along the axis of a tissue-equivalent phantom for
the neutron spectrum produced by 35-MeV 2Hon Be incident on
the phantom. In this section, calculated results are presented and
compared with these experimental data.
The geometry considered in all of the calculations is that of a point
source of neutrons a t a distance of 125 cm from the front surface of
a tissue- equivalent phantom. The phantom is a semi-infinite slab of
168 / APPENDIX B

tissue-equivalent material with a thickness of 30 cm. The composi-

tion of tissue-equivalent material is shown in Table B.l, and the
density of tissue-equivalent material was taken to be 1.1 g cm-$.
TABLEB.1-Composition of tissue-equivalent material used in the calculations
Element Percentage by weight

In all cases, the point source is assumed to emit isotropically, and

it is assumed that a collimator completely removes those neutrons
that would not be within the field size a t the surface of the phantom;
i.e., in all of the calculations, perfect collimation is assumed. Experi-
mentally, both circular and square fields were used and both are
considered here.
The energy distribution of neutrons from 35-MeV deuterons in a
thick Be target was obtained from measurements (Rogers, 1973,
Theus, 1973) and is shown in Figure B.1. The dashed portion of the
curve in Figure B.l represents an extrapolation because experimen-
tal data were not available a t all energies.
The calculated results presented here are taken from the work of
Alsmiller and Barish (1974). The calculations were carried out using
the high-energy Monte-Carlo transport code, HETC, developed a t
the Oak Ridge National Laboratory (Chandler and Armstrong, 1972;
Alsmiller, 1975; Alsmiller 1976). As explained in the discussion of

0 10 20 30 40 50 60 70
ENERGY, MeV

Fig. B.1. Energy distributions of neutrons from 35- and 50-MeV 2Hon Be and 67-
MeV 'H on Li. The distributions shown have been normalized to unity.
 B.l ABSORBED DOSE FROM NEUTRONS 1 169
this code in Section 8.3.3, the low-energy (515 MeV) neutrons are not
transported with HETC, but are transported with codes specifically
developed for these low-energy neutrons. In the calculations reported
here, this low-energy transport was done with the Monte-Carlo code
05R (Irving et al, 1965). The details of this low-energy transport are
given by Alsmiller (1975). The code 05R was not discussed in Section
8 because now it is not used often, but the results presented here
would not be appreciably different if the codes MORSE (Emmett,
1975) or MCNP (Thompson, 1979) had been used to transport the
low-energy neutrons.
The calculations were carried out by an analog Monte-Carlo
method (without use of variance-reduction techniques). All nuclear
reaction products (neutrons, protons, heavy nuclei, and photons) are
included in the calculations. The absorbed dose in a given volume is
defined as the energy deposited in the specified volume by charged
particles.
Rogers (1973) and Theus (1973) have measured the absorbed dose
along the axis of a n extended-slab, tissue-equivalent phantom for
various circular fields of radius R a t the surface of the phantom. The
experimental absorbed dose for R values of 17,13,9,5,and 3 cm are
shown as a function of depth in the phantom by the solid curves in
Figure B.2. The absolute normalization of the experimental data is
not known, but all of the experimental data in the figure have the
same normalization in terms of absorbed dose per incident neutron.
The histograms in Figure B.2 show the calculated absorbed doses
for the various field sizes used experimentally. To estimate absorbed
dose along the center line in the Monte-Carlo calculations, it was
necessary to average the calculations over a circular area about the
center line. The radial intervals used for this averaging process are
shown in the figure for each of the field sizes considered. Because the
calculations are statistical in nature, there is a statistical error
associated with each histogram value. In a few cases, this statistical
error-one standard deviation-has been indicated by an error bar
in Figure B.2.
To compare the calculations and experimental data in Figure B.2,
it was necessary to establish a normalization. This was done by
making the calculated and experimental results agree a t a depth of
5 cm for the case of R = 9 cm.It must be emphasized that all the
experimental data have the same normalization so only a single
normalization constant was used in obtaining all of the comparisons
shown in Figure B.2. The agreement between the calculated and
experimental absorbed doses is good for all depths and field sizes.
For neutrons from 35-MeV 2Hon Be, Rogers (1973) and Theus
(1973) also have measured the absorbed dose along the central axis
 8 RADIAL INTERVAL

f0

8
,-
0

9 to
X

$ 8
ui
$
0
to
0
m
a:
0
a
m to
u
8

0
0 5 40 (5 20 25
DEPTH, cm
Fig. B2. Center-line absorbed dose relative to depth for cylindrically collimated
neutrons from 35-MeV2Hon Be incident on a tissue-equivalent phantom. The absolute
calculated absorbed doses, in units of Gray per incident neutron on the phantom in
the field of radius R, may be obtained by multiplying the calculated values in the
figure by the normalization factor 1-cosrarctan(171125)1where the value 125 cm is
1-cos[arctan(R/125)1'
the distance between the neutron source and the front of the phantom.
 B.l ABSORBED DOSE FROM NEUTRONS 1 171
of a tissue-equivalent phantom with a collimator that produces a 10-
x 10-cm field a t the surface of the phantom. A comparison between
the calculations and experimental data for this case is shown in
Figure B.3. To estimate the center line absorbed dose, the calculated
results have been averaged over a 6- x 6-cm square about the center
line. The calculated results in the figure have an absolute normaliza-
tion. The experimental results have an unknown normalization and,
therefore,have been made to agree with the calculated absorbed dose
at a depth of 5 cm. The error bars on the calculated results are again
statistical only and represent one standard deviation. The agreement
between the calculated and experimental center-line absorbed dose
in the case of the square field is approximately the same as that
obtained in the case of the cylindrical fields.

h
c
e
C
3
(I)
C
C)
c
(I)
E
0
c
.-
Y

DEPTH, cm
Fig. B.3. Center-line absorbed dose relative to depth for neutrons from 35-MeV
'H on Be incident on a tissue-equivalent phantom. The collimator produced a square
10- x 10-cm field at the surface of the phantom, but the calculated results have been
averaged over a 6- x 6-cmsquare about the center line.
172 1 APPENDIXB
In the case of the square field discussed in the previous paragraph,
Todd et al. (in Rogers, 1973)measured the absorbed dose as a function
of distance from the beam center line at depths of 5 and 10 cm in the
phantom. In Figure B.4, the calculated doses are given with absolute
units and the experimental results have been normalized to agree
with the calculations at the point shown in the figure. The distance
from the centerline (abscissa in Figure B.4) is measured parallel to
the small distance of the field, that is, a t the surface of the phantom,
where the field is 10 x 10 cm, the distance from the centerline to the
edge of the field is 5 cm. The calculated results at depths of 5 and 10
cm shown in the figure were obtained by averaging over the depth
intervals 4 to 6 cm and 9 to 11 cm, respectively. The agreement
between the calculated and experimental absorbed doses is quite
good even at distances from the center line that are irradiated only
by scattered radiation.

B.2 Shielding of Manned Space Vehicles Against Galactic

Cosmic-Ray Protons and Alpha Particles

Galactic cosmic rays extend to very high energies and are conse-
quently very penetrating. This section considers the shielding of
manned space vehicles against galactic cosmic-ray protons and alpha
particles. The results presented are taken from the work of Santoro
et al. (1973).
The geometric model considered is a shield consisting ofa spherical
shell with a spherical tissue receptor a t the center. As indicated in
Figure B.5, the inside radius of the shield is 150 cm, the shield
thickness is re,and the receptor has a radius, r,, of 15 cm.
The hfferential omnidirectional flux density of protons a t solar
minimum and maximum and of alpha particles a t solar minimum
are shown in Fig. B.6. These flux density spectra are taken from
the review of McDonald (1969). The spectra applicable during solar
minimum were used in the calculations presented here. In the calcu-
lations, the radiations were assumed to be isotropically incident on
the spacecraft.
The calculations for incident protons were carried out using the
high-energy transport code HETC (Chandler and Armstrong, 1972;
Alsmiller, 1975; Alsmiller, 1976) (see also Section 8.3.3), and the
calculations for alpha particles were carried out using a slightly
modified version of HETC. The modifications in HETC to allow the
transport of incident alpha particles were those required to predict
the energy-angle distribution of particle production from alpha parti-
 B.2 SHIELDING OF MANNED SPACE VEHICLES 1 173

I I I I I
0 I 2 3 4 5 6 7 8 9
DISTANCE FROM CENTER LINE, crn
Fig. B.4. Absorbed dose relative to distance from the beam center line at two
depths when neutrons from 35-MeV 2Hon Be are incident on a tissue-equivalent
phantom. The collimator produced a square 10- x 10-cm field at the surface of the
phantom.

cle-nucleus collisions. The nuclear-reaction models used are those

described by Gabriel et al. (1971) and Armstrong and Chandler
(1973). In the case of both incident protons and alpha particles, the
secondary, low-energy (<I5 MeV) neutrons were transported using
the Monte Carlo code 05R (Irving et al., 1965) as described by Alsmil-
ler (1975).
The calculations were carried out with the use of biased rather
than analog Monte-Carlo techniques. The source-particle energies
and directions were sampled from biased distributions in order to
reduce the statistical fluctuations in the dose distributions in the
receptor. The biased distributions were constructed so that those
source-particle energies and directions that would produce relatively
large contributions to the absorbed dose and dose equivalent were
sampled more frequently. Statistical weighting fractions to account
for this biasing were then applied to each source particle so that the
original incident source spectral shape, angular distribution and
normalization were preserved. A detailed discussion of the biasing
formulas used are given in Appendix (3) of Alsmiller et al. (1972).
 Fig. B.5. A schematic diagram of the geometry used in the calculations for shield-
ing of manned space vehicles.

The absorbeddose rate was calculated by summing the energy

deposited by all charged particles per unit time in a specific tissue
volume. The dose equivalent rate was calculated by weighting the
energy deposited per unit time in a specific volume by a quality
factor as a function of linear energy transfer (LET) and summing
these weighted contributions. The quality factor as a function of LET
used in the calculations is that recommended by ICRP Publication
26 (ICRP, 1977).
The calculated particle contributions to the absorbeddose rate and
the dose-equivalent rate for incident galactic cosmic-ray protons and
alpha particles are shown in Figures B.7to B.lO, respectively. In all
of these figures the shield material is aluminum, the shield thickness
is 20 g/cm2, and the dose rates are shown a s a function of depth in
the receptor. In these figures, the total dose rate is not shown, but is
obtained by adding all the individual contributions. The contribution
labeled "primary protons" in Figures B.7 and B.8 and "primary
ionization" in Figures B.9 and B.10 is due to those incident particles
that have not undergone nuclear'collision and that deposit energy
 B.2 SHIELDING OF MANNED SPACE VEHICLES 1 175

Fig. B.6. Differential omnidirectional galactic proton flux density at solar maxi-
mum and minimum and alpha-particleflux density at solar minimum.

only by the excitation and ionization of atomic nuclei. All of the

other contributions are the result of particles produced by nuclear
collisions. Note that there is no contribution in the figures specifically
labeled "neutrons" because neutrons deposit energy by producing
charged particles and, thus, the contributions to the doses from neu-
trons are distributed among the charged-particle categories. The
curves labeled "heavy nuclei" indicate the doses due to all nuclear-
reaction products heavier than protons. Neutral pions decay rapidly
into two high-energy (270MeV) photons and these photons induce
176 1 APPENDIXB

C)PRIMARY PROTONS
a SECONDARY PROTONS
+ HEAVY NUCLEI
X CHARGED PIONS
PHOTONS FROM NEUTRAL PIONS
2 + ELECTRONS, POSITRONS, PHOTONS
X MUONS

1 0 - ' * 1 I I I I I I I I i I I l
0 4 2 3 4 5 6 7 8 9 4 0 4 4 4 2 4 3 4 4 4 5

DEPTH IN TISSUE, g ~ r n - ~
Fig. B.7. Contributions from the various types of particles ta the absorbed-dose
rate as a function of depth in tissue when a galajic, cosmic-ray, solar-minimum
proton spectrum is isotropically incident on a 20-g-~m-~-thick
aluminum shield.

an electron-photon cascade. Charged pions are unstable and decay

into muons, and muons are unstable and decay into electrons and
positrons.
In Figures B.7 and B.9 the major contributions to the absorbed-
dose rates are from primary ionization and secondary protons. In
Figs. B.8 and B.10 the major contributions to the dose-equivalent
rates are from primary ionization, secondary protons, and heavy
nuclei. The fact that heavy nuclei contribute appreciably more to the
dose-equivalent rate than to the absorbed-dose rate results from
the large quality factor associated with the heavily ionizing heavy
nuclei.
In Figures B.ll and B.12, the total absorbed dose rates and dose
equivalent rates, respectively, are shown as functions of depth in
tissue for incident galactic, cosmic-ray, solar-minimum protons and
 B.2 SHIELDING OF MANNED SPACE VEHICLES 1 177

O PRIMARY PROTONS
A SECONDARY PROTONS 0 PHOTONS FROM NEUTRAL PIONS
+ ELECTRONS. POSITRONS. PHOTONS
I X CHARGED PIONS X MUONS I
10-l2 1 I I I I I I I I I I I I I I
0 2 3 , 4 5 6 7 8 9 10 11 12 13 14 15
DEPTH IN TISSUE, g ~ r n - ~
Fig. B.8. Contributions from the various types of particles to the dose-equivalent
r a t . as a function of depth in tissue when a galactic, cosmic-ray, solar-minimum
proton spectrum is isotropically incident on a 20-g-~m-~-thick aluminum shield.

alpha particles. In these figures, results are given for aluminum

shield thicknesses of 5 g ~ m and - 20
~ g ~ m -The~ . total doses for the
shield with thickness of 20 g are those obtained by summing
all of the appropriate contributions given in Figures B.7 to B.lO.
Also shown for comparison (dashed histograms) in Figures B . l l and
B.12 are the contributions to the total dose rates from those primary
particles that have not undergone nuclear collisions. The difference
between the solid and dashed histograms in each case is the contribu-
tion to the total absorbed-dose and dose-equivalent rates from
nuclear reaction products. In all cases, the total absorbed-dose and
dose-equivalent rates are reasonably uniform over the tissue depth
considered, neglecting the statistical fluctuations that arise from the
Monte-Carlo nature of the calculation.
178 1 APPENDIX B

@ PRIMARY IONIZATION
A SECONDAY PROTONS
5 - + HEAVY NUCLEI -
X CHARGED PIONS
Q PHOTONS F R W NEUTRAL PIONS
2 - + ELECTRONS, POSITRONS. PHOTONS -
X MUONS
l l l l l l l 1 1 1 1~ 1 ~
0 2 4 6 8 40 12 44
DEPTH IN TISSUE, g ~ r n ' ~
Fig. B.9. Contribution of the varioua types of particles to the absorbed-dose rate
a8 a function of depth in tissue when a galactic, cosmic-ray, solar-minimum, alpha-
particle spectrum is isotropically incident on a 20-g-~rn-~-thickaluminum shield

B.3 Skyshine for Neutron Energies 5400 MeV

In designing the shielding for high-energy accelerators and for

fission and fusion reactors, i t is necessary to consider the radiation
that passes into the atmosphere from the top of the shield and scatters
back to earth to produce a radiation field at distances reasonably
removed from the devices (- 1000 m). This phenomenon, usually
called "skyshine", has been studied by several investigators (Linden-
baum, 1961;Distenfeld and Colvett, 1966;Rindi and Thomas, 1975).
Calculations are presented that enable estimation of the dose equiva-
lent from skyshine from an arbitrary, energy-angle-dependent neu-
tron source with neutron energies 4 0 0 MeV [Alsmiller, Barish and
Childs (1980)l.
 B.3 SKYSHINE FOR NEUTRONS 1 179
I I I I ~ I I I I ~ I I
I o PRIMARY loNlzATloN a PHOTONS FROM NEUTRAL PIONS'
5 - A SECONDARY PROTONS + ELECTRONS,POSITRONS, PHOTONS -
+ HEAVY NUCLEI x MUONS

-
X CHARGED PONS
-
V)

& 10-~ %
5
k
a:
I-
2 2
W
lo-,
1
f=?I 5

% 2
8 ,,-,.

0 2 4 6 8 10 12 14
DEPTH IN TISSUE, g cm7
Fig. B.lO. Contribution of the various types of particles to the dose-equivalent
rate as a function of depth in tissue when a galactic, cosmic-ray, solar-minimum,
alpha-particle spectrum is isotropically incident on a 20-g-~m-~-thick aluminum
shield.

A schematic diagram of the geometry considered here is shown in

Figure B.13. There is cylindrical symmetry about the z axis. The
point source is assumed to be located 15 m above the air-ground
interface and is assumed to emit neutrons only into the upward
hemisphere. In practice, the point source represents the neutrons
that emerge from the roof of an accelerator or a reactor. The source
should extend over the roof, but because the detector usually is
located far from the facility, it is a good approximation to assume
that the source may be represented by a point.
All of the calculations were carried out with the two-dimensional,
discrete-ordinates code, DOT (Rhoades and Mynatt, 1973), and the
first-collision source code, GRTUNCL (Lillie et al., 1978). The differ-
ential particle-production cross sections used in the calculations were
180 / APPENDIX B

SHIELD MATERIAL: A1
SHIELD THICKNESS= 59 cm2
GALACTIC COSMIC-RAY SOLAR-MINIMUM
PROTON SPECTRUM INCIENT
-
I
U)
X
I
I
I
1 SHIELD MATERIAL: Al
1 SHIELD THICKNESS- 5q cmZ
t - j - 4 = GALACTIC COSMIC-RAY SOLAR-MNIMJM
I

I
--4

I I - -- 7 P ---7----
-

.-
t u n a - PARTICLE SPECTRUM INUMNT

1 I I SHIELD MATERIAL: AI
0 '-I 1
I . SHELD THICKNESS = 2 0 p cm-2
GALACTIC COSMIC-RAY SOLAR-MINIMUM
PROTON SPECTRUM INCIDENT
t I ; -1

SHIELD MATERIAL: Al
SHIELD THICKNESS ~ 2 Q 0
I GALLCTIC COSMIC-RAY SOLAR-MINIMUM
. ALPHA-PARTICLE SPECTRUM INCIDENT

I I I I I I
0 2 4 6 8 10 12 14 !6

DEPTH IN TISSUE, g ~ r n - ~

Fig. B.ll. Absorbed-dose rate relative to depth in tissue for galactic, cosmic-ray,
proton and alpha-particle spectra isotropically incident on spherical-shell shields.
Solid lines: attenuated primary particles and all secondary particles; dashed lines:
attenuated primary particles; points: unattenuated primary particles.

taken from Alsmiller and Barish (1979) and (1981). The restriction
of consideration to neutron energies of 1400 MeV arises because a t
higher energies pion production becomes important and must be
taken into account. This limit on energy range is, however, sufficient
for most applications even around very high-energy accelerators
(Lindenbaum, 1961; Distenfeld, 1966; Rindi, 1975). The neutron and
photon flux densities were converted to dose-equivalent rates using
the data given in NCRP (1971) and in Claiborne et al. (1970)
If the energy and angular distribution of the neutron source are
known, the codes DOT and GRTUNCL can be used to calculate the
dose equivalent a t the detector position. The calculations, however,
require the use of a large computing machine, and they must be
repeated completely if results for a different source are required. It
would be convenient if these results could be used to estimate the
skyshine dose equivalent from different neutron-source, energy-
 B.3 SKYSHINE FOR NEUTRONS 1 181

SHIELD MATERIAL A1
I
SHIELD THICKNESS. 59 ~ r n - ~
I
GALACTIC COSMIC-RAY SOLAR-MINIMUM
PROTON SPECTRUM INCIOENT

SHIELD MATERIAL: A1
SHIELD THICKNESS- 59 ~ r n - ~
GALACTIC COSMIC-RAY SOLAR-MINIMUM
ALPHA-PARTICLE SPECTRUM INCIDENl

SHIELD MATERIAL: A1
SHIELD THICKNESS. X) g ~ r n - ~
GALACTIC COSMIC-RAY SOLAR-MINIMUM
PROTON SPECTRUM INCIDENT

SHIELD MATERIAL: A1
SHIELD THICKNESS= X) g ~ r n - ~
GALACTIC COSMIC-RAY SOLAR-MINIMUM
ALPHA PARTICLE SPECTRUM INCIDENT

I I I I

10-'O 1 1 I I
I

0 2 4 6 8 10 12 14 16

DEPTH IN TISSUE, g ~ r n - ~

Fig. B.12. Dose-equivalent rate relative to depth in tissue for galactic, cosmic-
ray, proton and alpha-particlespectra isotropically incident on spherical-shellshields.
Solid lines: attenuated primary particles and all secondary particles; dashed lines:
attenuated primary particles; points: unattenuated primary particles.

angle distributions without repeating the elaborate transport calcu-

lations. This can be done with thq use of adjoint techniques as dis-
cussed below.
Let us first introduce the concept of importance function. The dose
equivalent due to source neutrons may be written

where
H(r) = The dose equivalent a t a radial distance r from the source
and on the air-ground interface (i.e., a t z = O),
S(E,cos8) = The number of source neutrons per unit solid angle
a t energy E and polar angle 8 (the source emission is assumed
to be independent of the azimuthal angle),
182 / APPENDIX B

AIR

POINT SOURCE
I
(ASSUMED TO
EMIT ONLY INTO

Fig. B.13. Schematic diagram of skyshine geometry for which importance func-
tions are calculated.

I(E,cos8,r) = The importance of source neutrons a t energy E and

polar angle 8 for producing the dose equivalent a t r.
Equation (B.1) indicates that, if the importance function is known
as a function of energy, polar angle and radial distance, then, for any
source distribution, the dose equivalent may be obtained by carrying
out a double integration. The significant point is that, when used in
the adjoint mode, the codes GRTUNCL and DOT give this importance
function. It should be noted, however, that the importance function
is dependent on r and every distance r requires a complete transport
calculation. That is, in forward calculations, one obtains the dose
equivalent a t all radii for a given source with one transport calcula-
tion. In adjoint calculations, one obtains a source-independent impor-
tance function a t only one distance, r, with one transport calculation.
In Figure B.14, the neutron importance function is shown as a
function of energy for specific cosine intervals for a distance r of 1005
m. (Because of the manner in which discrete-ordinates calculations
B.3 SKYSHINE FOR NEUTRONS 1 183
184 / APPENDIX B
are done, the values in the figure have been obtained by averaging
over a radial interval of 1000 to 1010 m and over a z interval of 0 to
1 m.) The histogram intervals used in the figure correspond to the
energy group structure used in the calculations. The structure in the
histograms in the vicinity of 10 MeV is due to resonances in the
cross-section data. For the large r value considered in Figure B.14,
it is clear that the high-energy source neutrons are considerably
more important than the low-energy source neutrons. Also, for the
large r value considered, it is clear that high-energy neutrons emit-
ted at small values of cos 8 are more important than high-energy
neutrons that are emitted a t large values of cos 8.
Results similar to those shown in Figure B.14 for other values of
r are given by Alsmiller et al. (1981). Also, the same reference gives
a comparison between the calculated results obtained with the use
of the importance functions and the experimental data.
Appendix C
A Compilation of Geometric
Reduction Factors for
Standard Geometries

In the following, typical geometries are considered and the result-

ing geometric-reduction factors are given. The point-pair distance
distributions are not listed but can readily be obtained from the
geometric-reduction factor:

C.l The Autologous Case (A = B)

As source region A and receptor B coincide in the autologous case,

the simplified notation U(x)can be used instead of U,(x). Outside
the specified intervals for x, all functions U(x)or U
,(x) are zero.
a) Infinite slab of height h:

b> Sphere of radius r:

186 / APPENDIXC
C) Spherical shell with outer radius R and inner radius r:
For r 5 R/3:

For r e R13:

Limit of thin spherical shell with radius r and thickness 6 (6 <<r):

d) Spheroid with two axes d and the third axis ed (Kellerer, 1984)

for o < x 5 Max (d,ed)-

The expression in the first line of Equation ((2.7) applies for
xsMin(d,cd). In the case of e< 1, the expression in the second line
applies for x>e-d; in the case of e > l , the sum of the expressions in
the first and in the second lines is used for x>d.
In Equation (C.7)' E = is the eccentricity; then
 C.1 THE AUTOLOGOUS CASE (A = B) 1 187
and

cos - l(x), for O s x s l ,

ci(x) = (C.9)
cosh-l(x) = en(x + m),
for x > l .

The auxiliary function ci(x) is introduced to permit one common

formula for the oblate and the prolate spheroid (Kellerer, 1984).
Throughout Appendix C, fP1(x) is defined as the inverse function of
f(x).
e) Right cylinder with arbitrary cross section of width d and height
h (Kellerer, 1981):
U , ( V P )dz, x s v m , (C.10)

2, = , - d2) ; z2 = Min(x,h) .
d ~ a x ( ox2 ((2.11)
U,(x) is the geometric-reduction factor of the cross sedion ( x l d ) as
follows:
Circular cross section:
u c x [ c o s - ( ) - ] , d . (C.12)
'Tr

Square cross section (see Coleman, 1969):

Integrating the expression, one obtains for the unit cube the relation
derived by Piefke (1978):
188 / APPANDIX C
where

Figure C.l shows geometricreduction factors and point-pair dis-

tance distributions for the autologous case according to some of the
preceding equations.

C.2 The Heterologous Case (A f B)

a) Two concentric spheres A and B with radii R and r, respectively:

b) A spherical surface, A, of radius R and a concentric sphere, B, of

radius r:

!R-rJ<x<R+r (c.16)
R < r and x<r-R.

C) A spherical surface, A, of radius R and a concentric shell, B, of

inner radius r, and outer radius r,:
The geometric-reduction factor is equal to the preceding solution
with r = r, minus the solution with r = r,.
For R >r,, i.e., for A outside B:
 C.2 THE HETEROLOGOUS CASE (A f B) . 1 189

RELATIVE DISTANCE, x/D

Spheroids and Rod

-

b -

0
0 0.2 0.4 0.6 0.8 1 .O
RELATIVE DISTANCE, x/d
Fig. C.1. Geometric-reduction factore, U(x), or the point-pair distance distribu-
tions, p(x), for the autologoua case.
a: Geometric-reduction factors for spherical shells according to Equations ((3.4) and
(C.5). The distance, x, is given relative to the outer diameter, D. The ratio of inner
diameter, d, to outer diameter of the shell is indicated on the curves. The dashed curve
corresponds to the sphere [Equation (C.3)].
b: Geometric-reduction factors for ephemids according to Equation (C.7). Thedistance,
x, is given relative to the two equal diametere, d, of the spheroid. The ratio e of the
third diameter to d is indicated on the curves. The daahed curve represents the limit
of an infinitely extended prolate spheroid, while the dotted curve represents, for
comparison, an infinitely long circular rod of diameter d [Equations (C.10) and (C.12)I.
190 / APPENDIXC

RELATIVE DISTANCE, x/d

0 0.5 1.0 1.5

RELATIVE DISTANCE, x/d

Fig. C.1. continued

c: Geometric-reduction factors for the sphere of diameter d [Equation (C.3)1,the
circular right cylinder of diameter and height d [Equations (C.10)and (C.12)1,the
cube of side length d [Equation (C.14)],and the infinite slab of height d [Equation
(C.211.The distance, x, is given relative to d.
d: Point-pair distance densities for a disc of diameter d [Equation (C.12)], a aquare of
side length d [Equation (C.13)], and a spherical surface of diameter d [Equation (C.611.
The distance, x , is given relative to d.
 C.2 THE HETEROLOGOUS CASE (A f B) / 191
For R cr,, i.e., for A in the interior void of the shell for Rr (r2- rJ/
2:

For R 5 (r2 - rl)/2

Special cases:
An outer surface of the shell (R = r,):

An inner surface of the shell (R = r,) and R 2 r$3:

192 / APPENDIXC
d) Two concentric shells A and B of infinitesimal thickness and with
radii R and r:

where 6 is the (infinitesimal) thickness of the shell B.

Note: Solutions b) to d) apply also if A is part of the spherical surface,
e-g., a sector, a spherical cap, a ring within the shell or a point.
e) A plane A (or part of a plane) and a parallel plate B of thickness
h at distance a:

f) Two parallel infinite plates A and B of thickness H and h, respec-

tively, and separated by distance a (H<h)

The same solution applies if A is a section of the infinite plate, i.e.,

a right cylinder (finite plate).
g) A point A and a right cylinder B:
= & l:%Ac(ViG)& ,V&i? <x< V(b+ h12 + A (C.26)
z, = o 1 , z2 = Min (b+h,
Max(b, ~ M ~ J cf( -$) .
UAC(x)is the two-dimensional geometric-reduction factor for the
cross section C of the cylinder relative to the projection of point A,
 C.2 THE HETEROLOGOUS CASE (A # B) 1 193
P(A), into the plane which corresponds to C; y, and y, are the mini-
mum and maximum distances in this plane between P(A) and C; h
is the height of the cylinder; b is the projecteddistance on the cylinder
axis between A and P(A). Point A is assumed to lie below the cylinder
bottom. If A lies between the two planes through the faces of the
cylinder, one obtains the solution as a sum by suitable subdivision
of B into two cylinders.

For the circular area of radius r relative to the point a t distance a

from the center, one has:

[ (
U,,(x) = - cos-I Max
7r
- 1, "'z-31,
- Max ( a - r , o ) r x ~ a + r . (C.27)

h) A point A and a planar domain B of infinitesimal thickness 6:

Let h be the distance of A from the plane C of B, and let U,(y) be the
2-dimensional geometric-reduction factor (y,<y<y,) of B relative to
the projection of A in C. Then:

For a disc of radius r and a projected distance a of A from its center,

one obtains

For a rectangle with side lengths a and b, we restrict to a special

case for A: The straight lines which come out from the sides divide
the plane into 8 rectangles with infinite content; the projection of A,
P(A), is assumed to lie in the left lowest rectangle with infinite
content; s is the distance of P(A) to the next rectangle line parallel
to a, t is analogously related to b. (Different positions of A can
be reduced to the described special case.) For the two dimensional
geometrical factor, one has:

U2(y) = -
2n
1 [
sin-' i in (a +; up)
194 / APPENDIXC
therefore,

Min (a+s, v r (C.31)

4nx -

Max(s, v M a x ( O j 2- h2- ( b + t)2))

- sin-'

The diagrams in Figure (C.2) give geometric-reduction factors and

point-pair distance distributions for the heterologous case according
to some of the preceding equations.
 C.2 THE HETEROLOGOUS CASE (A f B) 195

X
3

0 1 2 3 4 5
RELATIVE DISTANCE, x/R

Spherical Surface

0 1 2 3 4 5
RELA-l7VE DISTANCE, x/R

Fig. C.2. Geometric reduction factors, U(x), or the point-pair distance distribu-
tions, p(x), for the heterologous case.
a: Geometric-reduction factors for a sphere, A, of radius R, and a concentric sphere,
B, of radius r according to Equation (C.15). The ratio rlR is indicated on the curves.
The distance, x, is given relative to R .
b: Geometric-reduction factors for a spherical surface, A, of radius R and a concentric
sphere, B, of radius r according to Equation (C.16). The ratio rlR is indicated on the
curves. The distance, x, is given relative to R.
196 / APPENDIX C

Plane and Plate

-
C
-

0 1 2 3 4 5
RELATIVE DISTANCE, x/h

Conc. Spherical Surfaces

-
d
-

-
I
0
0 1 2 3 4 5
RELATIVE DISTANCE, x/R

Fig. C.2. continued

c: Geometric-reductionfactors for a plane, A, separated by distance a from an infinite
plate, B, of height h according to Equation (C.23). The ratio alh is indicated on the
curves. The distance, x, is given relative to h.
d: Point-pair distance distributions between a spherical surface of radius R and a
concentric spherical surface of radius r according to Equation ((3.22). The ratio rlR is
indicated on the curves. The distance, x , is given relative to r.
References
AHRENS,J.,BORCHERT, H., CZOCK,K.H., EPPLER,H.B., GIMM,H., GUN-
DRUM, H., KROENING, M., RIEHM,P., SITA RAM,G. AND ZIEGER,A.,
(1975). "Total nuclear photon absorption cross sections for some
light elements," Nucl. Phys. A251,479.
ALLABY, J.V., BINON,F., DLDDENS, A.N., DUTEIL,P., KLOGINING.A.,
MEUNIER,R., PEIGNEUK, J., SACHARDIS, E.J., SCHLUPMANN, K.,
SPIGHEL, M., STROOT,J.P.,THORNDIKE, A.M.,AND WETHERELL, A.M.
(1970). "High energy particle spectra from proton interactions a t
19.2 GeVIc," CERN Report No. 70-20 (European Organization for
Nuclear Research, Geneva).
ALSMILLER,R.G., JR.(1975). "Calculations of the transport of fast
neutrons (<50 MeV) through matter," page 681 in Proceedings
of the Second Symposium on Neutron Dosimetry in Biology and
Medicine (NeuherbergIMiinchen).
ALSMILLER, R.G., JR.(1976). "Nucleon-meson transport calculation,"
in Spallation Nuclear Reactions a n d their Applications;Shen,
B.S.P. and Merker, M., Eds. (D. Reidel Publishing Co., Boston,
Massachusetts).
ALSMILLER, R.G.,JR.AND BARISH,J. (1974). "Calculations pertaining
to the use of fast (550 MeV) neutrons in cancer radiotherapy,"
Med. Phys. 1, 51.
ALSMILLER, R.G.,JR.,AND HERMAN, O.W. (1970). "Calculation of neu-
tron spectra from proton-nucleus nonelastic collisions in the
energy range 15 to 18 MeV and comparison with experiments,"
Nucl. Sci. Eng. 40, 254.
ALSMILLER, R.G., JR.,ARMSTRONG, T.W., AND COLEMAN, W.A. (1970).
"The absorbed dose and dose equivalent from neutrons in the
energy range 60 to 3000 MeV and protons in the energy range 400
to 3000 MeV," Nucl. Sci. Eng. 42,367.
AMMILLER.R.G., JR.,SANTORO, R.T., BARISH,J.,AND CWBORNE,H.C.
(1972). Shielding of Manned Space Vehicles Against Protons and
Alpha Particles, Report No. ORNL-RSIC-35 (Oak Ridge National
Laboratory, Oak Ridge, Tennessee).
AMMILLER, R.G., JR.,BARISH,J., AND CHILDS,R.L. (1981). "Skyshine
a t neutron energies 5400 MeV," Part. Accel. 11,131.
AMMILLER,R.G.,JR., BARNES, J.M. AND DRISCHLER, J.D. (1986). "Neu-
tron-photon multigroup cross section for neutron energies <400
MeV (Revision 11," Nucl. Inst. Methods A249,455.
198 / REFERENCES
AMALDI, E. (1958). "The production and slowing down of neutrons,"
page 1 in Handbuch &r Physik, Vol. XXXVIII, Part I1 (Springer-
Verlag, Berlin).
ANDERSEN, H.H. AND ZIEGLER, J.F. (1977).Hydrogen Stopping Powers
and Ranges in All Elements (Pergamon Press, New York).
ARMSTRONG, T.W. AND CHANDLER, K.C. (1972). "Monte Carlo calcula-
tions of the dose induced by charged pions and comparisons with
experiment," Radiat. Res. 52, 247.
ARMSTRONG, T.W. AND CHANDLER, K.C. (1973). "Calculation of the
neutron flux spectra induced in the earth's atmosphere by galactic
cosmic rays," J. Geophys. Res. 78, 2715.
ANSTRONG,T.W.,ALSMILLER,R.G.,JR., CHANDLER, K.C., AND BISHOP,
B.L. (1972). "Monte Carlo calculations of high-energy nucleon-
meson cascades and comparison with experiment," Nucl. Sci. Eng.
49, 82.
ASHLEY,J.C. (1982). "Stopping power of liquid water for low-energy
electrons," Radiat. Res. 89, 25.
ASHLEY, J.C.,RITCHIE,R.H.,ANDBRANDT, W. (1972). "Zl3 effect in the
stopping power of matter for charged particles," Phys. Rev. B5,
2393.
ASHLEY,J.C., RITCHIE,R.H., AND BRANDT, W. (1973). "ZI3 dependent
stopping power and range contributions," Phys. Rev. A8, 2402.
AUXER,J.A., SNYDER, W.S.,AND JONES, T.D. (1968). "Neutron inter-
actions and penetration in tissue," pages 275-316 in Radiation
Dosimetry, Vol. 1, Attix, F.H., Roesch, W.C., and Tochilin, E., Eds.
(Academic Press, New York, New York).
BALI,N.F., BROWN, L.S.,PECCEI, R.D.,AND PGNCYITI, A. (1970). "Scal-
ing behavior in pp + anything a t high energy," Phys. Rev. Lett.
25, 557.
BAMBYNEK, W., CRASEMANN, B., FINK,R.W.,FREUND, H.U., MARK, H.,
SWIFT,C.D.,PRICE, R.E., AND RAO,P.V. (1972). "X-ray fluorescence
yields, Auger and Coster-Kronig transition probabilities," Rev.
Mod. Phys. 44, 716.
BARASHENKOV, V.S.,GUDIMA, K.K., AND TONEEV, V.D. (1969)."Statis-
tical determination of inelastic collisions of fast particles with
intranuclear nucleons," Acta Phys. Polon. 36, 415.
BARASHENKOV, V.S., BERTINI,H.W., CHEN,K., FRIEDLANDER, G.,
HARP.G.D., ILJINOV,AS., MILLER,J.M., AND TONEEV, V.D. (1972).
"Medium energy intranuclear cascade calculations: A comparative
study," Nucl. Phys. A187, 531.
BARKAS. W.H. AND BERGER, M.J. (1964). "Tables of energy losses and
ranges of heavy charged particles," page 103 in National Academy
of Science-National Research Council Publication No. 1133
(National Academy of Sciences, Washington)
BECKURTS, K.H. AND WIRTZ,K. (1964). Neutron Physics (Springer,
New York, New York).
BERGER, M.J. (1963). "Monte Carlo calculation of the penetration and
diffusion of fast charged particles," in Methods in Computational
Physics, Vol. I , Adler, B., Fernback, S., and Rotenberg, M., Eds.
(Academic Press, New York, New York).
BERGER,M.J. (1968). Energy Deposition in Water by Photons from
Point Isotropic Sources, MIRD Pamphlet No. 2, J. Nucl. Med.,
Supplement No. 1, 15-25.
BERGER, M.J. (1970). "Beta-ray dosimetry calculations with the use of
point kernels," pages 63-86, in Medical Radionuclides: Radiation
Dose and Effects, USAEC Report CONF-691212 (National Techni-
cal Information Center, Springfield, Virgicia).
BERGER,M.J. (1971). Distribution of Absorbed Dose Around Point
Sources of Electron and Beta Particles in Water and Other Media,
MIRD Pamphlet No. 7, J. Nucl. Med. 12, Supplement No. 5,5-23.
BERGER,M.J. (1973). Improved Point Kernels for Electron and Beta
Ray Dosimetry, NBS Report NBSIR 73-107 (National Bureau of
Standards, Washington, D.C.).
BERGER,M.J. (1978). "Planned compilation of stopping power and
range data," page 639 in Proceedings of the Sixth Symposium on
Microdosimetry (Harwood Academic Publishers, London).
BERGER, M.J. (1984). "Discussion of electron cross sections for trans-
port calculations," pages 1-8 in Proceedings of the Workshop on
Electronic and Ionic Collision Cross Sections Needed in the Model-
ing of Radiation Interactions with Matter held a t Argonne National
Laboratory, 6-8 December 1983, Report No. ANL-84-28 (Argonne
National Laboratory, Argonne, Illinois).
BERGER, M.J. AND SELTZER, S.M. (1964). "Tables of energy losses and
ranges of electrons and positrons," page 205 in National Academy
of Sciences-National Research Council Publication No. 1133
(National Academy of Sciences, Washington).
BERGER,M.J. AND SELTZER, S.M. (1968). ETRAN Monte Carlo Code
System for Electron and Photon Transport through Mixed Media,
Radiation Shielding Information Center Code Package, CCC-107
(Oak Ridge National Laboratory, Oak Ridge, Tennessee).
BERGER, M.J. AND SELTZER, S.M. (1970). "Bremsstrahlung and photo-
neutrons from thick tungsten and tantalum targets," Phys. Rev.
C2,621.
BERGER,M.J. AND SELTZER,S.M. (1972). "Response functions for
sodium iodide scintillation detectors," Nucl. Inst. Meth. 104,317.
BERGER, M.J. AND SELTZER, S.M. (1982). Tables ofEnergy-Deposition
Distributions in Water Phantoms Irradiated by Point-Monodirec-
tional Electron Beams with Energies from 1 to 60 MeV, and Appli-
200 / REFERENCES
cations to Broad Beams, NBS Report No. NBSIR 82-2451(National
Bureau of Standards, Gaithersburg, Maryland).
BERGER, M.J. AND SPENCER, L.V. (1959). "General relation between
fluxes from collimated point and plane sources of radiation," Phys.
Rev. 113,408-312.
BERGER, M.J. SELTZER, S.M. AND MAEDA,K. (1970). "Energy deposi-
tion by auroral electrons in the atmosphere," J. Atmos. Terr. Phys.
32, 1015.
BERMAN, R.L. (1975). "Atlas of photoneutron cross sections obtained
with monoenergetic photons," Atomic Data and Nuclear Data
Tables 15, 319.
BERMAN, R.L. AND FULTZ,S.C. (1975). "Measurement of the giant
dipole resonance with monoenergetic photons," Rev. Mod. Phys.
47,713-761.
BERTINI,H.W. (1963). "Low energy intranuclear cascade calcula-
tions," Phys. Rev. 131, 1801.
BERTINI, H.W. (1967). "Secondary particle spectra from the interac-
tion of 30 to 340 MeV protons on complex nuclei; Experimental
data and comparison with theory," Phys. Rev. 162, 976.
BERTINI,H.W. (1969a). "Intranuclear-cascade calculation of the sec-
ondary nucleon spectra from nucleon-nucleus interactions in the
energy range 340 to 2900 MeV and comparisonswith experiment,"
Phys. Rev. 188,1711.
BERTINI,H.W. (1969b). "Calculation of nuclear reactions for incident
nucleons and 7~ mesons with photoemulsion," pages 42-62 in Pro-
ceedings of the Second International Conference on Accelerator
Dosimetry and Experience held at Stanford Linear Accelerator
Center, Stanford, CA, 5-7 November 1969, Report Conf-6911-1
(United States Atomic Energy Commission, Washington, DC).
BERTINI, H.W. (1972a). "Nonelastic interaction of nucleons and ll
mesons with complex nuclei a t energies below 3 GeV," Phys. Rev.
C6, 613.
BERTINI. H.W. (1972b). "Reaction cross section for 30 to 60 MeV
protonson various elements; comparison of theoretical results with
experiment," Phys. Rev. C5, 2118.
BERTINI, H.W., CULKOWSKI, H., HERMANN, O.W., GOVE,N.B., AND
GUTHRIE,M.P. (1978). "High energy (E> 1000 GeV)intranuclear
cascade model for nucleons and pions incident on nuclei and com-
parisons with experimental data," Phys. Rev. C17 1382-1395.
BETHE,H. (1930). "ZurTheorie des Durchgangs schneller Korpusku-
larstrahlen durch Materie," Ann. Physik. 5, 325.
BETHE,H. (1932). "Bremsformel fiir Elektronen relativistischer
Geschwindigkeit," Z. Phys. 76, 293.
 REFERENCES 1 201
BETHE.H. (1933). "Quantenmechanik der Ein-und Zweielektronen-
probleme," page 273 in Handbuch der Physik, Geiger, H. and
Scheel, K., Eds., Vol. 24, Part I (Springer-Verlag, Berlin).
BETHE.H. AND HEITLER.W. (1934). "On the stopping of fast particles
and on the creation of positive electrons," Proc. Roy. Soc. A146,
83.
BETHE.H. AND ASHKIN,J. (1953). "Passage of radiations through
matter," page 166 in Experimental Nuclear Physics, Segre, E., Ed.,
vol I (Wiley, New York, New York).
BETZ,H. (1972). "Charge states and charge-changing cross sections
of fast heavy ions penetrating through gaseous and solid media,"
Rev. Mod. Phys. 44, 465.
BHAT,M.R.AND PEARLSTEIN, S. (1980). Symposium on Neutron Cross
Sections from 1 0 to50 MeV, Report No. BNLNCS-51295 (Brookha-
ven National Laboratory, Upton, New York).
BICHSEL,H. (1968). "Charged particle interaction," pages 157-228
in Radiation Dosimetry, vol. 1, 2nd ed., Attix, F.H. and Roesch,
W.C., Eds. (Academic Press, New York).
BICHSEL, H. (1972). "Passage of charged particles through matter,"
page 142 in American Institute of Physics Handbook (3rd Ed),
Gray, D.E., Ed. (McGraw-Hill, New York, New York).
BIGGS,F. AND LIGHTHILL, R. (1971). Analytical Approximations for
X-ray Cross Sections, IZ, Report SC-RR-71-0507 (Sandia National
Laboratory, Albuquerque, New Mexico).
BIGGS, F. AND LIGHTHILL, R. (1972). Analytical Approximations for
Total and Energy Absorption Cross Sections for Photon-Atom Scat-
tering, Report SC-RR-72-0685(Sandia National Laboratory, Albu-
querque, New Mexico).
BIRKHOFF,R.D. (1958). 'The passage of fast electrons through mat-
ter," page 53 in Handbuch der Physik, Vol XXXIV, Fliigge, s., Ed.
(Springer-Verlag, Berlin).
BLOCH,F. (1933). "Zur Bremsung rasch bewegter Teilchen beim Dur-
chgang durch Materie," Ann. Phys. 16,285.
BOHR,N.(1948). "The penetration of atomic particles through rnat-
ter," Mag. Fys. Medd. Dan Vid. Selsk. 18, No. 8.
BONDERUP, E. (1967). "Stopping of swift protons evaluated from sta-
tistical atomic model," Mat. Fys. Medd. Dan Vid. Selsk 35, No. 17.
BROWN,R.T. (1975). "Differential cross section for coherent photon
scattering by elements Z = 2 through Z = 26," Atomic Data and
Nuclear Data Tables 15, 111.
BUCHE,G. AND PRZYBILLA, G. (1981). "Distributions of absorbed dose
from II mesons beams calculated from a new Monte Carlo pro-
gram," Nucl. Instr. Meth. 179, 321.
202 1 REFERENCES
BURHOP,E.H.S. AND ASAAD,W.N. (1972). "The Auger-Effect," page
163 in Advances in Atomic and Molecular Physics, Bates, D.R., Ed.
(Academic Press, New York, New York).
BURLIN,T.E. (1968). "Cavity chamber theory," pages 331-392 in
Radiation Dosimetry, Vol. I, Attix, F.H., and Roesch W.C., Eds.
(Academic Press, New York, New York).
BULOW,B. AND FORKMAN, B. (1974). "Photonuclear cross sections,"
page 475 in Handbook on Nuclear Activation Cross Sections, Tech-
nical Report Series No. 156 (International Atomic Energy Agency,
Vienna).
BURROWS, T.W. AND HOLDEN, N.E. (1978). "A source list of nuclear
data bibliographies, compilations and evaluations," Report No.
BNL-NCS-50702 (Brookhaven National Laboratory, Upton, New
York).
CASWELL, R.S., COYNE, J.J.AND RANDOLPH, M.L. (1980). "Kerma fac-
tors for neutron energies below 30 MeV," Radiat. Res. 83, 217.
CHANDLER, K.C. AND ARMSTRONG, T.W.(1972). Operating Instruc-
tions for the High-Energy Nucleon-Meson Transport Code HETC,
Report No. ORNL-4744 (Oak Ridge National Laboratory, Oak
Ridge, Tennessee).
CHEN,S. AND CHILTON,A.B. (1979). "Calculation of fast neutron depth
dose in the ICRU standard tissue phantom and the derivation of
neutron fluence-todose-index conversion factors," Radiat. Res. 78,
335.
CHEN,K., FRIEDLANDER, G., HARP,G.D., AND MILLER.J.M. (1971).
"Effects of nucleon-pair correlations on Monte Carlo intranuclear-
cascade simulations," Phys. Rev. C4, 2234-2240.
CHIRIKOV, B.V.,TAYURSKI, V.A., MOEHRING, H.J., RANFT,J., AND
SCHIRRMEZSTER, V. (1977). "Optimization of antiproton fluxes from
targets using hadronic cascade calculations," Nucl. Instr. Meth.
144, 129.
CINDA86 (1986a). The Index to Literature and Computer Files On
Microscopic Neutron Data (1982-1986), (International Atomic
Energy Agency, Vienna).
CINDA86 (1986b). The Index to Literature and Computer Files On
Microscopic Neutron Data (Supplement), (International Atomic
Energy Agency, Vienna).
CWBORNE,H.C. AND TRUBEY, D.K. (1970). "Gamma-ray dose rates
in a slab phantom,'' Nucl. Technol. 8, 450.
COLBERT, H.M., (1973). SANDYL: A Complete Code for Calculating
Combined PhotonlElectron Tmnsport in Complex Systems, Sandia
Laboratory (Livermore) Report SLL-74-0012 (Sandia National
Laboratory, Albuquerque, New Mexico).
COLEMAN, R. (1969). "Random paths through convex bodies," J. Appl.
Prob. 6,430.
CRISTY,M. AND ECKERMAN, K.F.(1985). Specific Absorbed F m t i o n s
of Energy a t Various Ages from Internal Photon Sources, I. Meth-
ods, Report ORNL/TM-8381 (Oak Ridge National Laboratory, Oak
Ridge, Tennessee).
CROSS,W.G. (1986). Requirements for a n d Status of Basic Nuclear
Data for Radiotherapy with Neutrons, Proceedings of Workshop,
September 1985, (International Atomic Energy Agency, Vienna).
CULLEN,D.E. AND HLAVAC,P.J. (1972). ENDFIB Cross Sections,
Report No. BNL-17100 (Brookhaven National Laboratory, Upton,
New York).
DAVTES,H., BETHE,H.A. AND MAXIMOM, L. (1954). "Theory of brems-
strahlung and pair production. 11, Integral cross section for pair
production," Phys. Rev. 93, 788.
DAVISSON,C.M. AND EVANS,R.D. (1952). "Gamma-ray absorption
coefficients," Rev. Mod. Phys. 24, 79.
DIMBYLOW, P.J. (1980). "Neutron cross-sections and kerma values
for carbon, nitrogen and oxygen from 20 to 50 MeV," Phys. Med.
Biol. 25, 637.
DIMBYLOW, P.J. (1981). "Neutron cross-sections and kerma values
for elements of biomedical importance," pages 341-352 in Proceed-
ings of the Fourth Symposium on Neutron Dosimetry, Report No.
EUR 7448 (Commission of European Community, Brussels).
DIMBYLOW, P.J. (1982). "Neutron cross-section and kerma value cal-
culations for C, N, 0,Mg, Al, P, S, Ar and Ca from 20 to 50 MeV,"
Phys. Med. Biol. 27, 989.
DISTENFELD, C.H. AND COLVETT,R.D. (1966). "Skyshine considera-
tions for accelerator shielding design," Nucl. Sci. Eng. 26, 117.
DOGGETT, J.A. AND SPENCER, L.V. (1956). "Elastic scattering of elec-
trons and positrons by point nuclei," Phys. Rev. 103, 1597.
DOUTHAT,D.A. (1975). "Electron degradation spectra in helium,"
Radiat. Res. 61, 1-20.
EMMETT, M.B. (1975). The MORSE Monte CarloRadiation Transport
Code System, Report No. ORNL-4972 (Oak Ridge National Labora-
tory, Oak Ridge, Tennessee).
ENGLE,W.W., JR.(1967). "A users manual for ANISN, a one-dirnen-
sional discrete ordinate transport code with a n isotropic scatter-
ing," (Computing Technology Center, Union Carbide Corporation,
K-1693).
EVANS,R.D. (1955). The Atomic Nucleus (McGraw-Hill, New York,
New York).
EVANS.R.D. (1958). "Compton effect," page 218 in Handbuch der
Physik, Vol. XXXIV (Springer-Verlag, Berlin).
204 1 REFERENCES
EVANS,R.D. (1968). "X-ray and gamma-ray interactions," pages
93-155 in Radiation Dosimetry, (2nd Ed) Vol. I, Attix, F. H. and
Roesch, W. C., Eds. (Academic Press, New York, New York).
FANO,U. (1953). Degradation and range straggling of high-energy
radiations, Phys. Rev. 92,328-349.
FANO,U. (1954). "Note on the Bragg-Gray cavity principle for mea-
suring energy dissipation," Radiat. Res. l , 237
FANO,U. (1963). "Penetration of protons, alpha particles and
mesons," Ann. Rev. Nucl. Sci. 13, 1.
FANO,U. AND COOPER,J.W. (1968). "Spectra distribution of atomic
oscillator strengths," Rev. Mod. Phys. 40, 441.
FANO,U., SPENCER, L.V., AND BERGER, M.J. (1959). "Penetration and
diffusion of x-rays," page 660-817 in Handbuch der Physik, Vol.
XXXVIII, Part 2 (Springer-Verlag, Berlin).
FERMI,E. AND TELLER, E. (1947). "The capture of negative mesotrons
in matter," Phys. Rev. 72, 399.
FEYNMAN, R.P. (1969). 'Very high-energy collisions of hadrons,"
Phys. Rev. Lett. 23, 1415.
FORD,R.L. AND NELSON,W.R. (1978). The EGS Code System: Com-
puter Programs for the Monte Carlo Simulation ofElectronagnetic
Cascade Showers (Version 3), Report No. 210 (Stanford Linear
Accelerator Center, Stanford, California).
FORD,W.E., III, SANTORO, R.T., ROUSSIN,R.W. AND PLASTER, D.M.
(1976). Modification Number One to the Coupled 100 n-21 y Cross
Section Library for EPR Calculations, Report No. RSIC DLC-37
(Radiation Shielding Information Center, Oak Ridge National
Laboratory, Oak Ridge, Tennessee).
FORKMAN, B. AND SCHRODER, B. (1972). "A review of intermediate
energy photofission," Physica Scripta 5, 105.
FRM, J.L. AND NEGELE,J.W. (1975). 'Theoretical and experimental
determination of nuclear charge distribution," Adv. Nucl. Phys. 8,
219.
FULLER,E.G. (1985). "The photonuclear reaction cross sections for
I2C, 14N and 160, Phys.
" Rep. 127,187-231.
FULLER,E., GERSTENBERG, H., VADERMOLEN,H..AND DUNN,T.C.
(1973). Photonuclear Reaction Data, page 380 in NBS Publication
No. 1 (National Bureau of Standards, Washington, D.C.).
GABRIEL, T.A., ALSMILLER, R.G., JR.,AND GUTHRIE,M.P. (1970). An
Extrapolation Method for Predicting Nucleon a n d Pion Differential
Production Cross Section from High-Energy (73 GEV) Nucleon-
Nucbus Collisions, Report No. ORNL-4542 (Oak Ridge National
Laboratory, Oak Ridge, Tennessee).
GABRIEL,T.A., SANFORD. R.T.. AND ALSMILLER, R.G., JR (1971). "An
approximate high energy alpha-particle-nucleus collision model,"
Nucl. Sci. Eng. 44, 104.
 REFERENCES 1 205
GARBER, D.G. (1975). ENDNB Summary Documentation, Report No.
BNL-17541 (Brookhaven National Laboratory, Upton, New York).
GARBER, D.I.AND KINSEY,R.R. (1976).Neutron Cross Section, Vo~ume
II. Curves Report No. BNL325 (Brookhaven National Laboratory,
Upton, New York).
GAFU, M. AND HEBACH, H. (1981). "Photonuclear reactions a t interme-
diate energies (40 MeVsEs400 MeV)," Phys. Reports 72, 1.
GOERTZEL.G. AND KALOS,M.H. (1958). "Monte Carlo methods in
transport problems," page 315 in Progress in Nuclear Energy, Vol.
11, Series I, Physics and Mathematics, Hughes, D.J., Sanders, J.E.,
and Horowitz, J., Eds. (Pergamon Press, New York, New York).
WLDBERGER, M.L. (1948). "The interaction of high energy neutrons
and heavy nuclei," Phys. Rev. 74, 1269.
GOLDSTEIN, H. (1971). Fundamental Aspects of Reactor Shielding
(Addison-Wesley, New York, New York).
GUTHRIE,M.P., ALSMILLER, R.G.,JR.,AND BERTINI, H.W. (1968). "Cal-
culation of the capture of negative pions in light elements and
comparison with experiments pertaining to cancer radiotherapy,"
Nucl. Instr. Meth. 66, 29.
HAGEDORN, R. AND RANFT,J. (1968). "Statistical thermodynamics
of strong interaction a t high energies. 11. Momentum spectra of
particles produced in p-p collision," Suppl. Nuova Cimento, 6,274.
HALBLEIB, J.A. (1978), "SPHERE: A spherical-geometry, multimate-
rial electrodphoton Monte Carlo transport code," Nucl. Sci. Eng.
66,269.
HALBLEIB, J.A. (1980). ''ACCEFT: A three-dimensional electrodpho-
ton Monte Carlo transport code using combinatorial geometry,"
Nud. Sci. Eng. 75, 200.
HALBLEIB, J.A. AND VANDE VENDER,W.H. (1975) "TIGER: A one
dimensional multimedia electrodphoton Monte Carlo transport
code," Nucl. Sci. Eng. 57, 94.
HALBLEIB, J.A. AND VAN DE VENDERW.H. (1976) "CYLTRAN: A
cylindrical-geometry multimaterial electrodphoton Monte Carlo
transport code," Nucl. Sci. Eng. 61, 288.
HEITLER, W. (1954). The Quantum Theory ofRadiation (Oxford Uni-
versity Press, London).
HENKE,B.L., LEE,P., TANKA, T.J., SHIMABUKURO, R.L., AND FUJI-
WAWA, B.K. (1982). "Low-energy x-ray interaction coefficients:
photoabsorption, scattering, and reflection," Atomic Data and
Nuclear Data Tables, 27, 1.
HERRINS,J.R. AND MERZBACHER, E. (1957). "Statistical problems in
energy loss and ionization processes," J. Elisha Mitchell Sci. Soc.
73, 267.
206 1 REFERENCES
HOFSTADTER, R. (1956). "Electron scattering and nuclear structure,
"Rev. Mod. Phys. 28,214.
HUBBELL, J.H. (1969).Photon Cross Sections, Attenuation Coefficients
and Energy Absorption Coefficients from 10 keV to 100 GeV, NBS
Report No. NSRD-NBS 29 (National Bureau of Standards, Wash-
ington, D.C.).
HUBBELL, J.H. (1977). "Photon mass attenuation and mass energy-
absorption coefficients for H, C, N, 0,Ar and seven mixtures from
1 keV to 20 MeV," Radiat. Res. 70, 58.
HUBBELL, J.H. (1982). "Photon mass attenuation and energy absorp-
tion coefficients from 1keV to 20 MeV," Int. J. Appl. Radiat. Isot.
33, 1269.
HUBBELL, J.H. AND BERGER, M.J. (1968). "Attenuation coefficients,
energy absorption coefficient and related quantities: Photon
atomic cross sections," page 167 in Engineering Compendium on
Radiation Shielding, Vol. I, Jaeger, R.J., Ed. (Springer, New
York).
HUBBELL, J.H., GRIMM, H.A.,AND QUERLIZ, J. (1980)."Pair triplet and
total atomic cross sections (and mass attenuation coefficients) for
1MeV-100 GeV photons in elements Z = 1to 100," J. Phys. Chem.
Ref. Data 9, 1023.
HUBBELL, J.H., VIEGELE, W.J.. BRIGGS, E.A.,BRowN,R.T., COMER, D.T.,
AND HOWERTON, R.J. (1975). "Atomic form factors, incoherent scat-
tering functions and photon scattering cross sections," J. Phys.
Chem. Ref. Data, 4,471.
HUBERT, F., FLEURY, A., BIMBOT, R., AND GARDES,D. (1980). "Range
and stopping power tables for 2.5-100 MeV nucleon heavy ions in
solids," Supplement aux Ann. de Physique, 5, 1.
ICRP (1973). International Commission on Radiological Protection.
Recommendations of the International Commission on Radiologi-
cal Protection, ICRP Publication 21 (Pergamon Press, New York).
ICRP (1977). International Commission on Radiological Protection,
Recommendations of the International Commission on Radiologi-
cal Protection, ICRP Publication 26, Annals of the ICRP 1, No. 3
(Pergamon Press, New York).
ICRP (1985). International Commission on Radiological Protection,
Statement from the 1985 Paris meeting of the ICRP, Annals of The
ICRP 15,(3),i-ii.
ICRP (1991). International Commission on Radiological Protection,
1990Recommendations of the International Commissionon Radio-
logical Protection, ICRP Publication 60, Annals of the ICRP 21,
Nos. 1-3, 1-193 (Pergamon Press, New York)
ICRU (1970). International Commission on Radiation Units and
Measurements, Linear Energy Transfer,ICRU Report 16 (Interna-
 REFERENCES 1 207
tional Commission on Radiation Units and Measurements, Wash-
ington, D.C.).
ICRU (1977). International Commission on Radiation Units and
Measurements, Neutron Dosimetry for Biology a n d Medicine,
ICRU Report 26 (International Commission on Radiation Units
and Measurements, Bethesda, Maryland).
ICRU (1978). International Commission on Radiation Units and
Measurements, Basic Aspects of High Energy Particle Interactions
and Radiation Dosimetry, ICRU Report 28 (International Commis-
sion on Radiation Units and Measurements, Bethesda, Maryland).
ICRU (1979). International Commission on Radiation Units and
Measurements, Methods ofAssessment ofAbsorbed Dose in Clini-
cal Use ofRadionuclides, ICRU Report 32 (International Commis-
sion on Radiation Units and Measurements, Bethesda, Maryland).
ICRU (1980). International Commission on Radiation Units and
Measurements, Radiation Quantities and Units, ICRU Report 33
(International Commission on Radiation Units and Measure-
ments, Washington, D.C.).
ICRU (1983). International Commission on Radiation Units and
Measurements, Microdosimetry, ICRU Report 36 (International
Commission on Radiation Units and Measurements, Bethesda,
Maryland).
ICRU (1984). International Commission on Radiation Units and
Measurements, Stopping Powers for Electrons a n d Positrons,
ICRU Report 37 (International Commission on Radiation Units
and Measurements, Bethesda, Maryland).
ICRU (1986). International Commission on Radiation Units and
Measurements, The Quality Factor in Radiation Protection, ICRU
Report 40 (International Commission on Radiation Units and Mea-
surements, Bethesda, Maryland).
IRVING,D.C., FREESTONE,R.M., JR., AND KAM,F.B.K. (1965). 05R,
A General Purpose Monte Carlo Neutron Transport Code, ORNL
Report No. ORNL-3622 (Oak Ridge National Laboratory, Oak
Ridge, Tennessee).
JACKSON, J.D. (1975). Classical Electrodynamics (Wiley, New York,
New York).
JANNI, J.J. (1966). Calculations of Energy Loss, Range, Pathlength,
Straggling, Multiple Scattering and the Probability of Inelastic
Nuclear Collisions for .1 to 1000 MeV Protons, Report NO.AFWL-
TR-65-150 (Air Force Weapons Laboratory, Albuquerque, New
Mexico).
JENKINS, T.M. NELSON,W.R., AND ALISSANDRO, R., Eds. (1988).
"Monte Carlo Transport of Electrons and Photons" (Plenum Press,
New York).
208 / REFERENCES
JOSEPH,P.M. (1969). "Range energy tables for high energy muons,"
Nucl. Instr. Meth. 75,13.
KALBACH, C. (1978a). "The Griffin model complex particles and direct
nucleon reactions," J. Phys. A283,401.
KALBACH, C. (1978b). "Exciton number dependence of the Griffin
model two-body matrix element," J. Phys. A2-87, 319.
KARR,P.R. AND LAMKIN, J.C. (1949). "Penetration and diffusion of
hard x-rays through thick barriers. 111.Studies of spectral distribu-
tions," Phys. Rev. 76, 1843.
KELLERER, A.M. (1971). "Considerations on the random traversal of
convex bodies and solutions for general cylinders," Radiat. Res.
47,359-376.
KELLERER, A.M. (1981). "Proximity functions for general right cylin-
ders," Radiat. Res. 86, 264-276.
KELLERER, A.M. (1984). "Chord-Length distributions and related
quantities for spheroids," Radiat. Res. 98,425-437.
KENDALL, M.G. AND MORAN, P.A.P. (1963). Geometrical Probability
(Griffin & Co., London).
KHARE,S.P. (1975). "Ionizing collisions of electrons with atoms and
molecules," Radiat. Res. 64, 106.
KINSEY,R. (1979). ENDF-102, Data Format and Procedures for the
Evaluated Nuclear Data File, Report No. BNL-NCS-50496
(National Nuclear Data Center).
KISSEL,L., FRATT, R.H., AND ROY,S.C. (1980). "Rayleigh scattering
by neutral atoms, 100 eV to 10 MeV," Phys. Rev. A22, 1970.
KISSEL,L., QUARLES. C.A., AND PRATI:R.H. (1983) "Shape functions
for atomic-field bremsstrahlung from electrons of kinetic energy
1-500 kev in selected neutral atoms ( 1 ~ Z s 9 2 )"Atomic
, Data and
Nuclear Data Tables 28, 381.
KLEIN,U.,BUCHE,G . ,MATTHAY, H., MECHTERSHEIMER. G . AND
MOLINE,A. (1979). "Energy spectra of neutrons emitted following
the absorption of stopped negative pions in light nuclei," Nucl.
Phys. A324, 339.
KOCH,H.W. AND M m , J.W. (1959). "Bremsstrahlung cross-section
formulas and related data," Rev. Mod. Phys. 31, 920.
KNUTH, D.E. (1981).The Art of ComputerProgramming, Vol. 2 (Addi-
son-Wesley, Reading, Massachusetts).
KRUMBEIN. A.D. AND O'REILLY, B.D. (1968). 'Transport method," in
Section 3.3 of Engineering Compendium on Radiation Shielding
in Vol. 1of ShieldingFundamentals and Methods, sponsored by the
International Atomic Energy Agency (Springer-Verlag, Berlin).
LATHROP, K.D.AND BRINKLEY, F.W. (1973). TWOTRAN-II:An Inter-
faced, Exportable Version of the TWOTRAN Code for Two-Dimen-
 REFERENCES / 209
sional Transport, Report No. LA-4848-MS (Los Alamos National
Laboratory, Los Alamos, New Mexico).
LILLIE,R.A.,ALSMILLER, R.G.,JR.AND MIHALCZO, J.T. (1978). "Design
calculations for a 14 MeV neutron collimator," Nucl. Sci. Eng. 43,
373.
LINDENBAUM, S.J. (1961). "Shielding of high-energy accelerators,"
Ann. Rev. Nucl. Sci. 11, 213-258.
LINHARD, J.,SCHARFF, M., AND SCHIOTT, H.E. (1963). "Range concepts
and heavy ion ranges," Kgl. Danske Vindenskab. Selskab. Mat.-
Fys Medd. 33, No. 14.
LIVINGSTON, M.S. AND BETHE,H.A. (1937). "Nuclear physics," Rev.
Mod. Phys. 9, 282.
MAGURNO, B.A.AND PEARLSTEIN, S. (1981).Proceedings of the Confer-
ence on Nuclear Data Evaluation Methods and Procedures, Report
No. BNL-NCS-51363 (Brookhaven National Laboratory, Upton,
New York).
MARSHAK,R.E. (1947). "Theory of the slowing down of neutrons by
elastic collisions with atomic nuclei," Rev. Mod. Phys. 19, 185.
MCDONALD, F.G. (1969). "IQSY observation of low-energy galactic
and cosmic-rays," page 187 in Annals ofIQSY, Vol. IV,Strickland,
A. C., Ed. (The MIT Press Cambridge, Massachusetts).
MCGINNIES, R.T. (1959). "Energy spectrum resulting from electron
slowing down," National Bureau of Standards, NBS circular 597
(National Bureau of Standards, Washington, D.C.).
MCMASTER, W.J., DELGRANDE, N.K., MALLETT, J.H., AND HUBBELL,
J.H. (1969). Compilation of X-Ray Cross Sections, Report No.
UCRL-50174 (Lawrence Livermore Laboratory, Livermore, Cali-
fornia).
MECHTERSHEIMER, G., BUCHE,G., KLEIN,U., KLUGE,W., MATTHY,H.,
MUNCHMEYER, D., A N D MOLINE,A. (1979). "Energy spectra of
charged particles emitted following the absorption of stopped nega-
tive pions in 12Cnuclei," Nucl. Phys. A324, 379.
MESSEL,H. AND CRAWFORD, D.F. (1970).Electron-Photon ShowerDis-
tribution Function. Tables for Lead, Copper and Air Absorbers
(Pergamon Press, New York).
MIRD (1988). Medical Internal Radiation Dose Committee. MIRD
Primer for Absorbed Dose Calculations (Society of Nuclear Medi-
cine, New York).
MISSONI,G., DICK.,C.E., PLACIOUS, R.C., A N D MOTZ,J.W. (1970).
"Inelastic atomic scattering of O.l,0.2,0.4 and 3.0 MeV electrons,"
Phys. Rev. A2,2309.
MOLI&RE, G. (1948). "Theorie der Streuung schneller geladener Teil-
chen. 11. Mehrsach- und Vielsachstreuung," Z. Naturforsch 3A, 78.
210 / REFERENCES
MOLLER,C. (1932). "Theorie der Durchgangs schneller Elektronen
durch Materie," Ann. Physik 14,531.
MORSE,P.M. AND FESHBACH,H. (1953). Methods of Theoretical Phys-
ics (McGraw Hill, New York).
Mo?T. N.F. (1929)."The scattering offast electrons by atomic nuclei,"
Proc. Roy. Soc. A124,475-442.
Mom, N.F. AND MASSEY,H.S.W. (1965). The Theory of Atomic Colli-
sions (Oxford University Press, London).
MOTZ,J.W., OLSEN,H.A., AND KOCH,H.W. (1964). "Electron scatter-
ing without atomic or nuclear excitation," Rev. Mod. Phys. 36,881.
MOTZ,J.W., OLSEN,H.A., AND KOCH,H.W. (1969). "Pair production
by photons," Rev. Mod. Phys. 41, 581.
MUGHAGBGHAB,S.F., DIVADEENAM,M., AND HOLDEN,N.E. (1981).
Neutron Cross Sections (Academic Press, New York, New York).
M~NCHMEYER, D.. AMOIS, H.I.,BUCHE,G., KLUGE,W., M A ~ H A YH., ,
RANDOLL, H. AND MOLINE,A. (1982). "Energy spectra of charged
particles emitted following the absorption of negative pions
stopped within oxygen-containing organic compounds," Phys.
Med. Biol. 27, 1131.
NCRP (1957). National Council on Radiation Protection and Mea-
surements, Protection Against Neutron Radiution up to 30 Million
Electron Volts, NCRP Report No. 20 (National Bureau of Stan-
dards Handbook 63), superseded by Protection Against Neutron
Radiation, NCRP Report No. 38 (National Council on Radiation
Protection and Measurements, Bethesda, Maryland).
NCRP (1971). National Council on Radiation Protection and Mea-
surements, Protection Against Neutron Radiation, NCRP Report
No. 38 (National Council on Radiation Protection and Measure-
ments, Bethesda, Maryland).
NCRP (1987). National Council on Radiation Protection and Mea-
surements, Recommendation on Limits for Exposure to Ionizing
Radiation, NCRP Report No. 91 (National Council on Radiation
Protection and Measurements, Bethesda, Maryland).
NELMS, A.T. (1953). Gmphs of the Compton Energy -Angle Relution-
ship and the Klein Nishina Formula from 10 keV to 500 MeV, NBS
Circular No. 542 (National Bureau of Standards, Washington,
D.C.).
NORTHCLIFFE, L.C. (1963). "Passage of heavy ions through matter,"
Ann. Rev. Nucl. Sci. 13, 67.
NORTHCLIFFE,L.C. AND SCHILLING, R.F. (1970). "Range and stopping-
power tables for heavy ions," Nuclear Data Tables A7, 233.
OLDENBURG, U. AND BOOZ,J. (1972).Mass Stopping Power and Path-
length of Neutron Produced Recoils in Tissue and Tissue Equiva-
 REFERENCES 1 211
lent Material I. Neutron Energy <6 MeV, Report No. EUR-4786e
(Commission of the European Communities, Brussels).
OLSEN,H. (1955). "Outgoing and ingoing waves in final states and
bremsstrahlung," Phys. Rev. 99, 1335.
PAGES, L., BERTEL,E., JOFFRE. H., AND SKLAVENITIS, L. (1972).
"Energy loss, range and bremsstrahlung yield for 10-KeV to 100-
Mev electrons in various elements and chemical counpounds,"
Atomic Data 4 , l .
PARTICLE DATAGROUP(1986). "Review of particle properties," Phys.
Lett. 170B,1-350.
PIEFKE,F. (1978). "Beziehungen zwischen der Sehnenllngenvertei-
lung und der Verteilung des Abstandes zweier zuflalliger Punkte
im Eikorper." Z. Wahrscheinlichkeitstheorie verw. Gebiete 43,
129-134.
PLACZEK, G. (1946). "On the theory of the slowing down of neutrons
in heavy substances," Phys. Rev. 69,423.
PLECHATY, E.F., CULLEN,D.E., AND HOWERTON, R.J. (1981). Tables
and Graphs of Photon-interaction Cross Sections from 0.1 keV to
100 MeV Derived from the LLL Evaluated-Nuclear-Datu Library,
UCRL-50400, Vol. 6 (Lawrence Livermore Laboratory, Livermore,
California).
PONOMAREV, L.I. (1973). "Molecular structure effects on atomic and
nuclear capture of mesons," Ann. Rev. Nucl. Sci. 23, 395-427.
PRATT, R.H., RON,A., AND TSENG,H.K. (1973). "Atomic photoeffect
above 10 keV," Rev. Mod. Phys. 45,273.
~ T TR.H.,
, TSENG,H.K., LEE,C.M. AND KISSELL,L. (1977). "Brems-
strahlung energy spectra from electrons of kinetic energy 1keV
to 2000 keV incident on neutral atoms 2 F Z F 92," Atomic Data
and Nuclear Data Tables 20,175.
PRICE, B.T., HORTON,C.C., AND SPINNEY. K.T. (1957). Radiation
Shielding (Pergamon Press, New York, New York).
RSIC (1981). Radiation Shielding Information Center. A Capsule
Review of the Computer Code Collection (CCC), Peripheral Shield-
ing Routines (PSR) and Data Library Collection (DLC), Packaged
by the Radiation Shielding Information Center, (Oak Ridge
National Laboratory, Oak Ridge, Tennessee).
RANFT, J. (1967). "Improved Monte Carlo calculation of nucleon-
meson cascade in shielding materials," Nucl. Inst. Meth. 48, 133.
RANFT,J. (1972). "Estimation of radiation problems around high
energy accelerators using calculations of the hadronic cascade in
matter," Particle Acceler. 3, 129.
RANIT, J. AND R o u ~J.T. , (1972). "Hadronic cascade calculations of
angular distributions of integrated secondary particle fluxes from
212 / REFERENCES
external targets and new empirical formulae describing particle
production in proton-nucleus collisions," Particle Acceler. 4, 101.
RANFT,J. AND ROUT~I,J.T. (1974). "Monte Carlo programs for calcu-
lating threedimensional high-energy (50 MeV-500 GeV) hadron
cascades in matter," Comput. Phys. Commun. 7,327.
RANFT,J. AND NELSON,W.R. (1979). "Monte Carlo calculation of
photon, electron, and positron production from primary proton
beams," Nucl. Instr. Meth. 167,443.
RAO,D.V. (1981). Experimental Studies on the Atomic Photoeffectand
Zncoherent Scattering of Low Energy Photons, (Ph.D. dissertation,
Department of Physics and Meteorology, Indian Institute of Tech-
nology, Kharagpur, India).
RAO,V.V. AND RAO,D.V. (1981). "Incoherent scattering of 84.4 and
123.6 keV gamrna-rays in Sc, Y, La, Dy and Pb," Comments At.
Mol. Phys. 14, 3397.
RHOADES. W.A. AND MYNAIT,F.R. (1973). The DOT IZI Two-Dimen-
sional Discrete Ordinates Transport Code, ORNL Report No.
ORNLSrM-4280 (Oak Ridge National Laboratory, Oak Ridge, Ten-
nessee).
RILEY, M.W., WCCALLUM, C.J., AND BIGGS,F. (1975). "Theoretical
electron-atom elastic scattering cross sections," Atomic Data and
Nuclear Data Tables 15, 443.
RINDI,A. AND THOMAS, R. (1975). "Skyshine--A paper tiger, " Parti-
cle Accel. 7, 23.
ROESCH,W.C. (1958). "Dose for nonelectronic equilibrium," Radiat.
Res. 9, 399.
ROESCH,W.C. (1968). "Mathematical theory of radiation fields,"
pages 229-274 in Radiation Dosimetry, Vol. 1,Attix, F.H., Roesch,
W.C., and Tochilin, E., Eds. (Academic Press, New York).
ROGERS,C.C. (1973). Middle Atlantic Neutron Therapy Association
(MANTA) Progress Report (Virginia Commonwealth University,
Medical College of Virginia, Richmond, Virginia).
R o s s ~B.
, (1952). High-Energy Particles (Prentice-Hall, New York,
New York).
Ross~,H.H. AND ROESCH,W.C. (1962). "Field equations in dosimetry,"
Radiat. Res. 16,783-795.
ROUSSW,R.W., WEISBIN, C.R., WHITE,J.E., GREENE,N.M., WRIGHT,
R.Q., AND WRIGHT,J.B. (1980). Vitamin-C: The CTR Processed
Multigroup Cross-Section Libmry for Neutronics Studies, Report
No. ORNL-RSIC-37 (Oak Ridge National Laboratory, Radiation
Shielding Information Center, Oak Ridge, Tennessee).
Ram,J.T. AND SANDBERG, J.V. (1981). "Fortran program, SPALL
for computing spallation reaction cross sections," Comp. Phys.
Comm. 23,411.
 REFERENCES 1 213
RUDSTAM, G. (1966). "Systematics of spallation yields," Z Naturfor-
sch 21a, 1027.
SANTORO, R.T., ALSMILLER, R.G., JR.,AND CHANDLER, K.C. (1973).
"Shielding of manned space vehicles against galactic cosmic ray
protons and alpha particles," Nucl. Sci. Eng. 52, 124.
SANTORO, R.T., ROUSSIN,R.W., AND BARNES,J.M. (1981). "FUNG:
Coupled 35-group neutron and 21-group gamma-ray, P, cross sec-
tions for fusion applications," RSIC No. RSIC-DLC-86 (Oak Ridge
National Laboratory Radiation Shielding Information Center,
Oak Ridge, Tennessee).
SCHUMACHER, M. (1971). "Compton scattering of gamma rays by
electrons bound in heavy atoms," Z. Physik 242,444.
SCOFIELD, J.H. (1973). Theoretical Photoionization Cross Sections
from 1 to 1500 keV, LLL Report No. UCRL-51326 (Lawrence Liver-
more Laboratory, Livermore, California).
SELTZER, S.M.AND BERGER, M.J. (1974). "Transmission and reflection
of electrons by foils," Nucl. Instr. Meth. 119, 157.
SELTZER, S.M.AND BERGER, M.J. (1978). Monte Carlo Studies ofElec-
tron and Photon Transport a t Energies up to 1000MeV, NBS Report
No. NBSIR 78-1534 (National Bureau of Standards, Gaithersburg,
Maryland).
SELTZER, S.M. AND BERGER, M.J. (1982). Status ofElectron Tmmport
Cross Sections, NBS Report No. NBSIR 82-2572 (National Bureau
of Standards, Washington, D.C.).
SELTZER,S.M. AND BERGER,M.J. (1985). "Bremsstrahlung spectra
from electron interactions with screened atomic nuclei and orbital
electrons," Nucl. Instr. Meth. Phys. Res. B12, 95-134.
SILVERBERG, R. AND TSAO,C.H. (1973a). "Partial cross sections in
high-energy nuclear reactions, and astrophysical applications. I.
Targets with Z 5 28," Astron. J. Suppl. 25, 315.
SILVERBERG, R. AND TSAO,C.H. (197313). "Partial cross sections in
high-energy nuclear reactions, and astrophysical applications, 11.
Targets heavier than nickel," Astron. J . Suppl. 25, 335.
SNYDER, W.S., FORD,M.R., WARNER, G.G. AND FISHER,H.L. (1969).
"Estimate of absorbed dose fractions for monoenergetic photon
sources uniformly distributed in various organs of a heterogeneous
phantom," J. Nucl. Med., Supplement No. 3.
SPANIER, J. AND GELBARD, E.M. (1969). Monte Carlo Principles and
Neutron Transport Problems (Addison-Welsey Publishing Co.,
Reading, Pennsylvania).
SPENCER, L.V. (1955). "Theory of electron penetration," Phys. Rev.
98,1597-1615.
SPENCER, L.V. (1959). Energy Dissipation by Fast Electrons, NBS
Monograph No. 1(Government Printing Office, Washington, D.C.).
SPENCER, L.V. (1971). "Bounds to truncation errors in biorthogonal
polynomial approximations, with illustrative applications to
gamma ray transport distributions," NBS J. Res. 75B, 1.
SPENCER, L.V. (1972). "Plural-series approximations of functions,"
NBS J. Res. 76B, 91-108.
SPENCER, L.V.AND FANO,U. (1951). "Penetration and ditfirsion of x-
rays. Calculation of spatial distributions by polynomial expan-
sion," J. Res. National Bureau of Standards 46,446.
SPENCER, L.V. AND FANO,U. (1954). "Energy spectrum resulting from
electron slowing down," Phys. Rev. 93, 1172.
SPENCER, L.V.AND SIMMONS, G.L. (1973)."Improved moment method
calculations of gamma-ray transport: Application to point isotropic
sources in water," Nucl. Sci. Eng. 50, 20.
SPENCER, L.V.. CHILTON,A.B.,AND EISENHAUER, C.M. (1980). Struc-
ture Shielding Against Fallout GammaRays from Nuclear Detonu-
tions, NBS Special Publication 570 (U. S. Government Printing
Office, Washington, D.C.).
STARACE, A.G. (1982). "Theory of atomic photoionization," Handbuch
der Physik 31, 1-122.
STERNHEIMER, R.M. (1952). "The density effect for ionization loss in
various materials," Phys. Rev. 88, 851.
STERNHEIMER, R.M., SELTZER, S.M., AND BERGER, M.J. (1982). "Den-
sity effect for the ionization loss of charged particles in various
substances," Phys. Rev. 26,6067-6076.
STORM, E. AND ISRAEL,H.I. (1970). "Photon cross sections from 1keV
to 100 MeV for elements Z = 1to Z = 100," Nuclear Data Tables
A7, 565.
STOYAN, D., KENDALL, W.S. AND MECKE,J. (1987). StochasticGeome-
try and its Applications, (John Wiley & Sons, New York).
THEUS.R.B. (1973). "Neutron Delivery," page 82 in Proceedings of
Conference on Particle Accelerators in Radiation Therapy, Report
No. LA-5180-C (Los Alamos National Laboratory, Los Alamos,
New Mexico).
THOMPSON, W.L. (1979). MCNP,A General Monte Carlo Code fir
Neutron and Photon Transport,Report No. LA-7396M (LosAlamos
National Laboratory, Los Alamos, New Mexico).
THWAITES, D.I. (1983). "Bragg's rule of stopping power additivity: a
compilation and summary of results," Radiat. Res. 95,495-518.
TSENG,H.K. AND PRATT, R.H. (1971). "Exact screened calculations of
atomic-field bremsstrahlung," Phys. Rev. A3, 100.
TSENG,H.K., PRATT, R.H., AND LEE,C.M. (1979). "Electron brems-
strahlung angular distributions in the 1-500 keV energy range,"
Phys. Rev. A19, 187.
 REFERENCES 1 215
TURNER,J.E., DUTRANNOIS, J., WRIGHT,H.A., HAMM,R.N., BAARLI,
J.,SULLIVAN, A.H., BERGER, M.J., AND SELTZER, S.M. (1972). "The
computation of pion depth-dose curves in water and comparison
with experiment," Radiat. Res. 52, 229.
VANGINNEKEN, A. AND AWSCHALOM, M. (1974). High energyparti-
cle interactions in large targets," Vol.I (Fermi National Accelerator
Laboratory, Batavia, Illinois).
VEIGELE, W.J. (1973). "Photon cross section from 0.1 keV to 1MeV
for elements Z = 1to Z = 94," Atomic Data Tables 5, 51.
WALSKE,M.C. (1952). "The stopping power of K electrons," Phys.
Rev. 88, 1283.
WALSKE,M.C. (1956). "Stopping power of L electrons," Phys. Rev.
101, 940.
WEIL, W. (1983). "Stereology: A survey for geometers," in Convexity
and Its Applications (Birkhauser Verlag, Basel).
WEINBERG, A.M. AND WIGNER,E.P. (1958). The Physical Theory of
Neutron Chain Reactors (University of Chicago Press, Chicago,
Illinois).
WHITE-GRODSTEIN, G. (1957).X-ray Attenuution Coefficientsfrom 10
keV to 100 MeV, NBS Circular 583 (National Bureau of Standards,
Gaithersburg, Maryland).
WILSON,W.E. AND TOBUREN, L.H. (1980). "Delta-ray production in
ion-atom collisions," page 435 in Proceedings of theseventh Sympo-
sium on Microdosimetry, Report No. 7147, Booz, J., Ebert, H. G.
and Hartfiel, H. D., Eds. (Harwood Academic Publishers, London).
WILSON, W.E., HAGGMARK, L.G.,AND BIERSACK, J.P. (1977)."Calcula-
tions of nuclear stopping, ranges and straggling in the low-energy
region," Phys. Rev. B15, 2458.
WRIGHT,H.A., HAMM,R.N., AND TURNER,J.E. (1979). "PION-1. A
Monte Carlo computer program for calculations with negative pion
beams," Radiat. Res. 79, 1-21.
ZEITLER, E. AND OLSEN,H. (1964). "Screening effects in elastic elec-
tron screening," Phys. Rev. 136, 1546.
ZIEGLER, J.F. (1977). Helium Stopping Powers and Ranges in All
Elements (Pergamon Press, New York, New York).
ZIEGLER, J.F. (1980). Handbook of Stopping Cross Sections for Ener-
getic Ions in A11 Elements, (Pergamon Press, New York, New
York).
ZHENGMING, L. (1980)."An electron transport theory of cavity ioniza-
tion," Radiat. Res. 84, 1.
The NCRP
The National Council on Radiation Protection and Measurements
is a nonprofit corporation chartered by Congress in 1964 to:
1. Collect, analyze, develop, and disseminate in the public interest
information and recommendations about (a) protection against
radiation and (b) radiation measurements, quantities, and
units, particularly those concerned with radiation protection;
2. Provide a means by which organizations concerned with the
scientific and related aspects of radiation protection and of
radiation quantities, units, and measurements may cooperate
for effective utilization of their combined resources, and to stim-
ulate the work of such organizations;
3. Develop basic concepts about radiation quantities, units, and
measurements, about the application of these concepts, and
about radiation protection;
4. Cooperate with the International Commission on Radiological
Protection, the International Commission on Radiation Units
and Measurements, and other national and international orga-
nizations, governmental and private, concerned with radiation
quantities, units, and measurements and with radiation pro-
tection.
The Council is the successor to the unincorporated association of
scientists known as the National Committee on Radiation Protection
and Measurements and was formed to carry on the work begun by
the Committee.
The Council is made up of the members and the participants who
serve on the over sixty scientific committees of the Council. The
scientific committees, composed of experts having detailed knowl-
edge and competence in the particular area of the committee's inter-
est draft proposed recommendations. These are then submitted to
the full membership of the Council for careful review and approval
before being published.
The following comprise the current officers and membership of the
Council:
OfFcers
President WARRENK . SINCLAIR
President Elect CHARLESB. MEINHOLD
Vice President S.JAMESADELSTEIN
Secretary and Treasurer W. ROGERNEY
Assistant Secretary CARLD. HOBELMAN
Assistant Treasurer JMES F. BERG
 THE NCRP / 217
Members
I. FABRXKANT
JACOB
R. J . MICHAEL FRY
THOMAS F. GESELL
ETHELS. GILBERT
ROBERTA. GOEPP
JOEL E. GRAY
ARTHURW. GUY
ERICJ. HALL
NAOMIH. HARLEY
WILLIAMR. HENDEE
DONALDG. JACOBS
A. EVERE~TEJAMES. JR.
BERNDWN
KENNETHR. KASE
HAROLDL. KUNDEL
CHARLESE. LAND
RAY D. LLOYD
HARRYR. MAXON
ROGER0.MCCLELLAN
B~RBARA J . MCNEIL
CHARLESB. MEWHOLD
MORTIMERL. MENDWOHN
FREDA. M E T ~ L E R
WILLIAMA. MILLS
DADEW. MOELLER

Honomry Members
LAUMSTON
S. TAYLOR,Honomry President

Currently, the following subgroups are actively engaged in for-

mulating recommendations:
SC 1 Basic Radiation Protection Criteria
SC 1-1 Probability of Causation for Genetic and Developmental
Effects
SC 1-2 The Assesernent of Risk for Radiation Protection huposes
SC 1-3 Collective Dose
SC 16 X-Ray Protection in Dental Offices
THE NCRP
Biological Aspects of Radiation Protection Criteria
SC 40-1 Atomic Bomb Survivor Dosimetry
Operational Radiation Safety
SC 46-2 Uranium Mining and Milling-Radiation Safety
l'wrams
SC 46-4 Calibration of Survey Instrumentation
SC 46-5 Maintaining Radiation Protection Records
SC 46-7 Emergency Planning
SC 46-8 Radiation Protection Design Guidelines for Particle
Accelerator Facilities
SC 46-9 ALARA a t Nuclear Plants
SC 46-10 Assessment of Occupational Dosee from Internal Emitters
SC 46-11 Radiation Protection During Special Medical Procedures
Lnternal Emitter Standards
SC 57-2 Respiratory Tract Model
SC 57-6 Bone Problems
SC 57-8 Leukemia Risk
SC 57-9 Lung Cancer Risk
SC 57-10 Liver Cancer Risk
SC 57-12 Strontium
SC 57-14 Placental 'hansfer
SC 57-15 Uranium
Human Population Exposure Experience
Radiation Exposure Control in a Nuclear Emergency
SC 63-1 Public Knowledge About Radiation
SC 63-2 Criteria for Radiation 1nstavment.afor the Public
Environmental Radioactivity and Waste Management
SC 64-6 Screening Models
SC 64-8 Ocean Disposal of Radioactive Waste
SC 64-9 Effects of Radiation on Aquatic Organism
SC 64-10 Xenon
SC 64-11 Disposal of Low Level Waste
SC 64-16 Uncertainties in Dose Assessment
Quality Assurance and Accuracy in Radiation Protection
Measurements
Biological Effects and Exposure Criteria for Ultrasound
Biological Effects of Magnetic Fields
Efficacy of Radiographic Procedures
Radiation Exposure and Potentially Related Injury
Guidance on Radiation Received in Space Activities
Effects of Radiation on the Embryo-Fetus
Guidance on Occupational and Public Exposure Resulting h m
Diagnostic Nuclear Medicine Procedures
Practical Guidance on the Evaluation of Human Exposures to
Radiofiequency Radiation
Extremely Low-Frequency Electric and Magnetic Fields
Radiation Biology of the Skin (Beta-Ray Dosimetry)
Assessment of Exposures from Therapy
Identification of M a r c h Needs
Radionuclide Contamination
SC 84-1 Decontamination and Decommissioning of Facilities
SC 84-2 Contaminated Soil
 THENCRP 1 219
SC 85 Risk of Lung Cancer from Radon
SC 86 Hot Particles in Eye, Ear and Lung
Ad Hoc Committee on Comparison of Radiation Exposures
Ad Hoc Group on Nuclear Medicine Misadministration
Ad Hoc Group on Plutonium
Ad Hoc Group on Radon
Ad Hoc Group on Video Display Terminals
Study Group on Comparative Risk
%ask Force on Occupational Exposure Levels

In recognition of its responsibility to facilitate and stimulate coop-

eration among organizations concerned with the scientific and re-
lated aspects of radiation protection and measurement, the Council
has created a category of NCRP Collaborating Organizations. Orga-
nizations or groups of organizations that are national or interna-
tional in scope and are concerned with scientific problems involving
radiation quantities, units, measurements, and effects, or radiation
protection may be admitted to collaborating status by the Council.
The present Collaborating Organizations with which the NCRP
maintains liaison are as follows:
American Academy of Dermatology
American Association of Physicists in Medicine
American College of Medical Physics
American College of Nuclear Physicians
American College of Radiology
American Dental Association
American Industrial Hygiene Association
American Institute of Ultrasound in Medicine
American Insurance Services Group
American Medical Association
American Nuclear Society
American Occupational Medical Association
American Pediatric Medical Association
American Public Health Association
American Radium Society
American Roentgen Ray Society
American Society of Radiologic Technologists
American Society for Therapeutic Radiology and Oncology
Association of University Radiologists
Bioelectromagnetics Society
College of American Pathologists
Conference of Radiation Control Program Directors
Electric Power Research Institute
Federal Communications Commission
Federal Emergency Management Agency
Genetics Society of America
Health Physics Society
Institute of Nuclear Power Operations
National Electrical Manufacturers Association
220 1 THENCRP
National Institute of Standards and Technology
Nuclear Management and Resources Council
Radiation Research Society
Radiological Society of North America
Society of Nuclear Medicine
United States Air Force
United States Army
United States Department of Energy
United States Department of Housing and Urban Development
United States Department of Labor
United States Environmental Protection Agency
United States Navy
United States Nuclear Regulatory Commission
United States Public Health Service

The NCRP has found its relationships with these organizations to

be extremely valuable to continued progress in its program.
Another aspect of the cooperative efforts of the NCRP relates to
the special liaison relationships established with various govern-
mental organizations that have an interest in radiation protection
and measurements. This liaison relationship provides: (1) an oppor-
tunity for participating organizations to designate an individual to
provide liaison between the organization and the NCRP; (2) that the
individual designated will receive copies of draft NCRP reports (at
the time that these are submitted to the members of the Council)
with an invitation to comment, but not vote; and (3)that new NCRP
efforts might be discussed with liaison individuals as appropriate,
so that they might have an opportunity to make suggestions on new
studies and related matters. The following organizations participate
in the special liaison program:
Australian Radiation Laboratory
Commissariat a 1'Energie Atomique (France)
Commission of the European Communities
Defense Nuclear Agency
Federal Emergency Management Agency
Japan Radiation Council
National Institute of Standards and Technology
National Radiological Protection hard (UnitedKingdom)
National Research Council (Canada)
Office of Science and Technology Policy
Office of Technology Assessment
Ultrasonics Institute of Australia
United States Air Force
United States A m y
United States Coast Guard
United States Department of Energy
United States Department of Health and Human Services
United States Department of Labor
United States Department of Ransportation
 THENCRP 1 221
United States Environmental Protection Agency
United States Navy
United States Nuclear Regulatory Commission

The NCRP values highly the participation of these organizations

in the liaison program.
The Council's activities are made possible by the voluntary contri-
bution of time and effort by its members and participants and the
generous support of the following organizations:
Alfred P. Sloan Foundation
Alliance of American Insurers
American Academy of Dental Radiology
American Academy of Dermatology
American Association of Physicists in Medicine
American College of Medical Physics
American College of Nuclear Physicians
American College of Radiology
American College of Radiology Foundation
American Dental Association
American Hospital Radiology Administrators
American Industrial Hygiene Association
American Insurance Services Group
American Medical Association
American Nuclear Society
American Occupational Medical Association
American Osteopathic College of Radiology
American Pediatric Medical kssociation
American Public Health Association
American Radium Society
American Roentgen Ray Society
American Society of Radiologic Technologists
American Society for Therapeutic Radiology and Oncolom
American Veterinary Medical Association
American Veterinary Radiology Society
Association of University Radiologists
Battelle Memorial Institute
Center for Devices and Radiological Health
College of American Pathologists
Committee on Radiation Research and Policy Coordination
Commonwealth of Pennsylvania
Defense Nuclear Agency
Eastman Kodak Company
Edison Electric Institute
Edward Mallincbrodt, Jr. Foundation
EG&G Idaho, Inc.
Electric Power Research Institute
Federal Emergency M w g e m e n t Agency
Florida Institute of Phosphate Research
Genetics Society of America
Health Effects Research Foundation (-Kyoto)
Health Physics Society
222 1 THENCRP
Institute of Nuclear Power Operations
James Picker Foundation
Martin Marietta Corporation
National Aeronautics and Space Administration
National Association of Photographic Manufacturers
National Cancer Institute
National Electrical Manufacturera Association
National Institute of Standards and Technology
Nuclear Management and Remurcea Council
Radiation Research Society
Radiological Society of North America
Richard Lounsbery Foundation
Sandia National Laboratory
Society of Nuclear Medicine
United States Department of Energy
United States Department of Labor
United States Environmental Protection Agency
United States Navy
United States Nuclear Regulatory & m m i ~ i o n
Victoreen, Incorporated

To all of these organizations the Council expresses its profound

appreciation for their support.
Initial funds for publication of NCRP reports were provided by a
grant from the James Picker Foundation and for this the Council
wishes to express its deep appreciation.
The NCRP seeks to promulgate information and recommendations
based on leading scientific judgment on matters of radiation protec-
tion and measurement and to foster cooperationamong organizations
concerned with these matters. These efforts are intended to serve
the public interest and the Council welcomes comments and sugges-
tions on its reports or activities from those interested in its work.
NCRP Publications
NCRP publications are distributed by the NCRP Publications
O h .Information on prices and how to order may be obtained by
directing an inquiry to:
NCRP Publications
7910 Woodmont Avenue
Suite 800
Bethesda, MD 20814-3095
The currently available publications are listed below.
NCFW Reports
No. Title
Control and Removal of Radioactive Contamination i n
Laboratories (1951)
Maximum Permissible Body Bur&ns and Maximum Per-
missible Concentrations of Radionuclides in Air and in
Waterfor OccupationalExposure (1959)[IncludesAdden-
dum 1 issued in August 19631
Measurement of Neutron Flux and Spectra for Physical and
Biolo@ Applications (1960)
Measurement of Absorbed Dose ofNeutrons, and of Mixtures
of Neutrons and Gamma: Rays (1961)
Stopping Powers fir Use with Cavity Chambers (1961)
Safe Handling of Radioactbe Materials (1964)
Radiation Protection in Edz~cationalInstitutions (1966)
Dental X-Ray Protection (1970)
Radiation Proction in Veterinary Medicine (1970)
Precautions i n the Management of Patients Who Have
Received Thenapelltic Amounts of Radionuclides (1970)
Protection Against Neutron Radiation (1971)
Protection Against Radiation j h m Bmchythempy Sources
(1972)
Specifiwtion of Gamma-Ray Bmchythempy Sources (1974)
Radiological Factors Affecting Decision-Making i n a
Nuclear Attach (1974)
Krypton-85 in the Atmosphere-Accumulation, Biolo@
Significance, and Control Technology (1975)
/ NCRP PUBLICATIONS

Alpha-Emitting Particles in Lungs (1975)

Tritium Measurement Techniques (1976)
Structuml Shielding Deslgn and Evaluation for Medioal Use
of X Rays and Gamma Rays of Energies Up to 10 MeV
(1976)
Environmental Radiation Measurements (1976)
Radiation Protection Design Guidelines for 0.1-100 MeV
Particle Accelerator Facilities (1977)
Cesium-137 from the Environment to Man: Metabolism and
Dose (1977)
Medical Radiation Exposure of Pregnunt and Potentially
Pregnant Women (1977)
Protection of the Thyroid Gland in the Event of Releases of
Radioiodine (1977)
Instrumentation and Monitoring Methods for Radiation
Protection (1978)
A Handbook of Radioactivity Measurements Procedures,
2nd ed. (1985)
Operational Radiation Safety Program (1978)
Physical, Chemical, and Biological Properties of Radioce-
rium Relevant to Radiation Protection Guidelines (1978)
Radiation Safety Training Criteria for Industrial Radio-
graphy (1978)
Tritium in the Environment (1979)
Tritium and Other Radionuclide Labeled Organic Com-
pounds Incorporated in Genetic Material (1979)
Influence of Dose and Its Distribution in Time on Dose-
Respome Relationships for Low-LET Radiations (1980)
Management of Persons Accidentally Contaminated with
Radionuclides (1980)
Radiofrequency Electromagnetic Fields-Prqaerties,
Quantities and Units, Biophysical Interaction, and
Measurements (1981)
Radiation Protection in Pediatric Radiology (1981)
Dosimetry of X-Ray and Gamma-Ray Beams for Radiation
Therapy in the Energy Range 10 keV to 50 MeV (1981)
Nuclear Medicine-Factors Influencing the Choice and Use
of Radionuclides in Diagnosis and Thempy (1982)
Operational Radiation Safety-Training (1983)
Radiation Protection and Measurement for Low-Voltage
Neutron Generators (1983)
Protection in Nuclear Medicine and Ultrasound Diagnostic
Procedures in Children (1983)
 NCRP PUBLICATIONS 1 225

Biological Effets of Ultrasound: Mechanisms and Clinical

Implications (1983)
Iodine-129: Evaluation ofReleases from Nuclear Power
Generation (1983)
R adwlogical Assessment: Predicting the Transport, Bioac-
cumulation, and Uptake by Man of Radionuclides
Released to the Environment (1984)
Exposures from the Uranium Series with Emphasis on
Radon and Its Daughters (1984)
Evalwtion of Occupational and Environmental Exposures
toRadon and Radon Daughters in the United States (1984)
Neutron Contaminationfrom Medical Electron Accelerators
(1984)
Induction of Thyroid Cancer by Ionizing Radiation (19851
Carbon-14 in the Environment (1985)
SI Units in Radiation Protection and Measurements (1985)
The Experimental Basis for Absorbed-Dose CalcuIations in
Medical Uses of Radionuclides (1985)
General Concepts for the Dosimetry of Internally Deposited
Radionuclides (1985)
Mammography-A User's Guide (1986)
Biological Efhects and Exposure Criteriafor Radwfrequency
Electromagnetic Fields (1986)
Use o f Bioassay Procedures for Assessment of Internal
Radionuclide Deposition (1987)
Radiation Alarms and Access Control Systems (1986)
Genetic Effects from Internally Deposited Radionuclides
(1987)
Neptunium: Radiation Protection Guiaklines (1988)
Public Radiation Exposure from Nuclear Power Genemtion
in the United States (1987)
Ionizing Radiation Exposure of the Populahn of the United
States (1987)
Exposure of the Population in the United States and Canada
from Natural Background Radiation (1987)
Radiation Exposure of the U.S. Population fiom Consumer
Products and Miscellaneous Sources (1987)
Comparative Carcinogenicity of Ionizing Radiation and
Chemicals (1989)
Measurement of Radon and Radon Daughters in Air (1988)
Guictance on Radiation Received in Spate Activities (1989)
Qwlity Assurance for Diagnostic Imaging (1988)
Exposure of the U.S. Population fiom Diagnostic Medical
Radiation (1989)
226 1 NCRP PUBLICATIONS

Exposure o f the U.S. Population from Occupational

Radiation (1989)
Medical X-Ray, Electron Beam and Gamma-Ray Protection
for Energies Up to 50 MeV (Equipment Design, Perfor-
mance and Use) (1989)
Control of Radon in Houses (1989)
The Relative Biological Effectiveness of Radiations ofDiffer-
ent Quality (1990)
Radiation Protection for Medical and Allied Health
Personnel (1989)
Limit for Exposure to "Hot Particles" on the Skin (1989)
Implementation of the Principle of As Low As Reasonably
Achievable (ALARA) for Medical and Dental Personnel
( 1990)
Conceptual Basis for Calculations of Absorbed-Dose
Distributions (1991)
Effects of Ionizing Radiation on Aquatic Organisms (1991)
Some Aspects of Strontium Radiobiology (1991)
Developing Radiation Emergency Phns fordcademic, Medi-
cal or Industrial Facilities (1991)
Calibration ofsurvey Instruments Used i n Radiation Protec-
tion for the Assessment of Ionizing Radiation Fields and
Radioactive Surface Contamination (1991)
Exposure Criteriu for Medical Diagnostic Ultrasound:I. Cri-
teria Based on Thermal Mechanisms (1992)
Maintaining Radiution Protection Records (1992)
Risk Estimutes for Radiation Protection (1993)
Limitation of Exposure to Ionizing Radiation (1993)
Research Needs for Radiation Protection ( 1993)
Radiation Protection in the Mineral Extraction Industry
(1993)
A Practical Guide to the Determination of Human Exposure
to Radiofrequency Fields (1993)
Binders for NCRP reports are available. Two sizes make it possible
to collect into small binders the "old series" of reports (NCRPReports
Nos. 8-30) and into large binders 'he more recent publications (NCRP
Reports Nos. 32-119). Each binaer will accommodate from five to
seven reports. The binders carry the identification WCRP Reports"
and come with label holders which permit the user to attach labels
showing the reports contained in each binder.
The following bound sets of NCRP reports are also available:
Volume I. NCRP Reports Nos. 8 , 2 2
Volume 11. NCRP Reports Nos. 23, 25, 27, 30
 NCRP PUBLICATIONS / 227

Volume III. NCRP Reports Nos. 32,35,36,37

Volume lV. NCRP Reports Nos. 38,40,41
Volume V. NCRP Reports Nos. 42,44,46
Volume VI. NCRP Reports Nos. 47,49,50,51
Volume W. NCRP Reports Nos. 52,53,54,55,57
Volume VIII. NCRP Report No. 58
Volume IX. NCRP Reports Nos. 59,60,61,62,63
Volume X. NCRP Reports Nos. 64,65,66,67
Volume XI. NCRP Reports Nos. 68,69,70,71,72
Volume XII. NCRP Reports Nos. 73,74,75,76
Volume Xm. NCRP Reports Nos. 77,78,79,80
Volume XIV. NCRP Reports Nos. 81,82,83,84,85
Volume XV. NCRP Reports Nos. 86,87,88,89
Volume XVI. NCRP Reports Nos. 90,91,92,93
Volume XVII. NCRP Reports Nos. 94,95,96,97
Volume XVIII. NCRP Reports Nos. 98,99,100
Volume XIX. NCRP Reports Nos. 101,102,103,104
Volume XX. NCRP Reports Nos. 105,106, 107,108
Volume XXI. NCRP Reports Nos. 109,110,111
Volume XXII. NCRP Reports Nos. 112,113,114
Volume XXIII. NCRP Reports Nos. 115,116,117,118

(Titles of the individual reports contained in each volume are

given above.)

NCRP Commentaries
No. Title
1 Krypton-85 in the Atmosphere-With Specific Reference to
the Public Health Significance of the Proposed Controlled
Release at Three Mile Island (1980)
2 Preliminary Evaluation of Criteria for the Disposal of Trans -
uranic Contaminated Waste (1982)
3 Screening Techniques for Determining Compliance with
Environmental Standards-Releases of Radionuclides to
the Atmosphere (1986),Revised (1989)
4 Guidelines for the Release of Waste Water from Nuclear
Facilities with Special Reference to the Public Health Sig-
nificance of the Proposed Release of Treated Waste Waters
at Three Mile Island (1987)
5 Review of the Publication, Living Without Landfills (1989)
6 Radon Exposure of the U.S. Population-Status of the
Problem (1991)
228 1 NCRP PUBLICATIONS

Misadministration of Radioactive Material in Medicine-

Scientific Background (1991)
Uncertainty in NCRP Screening Models Relating to A t m -
spheric Transport, Deposition and Uptake by Humans
(1993)

Proceedings of the Annual Meeting

No. Title
1 Perceptions of Risk, Proceedings of the Fifteenth Annual
Meeting held on March 14-15, 1979 (including Taylor
Lecture No. 3) (1980)
Critical Zss~lesin Setting Radiation Dose Limits, Proceed-
ings of the Seventeenth Annual Meeting held on April 8-
9, 1981 (including Taylor Lecture No. 5) (1982)
Radiation Protection a n d New Medical Diagnostic
Approaches, Proceedings of the Eighteenth Annual Meet-
ing held on April 6-7, 1982 (including Taylor Lecture
No. 6) (1983)
Environmental Radioactivity, Proceedings of the Nine-
teenth Annual Meeting held on April 6-7,1983 (including
Taylor Lecture No. 7) (1983)
Some Issues Important in Developing Basic Radiation Pro-
tection Recommendations, Proceedings of the Twentieth
Annual Meeting held on April 4-5,1984 (including Taylor
Lecture No. 8) (1985)
Radioactive Waste, Proceedings of the Twenty-first Annual
Meeting held on April 34,1985 (including Taylor Lecture
No. 9) (1986)
Nonionizing Electromagnetic Radiations and Ultrasound,
Proceedings of the Twenty-second Annual Meeting held
on April 2-3, 1986 (including Taylor Lecture No. 10)
(1988)
New Dosimetry at Hiroshima andNagasaki andZts Zmplica-
tions for Risk Estimates, Proceedings of the Twenty-third
Annual Meeting held on April 8-9,1987 (including Taylor
Lecture No. 11)(1988)
Radon, Proceedmgs of the Twenty-fourth Annual Meeting
held on March 30-31, 1988 (including Taylor Lecture
No. 12) (1989)
Radiation Protection Today-The NCRP at Sixty Years, Pro-
ceedings of the Twenty-fifth Annual Meeting held on
April 5-6, 1989 (including Taylor Lecture No. 13)(1990)
 NCRP PUBLICATIONS / 229

Health and Ecological Implications of Radioactively Con-

taminated Environments, Proceedings of the Twenty-
sixth Annual Meeting held on April 45,1990 (including
Taylor Lecture No. 14) (1991)
Genes, Cancer and Radiation Protection, Proceedings of the
Twenty-seventh Annual Meeting held on April 3-4,1991
(including Taylor Lecture No. 15) (1992)
Radiation Protection in Medicine, Proceedings of the Twen-
ty-eighthAnnual Meeting held on April 1-2,1992(includ-
ing Taylor Lecture No.16) (1993)

Lauriston S.Taylor Lectures

No. Title
1 Thesquares of tireNatudNumbers in Radiation Protection
by Herbert M. Parker (1977)
Why be Quantitative about Radiation Risk Estimates? by
Sir Edward Pochin (1978)
Radiaabn Pnnktion--Comqts and Tmde O j j i by Hymer L.
Friedell (1979) [Available also in Perceptions of Risk, see
above]
From "Quantity ofRadiation" and "Dose" to '%xposure" and
"Absorbed DoseJJ-An Historical Review by Harold 0.
Wyckoff (1980)
How Well Can We Assess Genetic Risk? Not Very by James F.
Crow (1981) [Available also in Critical Issues in Setting
Radiation Dose Limits, see abovel
Ethics, Trade-offs and Medical Radiation by Eugene L.
Saenger (1982) [Available also in Radiution Protection
and New Medical Diagnostic Approaches, see abovel
The Human Environment-Past, Present and Future by
Merril Eisenbud (1983) [Available also in Environmental
Radioactivity, see abovel
Limitation and Assessment i n Radiation Protection by
Harald H. Rossi (1984) [Available also in Some Isszces
Important in Developing Basic Radiation Protection Rec-
ommendations, see abovel
Truth (and Beauty) in Radiution Measurement by John H.
Harley (1985) [Available also in Radioactive Waste, see
abovel
Biological Effects of Non-ionizing Radiations: Cellular
Properties and Interactions by Herman P. Schwan (1987)
[Available also in Nonionizing Electromagnetic Radia-
tions and UZtmsound, see abovel
230 1 NCRP PUBLICATIONS

How to be Quantitative about Radiation Risk Estimates by

Seymour Jablon (1988) [Available also in New Dosimetry
at Hiroshima: and Nagasaki and its Implications for Risk
Estimates, see above]
How Safe is Safe Enough? by Bo Lindell(1988) [Available
also in Radon, see above]
Radiobiology and Radiation Protection: The Past Century
and Prospects for the Future by Arthur C. Upton (1989)
[Available also in Radiation Protection Today, see above]
Radiation Protection and the Internal Emitter Saga by
J. Newel1 Stannard (1990) [Available also in Health and
Ecological Implications of Radioactively Contaminated
Environments, see abovel
When is a Dose Not a Dose? by Vidor P. Bond (1992) [Avail-
able also in Genes, Cancer and Radiation Protection, see
above]
Dose and Risk in Diagnostic Radiology: How Big? How
Little? by Edward W. Webster (1992)[Available also in
Radiation Protection in Medicine, see abovel
Science, Radiation Protection and the NCRP by Warren K.
Sinclair (1993)

Symposium Proceedings
The Control of Exposure of the Public to Ionizing Radiation
i n the Event ofAccident or Attack, Proceedings of a Sympo-
sium held April 27-29, 1981 (1982)

NCRP Statements
No. Title
1 "Blood Counts, Statement of the National Committee on
Radiation Protection," Radiology 63,428 (1954)
"Statements on Maximum Permissible Dose h m Televi-
sion Receivers and Maximum Permissible Dose to the
Skin of the Whole Body," Am. J. Roentgenol., Radium
Ther. and Nucl. Med. 84, 152 (1960) and Radiology 75,
122 (1960)
X-Ray Protection Standards for Home Television Receivers,
Interim Statement of the National Council on Radiation
Protection and Measurements (1968)
Specification of Units ofNatural Uranium and Natural Tho-
rium, Statement of the National Council on Radiation
Protection and Measurements, (1973)
 NCRP PUBLICATIONS 1 231

5 NCRP Statement on Dose Limit for Nelltrons (1980)

6 Control of Air Emissions of Radionuclides (1984)
7 The Probability That a Particular Malignancy May Have
Been Caused by a Specified Irradiation (1992)

Other Documents
The following documents of the NCRP were published outside
of the NCRP Report, Cornmen- and Statement series:
Somatic Radiation Dose for the Geneml Population, Report
of the Ad Hoc Committee of the National Council on Radi-
ation Protection and Measurements, 6 May 1959,Science,
February 19,1960, Vol. 131, No. 3399, pages 482-486
Dose Effect Modifying Factors I n Radiation Protection,
Report of Subcommittee M-4 (Relative Biological Effec-
tiveness) of the National Council on Radiation Protection
and Measurements, Report BNL 50073 (T-471) (1967)
Brookhaven National Laboratory (National Technical
Information Service Springfield, Virginia)
The following documents are now superseded andlor out of print:

NCRP Reports
No. Title
1 X-Ray Protection (1931) [Superseded by NCRP Repart No. 31
2 R d m Pm&tion (1934) [Superseded by NCRP Report No. 41
3 X-Ray Protection (1936) [Superseded by NCRP Report No. 61
4 Radium Prdection (1938) [Supersededby NCRP Repart No. 131
5 Safe H a d i n g of Radioactive Luminow Compound (1941)
[Out of Print]
Medicul X-Ray Protection Up to Two Million Volts (1949)
[Superseded by NCRP Report No. 181
Safe Handling of Radioactive Isotopes (1949) [Superseded
by NCRP Report No. 301
Recommendations for WasteDisposal of Phosphorus32 and
Iodine-131 for Medical Users (1951) [Out of Print]
Radiological Monitoring Methods and Instruments (1952)
[Superseded by NCRP Report No. 571
Maximum Permissible Amounts of Radioisotopes i n the
Human Body and Maximum Permissible Concentmtion+
in Air and Water (1953) [Superseded by NCRP Report
No. 221
Recommendations for the Disposal of Carbon-14 Wastes
(1953) [Superseded by NCRP Report No. 81.1
232 / NCRP PUBLICATIONS

Protection Against Radiations from Radium, Cobalt-60 and

Cesium-I37 (1954)[Superseded by NCRP Report No. 241
Protection Against Betatron-Synchrotron Radiations Up to
100 Million Electron Volts (1954)[Superseded by NCRP
Report No. 511
Safe Handling of Cadavers Containing Radioactive Isotopes
(1953)[Superseded by NCRP Report No. 211
Radioactive-Waste Disposal i n the Ocean (1954)[Out of
Print]
Permissible Dose from External Sources of Ionizing Radia-
tion (1954)including Maximum Permissible Exposures to
Man, Addendum to National Bureau of Standards Hand-
book 59 (1958)[Superseded by NCRP Report No. 391
X-Ray Prdection (1955)[Supersededby NCRP Report No. 261
Regulation of Radiation Exposure by Legislative Means
(1955)[Out of Print]
ProtectionAgainst Neutron Radiation Up to30 Million Elec-
tron Volts (1957)[Superseded by NCRP Report No. 381
Safe Handling of Bodies Containing Radiwctive Isotopes
(1958)[Superseded by NCRP Report No. 371
Protection Against Radiations from Sealed Gamma Sources
(1960)[Superseded by NCRP Reports No. 33,34 and 401
Medical X-Ray Protection Up to Three Million Volts (1961)
[Superseded by NCRP Reports No. 33,34,35and 361
A Manual of Radioactivity Procedures (1961)[Superseded
by NCRP Report No. 581
Exposure to Radiation in an Emergency (1962)[Superseded
by NCRP Report No. 421
Shielding for High-Energy Electron Accelerator Znstalla-
tions (1964)[Superseded by NCRP Report No. 511
Medical X-Ray and Gamma-Ray Protection for Energies up
to 10 MeV-Equipment Design and Use (1968)[Super-
seded by NCRP Report No. 1021
Medical X-Ray and Gamma-Ray Protection for Energies Up
to 10 MeV-Structural Shielding Design and Evaluation
Handbook (1970)[Superseded by NCRP Report No. 491
Basic Radiation Protection Criteria (1971)[Superseded by
NCRP Report No. 911
Review of the Current State ofRadiation Protection Philoso-
phy (1975)[Superseded by NCRP Report No. 911
Natuml Background Radiation in the United States (1975)
[Superseded by NCRP Report No. 941
Radiation Protection for Medical and Allied Health Person-
nel (1976)[Superseded by NCRP Report No. 1051
 NCRP PUBLICATIONS 1 233

53 Review ofNCRPRadiation Dose Limit forEmbryo and Fetus

in Occupationally -Exposed Women (1977) [Out of Print]
56 R&n Exposwv from ConswnerProdudsan&M i s o e b u s
S o m (1977) [Superseded by NCRP Report No. 951
58 A Handbook of Radioactivity Measurements Procedures,
1st ed. (1978) [Superseded by NCRP Report No. 58,
2nd ed.]
66 Mammography (1980) [Out of Print]
91 Reoommeruhhns on Limits fir Eqmm to Ionizing RacEiation
(1987) [Superseded by NCRP Report No. 1161

NCRP Proceedings
No. Title
2 Quantitative Risk in Standards Setting, Proceedings of the
Sixteenth Annual Meeting held on April 2-3, 1980 [Out
of Print]
Index
Adjoint methods, 19, 31 Fluorescence radiation, 97
Attenuation coefficient, 111 Flux density, 7
Auger electrons, 97 flux-density distribution, 7
Bremsstrahlung, 122 Geometrical factors, 74
Charged-particle equilibrium, 13 autologous geometry, 82, 186
Compton scattering, 97 heterologous geometry, 82, 188
Continuous Slowing-Down Geometric reduction factor, 76, 81
Approximation, 23, 40, 126 compilation of, 186
Cross section, 9, 20 Green's theorem, 8
assumption for calculating, 21 Hadrons, 153
distributions of, 9 Hartree-Fock wave functions, 103
energy-transfer coefficient, 13 Importance function, 181
list of references, 25 Internal dosimetry. 68
macroscopic, 9 Kenna, 13
mean free path, 21 charged-particle equilibrium, 13
mirroecopic, 9 kerma factore. 146
scattering center. 21 Klein-Nishina equation, 103
Discrete-ordinate method, 36 Linear energy transfer (LET), 84
codes for, 49 unrestricted (LET), 84
dosimetry calculations, 60 Mean excitation energy, 131
neutron-photon transport, 49,67 Mean free path, 21
Dose, 1 Microdoeimetry, 3
absorbed-dose rate, 2 specific energy, 2
calculation of, 3, 6 Moment method, 24,47
deterministic approach, 4 Monte-Carlo technique, 6, 61
energy imparted, 1 and CSDA, 66
h m neutrons, 167 efficiency of, 63
geometric considerations, 74 eventby-event, 65
internal dosimetry, 58 transport codes, 70
ionizing radiation, 1 variance-reduction metho&, 66
kerma, 13 Neutron cmsa sections, 139
measurement of, 3 Neutron-photon transport, 49,67
Monte-Carlo method, 6.72 Nucleon-meeon tramport, 70
stochastic quantities, 1 Pair production. 103
total dose, 10 Particle current density, 7
banaport equatione, 5, 8 Particle density, 6
Dose equivalent, 84,167 Phase space,8
Doae rate, 2 Photoelectric effect, 93
Elastic .scattering, 113 Photonuclear effect, 107
electrons, 113 Pions, 147
neutrons, 140 Point-pair dietanee distribution, 76,
protons, 117 81
Electron-photon transport, 68 Point-source kernal, 76
Energy-absorption coefficient, 111 differential, 75
Energy conservation, 29 integral, 75
Energy-transfer coefficient, 13 isotropic, 78
Fano's theorem, 29 Probability density function, 61
 INDEX / 235
Quality factor, 84-85 plane muree. 16
Radiation equilibrium, 51 point aourc8,16
quasi-equilibrium, 56 reciprocity, 33
traneient equilibrium, 57 simplified repre~entation,15
Random (pseudorandom) number Specific Energy, 2
generator, 62 Stochastic quantities, 1
Rayleigh scattering, 106 Stopping power, 12, 128
Receptor, 17 denaity-effect correction, 132
absorbed dose in, 74 mass stopping power, 12
anthropomorphic, 17 'Ranslation operator, 27
receptor-free field, 17 lhmsport equations, 5,8,26
reciprocity, 33 demity d i n g , 28
for two kind8 of particles, 12
Reciprocity theorem, 77 in different coordinate eyatems, 34 .
Scattering operator, 27 integral form of, 26
Shielding, 172 iterative solution of, 27
Skyahine for neutrons, 178 linear and nonlinear, 10
fhurces, 6, 14 Neumann expansion, 28
fictitious s o m a . 15 radiation equilibrium, 38
monoenergetic, 16 euperposition, 30
non-uniform, 57 transport codes, 67