Professional Documents
Culture Documents
Electromagnetic Theory Bo Thide 2007
Electromagnetic Theory Bo Thide 2007
Field Theory
B O T HID
U PSILON B OOKS
Electromagnetic
Field Theory
B O T HID
Swedish Institute of Space Physics
Uppsala, Sweden
and
Department of Astronomy and Space Physics
Uppsala University, Sweden
and
LOIS Space Centre
School of Mathematics and Systems Engineering
Vxj University, Sweden
Also available
E LECTROMAGNETIC F IELD T HEORY
E XERCISES
by
Tobia Carozzi, Anders Eriksson, Bengt Lundborg,
Bo Thid and Mattias Waldenvik
Freely downloadable from
www.plasma.uu.se/CED
This book was typeset in LATEX 2 (based on TEX 3.141592 and Web2C 7.4.4) on an HP Visualize 90003600 workstation running HP-UX 11.11.
Copyright 19972006 by
Bo Thid
Uppsala, Sweden
All rights reserved.
Electromagnetic Field Theory
ISBN X-XXX-XXXXX-X
Contents
Contents
List of Figures
ix
xiii
Preface
xv
Classical Electrodynamics
1.1 Electrostatics
1.1.1 Coulombs law
1.1.2 The electrostatic field
1.2 Magnetostatics
1.2.1 Ampres law
1.2.2 The magnetostatic field
1.3 Electrodynamics
1.3.1 Equation of continuity for electric charge
1.3.2 Maxwells displacement current
1.3.3 Electromotive force
1.3.4 Faradays law of induction
1.3.5 Maxwells microscopic equations
1.3.6 Maxwells macroscopic equations
1.4 Electromagnetic duality
1.5 Bibliography
1.6 Examples
1
2
2
3
6
6
7
9
10
10
11
12
15
15
16
18
20
Electromagnetic Waves
2.1 The wave equations
2.1.1 The wave equation for E
2.1.2 The wave equation for B
2.1.3 The time-independent wave equation for E
25
26
26
27
27
ix
Contents
2.2
2.3
2.4
2.5
30
31
32
33
34
36
Plane waves
2.2.1 Telegraphers equation
2.2.2 Waves in conductive media
Observables and averages
Bibliography
Example
Electromagnetic Potentials
3.1 The electrostatic scalar potential
3.2 The magnetostatic vector potential
3.3 The electrodynamic potentials
3.4 Gauge transformations
3.5 Gauge conditions
3.5.1 Lorenz-Lorentz gauge
3.5.2 Coulomb gauge
3.5.3 Velocity gauge
3.6 Bibliography
3.7 Examples
39
39
40
40
41
42
43
47
49
49
51
53
53
53
56
58
58
59
62
63
65
67
69
71
74
74
75
76
77
77
78
6.2
6.3
6.4
6.5
81
85
85
89
91
92
93
94
96
104
108
115
122
124
Relativistic Electrodynamics
7.1 The special theory of relativity
7.1.1 The Lorentz transformation
7.1.2 Lorentz space
7.1.3 Minkowski space
7.2 Covariant classical mechanics
7.3 Covariant classical electrodynamics
7.3.1 The four-potential
7.3.2 The Linard-Wiechert potentials
7.3.3 The electromagnetic field tensor
7.4 Bibliography
131
131
132
134
139
142
143
143
144
147
150
153
153
153
159
160
167
169
Formul
F.1 The electromagnetic field
F.1.1 Maxwells equations
F.1.2 Fields and potentials
F.1.3 Force and energy
F.2 Electromagnetic radiation
171
171
171
171
172
172
xi
Contents
F.3
F.4
F.5
xii
172
172
172
173
173
173
174
174
174
174
174
174
175
175
175
175
175
176
176
178
M Mathematical Methods
M.1 Scalars, vectors and tensors
M.1.1 Vectors
M.1.2 Fields
M.1.3 Vector algebra
M.1.4 Vector analysis
M.2 Analytical mechanics
M.2.1 Lagranges equations
M.2.2 Hamiltons equations
M.3 Examples
M.4 Bibliography
179
179
179
181
184
186
188
188
189
190
198
Index
199
List of Figures
1.1
1.2
1.3
1.4
3
5
7
13
5.1
73
6.1
6.2
6.3
6.4
6.5
6.6
6.7
6.8
6.9
6.10
6.11
6.12
6.13
6.14
Linear antenna
Electric dipole antenna geometry
Loop antenna
Multipole radiation geometry
Electric dipole geometry
Radiation from a moving charge in vacuum
An accelerated charge in vacuum
Angular distribution of radiation during bremsstrahlung
Location of radiation during bremsstrahlung
Radiation from a charge in circular motion
Synchrotron radiation lobe width
The perpendicular field of a moving charge
Electron-electron scattering
Vavilov-Cerenkov
cone
79
80
82
87
89
94
96
105
106
109
111
113
115
120
7.1
7.2
7.3
133
139
140
8.1
160
190
xiii
Preface
This book is the result of a more than thirty year long love affair. In 1972, I took
my first advanced course in electrodynamics at the Department of Theoretical
Physics, Uppsala University. A year later, I joined the research group there and
took on the task of helping my supervisor, professor P ER -O LOF F RMAN, with
the preparation of a new version of his lecture notes on the Theory of Electricity.
These two things opened up my eyes for the beauty and intricacy of electrodynamics, already at the classical level, and I fell in love with it. Ever since that
time, I have on and off had reason to return to electrodynamics, both in my studies, research and the teaching of a course in advanced electrodynamics at Uppsala
University some twenty odd years after I experienced the first encounter with this
subject.
The current version of the book is an outgrowth of the lecture notes that I
prepared for the four-credit course Electrodynamics that was introduced in the
Uppsala University curriculum in 1992, to become the five-credit course Classical
Electrodynamics in 1997. To some extent, parts of these notes were based on
lecture notes prepared, in Swedish, by B ENGT L UNDBORG who created, developed
and taught the earlier, two-credit course Electromagnetic Radiation at our faculty.
Intended primarily as a textbook for physics students at the advanced undergraduate or beginning graduate level, it is hoped that the present book may be
useful for research workers too. It provides a thorough treatment of the theory
of electrodynamics, mainly from a classical field theoretical point of view, and
includes such things as formal electrostatics and magnetostatics and their unification into electrodynamics, the electromagnetic potentials, gauge transformations, covariant formulation of classical electrodynamics, force, momentum and
energy of the electromagnetic field, radiation and scattering phenomena, electromagnetic waves and their propagation in vacuum and in media, and covariant
Lagrangian/Hamiltonian field theoretical methods for electromagnetic fields, particles and interactions. The aim has been to write a book that can serve both as
an advanced text in Classical Electrodynamics and as a preparation for studies in
Quantum Electrodynamics and related subjects.
In an attempt to encourage participation by other scientists and students in
the authoring of this book, and to ensure its quality and scope to make it useful
in higher university education anywhere in the world, it was produced within a
xv
Preface
World-Wide Web (WWW) project. This turned out to be a rather successful move.
By making an electronic version of the book freely down-loadable on the net,
comments have been received from fellow Internet physicists around the world
and from WWW hit statistics it seems that the book serves as a frequently used
Internet resource.1 This way it is hoped that it will be particularly useful for
students and researchers working under financial or other circumstances that make
it difficult to procure a printed copy of the book.
Thanks are due not only to Bengt Lundborg for providing the inspiration to
write this book, but also to professor C HRISTER WAHLBERG and professor G RAN
FLDT, Uppsala University, and professor YAKOV I STOMIN, Lebedev Institute,
Moscow, for interesting discussions on electrodynamics and relativity in general
and on this book in particular. Comments from former graduate students M ATTIAS
WALDENVIK, T OBIA C AROZZI and ROGER K ARLSSON as well as A NDERS E RIKS SON , all at the Swedish Institute of Space Physics in Uppsala and who all have
participated in the teaching on the material covered in the course and in this book
are gratefully acknowledged. Thanks are also due to my long-term space physics
colleague H ELMUT KOPKA of the Max-Planck-Institut fr Aeronomie, Lindau,
Germany, who not only taught me about the practical aspects of high-power radio
wave transmitters and transmission lines, but also about the more delicate aspects
of typesetting a book in TEX and LATEX. I am particularly indebted to Academician
professor V ITALIY L AZAREVICH G INZBURG, 2003 Nobel Laureate in Physics, for
his many fascinating and very elucidating lectures, comments and historical notes
on electromagnetic radiation and cosmic electrodynamics while cruising on the
Volga river at our joint Russian-Swedish summer schools during the 1990s, and
for numerous private discussions over the years.
Finally, I would like to thank all students and Internet users who have downloaded and commented on the book during its life on the World-Wide Web.
I dedicate this book to my son M ATTIAS, my daughter K AROLINA, my
high-school physics teacher, S TAFFAN RSBY, and to my fellow members of the
C APELLA P EDAGOGICA U PSALIENSIS.
Uppsala, Sweden
December, 2006
1 At
xvi
B O T HID
www.physics.irfu.se/bt
the time of publication of this edition, more than 500 000 downloads have been recorded.
1
Classical
Electrodynamics
Classical electrodynamics deals with electric and magnetic fields and interactions
caused by macroscopic distributions of electric charges and currents. This means
that the concepts of localised electric charges and currents assume the validity of
certain mathematical limiting processes in which it is considered possible for the
charge and current distributions to be localised in infinitesimally small volumes of
space. Clearly, this is in contradiction to electromagnetism on a truly microscopic
scale, where charges and currents have to be treated as spatially extended objects
and quantum corrections must be included. However, the limiting processes used
will yield results which are correct on small as well as large macroscopic scales.
It took the genius of JAMES C LERK M AXWELL to unify electricity and magnetism into a super-theory, electromagnetism or classical electrodynamics (CED),
and to realise that optics is a subfield of this super-theory. Early in the 20th century, H ENDRIK A NTOON L ORENTZ took the electrodynamics theory further to the
microscopic scale and also laid the foundation for the special theory of relativity,
formulated by A LBERT E INSTEIN in 1905. In the 1930s PAUL A. M. D IRAC expanded electrodynamics to a more symmetric form, including magnetic as well
as electric charges. With his relativistic quantum mechanics, he also paved the
way for the development of quantum electrodynamics (QED) for which R ICHARD
P. F EYNMAN, J ULIAN S CHWINGER, and S IN -I TIRO T OMONAGA in 1965 received
their Nobel prizes in physics. Around the same time, physicists such as S HELDON
G LASHOW, A BDUS S ALAM, and S TEVEN W EINBERG were able to unify electrodynamics the weak interaction theory to yet another super-theory, electroweak
theory, an achievement which rendered them the Nobel prize in physics 1979.
The modern theory of strong interactions, quantum chromodynamics (QCD), is
influenced by QED.
In this chapter we start with the force interactions in classical electrostatics
1. Classical Electrodynamics
and classical magnetostatics and introduce the static electric and magnetic fields
to find two uncoupled systems of equations for them. Then we see how the conservation of electric charge and its relation to electric current leads to the dynamic
connection between electricity and magnetism and how the two can be unified
into one super-theory, classical electrodynamics, described by one system of
eight coupled dynamic field equationsthe Maxwell equations.
At the end of this chapter we study Diracs symmetrised form of Maxwells
equations by introducing (hypothetical) magnetic charges and magnetic currents
into the theory. While not identified unambiguously in experiments yet, magnetic charges and currents make the theory much more appealing, for instance by
allowing for duality transformations in a most natural way.
1.1 Electrostatics
The theory which describes physical phenomena related to the interaction between stationary electric charges or charge distributions in a finite space which
has stationary boundaries is called electrostatics. For a long time, electrostatics,
under the name electricity, was considered an independent physical theory of its
own, alongside other physical theories such as magnetism, mechanics, optics and
thermodynamics.1
F(x) =
=
=
(1.1)
40 |x x0 |3
40
40
|x x0 |
|x x0 |
1 The
The whole theory of electrostatics constitutes a group of abstract ideas and general propositions, formulated in the clear and concise language of geometry and algebra, and connected with one another by the rules of strict logic. This whole fully satisfies the reason of
a French physicist and his taste for clarity, simplicity and order. . . .
Electrostatics
q
x x0
x
q0
x0
O
F IGURE 1.1: Coulombs law describes how a static electric charge q, located at
a point x relative to the origin O, experiences an electrostatic force from a static
electric charge q0 located at x0 .
where in the last step formula (F.71) on page 177 was used. In SI units, which we
shall use throughout, the force F is measured in Newton (N), the electric charges q
and q0 in Coulomb (C) [= Ampre-seconds (As)], and the length |x x0 | in metres
(m). The constant 0 = 107 /(4c2 ) 8.8542 1012 Farad per metre (F/m) is
the vacuum permittivity and c 2.9979 108 m/s is the speed of light in vacuum.
In CGS units 0 = 1/(4) and the force is measured in dyne, electric charge in
statcoulomb, and length in centimetres (cm).
F
q0 q
Estat lim
(1.2)
1. Classical Electrodynamics
that surrounds it, regardless of the existence of a second charge anywhere in this
space.2
Using (1.1) and equation (1.2) on page 3, and formula (F.70) on page 177,
we find that the electrostatic field Estat at the field point x (also known as the
observation point), due to a field-producing electric charge q0 at the source point
x0 , is given by
q0 x x0
q0 0
q0
1
1
Estat (x) =
=
=
40 |x x0 |3
40
40
|x x0 |
|x x0 |
(1.3)
In the presence of several field producing discrete electric charges q0i , located
at the points x0i , i = 1, 2, 3, . . . , respectively, in an otherwise empty space, the assumption of linearity of vacuum3 allows us to superimpose their individual electrostatic fields into a total electrostatic field
Estat (x) =
1
40
0
0 x xi
q
i 0 3
x xi
i
(1.4)
If the discrete electric charges are small and numerous enough, we introduce
the electric charge density , measured in C/m3 in SI units, located at x0 within
a volume V 0 of limited extent and replace summation with integration over this
volume. This allows us to describe the total field as
Z
Z
1
1
x x0
1
3 0
0
=
Estat (x) =
d3x0 (x0 )
d
x
(x
)
40 V 0
40 V 0
|x x0 |
|x x0 |3
Z
1
(x0 )
d3x0
=
40
|x x0 |
V0
(1.5)
where we used formula (F.70) on page 177 and the fact that (x0 ) does not depend
on the unprimed (field point) coordinates on which operates.
2 In the preface to the first edition of the first volume of his book A Treatise on Electricity and Magnetism, first published in 1873, James Clerk Maxwell describes this in the following almost poetic manner
[9]:
For instance, Faraday, in his minds eye, saw lines of force traversing all space where the
mathematicians saw centres of force attracting at a distance: Faraday saw a medium where
they saw nothing but distance: Faraday sought the seat of the phenomena in real actions
going on in the medium, they were satisfied that they had found it in a power of action at
a distance impressed on the electric fluids.
3 In fact, vacuum exhibits a quantum mechanical nonlinearity due to vacuum polarisation effects manifesting themselves in the momentary creation and annihilation of electron-positron pairs, but classically
this nonlinearity is negligible.
Electrostatics
x x0i
x
q0i
x0i
V0
O
F IGURE 1.2:
We emphasise that under the assumption of linear superposition, equation (1.5) on page 4 is valid for an arbitrary distribution of electric charges, including discrete charges, in which case is expressed in terms of Dirac delta
distributions:
(x0 ) = q0i (x0 x0i )
(1.6)
as illustrated in figure 1.2. Inserting this expression into expression (1.5) on page 4
we recover expression (1.4) on page 4.
Taking the divergence of the general Estat expression for an arbitrary electric
charge distribution, equation (1.5) on page 4, and using the representation of the
Dirac delta distribution, formula (F.73) on page 177, we find that
1
x x0
d3x0 (x0 )
40 V 0
|x x0 |3
Z
1
1
=
d3x0 (x0 )
40 V 0
|x x0 |
Z
1
1
3 0
0
2
=
d x (x )
40 V 0
|x x0 |
Z
1
(x)
=
d3x0 (x0 ) (x x0 ) =
0 V 0
0
Estat (x) =
(1.7)
1. Classical Electrodynamics
(1.8)
(1.9a)
(1.9b)
1.2 Magnetostatics
While electrostatics deals with static electric charges, magnetostatics deals with
stationary electric currents, i.e., electric charges moving with constant speeds, and
the interaction between these currents. Here we shall discuss this theory in some
detail.
F(x) =
(1.10)
1
107
(F/m) 4 107 (H/m) = 2 (s2 /m2 )
4c2
c
(1.11)
Magnetostatics
C
I dl
x x0
I 0 dl0
x
C0
x0
O
F IGURE 1.3: Ampres law describes how a small loop C, carrying a static
electric current I through its tangential line element dl located at x, experiences
a magnetostatic force from a small loop C 0 , carrying a static electric current I 0
through the tangential line element dl0 located at x0 . The loops can have arbitrary
shapes as long as they are simple and closed.
dl
dl
4 C C 0 |x x0 |3
Since the integrand in the first integral is an exact differential, this integral vanishes and we can rewrite the force expression, equation (1.10) on page 6, in the
following symmetric way
F(x) =
0 II 0
4
I I
C
C0
x x0
dl dl0
|x x0 |3
(1.13)
1. Classical Electrodynamics
can be defined as
def
dBstat (x)
x x0
0 I 0 0
dl
4
|x x0 |3
(1.14)
which expresses the small element dBstat (x) of the static magnetic field set up at
the field point x by a small line element dl0 of stationary current I 0 at the source
point x0 . The SI unit for the magnetic field, sometimes called the magnetic flux
density or magnetic induction, is Tesla (T).
If we generalise expression (1.14) to an integrated steady state electric current
density j(x), measured in A/m2 in SI units, we obtain Biot-Savarts law:
Z
Z
0
0
x x0
1
3 0
0
Bstat (x) =
=
d3x0 j(x0 )
d
x
j(x
)
4 V 0
4 V 0
|x x0 |
|x x0 |3
Z
0
j(x0 )
=
d3x0
4
|x x0 |
V0
(1.15)
where we used formula (F.70) on page 177, formula (F.57) on page 177, and the
fact that j(x0 ) does not depend on the unprimed coordinates on which operates.
Comparing equation (1.5) on page 4 with equation (1.15), we see that there exists
a close analogy between the expressions for Estat and Bstat but that they differ
in their vectorial characteristics. With this definition of Bstat , equation (1.10) on
page 6 may we written
F(x) = I
dl Bstat (x)
(1.16)
In order to assess the properties of Bstat , we determine its divergence and curl.
Taking the divergence of both sides of equation (1.15) and utilising formula (F.63)
on page 177, we obtain
Z
0
0
stat
3 0 j(x )
B (x) =
dx
=0
(1.17)
4
|x x0 |
V0
since, according to formula (F.63) on page 177, ( a) vanishes for any vector
field a(x).
Applying the operator bac-cab rule, formula (F.64) on page 177, the curl of
equation (1.15) can be written
Z
0
j(x0 )
Bstat (x) =
d3x0
=
4
|x x0 |
V0
Z
Z
0
0
1
1
3 0
0
2
3 0
0
0
0
=
d x j(x )
+
d x [j(x ) ]
4 V 0
4 V 0
|x x0 |
|x x0 |
(1.18)
Electrodynamics
In the first of the two integrals on the right-hand side, we use the representation
of the Dirac delta function given in formula (F.73) on page 177, and integrate the
second one by parts, by utilising formula (F.56) on page 177 as follows:
Z
1
3 0
0
0
0
d x [j(x ) ]
|x x0 |
V0
Z
1
3 0 0
0
= x k d x j(x )
xk0 |x x0 |
V0
Z
1
d3x0 0 j(x0 ) 0
|x x0 |
V0
Z
Z
0
0
1
1
2 0 0
0
3 0
0
= x k d x n j(x ) 0
d x j(x )
xk |x x0 |
|x x0 |
S0
V0
(1.19)
Then we note that the first integral in the result, obtained by applying Gausss
theorem, vanishes when integrated over a large sphere far away from the localised
source j(x0 ), and that the second integral vanishes because j = 0 for stationary
currents (no charge accumulation in space). The net result is simply
B
stat
(x) = 0
Z
V0
(1.20)
1.3 Electrodynamics
As we saw in the previous sections, the laws of electrostatics and magnetostatics
can be summarised in two pairs of time-independent, uncoupled vector partial
differential equations, namely the equations of classical electrostatics
(x)
0
Estat (x) = 0
Estat (x) =
(1.21a)
(1.21b)
(1.22a)
(x) = 0 j(x)
(1.22b)
Since there is nothing a priori which connects Estat directly with Bstat , we must
consider classical electrostatics and classical magnetostatics as two independent
theories.
1. Classical Electrodynamics
(1.23)
which states that the time rate of change of electric charge (t, x) is balanced by a
divergence in the electric current density j(t, x).
= 0 j(t, x) + 0 0 E(t, x)
t
(1.24)
10
Electrodynamics
where, in the last step, we have assumed that a generalisation of equation (1.5) on
page 4 to time-varying fields allows us to make the identification4
Z
1
1
d3x0 (t, x0 )0
40 t V 0
|x x0 |
Z
1
1
3 0
0
d x (t, x )
=
(1.25)
t
40 V 0
|x x0 |
Z
(t, x0 )
d3x0
= E(t, x)
0
0
t
40
t
|x x |
V
The result is Maxwells source equation for the B field
1
(1.26)
where the last term 0 E(t, x)/t is the famous displacement current. This term
was introduced, in a stroke of genius, by Maxwell [8] in order to make the right
hand side of this equation divergence free when j(t, x) is assumed to represent the
density of the total electric current, which can be split up in ordinary conduction currents, polarisation currents and magnetisation currents. The displacement
current is an extra term which behaves like a current density flowing in vacuum.
As we shall see later, its existence has far-reaching physical consequences as it
predicts the existence of electromagnetic radiation that can carry energy and momentum over very long distances, even in vacuum.
(1.27)
where is the electric conductivity (S/m). In the most general cases, for instance
in an anisotropic conductor, is a tensor.
We can view Ohms law, equation (1.27) above, as the first term in a Taylor
expansion of the law j[E(t, x)]. This general law incorporates non-linear effects
4 Later,
11
1. Classical Electrodynamics
such as frequency mixing. Examples of media which are highly non-linear are
semiconductors and plasma. We draw the attention to the fact that even in cases
when the linear relation between E and j is a good approximation, we still have
to use Ohms law with care. The conductivity is, in general, time-dependent
(temporal dispersive media) but then it is often the case that equation (1.27) on
page 11 is valid for each individual Fourier component of the field.
If the current is caused by an applied electric field E(t, x), this electric field
will exert work on the charges in the medium and, unless the medium is superconducting, there will be some energy loss. The rate at which this energy is expended is j E per unit volume. If E is irrotational (conservative), j will decay
away with time. Stationary currents therefore require that an electric field which
corresponds to an electromotive force (EMF) is present. In the presence of such a
field EEMF , Ohms law, equation (1.27) on page 11, takes the form
j = (Estat + EEMF )
(1.28)
I
C
dl (Estat + EEMF )
(1.29)
dl EEMF
(1.30)
I
C
12
dl E(t, x) =
d
dt
Z
S
d
m (t, x)
dt
d2x n B(t, x) =
Z
S
d2x n
B(t, x)
t
(1.31)
Electrodynamics
B(x)
B(x)
d2x n
v
C
dl
F IGURE 1.4:
A loop C which moves with velocity v in a spatially varying magnetic field B(x) will sense a varying magnetic flux during the motion.
where m is the magnetic flux and S is the surface encircled by C which can
be interpreted as a generic stationary loop and not necessarily as a conducting
circuit. Application of Stokes theorem on this integral equation, transforms it
into the differential equation
E(t, x) =
B(t, x)
t
(1.32)
which is valid for arbitrary variations in the fields and constitutes the Maxwell
equation which explicitly connects electricity with magnetism.
Any change of the magnetic flux m will induce an EMF. Let us therefore
consider the case, illustrated if figure 1.4, that the loop is moved in such a way
that it links a magnetic field which varies during the movement. The convective
derivative is evaluated according to the well-known operator formula
d
=
+v
dt t
(1.33)
which follows immediately from the rules of differentiation of an arbitrary differentiable function f (t, x(t)). Applying this rule to Faradays law, equation (1.31)
on page 12, we obtain
E(t, x) =
d
dt
Z
S
d2x n B =
Z
S
d2x n
d2x n (v )B
(1.34)
13
1. Classical Electrodynamics
During spatial differentiation v is to be considered as constant, and equation (1.17) on page 8 holds also for time-varying fields:
B(t, x) = 0
(1.35)
(it is one of Maxwells equations) so that, according to formula (F.59) on page 177,
(B v) = (v )B
(1.36)
I
C
dl EEMF =
d
dt
= d x n
t
S
Z
d2x n B
(1.37)
2
d x n (B v)
S
I
C
dl EEMF =
d2x n
dl (B v)
(1.38)
dl (EEMF v B) =
Z
S
d2x n
B
t
(1.39)
where EEMF is the field which is induced in the loop, i.e., in the moving system.
The use of Stokes theorem backwards on equation (1.39) above yields
(EEMF v B) =
B
t
(1.40)
(1.41)
(1.42)
(1.43)
Hence, we can conclude that for a stationary observer, the Maxwell equation
E=
B
t
(1.44)
14
Electrodynamics
(1.45a)
E=
0
B
E=
(1.45b)
t
B=0
(1.45c)
E
B = 0 0
+ 0 j(t, x)
(1.45d)
t
In these equations (t, x) represents the total, possibly both time and space dependent, electric charge, i.e., free as well as induced (polarisation) charges, and j(t, x)
represents the total, possibly both time and space dependent, electric current, i.e.,
conduction currents (motion of free charges) as well as all atomistic (polarisation,
magnetisation) currents. As they stand, the equations therefore incorporate the
classical interaction between all electric charges and currents in the system and
are called Maxwells microscopic equations. Another name often used for them
is the Maxwell-Lorentz equations. Together with the appropriate constitutive relations, which relate and j to the fields, and the initial and boundary conditions
pertinent to the physical situation at hand, they form a system of well-posed partial
differential equations which completely determine E and B.
(1.46a)
H = H[t, x; E, B]
(1.46b)
Under certain conditions, for instance for very low field strengths, we may assume
that the response of a substance to the fields may be approximated as a linear one
15
1. Classical Electrodynamics
so that
D = E
(1.47)
H= B
(1.48)
i.e., that the derived fields are linearly proportional to the primary fields and that
the electric displacement (magnetising field) is only dependent on the electric
(magnetic) field.
The field equations expressed in terms of the derived field quantities D and H
are
D = (t, x)
B
E=
t
B=0
D
+ j(t, x)
H=
t
(1.49a)
(1.49b)
(1.49c)
(1.49d)
and are called Maxwells macroscopic equations. We will study them in more
detail in chapter 4.
(1.50a)
E=
(1.50b)
E=
B
0 jm
t
B = 0 m
E
B = 0 0
+ 0 je
t
16
(1.50c)
(1.50d)
Electromagnetic duality
We shall call these equations Diracs symmetrised Maxwell equations or the electromagnetodynamic equations.
Taking the divergence of (1.50b), we find that
( E) =
( B) 0 jm 0
t
(1.51)
where we used the fact that, according to formula (F.63) on page 177, the divergence of a curl always vanishes. Using (1.50c) to rewrite this relation, we obtain
the magnetic monopole equation of continuity
m
+ jm = 0
t
(1.52)
which has the same form as that for the electric monopoles (electric charges) and
currents, equation (1.23) on page 10.
We notice that the new equations (1.50) on page 16 exhibit the following symmetry (recall that 0 0 = 1/c2 ):
E cB
(1.53a)
cB E
(1.53b)
c
e
(1.53c)
c
m
e
(1.53d)
cj j
(1.53e)
j cj
(1.53f)
E = E cos + cB sin
c B = E sin + cB cos
? e
c = c cos + sin
e
? m
= c sin + cos
? e
c j = cj cos + j sin
? m
j = cj sin + j cos
e
(1.54a)
(1.54b)
(1.54c)
(1.54d)
(1.54e)
(1.54f)
which leaves the symmetrised Maxwell equations, and hence the physics they
describe (often referred to as electromagnetodynamics), invariant. Since E and je
are (true or polar) vectors, B a pseudovector (axial vector), e a (true) scalar, then
m and , which behaves as a mixing angle in a two-dimensional charge space,
must be pseudoscalars and jm a pseudovector.
17
1. Classical Electrodynamics
1.5 Bibliography
[1]
[2]
[3]
W. G REINER, Classical Electrodynamics, Springer-Verlag, New York, Berlin, Heidelberg, 1996, ISBN 0-387-94799-X.
[4]
[5]
J. D. JACKSON, Classical Electrodynamics, third ed., John Wiley & Sons, Inc.,
New York, NY . . . , 1999, ISBN 0-471-30932-X.
[6]
[7]
F. E. L OW, Classical Field Theory, John Wiley & Sons, Inc., New York, NY . . . , 1997,
ISBN 0-471-59551-9.
[8]
J. C. M AXWELL, A dynamical theory of the electromagnetic field, Royal Society Transactions, 155 (1864).
5 As
. . . there are strong theoretical reasons to believe that magnetic charge exists in nature,
and may have played an important role in the development of the universe. Searches for
magnetic charge continue at the present time, emphasising that electromagnetism is very
far from being a closed object.
18
Bibliography
[9]
J. C. M AXWELL, A Treatise on Electricity and Magnetism, third ed., vol. 1, Dover Publications, Inc., New York, NY, 1954, ISBN 0-486-60636-8.
[10] J. C. M AXWELL, A Treatise on Electricity and Magnetism, third ed., vol. 2, Dover Publications, Inc., New York, NY, 1954, ISBN 0-486-60637-8.
[11] D. B. M ELROSE AND R. C. M C P HEDRAN, Electromagnetic Processes in Dispersive Media, Cambridge University Press, Cambridge . . . , 1991, ISBN 0-521-41025-8.
[12] W. K. H. PANOFSKY AND M. P HILLIPS, Classical Electricity and Magnetism, second ed.,
Addison-Wesley Publishing Company, Inc., Reading, MA . . . , 1962, ISBN 0-201-057026.
[13] F. ROHRLICH, Classical Charged Particles, Perseus Books Publishing, L.L.C., Reading,
MA . . . , 1990, ISBN 0-201-48300-9.
[14] J. S CHWINGER, A magnetic model of matter, Science, 165 (1969), pp. 757761.
[15] J. S CHWINGER , L. L. D E R AAD , J R ., K. A. M ILTON , AND W. T SAI, Classical Electrodynamics, Perseus Books, Reading, MA, 1998, ISBN 0-7382-0056-5.
[16] J. A. S TRATTON, Electromagnetic Theory, McGraw-Hill Book Company, Inc., New
York, NY and London, 1953, ISBN 07-062150-0.
[17] J. VANDERLINDE, Classical Electromagnetic Theory, John Wiley & Sons, Inc., New
York, Chichester, Brisbane, Toronto, and Singapore, 1993, ISBN 0-471-57269-1.
19
1. Classical Electrodynamics
1.6 Examples
E XAMPLE 1.1
(x
)
= d3x0
4
|x x0 |
V0
[cf. equation (1.5) on page 4 for Estat ] and, if magnetic currents exist, a Biot-Savart-like law for
electric fields [cf. equation (1.15) on page 8 for Bstat ]:
Z
Z
0
x x0
0
1
3 0 m 0
Estat (x) =
d3x0 jm (x0 )
=
d
x
j
(x
)
4 V 0
4 V 0
|x x0 |
|x x0 |3
(1.56)
Z
m 0
0
j (x )
= d3x0
4
|x x0 |
V0
Taking the curl of the latter and using the operator bac-cab rule, formula (F.59) on page 177,
we find that
Z
0
jm (x0 )
Estat (x) = d3x0
=
4
|x x0 |
V0
(1.57)
Z
Z
0
1
0
1
3 0 m 0
0
0
d3x0 jm (x0 )2
d
x
[j
(x
)
]
=
4 V 0
4 V 0
|x x0 |
|x x0 |
Comparing with equation (1.18) on page 8 for Estat and the evaluation of the integrals there, we
obtain
Estat (x) = 0
Z
V0
(1.58)
We assume that formula (1.56) above is valid also for time-varying magnetic currents.
Then, with the use of the representation of the Dirac delta function, equation (F.73) on page 177,
the equation of continuity for magnetic charge, equation (1.52) on page 17, and the assumption
of the generalisation of equation (1.55) to time-dependent magnetic charge distributions, we
obtain, formally,
20
Examples
E(t, x) = 0
Z
V0
0
4 t
Z
V0
d3x0 m (t, x0 )0
1
|x x0 |
= 0 jm (t, x) B(t, x)
t
(1.59)
[cf. equation (1.24) on page 10] which we recognise as equation (1.50b) on page 16. A transformation of this electromagnetodynamic result by rotating into the electric realm of charge
space, thereby letting jm tend to zero, yields the electrodynamic equation (1.50b) on page 16,
i.e., the Faraday law in the ordinary Maxwell equations. This process also provides an alternative interpretation of the term B/t as a magnetic displacement current, dual to the electric
displacement current [cf. equation (1.26) on page 11].
By postulating the indestructibility of a hypothetical magnetic charge, we have thereby been
able to replace Faradays experimental results on electromotive forces and induction in loops as
a foundation for the Maxwell equations by a more appealing one.
C E ND OF EXAMPLE 1.1
E XAMPLE 1.2
e
=
cos + sin =
0
c
0
1
?B
= (E cos + cB sin ) +
E sin + B cos
?E +
t
t
c
B
1 E
= 0 jm cos
cos + c0 je sin +
sin
t
c t
1 E
B
sin +
cos = 0 jm cos + c0 je sin
c t
t
= 0 (cje sin + jm cos ) = 0 ?jm
e
1
sin + 0 m cos
?B = ( E sin + B cos ) =
c
c0
= 0 (ce sin + m cos ) = 0 ?m
(1.60)
(1.61)
(1.62)
21
1. Classical Electrodynamics
?B
1 ?E
1
1
= ( E sin + B cos ) 2 (E cos + cB sin )
c2 t
c
c t
1
1 B
1 E
= 0 jm sin +
cos + 0 je cos + 2
cos
c
c t
c t
1 E
1 B
2
cos
sin
c t
c t
1 m
= 0
j sin + je cos = 0 ?je
c
(1.63)
QED
C E ND OF EXAMPLE 1.2
E XAMPLE 1.3
(1.64a)
j = cj tan
(1.64b)
cos
cos
? m
= ce sin + ce tan cos = ce sin + ce sin = 0
1
1 e
? e
j = je cos + je tan sin =
(je cos2 + je sin2 ) =
j
cos
cos
? m
j = cje sin + cje tan cos = cje sin + cje sin = 0
? e
(1.65a)
(1.65b)
(1.65c)
(1.65d)
Hence, a fixed mixing angle, or, equivalently, a fixed ratio between the electric and magnetic
charges/currents, hides the magnetic monopole influence (m and jm ) on the dynamic equations.
We notice that the inverse of the transformation given by equation (1.54) on page 17 yields
E = ?E cos c?B sin
(1.66)
(1.67)
Furthermore, from the expressions for the transformed charges and currents above, we find that
?E =
22
? e
1 e
=
0
cos 0
(1.68)
Examples
and
?B = 0 ?m = 0
(1.69)
so that
E=
e
1 e
cos 0 =
cos 0
0
(1.70)
QED
C E ND OF EXAMPLE 1.3
E XAMPLE 1.4
(1.71)
3
(1.72)
(1.73a)
E B = Const
(1.73b)
2. The inner (scalar) product defined as G scalar multiplied with the complex conjugate of
itself
G G = (E + icB) (E icB) = E 2 + c2 B2
(1.74)
which is also an invariant scalar quantity. As we shall see later, this quantity is proportional to the electromagnetic field energy, which indeed is a conserved quantity.
3. As with any vector, the cross product of G with itself vanishes:
G G = (E + icB) (E + icB)
= E E c2 B B + ic(E B) + ic(B E)
(1.75)
= 0 + 0 + ic(E B) ic(E B) = 0
4. The cross product of G with the complex conjugate of itself
23
1. Classical Electrodynamics
G G = (E + icB) (E icB)
= E E + c2 B B ic(E B) + ic(B E)
(1.76)
E XAMPLE 1.5
(1.77)
(1.78)
while
?
G ?G = e2i G G
(1.79)
Furthermore, assuming that = (t, x), we see that the spatial and temporal differentiation
of ?G leads to
?G
= i(t )ei G + ei t G
t
?G ?G = iei G + ei G
t ?G
?
G G = ie G + e G
i
(1.80a)
(1.80b)
(1.80c)
24
2
Electromagnetic
Waves
E=
E
B = 0 j(t, x) + 0 0
t
(Gausss law)
(2.1a)
(Faradays law)
(2.1b)
(2.1c)
(Maxwells law)
(2.1d)
and can be viewed as an axiomatic basis for classical electrodynamics. They describe, in scalar and vector differential equation form, the electric and magnetic
fields E and B produced by given, prescribed charge distributions (t, x) and current distributions j(t, x) with arbitrary time and space dependences.
However, as is well known from the theory of differential equations, these four
first order, coupled partial differential vector equations can be rewritten as two uncoupled, second order partial equations, one for E and one for B. We shall derive
these second order equations which, as we shall see are wave equations, and then
discuss the implications of them. We show that for certain media, the B wave field
can be easily obtained from the solution of the E wave equation.
25
2. Electromagnetic Waves
( E) = ( B) = 0
j + 0 E
(2.2)
t
t
t
According to the operator triple product bac-cab rule equation (F.64) on page 177
( E) = ( E) 2 E
(2.3)
(2.4)
and since EEMF = 0, Ohms law, equation (1.28) on page 12, allows us to use the
approximation
j = E
(2.5)
2
E 0
E + 0 E = 0
t
t
(2.6)
E 1 2 E
2 2 =0
t
c t
(2.7)
1 2 E
= 2 E = 0
c2 t2
(2.8)
26
( E) = 0 E + 0 0 ( E)
t
t
(2.9)
which, with the use of equation (F.64) on page 177 and equation (2.1c) on page 25
can be rewritten
( B) 2 B = 0
B
2
0 0 2 B
t
t
(2.10)
Using the fact that, according to (2.1c), B = 0 for any medium and rearranging,
we can rewrite this equation as
2 B 0
B 1 2 B
2 2 =0
t
c t
(2.11)
This is the wave equation for the magnetic field. For waves propagating in vacuum
(no charges, no currents), the wave equation for B is
2 B
1 2 B
= 2 B = 0
c2 t2
(2.12)
We notice that for the simple propagation media considered here, the wave
equations for the magnetic field B has exactly the same mathematical form as the
wave equation for the electric field E, equation (2.7) on page 26. Therefore, it suffices to consider only the E field, since the results for the B field follow trivially.
For EM waves propagating in more complicated media, containing, eg., inhomogeneities, the wave equation for E and for B do not have the same mathematical
form.
27
2. Electromagnetic Waves
wave. When we insert this, in complex notation, into equation (2.7) on page 26
we find that
1 2
E0 (x)eit 2 2 E0 (x)eit
t
c t
1
2
it
= E0 (x)e
0 (i)E0 (x)eit 2 (i)2 E0 (x)eit
c
2 E0 (x)eit 0
2
E0 + 2 1 + i
E0 = 0
c
0
(2.14)
(2.15)
In the limit of very many frequency components the Fourier sum goes over
into a Fourier integral. To illustrate this general case, let us introduce the Fourier
transform of E(t, x)
def
1
2
dt E(t, x) eit
(2.17)
def
[E (x)] E(t, x) =
d E (x) eit
(2.18)
dt
e
F
t
2
t
Z
1
1
E(t, x) eit i
dt E(t, x) eit
=
2 |
2
{z
}
(2.19)
=0
= i E (x)
and that, consequently,
2 E(t, x)
F
t2
28
def
dt
2 E(t, x)
t2
eit = 2 E (x)
(2.20)
Fourier transforming equation (2.7) on page 26 and using (2.19) and (2.20), we
obtain
i
2
E = 0
(2.21)
2 E + 2 1 +
c
2
E=0
c2
(2.22)
which is a time-independent wave equation for E, representing undamped propagating waves. In the short limit we have instead
2 E + i0 E = 0
(2.23)
(2.24)
1
0
=
=
=
= R0
(2.25)
0 0 ck
k
0
k
where in the last step we introduced the characteristic impedance for vacuum
r
0
R0 =
376.7
(2.26)
0
29
2. Electromagnetic Waves
= n
= n
(2.27)
E
B
n
=
(2.28b)
t
B
n
=0
(2.28c)
B
E
E
n
= 0 j(t, x) + 0 0
= 0 E + 0 0
(2.28d)
t
t
Scalar multiplying (2.28d) by n,
we find that
B
= n 0 + 0 0
E
0 = n n
(2.29)
which simplifies to the first-order ordinary differential equation for the normal
component En of the electric field
dEn
(2.30)
+ En = 0
dt
0
with the solution
En = En0 et/0 = En0 et/
(2.31)
This, together with (2.28a), shows that the longitudinal component of E, i.e., the
component which is perpendicular to the plane surface is independent of and has
a time dependence which exhibits an exponential decay, with a decrement given
by the relaxation time in the medium.
Scalar multiplying (2.28b) by n,
we similarly find that
E
B
0 = n n
= n
(2.32)
t
or
B
n
=0
(2.33)
t
From this, and (2.28c), we conclude that the only longitudinal component of B
must be constant in both time and space. In other words, the only non-static
solution must consist of transverse components.
30
Plane waves
2 2 =0
(2.34)
0
2
t
c t
This equation, which describes the propagation of plane waves in a conducting
medium, is called the telegraphers equation. If the medium is an insulator so that
= 0, then the equation takes the form of the one-dimensional wave equation
2 E 1 2 E
=0
(2.35)
2 c2 t2
As is well known, each component of this equation has a solution which can be
written
Ei = f ( ct) + g( + ct),
i = 1, 2, 3
(2.36)
where f and g are arbitrary (non-pathological) functions of their respective arguments. This general solution represents perturbations which propagate along ,
where the f perturbation propagates in the positive direction and the g perturbation propagates in the negative direction.
If we assume that our electromagnetic fields E and B are time-harmonic,
i.e., that they can each be represented by a Fourier component proportional to
exp{it}, the solution of equation (2.35) above becomes
E = E0 ei(tk) = E0 ei(kt)
By introducing the wave vector
k = k n = n = k
c
c
this solution can be written as
E = E0 ei(kxt)
(2.37)
(2.38)
(2.39)
Let us consider the lower sign in front of k in the exponent in (2.37). This
corresponds to a wave which propagates in the direction of increasing . Inserting
this solution into equation (2.28b) on page 30, gives
E
= iB = ik n E
(2.40)
n
B = n E = k E = k E = 0 0 n E
(2.41)
c
Hence, to each transverse component of E, there exists an associated magnetic
field given by equation (2.41) above. If E and/or B has a direction in space which
is constant in time, we have a plane wave.
31
2. Electromagnetic Waves
E
+
i
E
=
+ K2E = 0
0
0
0
2
2
(2.42)
2
K = 0 0 1 + i
= 2 1+i
=k 1+i
0
c
0
0
(2.43)
where
2
where, in the last step, equation (2.24) on page 29 was used to introduce the wave
number k. Taking the square root of this expression, we obtain
r
K =k 1+i
= + i
(2.44)
0
Squaring, one finds that
2
= (2 2 ) + 2i
k 1+i
0
(2.45)
or
2 = 2 k2
=
(2.46)
k
20
2
(2.47)
Squaring the latter and combining with the former, one obtains the second order
algebraic equation (in 2 )
2 (2 k2 ) =
k 4 2
420 2
(2.48)
32
(2.49a)
(2.49b)
(2.50)
With the aid of equation (2.41) on page 31 we can calculate the associated magnetic field, and find that it is given by
B=
1
1
1
K k E = ( k E)( + i) = ( k E) |A| ei
(2.51)
2
0 2
K =k 1+i
=k i
1i
k(1 + i)
0
0
20
r
r
= (1 + i)
= 0 0 (1 + i)
20
2
(2.52)
In this limit we find that when the wave impinges perpendicularly upon the medium,
the fields are given, inside the medium, by
r
r
0
0
0
E = E0 exp
exp i
t
(2.53a)
2
2
r
0
0
B = (1 + i)
( n E0 )
(2.53b)
2
Hence, both fields fall off by a factor 1/e at a distance
s
2
=
0
(2.54)
33
2. Electromagnetic Waves
valued. I.e., Ephysical = Re {Emathematical }. It is particularly important to remember this when one works with products of physical quantities. For instance, if
we have two physical vectors F and G which both are time-harmonic, i.e., can be
represented by Fourier components proportional to exp{it}, then we must make
the following interpretation
F(t, x) G(t, x) = Re {F} Re {G} = Re F0 (x) eit Re G0 (x) eit
(2.55)
Furthermore, letting denote complex conjugate, we can express the real part of
the complex vector F as
1
(2.56)
Re {F} = Re F0 (x) eit = [F0 (x) eit + F0 (x) eit ]
2
and similarly for G. Hence, the physically acceptable interpretation of the scalar
product of two complex vectors, representing physical observables, is
F(t, x) G(t, x) = Re F0 (x) eit Re G0 (x) eit
1
1
= [F0 (x) eit + F0 (x) eit ] [G0 (x) eit + G0 (x) eit ]
2
2
1
= F0 G0 + F0 G0 + F0 G0 e2it + F0 G0 e2it
4
1
= Re F0 G0 + F0 G0 e2it
2
1
= Re F0 eit G0 eit + F0 G0 e2it
2
1
= Re F(t, x) G (t, x) + F0 G0 e2it
2
(2.57)
Often in physics, we measure temporal averages (h i) of our physical observables. If so, we see that the average of the product of the two physical quantities
represented by F and G can be expressed as
1
1
1
Re {F G } = F G = F G
(2.58)
2
2
2
since the temporal average of the oscillating function exp{2it} vanishes.
hF Gi hRe {F} Re {G}i =
2.4 Bibliography
[1] J. D. JACKSON, Classical Electrodynamics, third ed., John Wiley & Sons, Inc.,
New York, NY . . . , 1999, ISBN 0-471-30932-X.
34
Bibliography
[2] W. K. H. PANOFSKY AND M. P HILLIPS, Classical Electricity and Magnetism, second ed.,
Addison-Wesley Publishing Company, Inc., Reading, MA . . . , 1962, ISBN 0-201-057026.
35
2. Electromagnetic Waves
2.5 Example
E XAMPLE 2.1
B = 0 0
(2.59a)
(2.59b)
(2.59c)
E
+ 0 je
t
(2.59d)
under the assumption of vanishing net electric and magnetic charge densities and in the absence
of electromotive and magnetomotive forces. Interpret this equation physically.
Taking the curl of (2.59b) and using (2.59d), and assuming, for symmetry reasons, that
there exists a linear relation between the magnetic current density jm and the magnetic field B
(the magnetic dual of Ohms law for electric currents, je = e E)
jm = m B
(2.60)
( E) = 0 jm ( B) = 0 m B
t
1 E
E
1
= 0 m 0 e E + 2
0 e
2
c t
t
c
1 E
0 je + 2
c t
2 E
t2
(2.61)
1 0
= 2 E + 2
1 2 m e + i
E=0
c
0
0
Realising that, according to formula (2.26) on page 29, 0 /0 is the square of the vacuum
radiation resistance R0 , and rearranging a bit, we obtain the time-independent wave equation in
36
Example
R20 m e
2
e + m /c2
2
E = 0
E+ 2 1 2
1+i
R2
c
0 1 02 m e
(2.64)
From this equation we conclude that the existence of magnetic charges (magnetic monopoles),
and non-vanishing electric and magnetic conductivities would lead to a shift in the effective
wave number of the wave. Furthermore, even if the electric conductivity e vanishes, the
imaginary term does not necessarily vanish and the wave might therefore experience damping
(or growth) according as m is positive (or negative). This would happen in a hypothetical
medium which is a perfect insulator for electric currents but which can carry magnetic currents.
Finally, we note that in the particular case that = R0 m e , the wave equation becomes
a (time-independent) diffusion equation
m
(2.65)
2 E + i0 e + 2 E = 0
c
and, hence, no waves exist at all!
C E ND OF EXAMPLE 2.1
37
3
Electromagnetic
Potentials
As an alternative to expressing the laws of electrodynamics in terms of electric and
magnetic fields, it turns out that it is often more convenient to express the theory
in terms of potentials. This is particularly true for problems related to radiation. In
this chapter we will introduce and study the properties of such potentials and shall
find that they exhibit some remarkable properties which elucidate the fundamental
aspects of electromagnetism and lead naturally to the special theory of relativity.
(3.1)
Taking the divergence of this and using equation (1.7) on page 5, we obtain Poissons equation
2 stat (x) = Estat (x) =
(x)
0
(3.2)
A comparison with the definition of Estat , namely equation (1.5) on page 4, shows
that this equation has the solution
stat (x) =
1
40
Z
V0
d3x0
(x0 )
+
|x x0 |
(3.3)
39
3. Electromagnetic Potentials
where the integration is taken over all source points x0 at which the charge density
(x0 ) is non-zero and is an arbitrary quantity which has a vanishing gradient.
An example of such a quantity is a scalar constant. The scalar function stat (x) in
equation (3.3) on page 39 is called the electrostatic scalar potential.
(3.4)
0
4
Z
V0
d3x0
j(x0 )
+ a(x)
|x x0 |
(3.5)
where a(x) is an arbitrary vector field whose curl vanishes. From equation (F.62)
on page 177 we know that such a vector can always be written as the gradient of
a scalar field.
40
Gauge transformations
(3.6)
Inserting this expression into the other homogeneous Maxwell equation (1.32) on
page 13, we obtain
E(t, x) + A(t, x) = 0
t
E(t, x) =
(3.7)
(3.8)
A(t, x) = (t, x)
(3.9)
t
This means that in electrodynamics, E(t, x) is calculated from the potentials according to the formula
E(t, x) +
A(t, x)
(3.10)
t
and B(t, x) from formula (3.6) above. Hence, it is a matter of taste whether we
want to express the laws of electrodynamics in terms of the potentials (t, x) and
A(t, x), or in terms of the fields E(t, x) and B(t, x). However, there exists an important difference between the two approaches: in classical electrodynamics the
only directly observable quantities are the fields themselves (and quantities derived from them) and not the potentials. On the other hand, the treatment becomes
significantly simpler if we use the potentials in our calculations and then, at the
final stage, use equation (3.6) and equation (3.10) above to calculate the fields or
physical quantities expressed in the fields.
E(t, x) = (t, x)
41
3. Electromagnetic Potentials
terms of vanishing gradients and curls, to pick suitable forms for the potentials and
still get the same physical result. In fact, the way the electromagnetic scalar potential (t, x) and the vector potential A(t, x) are related to the physically observables
gives leeway for similar manipulation of them also in electrodynamics.
If we transform (t, x) and A(t, x) simultaneously into new ones 0 (t, x) and
0
A (t, x) according to the mapping scheme
(t, x)
t
0
A(t, x) 7 A (t, x) = A(t, x) (t, x)
(t, x) 7 0 (t, x) = (t, x) +
(3.11a)
(3.11b)
where (t, x) is an arbitrary, differentiable scalar function called the gauge function, and insert the transformed potentials into equation (3.10) on page 41 for the
electric field and into equation (3.6) on page 41 for the magnetic field, we obtain
the transformed fields
() A ()
A
A0
=
+
=
t
t
t
t
t
0
0
B = A = A () = A
E0 = 0
(3.12a)
(3.12b)
where, once again equation (F.62) on page 177 was used. We see that the fields
are unaffected by the gauge transformation (3.11). A transformation of the potentials and A which leaves the fields, and hence Maxwells equations, invariant
is called a gauge transformation. A physical law which does not change under a
gauge transformation is said to be gauge invariant. It is only those quantities (expressions) that are gauge invariant that have experimental significance. Of course,
the EM fields themselves are gauge invariant.
2 =
42
(3.13a)
(3.13b)
Gauge conditions
=
+
A
+
c2 t2
0
t
c2 t
1
1 2 A
2
A = 0 j(t, x) A + 2
c2 t2
c t
(3.14a)
(3.14b)
These two second order, coupled, partial differential equations, representing in all
four scalar equations (one for and one each for the three components Ai , i =
1, 2, 3 of A) are completely equivalent to the formulation of electrodynamics in
terms of Maxwells equations, which represent eight scalar first-order, coupled,
partial differential equations.
As they stand, equations (3.13) on page 42 and equations (3.14) look complicated and may seem to be of limited use. However, if we write equation (3.6) on
page 41 in the form A(t, x) = B(t, x) we can consider this as a specification
of A. But we know from Helmholtz theorem that in order to determine the
(spatial) behaviour of A completely, we must also specify A. Since this divergence does not enter the derivation above, we are free to choose A in whatever
way we like and still obtain the same physical results!
=
2 =
(3.16a)
c2 t2
c2 t2
0
def
1 2
1 2 A
2
2 A
A
=
2 A = 0 j(t, x)
(3.16b)
c2 t2
c2 t2
A+
where 2 is the dAlembert operator discussed in example M.5 on page 195. Each
of these four scalar equations is an inhomogeneous wave equation of the following
generic form:
2 (t, x) = f (t, x)
(3.17)
1 In fact, the Dutch physicist Hendrik Antoon Lorentz, who in 1903 demonstrated the covariance of
Maxwells equations, was not the original discoverer of this condition. It had been discovered by the Danish
physicist Ludvig V. Lorenz already in 1867 [6]. In the literature, this fact has sometimes been overlooked
and the condition was earlier referred to as the Lorentz gauge condition.
43
3. Electromagnetic Potentials
where is a shorthand for either or one of the components Ai of the vector potential A, and f is the pertinent generic source component, (t, x)/0 or 0 ji (t, x),
respectively.
We assume that our sources are well-behaved enough in time t so that the
Fourier transform pair for the generic source function f
def
F 1 [ f (x)] f (t, x) =
def
1
2
d f (x) eit
(3.18a)
dt f (t, x) eit
(3.18b)
exists, and that the same is true for the generic potential component :
(t, x) =
(x) =
1
2
d (x) eit
dt (t, x) eit
(3.19a)
(3.19b)
Inserting the Fourier representations (3.18a) and (3.19a) into equation (3.17) on
page 43, and using the vacuum dispersion relation for electromagnetic waves
= ck
(3.20)
the generic 3D inhomogeneous wave equation, equation (3.17) on page 43, turns
into
2 (x) + k2 (x) = f (x)
(3.21)
(3.22)
and the solution of equation (3.21) above which corresponds to the frequency
is given by the superposition
(x) =
44
Z
V0
(3.23)
Gauge conditions
where f (x0 ) is the wavelet amplitude at the source point x0 . The function G(x, x0 )
is called the Green function or the propagator.
Due to translational invariance in space, G(x, x0 ) = G(x x0 ). Furthermore, in
equation (3.22) on page 44, the Dirac generalised function (x x0 ), which represents the point source, depends only on x x0 and there is no angular dependence
in the equation. Hence, the solution can only be dependent on r = |x x0 | and not
on the direction of x x0 . If we interpret r as the radial coordinate in a spherically
polar coordinate system, and recall the expression for the Laplace operator in such
a coordinate system, equation (3.22) on page 44 becomes
d2
(rG) + k2 (rG) = r(r)
dr2
(3.24)
Away from r = |x x0 | = 0, i.e., away from the source point x0 , this equation takes
the form
d2
(rG) + k2 (rG) = 0
dr2
(3.25)
G = C+
eikr
eik|xx |
eik|xx |
eikr
+ C
+ C
C+
0
r
r
|x x |
|x x0 |
(3.26)
x x0 0
G(x x0 ) C +
+
C
,
(3.27)
|x x0 |
|x x0 |
The volume integrated equation (3.22) on page 44 can under this assumption be
approximated by
Z 3 0 1
Z 3 0 2
1
+
2
+
C +C
dx
+k C +C
dx
|x x0 |
|x x0 |
V0
V0
(3.28)
Z
= d3x0 (x x0 )
V0
In virtue of the fact that the volume element d3x0 in spherical polar coordinates
is proportional to |x x0 |2 , the second integral vanishes when |x x0 | 0. Furthermore, from equation (F.73) on page 177, we find that the integrand in the first
integral can be written as 4(|x x0 |) and, hence, that
C+ + C =
1
4
(3.29)
45
3. Electromagnetic Potentials
Insertion of the general solution equation (3.26) on page 45 into equation (3.23)
on page 44 gives
0
ik|xx |
eik|xx |
3 0
0 e
d
x
f
(x
)
+
C
(3.30)
|x x0 |
|x x0 |
V0
V0
The inverse Fourier transform of this back to the t domain is obtained by inserting
the above expression for (x) into equation (3.19a) on page 44:
h
i
0
|
Z
Z
exp i t k|xx
V0
i
h
(3.31)
0
|
Z
Z
exp i t + k|xx
+ C d3x0 d f (x0 )
|x x0 |
V0
0
0
If we introduce the retarded time tret
and the advanced time tadv
in the following
way [using the fact that in vacuum k/ = 1/c, according to equation (3.20) on
page 44]:
k |x x0 |
|x x0 |
0
0
tret
= tret
(t, x x0 ) = t
=t
(3.32a)
c
k |x x0 |
|x x0 |
0
0
=t+
(3.32b)
tadv
= tadv
(t, x x0 ) = t +
c
and use equation (3.18a) on page 44, we obtain
(x) = C +
d3x0 f (x0 )
0
0
0
f (tret
, x0 )
3 0 f (tadv , x )
+
C
(3.33)
d
x
|x x0 |
|x x0 |
V0
V0
This is a solution to the generic inhomogeneous wave equation for the potential
components equation (3.17) on page 43. We note that the solution at time t at the
field point x is dependent on the behaviour at other times t0 of the source at x0
and that both retarded and advanced t0 are mathematically acceptable solutions.
However, if we assume that causality requires that the potential at (t, x) is set up
0
by the source at an earlier time, i.e., at (tret
, x0 ), we must in equation (3.33) above
(t, x) = C +
d3x0
1
(t0 , x0 )
d3x0 ret 0
40 V 0
|x x |
Z
0
0
j(t
, x0 )
A(t, x) =
d3x0 ret 0
4 V 0
|x x |
(t, x) =
(3.34a)
(3.34b)
2 In fact, inspired by a discussion by Paul A. M. Dirac, John A. Wheeler and Richard P. Feynman
derived in 1945 a fully self-consistent electrodynamics using both the retarded and the advanced potentials
[8]; see also [4].
46
Gauge conditions
= 0 j(t, x)
(3.35b)
c2 t2
c2 t2
We notice that if we require that the gauge function (t, x) itself be restricted to
fulfil the wave equation
1 2
2 = 0
(3.36)
c2 t2
these transformed Lorenz-Lorentz equations will keep their original form. The
set of potentials which have been gauge transformed according to equation (3.11)
on page 42 with a gauge function (t, x) restricted to fulfil equation (3.36), or,
in other words, those gauge transformed potentials for which the Lorenz-Lorentz
equations (3.16) are invariant, comprise the Lorenz-Lorentz gauge.
(t, x) =
1
40
Z
V0
d3x0
(t, x0 )
+
|x x0 |
(3.38)
47
3. Electromagnetic Potentials
where has vanishing gradient. We note that in the scalar potential expression the
charge density source is evaluated at time t. The retardation (and advancement)
effects therefore occur only in the vector potential, which is the solution of the inhomogeneous wave equation equation (3.37b) on page 47 for the vector potential
A.
In order to solve this equation, one splits up j in a longitudinal (k) and transverse () part, j jk + j where j = 0 and jk = 0, and note that the
equation of continuity equation (1.23) on page 10 becomes
2
0 + jk = 0
+ jk =
+ jk = 0
t
t
t
(3.39)
Furthermore, since = 0 and jk = 0, one finds that
+ jk = 0
0
t
(3.40)
c2 t
1
= 0 jk + g
c2 t
(3.41a)
(3.41b)
(3.42a)
( f) = g = 0
(3.42b)
1 2 A
1
= 0 j + 2
= 0 j + 0 jk = 0 j
2
2
c t
c t
(3.43)
which shows that in Coulomb gauge the source of the vector potential A is the
transverse part of the current j . The longitudinal part of the current jk does not
contribute to the vector potential. The retarded solution is (cf. equation (3.34a) on
page 46):
A(t, x) =
0
4
Z
V0
d3x0
0
j (tret
, x0 )
|x x0 |
(3.44)
The Coulomb gauge condition is therefore also called the transverse gauge.
48
Bibliography
1
= 0,
c2 t
c2
v2
(3.45)
we obtain the Lorenz-Lorentz gauge condition for = 1 and the Coulomb gauge
condition for = 0, respectively. Hence, the velocity gauge is a generalisation
of both these gauges. Inserting equation (3.45) into the coupled inhomogeneous
wave equation (3.14) on page 43 they become
1 2
(t, x)
=
v2 t2
0
2
1A
1
2 A 2 2 = 0 j(t, x) + 2
c t
c
t
(3.46a)
(3.46b)
(3.47a)
(3.47b)
d E(t, x)
(3.48a)
d B(t, x) x
(3.48b)
3.6 Bibliography
[1] W. E. B RITTIN , W. R. S MYTHE , AND W. W YSS, Poincar gauge in electrodynamics, American Journal of Physics, 50 (1982), pp. 693696.
49
3. Electromagnetic Potentials
50
Examples
3.7 Examples
E XAMPLE 3.1
BE LECTROMAGNETODYNAMIC POTENTIALS
In Diracs symmetrised form of electrodynamics (electromagnetodynamics), Maxwells
equations are replaced by [see also equations (1.50) on page 16]:
E=
e
0
E = 0 jm
(3.49a)
B
t
(3.49b)
B = 0 m
(3.49c)
E
B = 0 je + 0 0
t
(3.49d)
In this theory, one derives the inhomogeneous wave equations for the usual electric scalar
and vector potentials (e , Ae ) and their magnetic counterparts (m , Am ) by assuming that the
potentials are related to the fields in the following symmetrised form:
e
A (t, x) Am
t
1
1
B = 2 m (t, x) 2 Am (t, x) + Ae
c
c t
E = e (t, x)
(3.50a)
(3.50b)
In the absence of magnetic charges, or, equivalently for m 0 and Am 0, these formulae
reduce to the usual Maxwell theory formula (3.10) on page 41 and formula (3.6) on page 41,
respectively, as they should.
Inserting the symmetrised expressions (3.50) above into equations (3.49), one obtains [cf.,
equations (3.13a) on page 42]
e (t, x)
( Ae ) =
t
0
m (t, x)
m
2 m
+ ( A ) =
t
0
1 2 Ae
1 e
2 e
e
= 0 je (t, x)
A
+
A
+
c2 t2
c2 t
1 2 Am
1 m
2 m
m
A
+
A
+
= 0 jm (t, x)
c2 t2
c2 t
2 e +
(3.51a)
(3.51b)
(3.51c)
(3.51d)
By choosing the conditions on the divergence of the vector potentials according to the LorenzLorentz condition [cf. equation (3.15) on page 43]
1 e
=0
c2 t
1
Am + 2 m = 0
c t
Ae +
(3.52)
(3.53)
51
3. Electromagnetic Potentials
1 2 e
2 e
c2 t2
1 2 Ae
2 Ae
c2 t2
1 2 m
2 m
c2 t2
1 2 Am
2 Am
c2 t2
e (t, x)
0
(3.54a)
= 0 je (t, x)
(3.54b)
m (t, x)
0
(3.54c)
= 0 jm (t, x)
(3.54d)
exhibiting, once again, the striking properties of Diracs symmetrised Maxwell theory.
C E ND OF EXAMPLE 3.1
52
4
Electromagnetic
Fields and
Matter
The microscopic Maxwell equations (1.45) derived in chapter 1 are valid on all
scales where a classical description is good. However, when macroscopic matter is present, it is sometimes convenient to use the corresponding macroscopic
Maxwell equations (in a statistical sense) in which auxiliary, derived fields are
introduced in order to incorporate effects of macroscopic matter when this is immersed fully or partially in an electromagnetic field.
53
monopole moment
q=
d3x0 (x0 )
(4.1)
V0
where the is the charge density introduced in equation (1.7) on page 5, the
electric dipole moment vector
p(x0 ) =
Z
V0
(4.2)
Z
V0
(4.3)
stat
1
q
(x x0 )i
1
(x) =
pi
+
2
40 |x x0 | |x x0 |
|x x0 |
3 (x x0 )i (x x0 ) j 1
1
Q
+
.
.
.
+
ij
ij
2 |x x0 | |x x0 |
2
|x x0 |3
(4.4)
p (x) =
Z
V0
d3x0 P(x0 )
1
|x x0 |
(4.5)
54
Using the expression equation (M.97) on page 196 and applying the divergence
theorem, we can rewrite this expression for the potential as follows:
Z
Z
0
0
1
P(x0 )
3 0 P(x )
p (x) =
d
x
d3x0 0
40 V 0
|x x0 |
|x x0 |
V0
I
(4.6)
Z
0
0
1
P(x0 )
3 0 P(x )
2 0 0
dx
=
d x n
40 S 0
|x x0 |
|x x0 |
V0
where the first term, which describes the effects of the induced, non-cancelling
dipole moment on the surface of the volume, can be neglected, unless there is a
discontinuity in n P at the surface. Doing so, we find that the contribution from
the electric dipole moments to the potential is given by
p =
1
40
d3x0
V0
0 P(x0 )
|x x0 |
(4.7)
Comparing this expression with expression equation (3.3) on page 39 for the electrostatic potential from a static charge distribution , we see that P(x) has
the characteristics of a charge density and that, to the lowest order, the effective
charge density becomes (x) P(x), in which the second term is a polarisation
term.
The version of equation (1.7) on page 5 where free, true charges and bound,
polarisation charges are separated thus becomes
E=
(4.8)
Rewriting this equation, and at the same time introducing the electric displacement vector (C/m2 )
D = 0 E + P
(4.9)
we obtain
(0 E + P) = D = true (x)
(4.10)
where true is the true charge density in the medium. This is one of Maxwells
equations and is valid also for time varying fields. By introducing the notation
pol = P for the polarised charge density in the medium, and total = true +
pol for the total charge density, we can write down the following alternative
version of Maxwells equation (4.22a) on page 58
E=
total (x)
0
(4.11)
55
Often, for low enough field strengths |E|, the linear and isotropic relationship
between P and E
P = 0 E
(4.12)
(4.13)
where, approximately,
= 0 (1 + )
(4.14)
56
order magnetic moment, corresponding to the electric dipole moment (4.2), is the
magnetic dipole moment
m=
1
2
Z
V0
(4.15)
(4.16)
In a stationary medium we therefore have a total current which is (approximately) the sum of the three currents enumerated above:
jtotal = jtrue +
P
+M
t
(4.17)
B
M
0
(4.18)
and using the definition for D, equation (4.9) on page 55, we can write this incorrect equation in the following form
H = jtrue +
P
D
E
= jtrue +
0
t
t
t
(4.19)
As we see, in this simplistic view, we would pick up a term which makes the
equation inconsistent; the divergence of the left hand side vanishes while the divergence of the right hand side does not. Maxwell realised this and to overcome
this inconsistency he was forced to add his famous displacement current term
which precisely compensates for the last term in the right hand side. In chapter 1,
we discussed an alternative way, based on the postulate of conservation of electric
charge, to introduce the displacement current.
We may, in analogy with the electric case, introduce a magnetic susceptibility
for the medium. Denoting it m , we can write
H=
(4.20)
57
where, approximately,
= 0 (1 + m )
(4.21)
(4.22a)
B=0
(4.22b)
E=
B
t
(4.22c)
D
(4.22d)
t
and are called Maxwells macroscopic equations. These equations are convenient
to use in certain simple cases. Together with the boundary conditions and the constitutive relations, they describe uniquely (but only approximately!) the properties
of the electric and magnetic fields in matter.
H = j(t, x) +
1
d3x0 (H B + E D) =
2
V0
Z
V0
d3x0 j E +
d2x0 n 0 (E H)
(4.24)
S0
We assume the validity of Ohms law so that in the presence of an electromotive force field, we make the linear approximation equation (1.28) on page 12:
j = (E + EEMF )
58
(4.25)
d3x0 j E =
d3x0
V0
j2
d3x0 j EEMF
(4.26)
V0
|V
3 0
1
j2
d3x0 (E D + H B)
+
=
dx
0
0
t
2
|V
{z
}
} | V {z }
EMF
dx jE
{z
3 0
Joule heat
(4.27)
Field energy
d2x0 n 0 (E H)
|
{z
}
(4.28)
S0
Radiated power
which is the energy theorem in Maxwells theory also known as Poyntings theorem.
It is convenient to introduce the following quantities:
1
Ue =
d3x0 E D
2 V0
Z
1
d3x0 H B
Um =
2 V0
S=EH
Z
(4.29)
(4.30)
(4.31)
where Ue is the electric field energy, Um is the magnetic field energy, both measured in J, and S is the Poynting vector (power flux), measured in W/m2 .
59
(D B) + D
t
t
= E( D) B ( H)
(D B) D ( E) + H(
B})
| {z
t
=0
(D B)
t
(4.32)
One verifies easily that the ith vector components of the two terms in square
brackets in the right hand member of (4.32) can be expressed as
1
2
D
E
1
E
D
+
E i D j E D i j
xi
xi
x j
2
(4.33)
1
[H( B) B ( H)]i =
2
B
H
1
H
B
+
Hi B j B H i j
xi
xi
x j
2
(4.34)
[E( D) D ( E)]i =
and
respectively.
Using these two expressions in the ith component of equation (4.32) and reshuffling terms, we get
D
E
B
H
1
D
B
(E + j B)i
E
+ H
+ (D B)i
2
xi
xi
xi
xi
t
1
1
=
E i D j E D i j + Hi B j H B i j
x j
2
2
(4.35)
60
Introducing the electric volume force Fev via its ith component
B
D
E
H
1
+ H
D
B
E
(Fev )i = (E + j B)i
2
xi
xi
xi
xi
(4.36)
and the Maxwell stress tensor T with components
1
1
T i j = E i D j E D i j + Hi B j H B i j
2
2
we finally obtain the force equation
T i j
Fev + (D B) =
= ( T)i
t
x j
i
(4.37)
(4.38)
(4.39)
B = m 0 H = H
(4.40)
(4.41)
where S is the Poynting vector defined in equation (4.31) on page 59. Integration
over the entire volume V yields
e m
d
d3x0 2 S =
d3x0 Fev +
d2x0 T n
0
0
dt V
c
V
S0
| {z }
|
{z
} | {z }
Z
Field momentum
(4.42)
Maxwell stress
which expresses the balance between the force on the matter, the rate of change
of the electromagnetic field momentum and the Maxwell stress. This equation is
called the momentum theorem in Maxwells theory.
In vacuum (4.42) becomes
Z
V0
d3x0 (E + v B) +
1 d
c2 dt
Z
V0
d3x0 S =
I
S0
d2x0 T n
(4.43)
or
d mech d field
p
+ p
=
dt
dt
I
S0
d2x0 T n
(4.44)
61
4.4 Bibliography
[1] E. H ALLN, Electromagnetic Theory, Chapman & Hall, Ltd., London, 1962.
[2] J. D. JACKSON, Classical Electrodynamics, third ed., John Wiley & Sons, Inc.,
New York, NY . . . , 1999, ISBN 0-471-30932-X.
[3] W. K. H. PANOFSKY AND M. P HILLIPS, Classical Electricity and Magnetism, second ed.,
Addison-Wesley Publishing Company, Inc., Reading, MA . . . , 1962, ISBN 0-201-057026.
[4] J. A. S TRATTON, Electromagnetic Theory, McGraw-Hill Book Company, Inc., New
York, NY and London, 1953, ISBN 07-062150-0.
62
Example
4.5 Example
E XAMPLE 4.1
stat (x) =
Z
V0
d3x0
(x0 )
|x x0 |
(4.45)
For a charge distribution source (x0 ), well localised in a small volume V 0 around x0 , we Taylor
expand the inverse distance 1/ |x x0 | with respect to x0 to obtain
1
1
=
|x x0 | |(x x0 ) (x0 x0 )|
=
1
3
n |xx
1
1 3
0|
+
xi xi [(xi01 x0i1 )] [(xi0n x0in )]
|x x0 | n=1 n! i
n
1
=1
i
=1
n
1
n |xx
1
(1)n
0|
n3
n2 0
0
n1 0
+
n (x x01 ) (x2 x02 ) (x3 x03 )
n2
n1
|x x0 | n=1 n1 +n2 +n3 =n n1 !n2 !n3 ! x1 x2 x33 1
ni 0
(4.46)
Inserting this expansion into the integrand of equation (4.45), we get
R 3 0
0
1
stat
V 0 d x (x )
(x) =
40
|x x0 |
1
Z
n |xx
(1)n
0|
0
n3
n2 0
3 0
0
n1 0
(x
)
(x
x
)
(x
x
)
d
x
(x
x
)
0
0
0
n3
n2
n1
3
2
1
3
2
1
V0
n=1 n1 +n2 +n3 =n n1 !n2 !n3 ! x1 x2 x3
ni 0
(4.47)
Limiting ourselves to the first three terms
"
#
1
1
3
3 3
|xx
2 |xx
q
1
1
0|
0|
stat
(x) =
pi
+ Qi j
+ ...
40 |x x0 |
xi
xi x j
i=1 j=1 2
i=1
(4.48)
(4.49)
1
2 |xx
3(xi x0i )(x j x0 j ) |x x0 |2 i j
0|
=
xi x j
|x x0 |5
(4.50)
and
63
5
Electromagnetic
Fields from
Arbitrary Source
Distributions
While, in principle, the electric and magnetic fields can be calculated from the
Maxwell equations in chapter 1, or even from the wave equations in chapter 2, it is
often physically more lucid to calculate them from the electromagnetic potentials
derived in chapter 3. In this chapter we will derive the electric and magnetic fields
from the potentials.
We recall that in order to find the solution (3.33) for the generic inhomogeneous wave equation (3.17) on page 43 we presupposed the existence of a Fourier
transform pair (3.18a) on page 44 for the generic source term
f (t, x) =
f (x) =
d f (x) eit
(5.1a)
1
2
dt f (t, x) eit
(5.1b)
That such transform pairs exist is true for most physical variables which are neither strictly monotonically increasing nor strictly monotonically decreasing with
time. For charge and current densities varying in time we can therefore, without
loss of generality, work with individual Fourier components (x) and j (x), respectively. Strictly speaking, the existence of a single Fourier component assumes
a monochromatic source (i.e., a source containing only one single frequency component), which in turn requires that the electric and magnetic fields exist for infinitely long times. However, by taking the proper limits, we may still use this
approach even for sources and fields of finite duration.
65
This is the method we shall utilise in this chapter in order to derive the electric
and magnetic fields in vacuum from arbitrary given charge densities (t, x) and
current densities j(t, x), defined by the temporal Fourier transform pairs
(t, x) =
(x) =
1
2
d (x) eit
(5.2a)
dt (t, x) eit
(5.2b)
and
j(t, x) =
j (x) =
1
2
d j (x) eit
(5.3a)
dt j(t, x) eit
(5.3b)
under the assumption that only retarded potentials produce physically acceptable
solutions.
The temporal Fourier transform pair for the retarded scalar potential can then
be written
(t, x) =
(x) =
1
2
d (x) eit
(5.4a)
dt (t, x) eit =
1
40
Z
V0
d3x0 (x0 )
eik|xx |
|x x0 |
(5.4b)
where in the last step, we made use of the explicit expression for the temporal
Fourier transform of the generic potential component (x), equation (3.30) on
page 46. Similarly, the following Fourier transform pair for the vector potential
must exist:
A(t, x) =
A (x) =
1
2
d A (x) eit
(5.5a)
dt A(t, x) eit =
0
4
Z
V0
d3x0 j (x0 )
eik|xx |
|x x0 |
(5.5b)
(5.6a)
j(t, x) = j0 (x)e
(5.6b)
i0 t
(t, x) = 0 (x)e
i0 t
A(t, x) = A0 (x)e
66
(5.6c)
i0 t
(5.6d)
(5.7)
The calculations are much simplified if we work in space and, at the final
stage, inverse Fourier transform back to ordinary t space. We are working in
the Lorenz-Lorentz gauge and note that in space the Lorenz-Lorentz condition,
equation (3.15) on page 43, takes the form
k
A i = 0
c
(5.8)
which provides a relation between (the Fourier transforms of) the vector and scalar
potentials.
Using the Fourier transformed version of equation (5.7) and equation (5.5b)
on page 66, we obtain
B (x) = A (x) =
Z
V0
d3x0 j (x0 )
eik|xx |
|x x0 |
(5.9)
67
Utilising formula (F.57) on page 177 and recalling that j (x0 ) does not depend on
x, we can rewrite this as
ik|xx0 |
Z
e
0
3 0
0
d x j (x )
B (x) =
4 V 0
|x x0 |
Z
0
x x0
0
3 0
0
eik|xx |
=
d x j (x )
3
0
4 V 0
|x x |
Z
x x0 ik|xx0 |
1
3 0
0
+ d x j (x ) ik
e
(5.10)
|x x0 |
|x x0 |
V0
Z
0
0
j (x0 )eik|xx | (x x0 )
=
d3x0
4 V 0
|x x0 |3
Z
0 ik|xx0 |
(x x0 )
3 0 (ik)j (x )e
+ dx
|x x0 |2
V0
From this expression for the magnetic field in the frequency () domain, we
obtain the total magnetic field in the temporal (t) domain by taking the inverse
Fourier transform (using the identity ik = i/c):
B(t, x) =
d B (x) eit
R
Z
0 i(tk|xx0 |)
(x x0 )
0
d j (x )e
3 0
=
dx
4
|x x0 |3
V0
R
Z
0 i(tk|xx0 |)
(x x0 )
1
d (i)j (x )e
3 0
dx
+
c V0
|x x0 |2
Z
Z
0
0 0
0
j(t0 , x0 ) (x x0 ) 0
3 0 j(tret , x ) (x x )
=
d3x0 ret
+
d
x
3
2
4 V 0
4c V 0
|x x0 |
|x x0 |
|
{z
} |
{z
}
Induction field
Radiation field
(5.11)
where
def
0
j(tret
, x0 )
j
t
(5.12)
0
t=tret
0
and tret
is given in equation (3.32) on page 46. The first term, the induction field,
dominates near the current source but falls off rapidly with distance from it, is the
electrodynamic version of the Biot-Savart law in electrostatics, formula (1.15) on
page 8. The second term, the radiation field or the far field, dominates at large
distances and represents energy that is transported out to infinity. Note how the
spatial derivatives () gave rise to a time derivative ()!
68
ik|xx |
eik|xx | i0
1
3 0
0 e
d3x0 (x0 )
+
d
x
j
(x
)
=
40
4 V 0
|x x0 |
|x x0 |
V0
Z
0
0 ik|xx0 |
(x )e
1
(x x )
=
d3x0
0
0
40 V
|x x |3
0
Z
(x0 )(x x0 ) j (x0 ) eik|xx |
ik d3x0
c
|x x0 |
|x x0 |
V0
(5.13)
Using the Fourier transform of the continuity equation (1.23) on page 10
0 j (x0 ) i (x0 ) = 0
(5.14)
Doing so in the last term of equation (5.13) above, and also using the fact that
k = /c, we can rewrite this equation as
Z
0 ik|xx0 |
1
(x x0 )
3 0 (x )e
E (x) =
dx
40 V 0
|x x0 |3
0
ik|xx0 |
Z
1
[ j (x0 )](x x0 )
e
(5.16)
0
d3x0
ikj
(x
)
c V0
|x x0 |
|x x0 |
|
{z
}
I
The last vector-valued integral can be further rewritten in the following way:
ik|xx0 |
0
Z
e
[ j (x0 )](x x0 )
0
3 0
ikj (x )
I =
dx
0
0
|x x |
|x x0 |
V
(5.17)
0
Z
jm xl xl0
eik|xx |
0
=
d3x0
ik
j
(x
)
x
l
l
0 |x x0 |
xm
|x x0 |
V0
But, since
xl xl0 ik|xx0 |
jm
xl xl0 ik|xx0 |
j
e
=
e
m
0
0
xm
xm
|x x0 |2
|x x0 |2
xl xl0 ik|xx0 |
+ jm 0
e
xm |x x0 |2
(5.18)
69
we can rewrite I as
0
eik|xx |
xl xl0
ik|xx0 |
x l e
I = d x jm 0
+ ikj
xm |x x0 |2
|x x0 |
V0
Z
xl xl0
0
+ d3x0 0 jm
x l eik|xx |
xm
|x x0 |2
V0
Z
3 0
(5.19)
3 0
Using the triple product bac-cab formula (F.51) on page 176 backwards, and
inserting the resulting expression for I into equation (5.16) on page 69, we arrive
at the following final expression for the Fourier transform of the total E field:
0
1
eik|xx | i0
d3x0 (x0 )
+
40
4
|x x0 |
V0
Z
0
0 ik|xx0 |
1
(x
x
)
(x
)e
=
d3x0
3
0
40 V 0
|x x |
Z
E (x) =
Z
V0
d3x0 j (x0 )
eik|xx |
|x x0 |
1
[j (x0 )eik|xx | (x x0 )](x x0 )
+
d3x0
c V0
|x x0 |4
0
Z
[j (x0 )eik|xx | (x x0 )] (x x0 )
1
+
d3x0
c V0
|x x0 |4
0
Z
0 ik|xx |
ik
(x x0 )] (x x0 )
3 0 [j (x )e
dx
c V0
|x x0 |3
Z
(5.21)
Taking the inverse Fourier transform of equation (5.21), once again using the
vacuum relation = kc, we find, at last, the expression in time domain for the
70
d E (x) eit
1
40
|
d3x0
V0
0
(tret
, x0 )(x x0 )
|x x0 |3
{z
}
1
40 c
|
1
40 c
|
d3x0
V0
[j(tret , x ) (x x0 )](x x0 )
|x x0 |4
{z
}
Intermediate field
1
+
4 c2
| 0
d3x0
V0
(5.22)
0
[j(tret
, x0 ) (x x0 )] (x x0 )
|x x0 |4
{z
}
Intermediate field
d3x0
V0
0
[j(tret
, x0 ) (x x0 )] (x x0 )
|x x0 |3
{z
}
Radiation field
Here, the first term represents the retarded Coulomb field and the last term represents the radiation field which carries energy over very large distances. The other
two terms represent an intermediate field which contributes only in the near zone
and must be taken into account there.
With this we have achieved our goal of finding closed-form analytic expressions for the electric and magnetic fields when the sources of the fields are completely arbitrary, prescribed distributions of charges and currents. The only assumption made is that the advanced potentials have been discarded; recall the
discussion following equation (3.33) on page 46 in chapter 3.
71
Erad (t, x) =
0
4c
d3x0
V0
0
j(tret
, x0 ) (x x0 )
|x x0 |2
(5.23a)
1
=
40 c2
0
[j(tret
, x0 ) (x x0 )] (x x0 )
dx
|x x0 |3
V0
(5.23b)
3 0
where
def
0
j(tret
, x0 )
j
t
(5.24)
0
t=tret
Instead of studying the fields in the time domain, we can often make a spectrum analysis into the frequency domain and study each Fourier component separately. A superposition of all these components and a transformation back to the
time domain will then yield the complete solution.
The Fourier representation of the radiation fields equation (5.23a) and equation (5.23b) above were included in equation (5.10) on page 68 and equation (5.21)
on page 70, respectively and are explicitly given by
1
=
dt Brad (t, x) eit
2
Z
k0
j (x0 ) (x x0 ) ik|xx0 |
e
= i
d3x0
4 V 0
|x x0 |2
Z
0
j (x0 ) k ik|xx0 |
= i
e
d3x0
4 V 0
|x x0 |
Z
1
Erad
(x)
=
dt Erad (t, x) eit
2
Z
k
[j (x0 ) (x x0 )] (x x0 ) ik|xx0 |
= i
e
d3x0
40 c V 0
|x x0 |3
Z
1
[j (x0 ) k] (x x0 ) ik|xx0 |
= i
d3x0
e
40 c V 0
|x x0 |2
Brad
(x)
(5.25a)
(5.25b)
72
dS = d2x n
S (x0 )
x x0
x x0
x0
x0 x0
x0
V0
O
F IGURE 5.1: Relation between the surface normal and the k vector for radiation
generated at source points x0 near the point x0 in the source volume V 0 . At distances much larger than the extent of V 0 , the unit vector n,
normal to the surface
S which has its centre at x0 , and the unit vector k of the radiation k vector from
x0 are nearly coincident.
(5.26)
Recalling from Formula (F.45) and formula (F.46) on page 176 that
dS = |x x0 |2 d = |x x0 |2 sin d d
and noting from figure 5.1 that k and n are nearly parallel, we see that we can
approximate
k dS
d2x
k n d
|x x0 |2 |x x0 |2
(5.27)
73
i
e
e
d
x
40 c
|x x0 |2
V0
Z
1 eik|xx0 | (x x0 )
0
i
d3x0 [j (x0 ) k] eik(x x0 )
0
40 c |x x0 | |x x0 |
V
(5.28b)
Brad
(x) i
hSi = hE Hi =
Using the far-field approximations (5.28a) and (5.28b) and the fact that 1/c =
74
(5.30)
Radiated energy
(5.31)
dt S(t) =
=
dt (E H)
Z
dt (E H ) e
(5.32)
i(+0 )t
If we carry out the temporal integration first and use the fact that
Z
dt ei(+ )t = 2( + 0 )
(5.33)
dt S(t) = 2
d (E H )
Z
Z 0
= 2
d (E H ) +
d (E H )
0
Z
Z
= 2
d (E H )
d (E H )
0
0
Z
Z
= 2
d (E H ) +
d (E H )
0
(5.34)
2
d (E B + E B )
=
0 0
Z
2
=
d (E B + E B )
0 0
Z
75
integration over the area S of a large sphere which encloses the source volume V 0 ,
is
2
Z
r I
Z
0
1
j k ik|xx0 |
e
(5.35)
d2x n
d d3x0
U=
k
4 0 S
|x x0 |
0
V0
Inserting the approximations (5.26) and (5.27) into equation (5.35) above and
also introducing
U=
dU
(5.36)
and recalling the definition (2.26) on page 29 for the vacuum resistance R0 we
obtain
2
Z
dU
1
0
3
0
ik(x
x
)
0
d
R0 d x (j k)e
(5.37)
d
d
4
V0
which, at large distances, is a good approximation to the energy that is radiated
per unit solid angle d in a frequency band d. It is important to notice that
Formula (5.37) includes only source coordinates. This means that the amount of
energy that is being radiated is independent on the distance to the source (as long
as it is large).
5.5 Bibliography
[1] F. H OYLE , S IR AND J. V. NARLIKAR, Lectures on Cosmology and Action at a Distance
Electrodynamics, World Scientific Publishing Co. Pte. Ltd, Singapore, New Jersey, London and Hong Kong, 1996, ISBN 9810-02-2573-3(pbk).
[2] J. D. JACKSON, Classical Electrodynamics, third ed., John Wiley & Sons, Inc.,
New York, NY . . . , 1999, ISBN 0-471-30932-X.
[3] L. D. L ANDAU AND E. M. L IFSHITZ, The Classical Theory of Fields, fourth revised English ed., vol. 2 of Course of Theoretical Physics, Pergamon Press, Ltd., Oxford . . . , 1975,
ISBN 0-08-025072-6.
[4] W. K. H. PANOFSKY AND M. P HILLIPS, Classical Electricity and Magnetism, second ed.,
Addison-Wesley Publishing Company, Inc., Reading, MA . . . , 1962, ISBN 0-201-057026.
[5] J. A. S TRATTON, Electromagnetic Theory, McGraw-Hill Book Company, Inc., New
York, NY and London, 1953, ISBN 07-062150-0.
76
6
Electromagnetic
Radiation and
Radiating
Systems
In chapter 3 we were able to derive general expressions for the scalar and vector
potentials from which we then, in chapter 5, calculated the total electric and magnetic fields from arbitrary distributions of charge and current sources. The only
limitation in the calculation of the fields was that the advanced potentials were
discarded.
Thus, one can, at least in principle, calculate the radiated fields, Poynting
flux, energy and other electromagnetic quantities for an arbitrary current density
Fourier component and then add these Fourier components together to construct
the complete electromagnetic field at any time at any point in space. However, in
practice, it is often difficult to evaluate the source integrals unless the current has
a simple distribution in space. In the general case, one has to resort to approximations. We shall consider both these situations.
77
(6.1)
and where the spatially varying Fourier amplitude I(x30 ) of the antenna current
fulfils the time-independent wave equation (Helmholtz equation)
d2 I
+ k2 I(x30 ) = 0 ,
dx302
I(L/2) = I(L/2) = 0 ,
3
I(x30 ) = I0
sin(kL/2)
I(0) = I0
(6.2)
(6.3)
78
sin[k(L/2 x30 )]
j(t0 , x0 )
L2
L
2
F IGURE 6.1: A linear antenna used for transmission. The current in the feeder
and the antenna wire is set up by the EMF of the generator (the transmitter).
At the ends of the wire, the current is reflected back with a 180 phase shift to
produce a antenna current in the form of a standing wave.
x30 )]
x 3
sin(kL/2)
sin[k(L/2
I0 (x10 )(x20 )
(6.4)
For a halfwave dipole antenna (L = /2), the antenna current density is simply
j0 (x0 ) = I0 (x10 )(x20 ) cos(kx30 )
(6.5)
(6.6)
In the case of a travelling wave antenna, in which one end of the antenna is connected to ground via a resistance so that the current at this end does not vanish,
the Fourier amplitude of the antenna current density is
j0 (x0 ) = I0 (x10 )(x20 ) exp(kx30 )
(6.7)
79
x3 = z
L
2
j (x )
x2
x1
L2
F IGURE 6.2: We choose a spherical polar coordinate system (r = |x| , , ) and
arrange it so that the linear electric dipole antenna axis (and thus the antenna
current density j ) is along the polar axis with the feed point at the origin.
2
0
Z L/2
0 sin[k(L/2 x3 )]
ikx30 cos ikx0 cos
=
k sin e
e
dx3 I0
L/2
sin(kL/2)
2
2
2
Z L/2 0
2 k sin ikx0 cos 2
0
0
e
2
= I0 2
dx3 sin[k(L/2 x3 )] cos(kx3 cos )
sin (kL/2)
0
2
cos[(kL/2) cos ] cos(kL/2)
2
= 4I0
sin sin(kL/2)
(6.8)
Inserting this expression and d = 2 sin d into formula (5.31) on page 75 and
integrating over , we find that the total radiated power from the antenna is
2
Z
cos[(kL/2) cos ] cos(kL/2)
2 1
P(L) = R0 I0
d
sin
(6.9)
4 0
sin sin(kL/2)
80
lim P(L) =
kL0
12
2
L
R0 I02
(6.10)
R (L) = 2 = 1 2 = R0
6
Ieff
2 I0
rad
2
2
L
L
197
(6.11)
Z
1 2 1 cos v
1
=
dv = [ + ln 2 Ci(2)]
2 0
v
2
1.22
(6.13)
where in the last step the Euler-Mascheroni constant = 0.5772 . . . and the cosine
integral Ci(x) were introduced. Inserting this into the expression equation (6.12)
we obtain the value Rrad (/2) 73 .
81
x3 = z = z 0
k
x2
z 0
j (x0 )
x0
x1
0
0
F IGURE 6.3: For the loop antenna the spherical coordinate system (r, , ) describes the field point x (the radiation field) and the cylindrical coordinate system
(0 , 0 , z0 ) describes the source point x0 (the antenna current).
the Cartesian coordinate system x1 x2 x3 with its origin at the centre of the loop as
in figure 6.3
According to equation (5.28a) on page 74 the Fourier component of the radiation part of the magnetic field generated by an extended, monochromatic current
source is
Brad
=
i0 eik|x|
4 |x|
Z
V0
d3x0 eikx j k
(6.14)
In our case the generator produces a single frequency and we feed the antenna
across a small gap where the loop crosses the positive x1 axis. The circumference
of the loop is chosen to be exactly one wavelength = 2c/. This means that
the antenna current oscillates in the form of a sinusoidal standing current wave
around the circular loop with a Fourier amplitude
j = I0 cos 0 (0 a)(z0 ) 0
(6.15)
For the spherical coordinate system of the field point, we recall from subsec-
82
tion F.4.1 on page 176 that the following relations between the base vectors hold:
r = sin cos x 1 + sin sin x 2 + cos x 3
= cos cos x 1 + cos sin x 2 sin x 3
= sin x 1 + cos x 2
and
x 1 = sin cos r + cos cos sin
x 2 = sin sin r + cos sin + cos
x 3 = cos r sin
With the use of the above transformations and trigonometric identities, we obtain
for the cylindrical coordinate system which describes the source:
0 = cos 0 x 1 + sin 0 x 2
= sin cos(0 )r + cos cos(0 ) + sin(0 )
0 = sin 0 x 1 + cos 0 x 2
= sin sin(0 )r cos sin(0 ) + cos(0 )
z 0 = x 3 = cos r sin
(6.16)
(6.17)
(6.18)
(6.19)
(6.20)
and
V0
d3x0 eikx j k = a
= I0 ak
0
eika sin cos( ) cos(0 ) cos 0 d0
(6.21)
+ I0 ak cos
Utilising the periodicity of the integrands over the integration interval [0, 2],
introducing the auxiliary integration variable 00 = 0 , and utilising standard
83
trigonometric identities, the first integral in the RHS of (6.21) can be rewritten
Z
00
(6.22)
2
1
00
sin
eika sin cos d00
2
0
Z 2
1
00
sin
eika sin cos cos 200 d00
2
0
(6.23)
e
0
i cos
in
cos n d =
2
ei cos cos n d
(6.24)
(6.25)
eika sin cos cos 200 d00 = 2J2 (ka sin )
00
0
d3x0 eikx j k = I + I
(6.26)
84
i0 eikr
I + I
4r
(6.27)
To obtain the desired physical magnetic field in the radiation (far) zone we
must Fourier transform back to t space and take the real part and evaluate it at the
retarded time:
0
i0 e(ikrt )
I + I
Brad (t, x) = Re
4r
0
sin(kr t0 ) I + I
=
4r
I0 ak0
0
=
sin(kr t ) cos [J0 (ka sin ) J2 (ka sin )]
4r
+ cos sin [J0 (ka sin ) + J2 (ka sin )]
(6.28)
From this expression for the radiated B field, we can obtain the radiated E field
with the help of Maxwells equations.
(6.29)
= jtrue
t
(6.30a)
(6.30b)
85
and compare with equation (6.29) on page 85, we see that (t, x) satisfies this
equation of continuity. Furthermore, if we compare with the electric polarisation
[cf. equation (4.9) on page 55], we see that the quantity is related to the true
charges in the same way as P is related to polarised charge, namely as a dipole
moment density. The quantity is referred to as the polarisation vector since,
formally, it treats also the true (free) charges as polarisation charges so that
E=
true + pol P
=
0
0
(6.31)
(6.32a)
(6.32b)
where and A are the electromagnetic scalar and vector potentials, respectively.
As we see, e acts as a super-potential in the sense that it is a potential from
which we can obtain other potentials. It is called the Hertz vector or polarisation
potential. Requiring that the scalar and vector potentials and A, respectively,
fulfil their inhomogeneous wave equations, one finds, using (6.30) and (6.32),
that Hertz vector must satisfy the inhomogeneous wave equation
2 e =
1 2 e
2 e =
c2 t2
0
(6.33)
This equation is of the same type as equation (3.17) on page 43, and has therefore the retarded solution
e (t, x) =
1
40
d3x0
V0
0
(tret
, x0 )
|x x0 |
(6.34)
1
=
40
(x0 )eik|xx |
dx
|x x0 |
V0
3 0
(6.35)
(6.36)
we see that we can calculate the magnetic and electric fields, respectively, as follows
1 C
c2 t
(6.37a)
E=C
(6.37b)
B=
86
x x0
x0
x x0
x
x0 x0
x0
V0
O
F IGURE 6.4: Geometry of a typical multipole radiation problem where the field
point x is located some distance away from the finite source volume V 0 centred
around x0 . If k |x0 x0 | 1 k |x x0 |, then the radiation at x is well approximated by a few terms in the multipole expansion.
Clearly, the last equation is valid only outside the source volume, where E = 0.
Since we are mainly interested in the fields in the far zone, a long distance from
the source region, this is no essential limitation.
Assume that the source region is a limited volume around some central point
x0 far away from the field (observation) point x illustrated in figure 6.4. Under
these assumptions, we can expand the Hertz vector, expression (6.35) on page 86,
0
due to the presence of non-vanishing (tret
, x0 ) in the vicinity of x0 , in a formal
series. For this purpose we recall from potential theory that
0
eik|xx |
eik|(xx0 )(x x0 )|
0
|x x | |(x x0 ) (x0 x0 )|
(6.38)
= ik (2n + 1)Pn (cos ) jn (k x0 x0 )h(1)
n (k |x x0 |)
n=0
eik|xx |
is a Green function
|x x0 |
is the angle between x0 x0 and x x0
Pn (cos ) is the Legendre polynomial of order n
jn (k x0 x0 ) is the spherical Bessel function of the first kind of order n
h(1)
n (k |x x0 |) is the spherical Hankel function of the first kind of order n
According to the addition theorem for Legendre polynomials
Pn (cos ) =
0 im( )
(1)m Pmn (cos )Pm
n (cos )e
(6.39)
m=n
87
where Pm
n is an associated Legendre polynomial and, in spherical polar coordinates,
x0 x0 = (x0 x0 , 0 , 0 )
(6.40a)
x x0 = (|x x0 | , , )
(6.40b)
Inserting equation (6.38) on page 87, together with formula (6.39) on page 87,
into equation (6.35) on page 86, we can in a formally exact way expand the Fourier
component of the Hertz vector as
e =
ik
40
Z
V0
m
im
(2n + 1)(1)m h(1)
n (k |x x0 |) Pn (cos ) e
n=0 m=n
(6.41)
0 im0
d x (x ) jn (k x0 x0 ) Pm
n (cos ) e
3 0
eik|xx0 |
k |x x0 |
(6.43)
and replace jn with the first term in its power series expansion:
jn (k x0 x0 )
n
2n n!
k x0 x0
(2n + 1)!
(6.44)
Inserting these expansions into equation (6.41), we obtain the multipole expansion
of the Fourier component of the Hertz vector
e(n)
(6.45a)
n=0
where
e(n)
1 eik|xx0 | 2n n!
= (i)
40 |x x0 | (2n)!
n
Z
V0
d3x0 (x0 ) (k x0 x0 )n Pn (cos )
(6.45b)
88
k
x3
Brad
x
Erad
x2
x1
F IGURE 6.5: If a spherical polar coordinate system (r, , ) is chosen such that
the electric dipole moment p (and thus its Fourier transform p ) is located at the
origin and directed along the polar axis, the calculations are simplified.
eik|xx0 |
40 |x x0 |
Z
V0
d3x0 (x0 ) =
1 eik|xx0 |
p
40 |x x0 |
(6.46)
89
Since represents a dipole moment density for the Rtrue charges (in the same
vein as P does so for the polarised charges), p = V 0 d3x0 (x0 ) is the Fourier
component of the electric dipole moment
p(t, x0 ) =
Z
V0
d x (t , x ) =
3 0
Z
V0
(6.47)
[cf. equation (4.2) on page 54 which describes the static dipole moment]. If a
spherical coordinate system is chosen with its polar axis along p as in figure 6.5
on page 89, the components of e(0) are
1 eik|xx0 |
p cos
40 |x x0 |
def
1 eik|xx0 |
p sin
e e(0) =
40 |x x0 |
def
er e(0) r =
def
e e(0) = 0
(6.48a)
(6.48b)
(6.48c)
Evaluating formula (6.36) on page 86 for the help vector C, with the spherically polar components (6.48) of e(0) inserted, we obtain
ik|xx0 |
1
1
e
(0)
C = C, =
ik
p sin
(6.49)
40 |x x0 |
|x x0 |
Applying this to equations (6.37) on page 86, we obtain directly the Fourier components of the fields
ik|xx0 |
1
0
e
B = i
ik
p sin
(6.50a)
4
|x x0 |
|x x0 |
ik
x x0
1
1
E =
2
cos
2
40
|x x0 |
|x x0 |
|x x0 |
(6.50b)
ik|xx0 |
ik
1
e
2
k sin
p
|x x0 |
|x x0 |2 |x x0 |
Keeping only those parts of the fields which dominate at large distances (the
radiation fields) and recalling that the wave vector k = k(x x0 )/ |x x0 | where
k = /c, we can now write down the Fourier components of the radiation parts of
the magnetic and electric fields from the dipole:
0 eik|xx0 |
0 eik|xx0 |
p k sin =
(p k)
(6.51a)
4 |x x0 |
4 |x x0 |
1 eik|xx0 |
1 eik|xx0 |
=
p k2 sin =
[(p k) k] (6.51b)
40 |x x0 |
40 |x x0 |
Brad
=
Erad
These fields constitute the electric dipole radiation, also known as E1 radiation.
90
e(1) = i
(6.52)
(6.53)
and introducing
i = xi x0,i
(6.54a)
0i
(6.54b)
xi0
x0,i
1
i , j 0i + ,i 0j
2
1
+ i , j 0i ,i 0j
2
(6.55)
i.e., as the sum of two parts, the first being symmetric and the second antisymmetric in the indices i, j. We note that the antisymmetric part can be written as
1
1
i , j 0i ,i 0j = [, j (i 0i ) 0j (i ,i )]
2
2
1
= [ ( 0 ) 0 ( )] j
2
1
=
(x x0 ) [ (x0 x0 )] j
2
(6.56)
The utilisation of equations (6.30) on page 85, and the fact that we are considering a single Fourier component,
(t, x) = eit
(6.57)
allow us to express in j as
= i
(6.58)
91
Hence, we can write the antisymmetric part of the integral in formula (6.52) on
page 91 as
1
(x x0 ) d3x0 (x0 ) (x0 x0 )
2
V0
Z
1
= i (x x0 ) d3x0 j (x0 ) (x0 x0 )
2
V0
1
= i (x x0 ) m
(6.59)
Z
V0
(6.60)
The final result is that the antisymmetric, magnetic dipole, part of e(1) can be
written
e,antisym
(1)
k
eik|xx0 |
(x x0 ) m
40 |x x0 |2
(6.61)
In analogy with the electric dipole case, we insert this expression into equation (6.36) on page 86 to evaluate C, with which equations (6.37) on page 86
then gives the B and E fields. Discarding, as before, all terms belonging to the
near fields and transition fields and keeping only the terms that dominate at large
distances, we obtain
0 eik|xx0 |
(m k) k
4 |x x0 |
k eik|xx0 |
Erad
m k
(x) =
40 c |x x0 |
Brad
(x) =
(6.62a)
(6.62b)
which are the fields of the magnetic dipole radiation (M1 radiation).
92
Z
V0
0
d3x0 (x0 x0 )(x0 x0 )(tret
, x0 )
(6.63)
Again we use this expression in equation (6.36) on page 86 to calculate the fields
via equations (6.37) on page 86. Tedious, but fairly straightforward algebra (which
we will not present here), yields the resulting fields. The radiation components of
the fields in the far field zone (wave zone) are given by
i0 eik|xx0 |
(k Q ) k
8 |x x0 |
i eik|xx0 |
[(k Q ) k] k
Erad
(x)
=
80 |x x0 |
Brad
(x) =
(6.64a)
(6.64b)
(6.65a)
(6.65b)
and consider a source region with such a limited spatial extent that the charges
and currents are well localised. Specifically, we consider a charge q0 , for instance
an electron, which, classically, can be thought of as a localised, unstructured and
rigid charge distribution with a small, finite radius. The part of this charge
distribution dq0 which we are considering is located in dV 0 = d3x0 in the sphere in
figure 6.6 on page 94. Since we assume that the electron (or any other other similar
electric charge) moves with a velocity v whose direction is arbitrary and whose
magnitude can even be comparable to the Rspeed of light, we cannot
say that the
R
0
0
charge and current to be used in (6.65) is V 0 d3x0 (tret
, x0 ) and V 0 d3x0 v(tret
, x0 ),
respectively, because in the finite time interval during which the observed signal
is generated, part of the charge distribution will leak out of the volume element
d3x0 .
93
x(t)
dr0
0
v(t )
x x0
dS0
0 0
x (t )
dV 0
q
F IGURE 6.6: Signals which are observed at the field point x at time t were
generated at source points x0 (t0 ) on a sphere, centred on x and expanding, as time
increases, with the velocity c outward from the centre. The source charge element
moves with an arbitrary velocity v and gives rise to a source leakage out of the
source volume dV 0 = d3x0 .
(x x0 ) v 0 0
dS dt
|x x0 |
(6.66)
where the last term represents the amount of source leakage due to the fact that
the charge distribution moves with velocity v(t0 ) = dx0 /dt0 . Since dt0 = dr0 /c and
dS 0 dr0 = d3x0 we can rewrite the expression for the net charge as
(x x0 ) v 3 0
0
0
dx
dq0 = (tret
, x0 ) d3x0 (tret
, x0 )
c |x x0 |
(x x0 ) v
0
= (tret
, x0 ) 1
d3x0
c |x x0 |
94
(6.67)
or
0
(tret
, x0 ) d3x0 =
dq0
1
(6.68)
(xx0 )v
c|xx0 |
(6.69)
This is the expression to be used in the formulae (6.65) on page 93 for the retarded
potentials. The result is (recall that j = v)
dq0
0
|x x0 | (xxc )v
Z
0
v dq0
A(t, x) =
0
4 |x x0 | (xxc )v
1
(t, x) =
40
(6.70a)
(6.70b)
For a sufficiently small and well localised charge distribution we can, assuming
that the integrands do not change sign in the integration volume, use the mean
value theorem to evaluate these expressions to become
1
1
q0 1
3 0
0
(6.71a)
d
x
dq
=
0
40 |x x0 | (xxc )v V 0
40 s
Z
1
v
v
q0 v
A(t, x) =
=
(t, x)
d3x0 dq0 =
0 )v
(xx
2
40 c |x x0 | c
40 c2 s c2
V0
(6.71b)
(t, x) =
where
[x x0 (t0 )] v(t0 )
s = s(t0 , x) = x x0 (t0 )
c
0 0
0
v(t
)
x
x
(t
)
0
0
= x x (t ) 1
c
|x x0 (t0 )|
0 0
0
x x (t )
v(t )
= [x x0 (t0 )]
c
|x x0 (t0 )|
(6.72a)
(6.72b)
(6.72c)
is the retarded relative distance. The potentials (6.71) are precisely the LinardWiechert potentials which will be derived in section 7.3.2 on page 144 by using a
relativistically covariant formalism.
It should be noted that in the complicated derivation presented above, the observer is in a coordinate system which has an absolute meaning and the velocity
v is that of the localised charge q0 , whereas, as we shall see later, in the covariant derivation, two reference frames of equal standing are moving relative to each
other with v.
95
?
|x x0 |
v
c
q0
x0 (t0 )
v(t0 )
x0 (t)
x x0
x x0
x(t)
F IGURE 6.7: Signals which are observed at the field point x at time t were
generated at the source point x0 (t0 ). After time t0 the particle, which moves with
nonuniform velocity, has followed a yet unknown trajectory. Extrapolating tangentially the trajectory from x0 (t0 ), based on the velocity v(t0 ), defines the virtual
simultaneous coordinate x0 (t).
The Linard-Wiechert potentials are applicable to all problems where a spatially localised charge in arbitrary motion emits electromagnetic radiation, and
we shall now study such emission problems. The electric and magnetic fields are
calculated from the potentials in the usual way:
B(t, x) = A(t, x)
E(t, x) = (t, x)
(6.73a)
A(t, x)
t
(6.73b)
(6.74)
(in the interest of simplifying our notation, we drop the subscript ret on t0 from
now on). This means that we know the trajectory of the charge q0 , i.e., x0 , for all
times up to the time t0 at which a signal was emitted in order to precisely arrive at
the field point x at time t. Because of the finite speed of propagation of the fields,
the trajectory at times later than t0 cannot be known at time t.
96
dx0
dt0
a(t0 ) = v (t0 ) =
(6.75a)
dv d2 x0
= 02
dt0
dt
(6.75b)
(6.76a)
d2
[x x0 (t0 )] = v(t0 )
dt0 2
(6.76b)
The retarded time t0 can, at least in principle, be calculated from the implicit
relation
|x x0 (t0 )|
(6.77)
c
and we shall see later how this relation can be taken into account in the calculations.
According to formulae (6.73) on page 96 the electric and magnetic fields are
determined via differentiation of the retarded potentials at the observation time t
and at the observation point x. In these formulae the unprimed , i.e., the spatial
derivative differentiation operator = x i /xi means that we differentiate with
respect to the coordinates x = (x1 , x2 , x3 ) while keeping t fixed, and the unprimed
time derivative operator /t means that we differentiate with respect to t while
keeping x fixed. But the Linard-Wiechert potentials and A, equations (6.71) on
page 95, are expressed in the charge velocity v(t0 ) given by equation (6.75a) above
and the retarded relative distance s(t0 , x) given by equation (6.72) on page 95.
This means that the expressions for the potentials and A contain terms which
are expressed explicitly in t0 , which in turn is expressed implicitly in t via equation (6.77) above. Despite this complication it is possible, as we shall see below,
to determine the electric and magnetic fields and associated quantities at the time
of observation t. To this end, we need to investigate carefully the action of differentiation on the potentials.
t0 = t0 (t, x) = t
97
(x x0 ) v(t0 )
0 0
x x0 (t0 ) = x x
x
x
(t
)
=
t0 x
t0 x
|x x0 |
|x x0 |
(6.78)
Furthermore, by applying the operator (/t)x to equation (6.77) on page 97 we
find that
0
t
t
|x x0 |
=1
(6.79)
t0 x c
t x
(x x0 ) v(t0 ) t0
=1+
c |x x0 |
t x
This is an algebraic equation in (t0 /t)x which we can solve to obtain
0
t
|x x0 |
|x x0 |
=
=
t x |x x0 | (x x0 ) v(t0 )/c
s
(6.80)
t
|x x0 |
=
=
(6.81)
t x
t x t0 x
s
t0 x
Likewise, by applying ()t to equation (6.77) on page 97 we obtain
x x0
|x x0 (t0 (t, x))|
=
()t (x x0 )
c
c |x x0 |
(x x0 ) v(t0 )
x x0
+
()t t0
=
0
c |x x |
c |x x0 |
()t t0 = ()t
(6.82)
x x0
cs
(6.83)
which gives the following operator relation when ()t is acting on an arbitrary
function of t0 and x:
x x0
0
()t = ()t t
+ ()t0 =
+ ()t0
(6.84)
t0 x
cs
t0 x
98
With the help of the rules (6.84) and (6.81) we are now able to replace t by t0 in
the operations which we need to perform. We find, for instance, that
1 q0
()t =
40 s
0
q
x x0
v(t0 ) x x0 s
=
40 s2 |x x0 |
c
cs
t0 x
A
0 q0 v(t0 )
A
=
t
t x t 4 s
x
0 s
q0
0
0
0
=
x x sv(t ) x x v(t )
40 c2 s3
t0 x
(6.85a)
(6.85b)
Utilising these relations in the calculation of the E field from the Linard-Wiechert
potentials, equations (6.71) on page 95, we obtain
cs(t0 , x)
t0
c2
x
(6.86)
Starting from expression (6.72a) on page 95 for the retarded relative distance
s(t0 , x), we see that we can evaluate (s/t0 )x in the following way
s
t0
x
(x x0 ) v(t0 )
0
=
xx
t0 x
c
1 [x x0 (t0 )]
v(t0 )
0 0
0
0 0
= 0 x x (t )
v(t ) + [x x (t )]
t
c
t0
t0
(x x0 ) v(t0 ) v2 (t0 ) (x x0 ) v (t0 )
=
+
c
c
|x x0 |
(6.87)
where equation (6.78) on page 98 and equations (6.75) on page 97, respectively,
were used. Hence, the electric field generated by an arbitrarily moving charged
99
x
(t
)]
E(t, x) =
40 s3 (t0 , x)
c
c2
|
{z
}
Coulomb field when v 0
x x0 (t0 )
q0
|x x0 (t0 )| v(t0 )
0 0
0
[x x (t )]
v (t )
+
40 s3 (t0 , x)
c2
c
|
{z
}
Radiation (acceleration) field
(6.88)
The first part of the field, the velocity field, tends to the ordinary Coulomb field
when v 0 and does not contribute to the radiation. The second part of the field,
the acceleration field, is radiated into the far zone and is therefore also called the
radiation field.
From figure 6.7 on page 96 we see that the position the charged particle would
have had if at t0 all external forces would have been switched off so that the trajectory from then on would have been a straight line in the direction of the tangent
at x0 (t0 ) is x0 (t), the virtual simultaneous coordinate. During the arbitrary motion,
we interpret x x0 (t) as the coordinate of the field point x relative to the virtual
simultaneous coordinate x0 (t). Since the time it takes for a signal to propagate (in
the assumed vacuum) from x0 (t0 ) to x is |x x0 | /c, this relative vector is given by
x x0 (t) = x x0 (t0 )
|x x0 (t0 )| v(t0 )
c
(6.89)
(6.90)
cs
q0
x x0
x x0
A
=
v
40 c2 s2 |x x0 |
c |x x0 |
t x
t0
A
x
(6.91)
where we made use of equation (6.71) on page 95 and formula (6.81) on page 98.
But, according to (6.85a),
x x0
q0
x x0
()t =
v
0
2
2
c |x x |
40 c s |x x0 |
100
(6.92)
so that
A
x x0
()t
B(t, x) =
0
c |x x |
t x
x x0
=
E(t, x)
c |x x0 |
(6.93)
The radiation part of the electric field is obtained from the acceleration field in
formula (6.88) on page 100 as
Erad (t, x) =
lim E(t, x)
|xx0 |
0
q
|x x0 | v
0
0
v
(x x ) (x x )
=
40 c2 s3
c
q0
[x x0 (t0 )] {[x x0 (t)] v (t0 )}
=
40 c2 s3
(6.94)
where in the last step we again used formula (6.89) on page 100. Using this
formula and formula (6.93), the radiation part of the magnetic field can be written
Brad (t, x) =
x x0
Erad (t, x)
c |x x0 |
(6.95)
(6.96)
(6.97)
(x x0 ) v
c
2
(x x0 ) v
c
2
|x x0 |2 v2 2 0
|x x0 |2 v2
2 0
cos
+
sin
c2
c2
|x x0 |2 v2
|x x0 |2 v2
2 0
2 0
=
(cos
+
sin
)
=
c2
c2
=
(6.98)
101
we find that
2
2
(x x0 ) v
(x x0 ) v
|x x0 |2 v2
=
c
c2
c
(6.99)
Furthermore, from equation (6.89) on page 100, we obtain the following identity:
[x x0 (t0 )] v = [x x0 (t)] v
(6.100)
which, when inserted into equation (6.99) above, yields the relation
2
2
(x x0 ) v
(x x0 ) v
|x x0 |2 v2
=
c
c2
c
(6.101)
Inserting the above into expression (6.96) on page 101 for s2 , this expression
becomes
2
2
(x x0 ) v |x x0 |2 v2
(x x0 ) v
s2 = x x0 2 x x0
+
c
c2
c
2
2
0
(x x0 ) v
|x x | v
= (x x0 )
c
c
2
(x x0 ) v
= (x x0 )2
c
2
[x x0 (t)] v(t0 )
2
|x x0 (t)|
c
(6.102)
where in the penultimate step we used equation (6.89) on page 100.
What we have just demonstrated is that if the particle velocity at time t can be
calculated or projected from its value at the retarded time t0 , the retarded distance
s in the Linard-Wiechert potentials (6.71) can be expressed in terms of the virtual
simultaneous coordinate x0 (t), viz., the point at which the particle will have arrived
at time t, i.e., when we obtain the first knowledge of its existence at the source
point x0 at the retarded time t0 , and in the field coordinate x = x(t), where we
make our observations. We have, in other words, shown that all quantities in the
definition of s, and hence s itself, can, when the motion of the charge is somehow
known, be expressed in terms of the time t alone. I.e., in this special case we are
able to express the retarded relative distance as s = s(t, x) and we do not have
to involve the retarded time t0 or any transformed differential operators in our
calculations.
Taking the square root of both sides of equation (6.102), we obtain the following alternative final expressions for the retarded relative distance s in terms of the
102
(6.103a)
(6.103b)
(6.103c)
If we know what velocity the particle will have at time t, expression (6.103) above
for s will not be dependent on t0 .
Using equation (6.103c) and standard vector analytic formulae, we obtain
"
2 #
(x x0 ) v
v2
2
2
s = |x x0 | 1 2 +
c
c
vv
v2
(6.104)
= 2 (x x0 ) 1 2 + 2 (x x0 )
c
c
h
i
v v
= 2 (x x0 ) +
(x x0 )
c
c
which we shall use in example 6.1 on page 124 for a uniform, unaccelerated motion of the charge.
Erad (t, x) =
q0
(x x0 ) [(x x0 ) v ],
40 c2 |x x0 |3
vc
(6.107)
from which we obtain, with the use of formula (6.93) on page 101, the magnetic
field
q0
Brad (t, x) =
[v (x x0 )], v c
(6.108)
40 c3 |x x0 |2
103
It is interesting to note the close correspondence which exists between the nonrelativistic fields (6.107) and (6.108) and the electric dipole field equations (6.51)
on page 90 if we introduce
p = q0 x0 (t0 )
(6.109)
(6.110a)
x x = x x0
0
(6.110b)
The power flux in the far zone is described by the Poynting vector as a function
of Erad and Brad . We use the close correspondence with the dipole case to find that
it becomes
S=
0 q0 2 (v)2
x x0
2
sin
|x x0 |
162 c |x x0 |2
(6.111)
where is the angle between v and x x0 . The total radiated power (integrated
over a closed spherical surface) becomes
P=
0 q0 2 (v)2
q0 2 v2
=
6c
60 c3
(6.112)
which is the Larmor formula for radiated power from an accelerated charge. Note
that here we are treating a charge with v c but otherwise totally unspecified motion while we compare with formulae derived for a stationary oscillating
dipole. The electric and magnetic fields, equation (6.107) on page 103 and equation (6.108) on page 103, respectively, and the expressions for the Poynting flux
and power derived from them, are here instantaneous values, dependent on the instantaneous position of the charge at x0 (t0 ). The angular distribution is that which
is frozen to the point from which the energy is radiated.
6.3.3 Bremsstrahlung
An important special case of radiation is when the velocity v and the acceleration
v are collinear (parallel or anti-parallel) so that v v = 0. This condition (for
an arbitrary magnitude of v) inserted into expression (6.94) on page 101 for the
radiation field, yields
Erad (t, x) =
104
q0
(x x0 ) [(x x0 ) v ],
40 c2 s3
v k v
(6.113)
v = 0.5c
v=0
v = 0.25c
v
F IGURE 6.8:
Polar diagram of the energy loss angular distribution factor
sin2 /(1 v cos /c)5 during bremsstrahlung for particle speeds v = 0, v = 0.25c,
and v = 0.5c.
from which we obtain, with the use of formula (6.93) on page 101, the magnetic
field
Brad (t, x) =
q0 |x x0 |
[v (x x0 )],
40 c3 s3
v k v
(6.114)
The difference between this case and the previous case of v c is that the approximate expression (6.105) on page 103 for s is no longer valid; we must instead use
the correct expression (6.72) on page 95. The angular distribution of the power
flux (Poynting vector) therefore becomes
S=
sin2
x x0
0 q0 2 v2
162 c |x x0 |2 1 v cos 6 |x x0 |
c
(6.115)
It is interesting to note that the magnitudes of the electric and magnetic fields are
the same whether v and v are parallel or anti-parallel.
We must be careful when we compute the energy (S integrated over time). The
Poynting vector is related to the time t when it is measured and to a fixed surface
in space. The radiated power into a solid angle element d, measured relative to
the particles retarded position, is given by the formula
dU rad ()
0 q0 2 v2
sin2
d = S (x x0 ) x x0 d =
d
dt
162 c 1 v cos 6
c
(6.116)
105
dS
dr
x
d
0
q
0
x02 vdt x01
x x0 + c dt0
2
F IGURE 6.9: Location of radiation between two spheres as the charge moves
with velocity v from x01 to x02 during the time interval (t0 , t0 + dt0 ). The observation
point (field point) is at the fixed location x.
On the other hand, the radiation loss due to radiation from the charge at retarded
time t0 :
dU rad t
dU rad
d
=
d
(6.117)
dt0
dt
t0 x
Using formula (6.80) on page 98, we obtain
dU rad
dU rad s
d
=
d = S (x x0 )s d
dt0
dt |x x0 |
(6.118)
Inserting equation (6.115) on page 105 for S into (6.118), we obtain the explicit expression for the energy loss due to radiation evaluated at the retarded time
0 q0 2 v2
sin2
dU rad ()
d
=
d
dt0
162 c 1 v cos 5
c
(6.119)
The angular factors of this expression, for three different particle speeds, are plotted in figure 6.8 on page 105.
Comparing expression (6.116) on page 105 with expression (6.119) above, we
see that they differ by a factor 1 v cos /c which comes from the extra factor
s/ |x x0 | introduced in (6.118). Let us explain this in geometrical terms.
During the interval (t0 , t0 + dt0 ) and within the solid angle element d the
particle radiates an energy [dU rad ()/dt0 ] dt0 d. As shown in figure 6.9 this energy
106
is at time t located between two spheres, one outer with its origin at x01 (t0 ) and
radius c(t t0 ), and one inner with its origin at x02 (t0 + dt0 ) = x01 (t0 ) + v dt0 and
radius c[t (t0 + dt0 )] = c(t t0 dt0 ).
From Figure 6.9 we see that the volume element subtending the solid angle
element
dS
d =
x x0 2
2
(6.120)
2
d3x = dS dr = x x02 d dr
(6.121)
is
Here, dr denotes the differential distance between the two spheres and can be
evaluated in the following way
x x02
v dt0
dr = x x02 + c dt0 x x02
x x02
| {z }
v cos
!
0
x x2
cs
v dt0 =
dt0
= c
0
x x2
x x02
(6.122)
where formula (6.72) on page 95 was used in the last step. Hence, the volume
element under consideration is
s
dS cdt0
d3x = dS dr =
x x02
(6.123)
We see that the energy which is radiated per unit solid angle during the time
interval (t0 , t0 + dt0 ) is located in a volume element whose size is dependent.
This explains the difference between expression (6.116) on page 105 and expression (6.119) on page 106.
Let the radiated energy, integrated over , be denoted U rad . After tedious, but
relatively straightforward integration of formula (6.119) on page 106, one obtains
dU rad 0 q0 2 v2
=
dt0
6c
1
1
v2
c2
3 =
3
2 q0 2 v2
v2
1
3 40 c3
c2
(6.124)
If we know v(t0 ), we can integrate this expression over t0 and obtain the total energy radiated during the acceleration or deceleration of the particle. This way we
obtain a classical picture of bremsstrahlung (braking radiation, free-free radiation). Often, an atomistic treatment is required for obtaining an acceptable result.
107
(6.125a)
x (t ) = a[ x 1 cos (t ) + x 2 sin (t )]
0 0
(6.125b)
0 0
v = |v| = a0
v (t ) =
0
v =
(6.125d)
x (t ) = a20 [ x 1
|v| = a20
0 0
(6.125c)
cos (t ) + x 2 sin (t )]
0
(6.125e)
(6.125f)
Because of the rotational symmetry we can, without loss of generality, rotate our
coordinate system around the x3 axis so the relative vector x x0 from the source
point to an arbitrary field point always lies in the x2 x3 plane, i.e.,
x x0 = x x0 ( x 2 sin + x 3 cos )
(6.126)
where is the angle between x x0 and the normal to the plane of the particle
orbit (see Figure 6.10). From the above expressions we obtain
(x x0 ) v = x x0 v sin cos
(6.127a)
0
0
0
1
1
x x0
(E B) =
|E|2
0
c0
|x x0 |
(6.128)
(6.129)
where the retarded distance s is given by expression (6.72) on page 95. With the
radiation part of the electric field, expression (6.94) on page 101, inserted, and
108
x2
(t, x)
x x0
x
v
q0 0 0
(t , x )
v
(t0 )
x1
x3
F IGURE 6.10: Coordinate system for the radiation from a charged particle at
x0 (t0 ) in circular motion with velocity v(t0 ) along the tangent and constant acceleration v (t0 ) toward the origin. The x1 x2 axes are chosen so that the relative field
point vector x x0 makes an angle with the x3 axis which is normal to the plane
of the orbital motion. The radius of the orbit is a.
using (6.127a) and (6.127b) on page 108, one finds, after some algebra, that
2
2
2
v
v2
dU rad (, ) 0 q0 2 v2 1 c sin cos 1 c2 sin sin
=
5
dt0
162 c
1 cv sin cos
(6.130)
The angles and vary in time during the rotation, so that refers to a moving
coordinate system. But we can parametrise the solid angle d in the angle and
the (fixed) angle so that d = sin d d. Integration of equation (6.130) over
this d gives, after some cumbersome algebra, the angular integrated expression
dU rad 0 q0 2 v2
1
=
2
0
2
dt
6c
1 cv2
(6.131)
109
Cyclotron radiation
For a non-relativistic speed v c, equation (6.130) on page 109 reduces to
dU rad (, ) 0 q0 2 v2
=
(1 sin2 sin2 )
dt0
162 c
But, according to equation (6.127b) on page 108
sin2 sin2 = cos2
(6.132)
(6.133)
where is defined in figure 6.10 on page 109. This means that we can write
dU rad () 0 q0 2 v2
0 q0 2 v2
(1 cos2 ) =
sin2
(6.134)
=
0
2
dt
16 c
162 c
Consequently, a fixed observer near the orbit plane ( /2) will observe
cyclotron radiation twice per revolution in the form of two equally broad pulses
of radiation with alternating polarisation.
Synchrotron radiation
When the particle is relativistic, v . c, the denominator in equation (6.130) on
page 109 becomes very small if sin cos 1, which defines the forward direction of the particle motion ( /2, 0). The equation (6.130) on page 109
becomes
1
dU rad (/2, 0) 0 q0 2 v2
=
0
2
dt
16 c 1 v 3
c
(6.135)
which means that an observer near the orbit plane sees a very strong pulse followed, half an orbit period later, by a much weaker pulse.
The two cases represented by equation (6.134) above and equation (6.135) are
very important results since they can be used to determine the characteristics of
the particle motion both in particle accelerators and in astrophysical objects where
a direct measurement of particle velocities are impossible.
In the orbit plane ( = /2), equation (6.130) on page 109 gives
2
v
v2
rad
0 2 2 1 c cos
sin2
2
c
dU (/2, ) 0 q v
=
(6.136)
5
dt0
162 c
1 v cos
c
110
(6.137a)
(6.137b)
x2
(t, x)
x x0
v
q0 0 0
(t , x )
v
(t0 )
0
x1
x3
F IGURE 6.11:
Hence, the angle 0 is a measure of the synchrotron radiation lobe width ; see
figure 6.11. For ultra-relativistic particles, defined by
r
v2
1
1 2 1,
(6.138)
= q
1,
2
c
1 v
c2
v2
1
=
2
c
(6.139)
(6.140)
This angular interval is swept by the charge during the time interval
t0 =
(6.141)
(6.142)
111
in the direction toward the observer who therefore measures a compressed pulse
width of length
vt0
v 0
v
v 1
l0
= t0
= 1
t = 1
1
t = t0
c
c
c
c 0
c 0
v
v
2
1 c 1+ c
1
v
1
1 1
=
1 2
= 3
v
0
c
20 2 0
1+
| {z }
| {z c}
1/2
2
(6.143)
Typically, the spectral width of a pulse of length t is . 1/t. In the ultrarelativistic synchrotron case one can therefore expect frequency components up
to
1
max
= 23 0
(6.144)
t
A spectral analysis of the radiation pulse will therefore exhibit a (broadened) line
spectrum of Fourier components n0 from n = 1 up to n 23 .
When many charged particles, N say, contribute to the radiation, we can have
three different situations depending on the relative phases of the radiation fields
from the individual particles:
1. All N radiating particles are spatially much closer to each other than a typical wavelength. Then the relative phase differences of the individual electric and magnetic fields radiated are negligible and the total radiated fields
from all individual particles will add up to become N times that from one
particle. This means that the power radiated from the N particles will be N 2
higher than for a single charged particle. This is called coherent radiation.
2. The charged particles are perfectly evenly distributed in the orbit. In this
case the phases of the radiation fields cause a complete cancellation of the
fields themselves. No radiation escapes.
3. The charged particles are somewhat unevenly distributed in the orbit. This
happens for an open ring current, carried initially by evenly distributed
charged particles, which is subject to thermal fluctuations. From statistical mechanics we know that this happens for all
open systems and that the
particle densitiesexhibit fluctuations of order N. This means that out of
the N particles, N will exhibit deviation from perfect randomnessand
thereby perfect radiation field cancellationand
give rise to net radiation
112
vt
q0
v = v x 1
|x x0 |
B
E x 3
F IGURE 6.12:
(6.146)
Virtual photons
Let us consider a charge q0 moving with constant, high velocity v(t0 ) along the x1
axis. According to formula (6.194) on page 125 and figure 6.12, the perpendicular
component along the x3 axis of the electric field from this moving charge is
q0
v2
E = E3 =
1 2 (x x0 ) x 3
(6.147)
40 s3
c
113
Utilising expression (6.103) on page 103 and simple geometrical relations, we can
rewrite this as
b
q0
(6.148)
E =
40 2 (vt)2 + b2 /2 3/2
This represents a contracted Coulomb field, approaching the field of a plane wave.
The passage of this field pulse corresponds to a frequency distribution of the field
energy. Fourier transforming, we obtain
Z
1
q0
b
b
it
E, =
dt E (t) e = 2
K1
(6.149)
2
4 0 bv
v
v
Here, K1 is the Kelvin function (Bessel function of the second kind with imaginary
argument) which behaves in such a way for small and large arguments that
E,
q0
42 0 bv
b v
b v
E, 0,
b
1
v
b
1
v
(6.150a)
(6.150b)
Z
V
d3x E2 = 0
bmax
db 2b
bmin
dt vE2
(6.151)
where the volume integration is over the plane perpendicular to v. With the use
of Parsevals identity for Fourier transforms, formula (5.34) on page 75, we can
rewrite this as
U =
Z
0
d U = 40 v
q02
2
2 0 v
bmax
db 2b
bmin
v/ db
bmin
d E,
0
(6.152)
ln
U 2
2 0 v
bmin
(6.153)
bmin
114
p02
p2
p1
F IGURE 6.13:
p01
d
2 c
ln
p1 p01
(6.155)
m0 c2 E1 E10
115
(6.157)
B = (t, x)H = m 0 H
(6.158)
Expressed in terms of these derived field quantities, the Maxwell equations, often
called macroscopic Maxwell equations, take the form
D = (t, x)
B
E=
t
B=0
D
H=
+ j(t, x)
t
(6.159a)
(6.159b)
(6.159c)
(6.159d)
Assuming for simplicity that the electric permittivity and the magnetic permeability , and hence the relative permittivity e and the relative permeability m
all have fixed values, independent on time and space, for each type of material we
consider, we can derive the general telegraphers equation [cf. equation (2.34) on
page 31]
2 E
E
2 E
=0
2
t
t2
(6.160)
=0
2
t2
(6.161)
e 0 m 0
e m
(6.162)
116
i = 1, 2, 3
(6.163)
c
def
= e m = c n
v
(6.164)
(6.165)
(6.166)
where now k is the wave vector in the medium given by equation (6.165). With
these definitions, the vacuum formula for the associated magnetic field, equation (2.41) on page 31,
B=
k E =
1
1
kE= kE
v
(6.167)
c
=
n
k
(6.168)
(6.169)
117
has a unique value for each frequency component, and is different from v . Except
in regions of anomalous dispersion, vg is always smaller than c. In a gas of free
charges, such as a plasma, the refractive index is given by the expression
n2 () = 1
2p
2
(6.170)
where
2p =
N q2
0 m
(6.171)
is the square of the plasma frequency p . Here m and N denote the mass and
number density, respectively, of charged particle species . In an inhomogeneous
plasma, N = N (x) so that the refractive index and also the phase and group
velocities are space dependent. As can be easily seen, for each given frequency,
the phase and group velocities in a plasma are different from each other. If the
frequency is such that it coincides with p at some point in the medium, then at
that point v while vg 0 and the wave Fourier component at is reflected
there.
Vavilov-Cerenkov
radiation
As we saw in subsection 6.1, a charge in uniform, rectilinear motion in vacuum
does not give rise to any radiation; see in particular equation (6.192a) on page 124.
Let us now consider a charge in uniform, rectilinear motion in a medium with electric properties which are different from those of a (classical) vacuum. Specifically,
consider a medium where
= Const > 0
(6.172a)
= 0
(6.172b)
c
1
=
<c
n
0
(6.173)
Hence, in this particular medium, the speed of propagation of (the phase planes of)
electromagnetic waves is less than the speed of light in vacuum, which we know
is an absolute limit for the motion of anything, including particles. A medium
of this kind has the interesting property that particles, entering into the medium
at high speeds |v|, which, of course, are below the phase speed in vacuum, can
experience that the particle speeds are higher than the phase speed in the medium.
118
If we recall the general derivation, in the vacuum case, of the retarded (and advanced) potentials in chapter 3 and the Linard-Wiechert potentials, equations (6.71)
on page 95, we realise that we obtain the latter in the medium by a simple formal replacement c c/n in the expression (6.72) on page 95 for s. Hence, the
Linard-Wiechert potentials in a medium characterized by a refractive index n,
are
1 q0
q0
1
(t, x) =
=
(6.174a)
0
40 |x x0 | n (xxc )v 40 s
1
q0 v
1
q0 v
A(t, x) =
=
(6.174b)
0
40 c2 |x x0 | n (xxc )v 40 c2 s
where now
(x x0 ) v
s = x x0 n
c
(6.175)
The need for the absolute value of the expression for s is obvious in the case
when v/c 1/n because then the second term can be larger than the first term;
if v/c 1/n we recover the well-known vacuum case but with modified phase
speed. We also note that the retarded and advanced times in the medium are [cf.
equation (3.32) on page 46]
k |x x0 |
|x x0 | n
0
0
=t
(6.176a)
tret
= tret
(t, x x0 ) = t
c
k |x x0 |
|x x0 | n
0
0
=t+
(6.176b)
tadv
= tadv
(t, x x0 ) = t +
c
so that the usual time interval t t0 between the time measured at the point of
observation and the retarded time in a medium becomes
|x x0 | n
t t0 =
(6.177)
c
For v/c 1/n, the retarded distance s, and therefore the denominators in
equations (6.174) above, vanish when
nv
v
n(x x0 ) = x x0 cos c = x x0
(6.178)
c
c
or, equivalently, when
c
cos c =
(6.179)
nv
In the direction defined by this angle c , the potentials become singular. During
the time interval t t0 given by expression (6.177), the field exists within a sphere
of radius |x x0 | around the particle while the particle moves a distance
l0 = (t t0 )v
(6.180)
119
x(t)
q0
x0 (t0 )
F IGURE 6.14: Instantaneous picture of the expanding field spheres from a point
charge moving with constant speed v/c > 1/n in a medium where n > 1. This
generates a Vavilov-Cerenkov
shock wave in the form of a cone.
c
nv
(6.181)
The Vavilov-Cerenkov
cone is similar in nature to the Mach cone in acoustics.
1 The first systematic exploration of this radiation was made by P. A. Cerenkov
surprised that his old 1904 ideas were now becoming interesting. Tamm, Frank and Cerenkov
received
the Nobel Prize in 1958 for the discovery and the interpretation of the Cerenkov
effect [V. L. Ginzburg,
private communication].
The first observation of this type of radiation was reported by Marie Curie in 1910, but she never pursued
the exploration of it [8].
120
(6.182)
Z
0
Urad () sin d
q0 2 n2 sin2 c
22 0 c3
= dcos = c = 2
X
sin2 1 + nv
c
v
d
nv 2
1+ c v
Urad () d
(6.187)
121
sin c
2
n2 v2 n x2
1
1 + nv
c
v
(6.188)
cX
c2
=
1 2 2
n
nv
leading to the final approximate result for the total energy loss in the frequency
interval (, + d)
q0 2 X
c2
rad
U d =
1 2 2 d
(6.189)
20 c2
nv
As mentioned earlier, the refractive index is usually frequency dependent. Realising this, we find that the radiation energy per frequency unit and per unit length
is
U rad d
q0 2
c2
=
1 2
d
(6.190)
2X
40 c2
n ()v2
This result was derived under the assumption that v/c > 1/n(), i.e., under the
condition that the expression inside the parentheses in the right hand side is positive. For all media it is true that n() 1 when , so there exist always
6.4 Bibliography
[1]
H. A LFVN
AND
(1950), p. 616.
122
[2]
[3]
M. B ORN AND E. W OLF, Principles of Optics. Electromagnetic Theory of Propagation, Interference and Diffraction of Light, sixth ed., Pergamon Press, Oxford,. . . , 1980,
ISBN 0-08-026481-6.
Bibliography
[4]
V. L. G INZBURG, Applications of Electrodynamics in Theoretical Physics and Astrophysics, Revised third ed., Gordon and Breach Science Publishers, New York, London,
Paris, Montreux, Tokyo and Melbourne, 1989, ISBN 2-88124-719-9.
[5]
J. D. JACKSON, Classical Electrodynamics, third ed., John Wiley & Sons, Inc.,
New York, NY . . . , 1999, ISBN 0-471-30932-X.
[6]
[7]
[8]
J. S CHWINGER , L. L. D E R AAD , J R ., K. A. M ILTON , AND W. T SAI, Classical Electrodynamics, Perseus Books, Reading, MA, 1998, ISBN 0-7382-0056-5.
[9]
[10] J. VANDERLINDE, Classical Electromagnetic Theory, John Wiley & Sons, Inc., New
York, Chichester, Brisbane, Toronto, and Singapore, 1993, ISBN 0-471-57269-1.
123
6.5 Examples
E XAMPLE 6.1
v
t
(6.191)
Hence, the E and B fields can be obtained from formulae (6.73) on page 96, with the potentials
given by equations (6.71) on page 95 as follows:
A
1 v
v
= 2
= 2
t
c
t
c
t
v v
vv
= +
= 1 2
c
vv c c
=
1
c2
v
v
v
B = A = 2 = 2 = 2
c v
i vc vvc
v h v
= 2
= 2 2 1
c
c
c
c
c
v
= 2 E
c
E =
(6.192a)
(6.192b)
Here 1 = x i x i is the unit dyad and we used the fact that v v 0. What remains is just to
express in quantities evaluated at t and x.
From equation (6.71a) on page 95 and equation (6.104) on page 103 we find that
1
q0
q0
=
s2
40
s
80 s3
i
q0 h
v v
=
(x
x
)
+
(x
x
)
0
0
40 s3
c
c
(6.193)
When this expression for is inserted into equation (6.192a), the following result
124
Examples
q0 vv
1 =
1 s2
3
2
80 s
c
q0
v v
=
(x
x
)
+
(x
x
)
0
0
40 s3
c
c
i
v v
vv h v v
(x x0 ) 2
(x x0 )
c c
c
c
c
v v
v2
q0
=
(x x0 ) +
(x x0 ) (x x0 ) 2
3
40 s
c c
c
v v
(x x0 )
c c
q0
v2
=
(x x0 ) 1 2
40 s3
c
E(t, x) =
vv
c2
(6.194)
follows. Of course, the same result also follows from equation (6.90) on page 100 with v 0
inserted.
From equation (6.194) we conclude that E is directed along the vector from the simultaneous coordinate x0 (t) to the field (observation) coordinate x(t). In a similar way, the magnetic
field can be calculated and one finds that
v2
1
0 q0
1
v (x x0 ) = 2 v E
(6.195)
B(t, x) =
3
2
4s
c
c
From these explicit formulae for the E and B fields and formula (6.103b) on page 103 for s, we
can discern the following cases:
1. v 0 E goes over into the Coulomb field ECoulomb
2. v 0 B goes over into the Biot-Savart field
3. v c E becomes dependent on 0
4. v c, sin 0 0 E (1 v2 /c2 )ECoulomb
5. v c, sin 0 1 E (1 v2 /c2 )1/2 ECoulomb
C E ND OF EXAMPLE 6.1
E XAMPLE 6.2
Let us consider in more detail the treatment of the radiation from a uniformly moving rigid
charge distribution.
If we return to the original definition of the potentials and the inhomogeneous wave equation, formula (3.17) on page 43, for a generic potential component (t, x) and a generic source
component f (t, x),
1 2
2
2 (t, x) =
(t, x) = f (t, x)
(6.196)
c2 t2
125
we find that under the assumption that v = v x 1 , this equation can be written
v2 2 2 2
+ 2 + 2 = f (x)
1 2
c
x12
x2
x3
(6.197)
(6.198b)
3 = x3
(6.198c)
(6.198a)
and introducing the vectorial nabla operator in space, def (/1 , /2 , /3 ), the timeindependent equation (6.197) reduces to an ordinary Poisson equation
p
2 () = f ( 1 v2 /c2 1 , 2 , 3 ) f ()
(6.199)
in this space. This equation has the well-known Coulomb potential solution
() =
1
4
Z
V
f (0 ) 3 0
d
| 0 |
(6.200)
1
4
Z
V
f (x0 ) 3 0
dx
s
(6.201)
(6.202)
Applying this to the explicit scalar and vector potential components, realising that for a rigid
charge distribution moving with velocity v the current is given by j = v, we obtain
1
(x0 ) 3 0
dx
40 V s
Z
1
v(x0 ) 3 0
v
A(t, x) =
d x = 2 (t, x)
2
40 c V s
c
(t, x) =
(6.203a)
(6.203b)
q0
40 s
q0 v
A(t, x) =
40 c2 s
(t, x) =
(6.204a)
(6.204b)
which we recognise as the Linard-Wiechert potentials; cf. equations (6.71) on page 95. We
notice, however, that the derivation here, based on a mathematical technique which in fact is a
Lorentz transformation, is of more general validity than the one leading to equations (6.71) on
page 95.
Let us now consider the action of the fields produced from a moving, rigid charge distribution represented by q0 moving with velocity v, on a charged particle q, also moving with
velocity v. This force is given by the Lorentz force
126
Examples
F = q(E + v B)
(6.205)
With the help of equation (6.195) on page 125 and equations (6.203) on page 126, and the
fact that t = v [cf. formula (6.191) on page 124], we can rewrite expression (6.205) as
i
h v
h
v
v
i
v v
F=q E+v 2 E =q
(6.206)
c
c
c
c
c
Applying the bac-cab rule, formula (F.51) on page 176, on the last term yields
v
v v2
v v
=
c
c
c
c c2
(6.207)
(6.208)
where
=
1
v2
c2
(6.209)
The scalar function is called the convection potential or the Heaviside potential. When the
rigid charge distribution is well localised so that we can use the potentials (6.204) the convection
potential becomes
v2
q0
= 1 2
(6.210)
c
40 s
The convection potential from a point charge is constant on flattened ellipsoids of revolution,
defined through equation (6.202) on page 126 as
2
x1 x10
+ (x2 x20 )2 + (x3 x30 )2
1 v2 /c2
= 2 (x1 x10 )2 + (x2 x20 )2 + (x3 x30 )2 = Const
(6.211)
These Heaviside ellipsoids are equipotential surfaces, and since the force is proportional to the
gradient of , which means that it is perpendicular to the ellipsoid surface, the force between
two charges is in general not directed along the line which connects the charges. A consequence
of this is that a system consisting of two co-moving charges connected with a rigid bar, will
experience a torque. This is the idea behind the Trouton-Noble experiment, aimed at measuring
the absolute speed of the earth or the galaxy. The negative outcome of this experiment is
explained by the special theory of relativity which postulates that mechanical laws follow the
same rules as electromagnetic laws, so that a compensating torque appears due to mechanical
stresses within the charge-bar system.
C E ND OF EXAMPLE 6.2
127
E XAMPLE 6.3
(6.212)
dt0 v
(6.213)
(6.214)
(6.215)
B=
(6.216)
E
c
(6.217)
q0 sin 0
v eit0
82 0 c2 |x x0 |
(6.218)
We note that the magnitude of this Fourier component is independent of . This is a consequence of the infinitely short impulsive step (t0 t0 ) in the time domain which produces an
infinite spectrum in the frequency domain.
The total radiation energy is given by the expression
Z I
Z
B
dU rad
0
2 0 0
=
dt
d
x
n
U rad =
dt0
dt0
0
S0
I
Z
I
Z
1
1
=
d2x0
dt0 EB =
d2x0
dt0 E 2
0 S 0
0 c S 0
= 0 c
I
S0
d2x0
(6.219)
dt0 E 2
According to Parsevals identity [cf. equation (5.34) on page 75] the following equality
holds:
128
Examples
dt0 E 2 = 4
d |E |2
(6.220)
(6.221)
I
S0
d2x0
sin2 0
d
|x x0 |2
q0 2 (v)2 2 0 0
d
d sin 0 sin2 0 d
163 0 c3 0
0
2
v
d
q0 2
=
30 c c
2
(6.222)
We see that the energy spectrum U rad is independent of frequency . This means that if we
would integrate it over all frequencies [0, ), a divergent integral would result.
In reality, all spectra have finite widths, with an upper cutoff limit set by the quantum
condition
~max =
1
1
m(v + v)2 mv2
2
2
(6.223)
which expresses that the highest possible frequency max in the spectrum is that for which all
kinetic energy difference has gone into one single field quantum (photon) with energy ~max .
If we adopt the picture that the total energy is quantised in terms of N photons radiated during
the process, we find that
U rad d
= dN
~
(6.224)
40 ~c 3 c
137 3 c
(6.225)
where we used the value of the fine structure constant = e2 /(40 ~c) 1/137.
Even if the number of photons becomes infinite when 0, these photons have negligible
energies so that the total radiated energy is still finite.
C E ND OF EXAMPLE 6.3
129
7
Relativistic
Electrodynamics
We saw in chapter 3 how the derivation of the electrodynamic potentials led, in
a most natural way, to the introduction of a characteristic, finite speed of propa
gation in vacuum that equals the speed of light c = 1/ 0 0 and which can be
considered as a constant of nature. To take this finite speed of propagation of
information into account, and to ensure that our laws of physics be independent
of any specific coordinate frame, requires a treatment of electrodynamics in a relativistically covariant (coordinate independent) form. This is the object of this
chapter.
Special Theory of Relativity, by the American physicist and philosopher David Bohm, opens
with the following paragraph [4]:
The theory of relativity is not merely a scientific development of great importance in its
own right. It is even more significant as the first stage of a radical change in our basic
concepts, which began in physics, and which is spreading into other fields of science,
and indeed, even into a great deal of thinking outside of science. For as is well known,
the modern trend is away from the notion of sure absolute truth, (i.e., one which holds
independently of all conditions, contexts, degrees, and types of approximation etc..) and
toward the idea that a given concept has significance only in relation to suitable broader
131
7. Relativistic Electrodynamics
inertial systems in uniform, rectilinear motion relative to each other and is based
on two postulates:
Postulate 7.1 (Relativity principle; Poincar, 1905). All laws of physics (except
the laws of gravitation) are independent of the uniform translational motion of the
system on which they operate.
Postulate 7.2 (Einstein, 1905). The velocity of light in empty space is independent
of the motion of the source that emits the light.
A consequence of the first postulate is that all geometrical objects (vectors,
tensors) in an equation describing a physical process must transform in a covariant
manner, i.e., in the same way.
v
c
(7.1)
1
= p
(7.2)
1 2
where v = |v|. In the following, we shall make frequent use of these shorthand
notations.
As shown by Einstein, the two postulates of special relativity require that the
spatial coordinates and times as measured by an observer in and 0 , respectively,
are connected by the following transformation:
ct0 = (ct x)
(7.3a)
x = (x vt)
(7.3b)
y =y
(7.3c)
z =z
(7.3d)
0
0
forms of reference, within which that concept can be given its full meaning.
132
vt
y
y0
v
P(t, x, y, z)
P(t0 , x0 , y0 , z0 )
O
z
x0
O0
x
z0
F IGURE 7.1:
Taking the difference between the square of (7.3a) and the square of (7.3b) we
find that
c2 t02 x02 = 2 c2 t2 2xct + x2 2 x2 + 2xvt v2 t2
v2
v2
1
2 2
2
c
t
1
x
1
=
c2
c2
v2
1 2
c
= c2 t 2 x 2
(7.4)
From equations (7.3) on page 132 we see that the y and z coordinates are unaffected by the translational motion of the inertial system 0 along the x axis of system . Using this fact, we find that we can generalise the result in equation (7.4)
to
c2 t2 x2 y2 z2 = c2 t02 x02 y02 z02
(7.5)
which means that if a light wave is transmitted from the coinciding origins O and
O0 at time t = t0 = 0 it will arrive at an observer at (x, y, z) at time t in and an
observer at (x0 , y0 , z0 ) at time t0 in 0 in such a way that both observers conclude
that the speed (spatial distance divided by time) of light in vacuum is c. Hence, the
speed of light in and 0 is the same. A linear coordinate transformation which
has this property is called a (homogeneous) Lorentz transformation.
133
7. Relativistic Electrodynamics
(7.6)
We want to interpret this quadruple x as (the component form of) a radius fourvector in a real, linear, four-dimensional vector space.2 We require that this fourdimensional space be a Riemannian space, i.e., a metric space where a distance
and a scalar product are defined. In this space we therefore define a metric tensor,
also known as the fundamental tensor, which we denote by g .
(7.7)
where the upper index in x is summed over and is therefore a dummy index
and may be replaced by another dummy index This summation process is an
example of index contraction and is often referred to as index lowering.
(7.8)
I regret that it has been necessary for me in this lecture to administer a large dose of
four-dimensional geometry. I do not apologise, because I am really not responsible for the
fact that nature in its most fundamental aspect is four-dimensional. Things are what they
are. . . .
134
to perceive it as the manifestation of the conservation of the norm in a 4D Riemannian space. Then the explicit expression for the scalar product of x with itself in
this space must be
x x = c2 t2 x2 y2 z2
(7.9)
We notice that our space will have an indefinite norm which means that we deal
with a non-Euclidean space. We call the four-dimensional space (or space-time)
with this property Lorentz space and denote it L4 . A corresponding real, linear 4D
space with a positive definite norm which is conserved during ordinary rotations
is a Euclidean vector space. We denote such a space R4 .
Metric tensor
By choosing the metric tensor in L4 as
if = = 0
1
g = 1 if = = i = j = 1, 2, 3
0
if ,
or, in matrix notation,
1 0
0
0
0 1 0
0
(g ) =
0 0 1 0
0 0
0 1
(7.10)
(7.11)
i.e., a matrix with a main diagonal that has the sign sequence, or signature,
{+, , , }, the index lowering operation in our chosen flat 4D space becomes
nearly trivial:
x = g x = (ct, x)
(7.12)
=
x2 0 0 1 0 x2 = x2
x3
x3
0 0
0 1
x3
(7.13)
Hence, if the metric tensor is defined according to expression (7.10) the covariant radius four-vector x is obtained from the contravariant radius four-vector x
simply by changing the sign of the last three components. These components are
135
7. Relativistic Electrodynamics
(7.14)
which indeed is the desired Lorentz transformation invariance as required by equation (7.9) on page 135. Without changing the physics, one can alternatively choose
a signature {, +, +, +}. The latter has the advantage that the transition from 3D
to 4D becomes smooth, while it will introduce some annoying minus signs in the
theory. In current physics literature, the signature {+, , , } seems to be the most
commonly used one.
The L4 metric tensor equation (7.10) on page 135 has anumber of interesting
properties: firstly, we see that this tensor has a trace Tr g = 2 whereas in R4 ,
as in any vector space with definite norm, the trace equals the space dimensionality. Secondly, we find, after trivial algebra, that the following relations between
the contravariant, covariant and mixed forms of the metric tensor hold:
g = g
g g
(7.15a)
= g
=
g g =
g
g
(7.15b)
=
=
(7.15c)
(7.15d)
Here we have introduced the 4D version of the Kronecker delta , a mixed fourtensor of rank 2 which fulfils
(
1 if =
(7.16)
= =
0 if ,
(7.17)
where the metric tensor is as in equation (7.10) on page 135. As we see, this
form is indefinite as expected for a non-Euclidean space. The square root of this
136
(7.18)
where we introduced
d = dt/
(7.19)
Since d measures the time when no spatial changes are present, it is called the
proper time.
Expressing the property of the Lorentz transformation described by equations (7.5) on page 133 in terms of the differential interval ds and comparing
with equation (7.17) on page 136, we find that
ds2 = c2 dt2 dx2 dy2 dz2
(7.20)
(7.21)
(7.22)
(7.23)
is a light-like interval; we may also say that in this case we are on the light cone.
A vector which has a light-like interval is called a null vector. The time-like,
space-like or light-like aspects of an interval ds are invariant under a Lorentz
transformation. I.e., it is not possible to change a time-like interval into a spacelike one or vice versa via a Lorentz transformation.
137
7. Relativistic Electrodynamics
Four-vector fields
Any quantity which relative to any coordinate system has a quadruple of real
numbers and transforms in the same way as the radius four-vector x does, is
called a four-vector. In analogy with the notation for the radius four-vector we
introduce the notation a = (a0 , a) for a general contravariant four-vector field in
L4 and find that the lowering of index rule, formula (7.7) on page 134, for such
an arbitrary four-vector yields the dual covariant four-vector field
a (x ) = g a (x ) = (a0 (x ), a(x ))
(7.24)
The scalar product between this four-vector field and another one b (x ) is
g a (x )b (x ) = (a0 , a) (b0 , b) = a0 b0 a b
(7.25)
0 0
0 0
=
0
0
1 0
0
0
0 1
(7.26)
the linear Lorentz transformation (7.3) on page 132, i.e., the coordinate transformation x x0 = x0 (x0 , x1 , x2 , x3 ), from one inertial system to another inertial
system 0 in the standard configuration, can be written
x0 = x
(7.27)
1 + 2
1 + 1 2
(7.28)
This means that the nonempty set of Lorentz transformations constitutes a closed
algebraic structure with a binary operation which is associative. Furthermore,
one can show that this set possesses at least one identity element and at least one
inverse element. In other words, this set of Lorentz transformations constitutes
a mathematical group. However tempting, we shall not make any further use of
group theory.
138
X0
X
00
x01
x1
F IGURE 7.2: Minkowski space can be considered an ordinary Euclidean space
where a Lorentz transformation from (x1 , X 0 = ict) to (x01 , X 00 = ict0 ) corresponds
to an ordinary rotation through an angle . This rotation leaves the Euclidean
2
2
distance x1 + X 0 = x2 c2 t2 invariant.
(7.29a)
X =x
(7.29b)
X =x
(7.29c)
X =x
(7.29d)
dS = ids
(7.29e)
(7.30)
fact that our Riemannian space can be transformed in this way into a Euclidean one means that
it is, strictly speaking, a pseudo-Riemannian space.
139
7. Relativistic Electrodynamics
x = ct
0
x00
x0 = x1
P0
O=O
x01
ct
x1 = x
140
As before, it suffices to consider the simplified case where the relative motion
between and 0 is along the x axes. Then
dS 2 = (dX 0 )2 + (dX 1 )2 = (dX 0 )2 + (dx1 )2
(7.31)
(7.32a)
(7.32b)
(7.33a)
x = x cosh ct sinh
(7.33b)
which are identical to the transformation equations (7.3) on page 132 if we let
sinh =
(7.34a)
cosh =
(7.34b)
tanh =
(7.34c)
tanh 1 + tanh 2
1 + tanh 1 tanh 2
(7.35)
The use of ict and M4 , which leads to the interpretation of the Lorentz transformation as an ordinary rotation, may, at best, be illustrative, but is not very
physical. Besides, if we leave the flat L4 space and enter the curved space of
general relativity, the ict trick will turn out to be an impasse. Let us therefore
immediately return to L4 where all components are real valued.
141
7. Relativistic Electrodynamics
dx
v
c
= (u0 , u)
u =
= (c, v) = q
, q
(7.36)
d
v2
v2
1
1
c2
c2
which, when multiplied with the scalar invariant m0 yields the four-momentum
dx
m0 c
m0 v
p = m0
= m0 (c, v) = q
, q
= (p0 , p)
(7.37)
2
d
v
v2
1
1
c2
c2
(7.38)
where
m0
m = m0 = q
1
(7.39)
v2
c2
We can interpret this such that the Lorentz covariance implies that the mass-like
term in the ordinary 3D linear momentum is not invariant. A better way to look
at this is that p = mv = m0 v is the covariantly correct expression for the kinetic
three-momentum.
Multiplying the zeroth (time) component of the four-momentum p with the
scalar invariant c, we obtain
m0 c 2
= mc2
cp0 = m0 c2 = q
v2
1 c2
(7.40)
Since this component has the dimension of energy and is the result of a covariant
description of the motion of a particle with its kinetic momentum described by
the spatial components of the four-momentum, equation (7.37), we interpret cp0
as the total energy E. Hence,
cp = (cp0 , cp) = (E, cp)
142
(7.41)
(7.42)
Since this is an invariant, this equation holds in any inertial frame, particularly in
the frame where p = 0 and there we have
E = m0 c 2
(7.43)
(7.44)
= 0
(7.45)
2 = = =
A =
,A
c
(7.47)
143
7. Relativistic Electrodynamics
where is the scalar potential and A the vector potential, defined in section 3.3
on page 40, we can write the uncoupled inhomogeneous wave equations, equations (3.16) on page 43, in the following compact (and covariant) way:
2 A = 0 j
(7.48)
With the help of the above, we can formulate our electrodynamic equations
covariantly. For instance, the covariant form of the equation of continuity, equation (1.23) on page 10 is
j = 0
(7.49)
and the Lorenz-Lorentz gauge condition, equation (3.15) on page 43, can be written
A = 0
(7.50)
(7.51)
If only one dimension Lorentz contracts (for instance, due to relative motion
along the x direction), a 3D spatial volume element transforms according to
r
p
1
v2
3
2
dV = d x = dV0 = dV0 1 = dV0 1 2
(7.52)
c
where dV0 denotes the volume element as measured in the rest system, then from
equation (7.45) on page 143 we see that
dV = 0 dV0
(7.53)
i.e., the charge in a given volume is conserved. We can therefore conclude that
the elementary charge is a universal constant.
(A )0 =
,A
=
,0
(7.54)
c
40 c |x x0 |0
v=0
144
where |x x0 |0 is the usual distance from the source point to the field point, evaluated in the rest system (signified by the index 0).
Let us introduce the relative radius four-vector between the source point and
the field point:
R = x x0 = (c(t t0 ), x x0 )
(7.55)
(7.58)
(7.59)
Now we want to find the correspondence to the rest system solution, equation (7.54) on page 144, in an arbitrary inertial system. We note from equation (7.36) on page 142 that in the rest system
c
v
= (c, 0)
(u )0 = q
(7.60)
, q
v2
v2
1 c2
1 c2
v=0
and
(R )0 = (x x0 , x x0 )0 = (x x0 0 , (x x0 )0 )
(7.61)
145
7. Relativistic Electrodynamics
u
q0
40 cu R
(7.63)
= v
v
c
(7.65)
and introducing
v (x x0 )
def
s x x0
x x0 (x x0 )
c
(7.66)
we can write
u R = cs
(7.67)
and
u
=
cu R
1 v
,
cs c2 s
(7.68)
q0
,
,A
=
A (x ) =
40 cs c2 s
c
(7.69)
where in the last step the definition of the four-potential, equation (7.47) on page 143,
was used. Writing the solution in the ordinary 3D way, we conclude that for a very
localised charge volume, moving relative an observer with a velocity v, the scalar
and vector potentials are given by the expressions
q0 1
q0
1
=
0
40 s 40 |x x | (x x0 )
q0 v
q0
v
A(t, x) =
=
2
2
0
40 c s 40 c |x x | (x x0 )
(t, x) =
(7.70a)
(7.70b)
146
(7.71)
i, j , k
(7.72)
B
t
(7.73)
(7.74)
We know from chapter 3 that the fields can be derived from the electromagnetic potentials in the following way:
B=A
(7.75a)
A
E =
t
(7.75b)
= i A j j Ai
xi x j
Ai
Ei = i
= i t Ai
x
t
Bi j =
(7.76a)
(7.76b)
From this, we notice the clear difference between the axial vector (pseudovector)
B and the polar vector (ordinary vector) E.
Our goal is to express the electric and magnetic fields in a tensor form where
the components are functions of the covariant form of the four-potential, equation (7.47) on page 143:
A =
,A
(7.77)
c
147
7. Relativistic Electrodynamics
Inspection of (7.77) and equation (7.76) on page 147 makes it natural to define
the four-tensor
F =
A A
= A A
x x
(7.78)
This anti-symmetric (skew-symmetric), four-tensor of rank 2 is called the electromagnetic field tensor. In matrix representation, the contravariant field tensor can
be written
0
E x /c Ey /c Ez /c
E x /c
0
Bz
By
(F ) =
(7.79)
Ey /c
Bz
0
Bx
Ez /c By
Bx
0
We note that the field tensor is a sort of four-dimensional curl of the four-potential
vector A .
The covariant field tensor is obtained from the contravariant field tensor in the
usual manner by index lowering
F = g g F = A A
which in matrix representation becomes
0
E x /c Ey /c
E x /c
0
Bz
F =
Ey /c Bz
0
Ez /c By
Bx
(7.80)
Ez /c
By
Bx
0
(7.81)
Comparing formula (7.81) above with formula (7.79) we see that the covariant
field tensor is obtained from the contravariant one by a transformation E E.
That the two Maxwell source equations can be written
F = 0 j
(7.82)
148
E=
0
(7.84)
which we recognise at the Maxwell source equation for the electric field, equation (1.45a) on page 15.
For = 1 (the second column in equation (7.79) on page 148), equation (7.82)
on page 148 yields
1 E x
Bz By
F 01 F 11 F 21 F 31
+
+
+
= 2
+0+
= 0 j1 = 0 v x
0
1
2
3
x
x
x
x
c t
y
z
(7.85)
This result can be rewritten as
Bz By
E x
0 0
= 0 j x
y
z
t
(7.86)
or, equivalently, as
E x
(7.87)
t
and similarly for = 2, 3. In summary, we can write the result in three-vector
form as
E
B = 0 j(t, x) + 0 0
(7.88)
t
which we recognise as the Maxwell source equation for the magnetic field, equation (1.45d) on page 15.
With the help of the fully antisymmetric rank-4 pseudotensor
= F
(7.90)
0 cBx cBy
cB
0
Ez
x
?
F
=
cBy Ez
0
cBz
Ey
E x
cBz
Ey
Ex
0
(7.91)
149
7. Relativistic Electrodynamics
i.e., the dual field tensor is obtained from the ordinary field tensor by the duality
transformation E c2 B and B E.
The covariant form of the two Maxwell field equations
E=
B
t
(7.92)
B=0
(7.93)
(7.94)
(7.95)
sometimes referred to as the Jacobi identity. Hence, equation (7.82) on page 148
and equation (7.95) constitute Maxwells equations in four-dimensional formalism.
It is interesting to note that equation (7.82) on page 148 and
?F = 0 jm
(7.96)
7.4 Bibliography
150
[1]
J. A HARONI, The Special Theory of Relativity, second, revised ed., Dover Publications, Inc., New York, 1985, ISBN 0-486-64870-2.
[2]
A. O. BARUT, Electrodynamics and Classical Theory of Fields and Particles, Dover Publications, Inc., New York, NY, 1980, ISBN 0-486-64038-8.
[3]
[4]
D. B OHM, The Special Theory of Relativity, Routledge, New York, NY, 1996, ISBN 0415-14809-X.
[5]
Bibliography
[6]
[7]
F. E. L OW, Classical Field Theory, John Wiley & Sons, Inc., New York, NY . . . , 1997,
ISBN 0-471-59551-9.
[8]
H. M UIRHEAD, The Special Theory of Relativity, The Macmillan Press Ltd., London,
Beccles and Colchester, 1973, ISBN 333-12845-1.
[9]
C. M LLER, The Theory of Relativity, second ed., Oxford University Press, Glasgow . . . ,
1972.
[10] W. K. H. PANOFSKY AND M. P HILLIPS, Classical Electricity and Magnetism, second ed.,
Addison-Wesley Publishing Company, Inc., Reading, MA . . . , 1962, ISBN 0-201-057026.
[11] J. J. S AKURAI, Advanced Quantum Mechanics, Addison-Wesley Publishing Company, Inc., Reading, MA . . . , 1967, ISBN 0-201-06710-2.
[12] B. S PAIN, Tensor Calculus, third ed., Oliver and Boyd, Ltd., Edinburgh and London,
1965, ISBN 05-001331-9.
[13] A. N. W HITEHEAD, Concept of Nature, Cambridge University Press, Cambridge . . . ,
1920, ISBN 0-521-09245-0.
151
8
Electromagnetic
Fields and
Particles
In previous chapters, we calculated the electromagnetic fields and potentials from
arbitrary, but prescribed distributions of charges and currents. In this chapter we
study the general problem of interaction between electric and magnetic fields and
electrically charged particles. The analysis is based on Lagrangian and Hamiltonian methods, is fully covariant, and yields results which are relativistically
correct.
153
Lagrange formalism
Let us now introduce a generalised action
S4 =
L4 (x , u ) d
(8.1)
where d is the proper time defined via equation (7.18) on page 137, and L4 acts
as a kind of generalisation to the common 3D Lagrangian so that the variational
principle
S 4 =
L4 (x , u ) d = 0
(8.2)
1
m0 u u
2
(8.3)
For an interaction with the electromagnetic field we can introduce the interaction
with the help of the four-potential given by equation (7.77) on page 147 in the
following way
L4 =
1
m0 u u + qu A (x )
2
(8.4)
We call this the four-Lagrangian and shall now show how this function, together
with the variation principle, formula (8.2), yields covariant results which are physically correct.
The variation principle (8.2) with the 4D Lagrangian (8.4) inserted, leads to
Z 1
m0
S 4 =
u u + qu A d
2
0
Z 1
m0 (u u )
(8.5)
=
u + q A u + u
x
d
2 u
x
0
Z 1
=
m0 u u + q A u + u A x d = 0
0
154
dx
d
(8.6)
which means that we can write the variation of u as a total derivative with respect
to :
d
dx
(x )
=
(8.7)
u =
d
d
Inserting this into the first two terms in the last integral in equation (8.5) on
page 154, we obtain
Z 1
d
d
S 4 =
m0 u (x ) + qA (x ) + qu A x d
(8.8)
d
d
0
Partial integration in the two first terms in the right hand member of (8.8) gives
Z 1
du
dA
S 4 =
m0
x q
x + qu A x d
(8.9)
d
d
0
where the integrated parts do not contribute since the variations at the endpoints
vanish. A change of irrelevant summation index from to in the first two terms
of the right hand member of (8.9) yields, after moving the ensuing common factor
x outside the parenthesis, the following expression:
Z 1
dA
du
(8.10)
q
+ qu A x d
S 4 =
m0
d
d
0
Applying well-known rules of differentiation and the expression (7.36) for the
four-velocity, we can express dA /d as follows:
dA A dx
=
= A u
d
x d
(8.11)
By inserting this expression (8.11) into the second term in right-hand member of
equation (8.10) above, and noting the common factor qu of the resulting term
and the last term, we obtain the final variational principle expression
Z 1
du
S 4 =
m0
+ qu A A x d
(8.12)
d
0
Since, according to the variational principle, this expression shall vanish and
x
is arbitrary between the fixed end points 0 and 1 , the expression inside
in
the integrand in the right hand member of equation (8.12) must vanish. In other
words, we have found an equation of motion for a charged particle in a prescribed
electromagnetic field:
m0
du
= qu A A
d
(8.13)
155
With the help of formula (7.80) on page 148 for the covariant component form of
the field tensor, we can express this equation in terms of the electromagnetic field
tensor in the following way:
m0
du
= qu F
d
(8.14)
This is the sought-for covariant equation of motion for a particle in an electromagnetic field. It is often referred to as the Minkowski equation. As the reader
can easily verify, the spatial part of this 4-vector equation is the covariant (relativistically correct) expression for the Newton-Lorentz force equation.
Hamiltonian formalism
The usual Hamilton equations for a 3D space are given by equation (M.55) on
page 189 in appendix M. These six first-order partial differential equations are
H dqi
=
pi
dt
H
dpi
=
qi
dt
(8.15a)
(8.15b)
where H(pi , qi , t) = pi q i L(qi , q i , t) is the ordinary 3D Hamiltonian, qi is a generalised coordinate and pi is its canonically conjugate momentum.
We seek a similar set of equations in 4D space. To this end we introduce a
canonically conjugate four-momentum p in an analogous way as the ordinary
3D conjugate momentum:
p =
L4
u
(8.16)
and utilise the four-velocity u , as given by equation (7.36) on page 142, to define
the four-Hamiltonian
H4 = p u L4
(8.17)
With the help of these, the radius four-vector x , considered as the generalised
four-coordinate, and the invariant line element ds, defined in equation (7.18) on
page 137, we introduce the following eight partial differential equations:
H4 dx
=
p
d
dp
H4
=
x
d
156
(8.18a)
(8.18b)
(8.19)
1
L4
=
m0 u u + qu A (x ) = m0 u + qA
(8.20)
p =
u u 2
Inserting this into (8.19), we obtain
1
1
H4 = m0 u u + qA u m0 u u qu A (x ) = m0 u u
2
2
(8.21)
m0 c 2
2
(8.22)
1
(p qA )
m0
(8.23)
(8.24)
157
That this four-Hamiltonian yields the correct covariant equation of motion can
be seen by inserting it into the four-dimensional Hamiltons equations (8.18) and
using the relation (8.23):
q
H4
A
= (p qA )
x
m0
x
q
A
= m0 u
m0
x
A
= qu
x
du
A
dp
= m0
q u
=
d
d
x
(8.25)
where in the last step equation (8.20) on page 157 was used. Rearranging terms,
and using equation (7.80) on page 148, we obtain
m0
du
= qu A A = qu F
d
(8.26)
which is identical to the covariant equation of motion equation (8.14) on page 156.
We can then safely conclude that the Hamiltonian in question is correct.
Recalling expression (7.47) on page 143 and representing the canonically conjugate four-momentum as p = (p0 , p), we obtain the following scalar products:
p p = (p0 )2 (p)2
1
A p = p0 (p A)
c
1
A A = 2 2 (A)2
c
(8.27a)
(8.27b)
(8.27c)
Inserting these explicit expressions into equation (8.24) on page 157, and using the
fact that for H4 is equal to the scalar value m0 c2 /2, as derived in equation (8.22)
on page 157, we obtain the equation
m0 c 2
1
2
q2 2
0 2
2
0
2
2
=
(p ) (p) qp + 2q(p A) + 2 q (A)
(8.28)
2
2m0
c
c
which is the second order algebraic equation in p0 :
(p0 )2
q2
2q 0 2
p (p) 2qp A + q2 (A)2 + 2 2 m20 c2 = 0
c
|
{z
} c
(8.29)
(pqA)2
158
(8.30)
(8.32)
Using the explicit expressions (equation (8.31) above) and (equation (8.32)), we
obtain the explicit expression for the ordinary 3D Lagrange function
q
L = p v q c (p qA)2 + m20 c2
(8.33)
and if we make the identification
m0 v
p qA = q
= mv
2
1 cv2
(8.34)
where the quantity mv is the usual kinetic momentum, we can rewrite this expression for the ordinary Lagrangian as follows:
q
2
L = qA v + mv q c m2 v2 + m20 c2
r
(8.35)
v2
2
2
2
= mv q( A v) mc = q + qA v m0 c
1 2
c
What we have obtained is the relativistically correct (covariant) expression for
the Lagrangian describing the motion of a charged particle in scalar and vector
potentials associated with prescribed electric and magnetic fields.
159
i1
m
m
k
a
i+1
m
k
a
m
k
a
m
k
a
derived the equations of motion for charged particles in given, prescribed fields.
Let us now put the fields and the particles on an equal footing and present a theoretical description which treats the fields, the particles, and their interactions in a
unified way. This involves transition to a field picture with an infinite number of
degrees of freedom. We shall first consider a simple mechanical problem whose
solution is well known. Then, drawing inferences from this model problem, we
apply a similar view on the electromagnetic problem.
L(i , i , t) dt = 0
(8.36)
160
1 N 2
mi k(i+1 i )2
2
i=1
(8.37)
Let us write the Lagrangian, as given by equation (8.37), in the following way:
N
L = aLi
(8.38)
2
1 m 2
i+1
i
Li =
ka
2 a i
a
(8.39)
i=1
Here,
=
a
dx
ka Y
i+1 i
a
x
(8.40a)
linear mass density
(8.40b)
Youngs modulus
(8.40c)
(8.40d)
we obtain
L=
L dx
(8.41)
where
"
2 #
2
1
L , , , t =
Y
t x
2
t
x
(8.42)
Notice how we made a transition from a discrete description, in which the mass
points were identified by a discrete integer variable i = 1, 2, . . . , N, to a continuous description, where the infinitesimal mass points were instead identified by a
continuous real parameter x, namely their position along x .
A consequence of this transition is that the number of degrees of freedom for
the system went from the finite number N to infinity! Another consequence is
that L has now become dependent also on the partial derivative with respect to
x of the field coordinate . But, as we shall see, the transition is well worth the
161
cost because it allows us to treat all fields, be it classical scalar or vectorial fields,
or wave functions, spinors and other fields that appear in quantum physics, on an
equal footing.
Under the assumption of time independence and fixed endpoints, the variation
principle (8.36) on page 160 yields:
Z
L dt
= L , ,
dx dt
t x
ZZ
dx dt
+
=
+
t
x
t
x
ZZ
(8.43)
=0
The last integral can be integrated by parts. This results in the expression
ZZ
L L L dx dt = 0
(8.44)
t
x
t
where the variation is arbitrary (and the endpoints fixed). This means that the
integrand itself must vanish. If we introduce the functional derivative
L
L
L
=
(8.45)
x
x
L
L
= 0
(8.46)
=0
(8.47)
t
x
Y t2 x2
i.e., the one-dimensional wave equation for compression waves which propagate
162
ZZ
L d3x dt
Z
= L , d4x
x
ZZ
L
L
d4x
=
L dt =
(8.48)
=0
where the variation is arbitrary and the endpoints are fixed. This means that
the integrand itself must vanish:
L
L
= 0
(8.49)
x
x
L
L
L
=
i
x
i
(8.50)
L
L
=0
(8.51)
L
H , , i ; t = L , , i
x
t
t x
(8.52)
(8.53)
163
If, as usual, we differentiate this expression and identify terms, we obtain the
following Hamilton density equations
H
=
t
H
(8.54a)
(8.54b)
The Hamilton density functions are in many ways similar to the ordinary Hamilton
functions and lead to similar results.
(8.55)
where the mechanical part has to do with the particle motion (kinetic energy). It
is given by L4 /V where L4 is given by equation (8.3) on page 154 and V is the
volume. Expressed in the rest mass density %0 , the mechanical Lagrange density
can be written
1
L mech = %0 u u
2
(8.56)
The L inter part describes the interaction between the charged particles and
the external electromagnetic field. A convenient expression for this interaction
Lagrange density is
L inter = j A
(8.57)
For the field part L field we choose the difference between magnetic and electric energy density (in analogy with the difference between kinetic and potential
energy in a mechanical field). Using the field tensor, we express this field Lagrange density as
L field =
164
1
F F
40
(8.58)
F F
( A )
40
( A )
( A A )( A A )
=
40
( A )
1
=
A A A A
(8.62)
40
( A )
A A + A A
1
A A A A
=
20
( A )
But
A A = A
A + A
A
( A )
( A )
( A )
= A
A + A
g g A
( A )
( A )
= A
A + g g A
A
( A )
( A )
= A
A + A
A
( A )
( A )
= 2 A
(8.63)
165
Similarly,
A A = 2 A
( A )
(8.64)
so that
L EM
1
1
= ( A A ) = F
( A )
0
0
(8.65)
This means that the Euler-Lagrange equations, expression (8.49) on page 163,
for the Lagrangian density L EM and with A as the field quantity become
L EM
1
L EM
= j F = 0
(8.66)
A
( A )
0
or
F = 0 j
(8.67)
Other fields
In general, the dynamic equations for most any fields, and not only electromagnetic ones, can be derived from a Lagrangian density together with a variational
principle (the Euler-Lagrange equations). Both linear and non-linear fields are
studied with this technique. As a simple example, consider a real, scalar field
which has the following Lagrange density:
L =
1
m2 2
2
(8.68)
Insertion into the 1D Euler-Lagrange equation, equation (8.46) on page 162, yields
the dynamic equation
(2 m2 ) = 0
(8.69)
em|x|
|x|
(8.70)
which describes the Yukawa meson field for a scalar meson with mass m. With
=
166
1
c2 t
(8.71)
Bibliography
H =
1 2 2
c + ()2 + m2 2
2
(8.72)
L EM = L inter + L field = j A +
1
F F + m2 A A
40
(8.73)
(8.74)
This equation describes an electromagnetic field with a mass, or, in other words,
massive photons. If massive photons would exist, large-scale magnetic fields,
including those of the earth and galactic spiral arms, would be significantly modified to yield measurable discrepancies from their usual form. Space experiments
of this kind on board satellites have led to stringent upper bounds on the photon
mass. If the photon really has a mass, it will have an impact on electrodynamics
as well as on cosmology and astrophysics.
8.3 Bibliography
[1] A. O. BARUT, Electrodynamics and Classical Theory of Fields and Particles, Dover Publications, Inc., New York, NY, 1980, ISBN 0-486-64038-8.
[2] V. L. G INZBURG, Applications of Electrodynamics in Theoretical Physics and Astrophysics, Revised third ed., Gordon and Breach Science Publishers, New York, London,
Paris, Montreux, Tokyo and Melbourne, 1989, ISBN 2-88124-719-9.
[3] H. G OLDSTEIN, Classical Mechanics, second ed., Addison-Wesley Publishing Company, Inc., Reading, MA . . . , 1981, ISBN 0-201-02918-9.
[4] W. T. G RANDY, Introduction to Electrodynamics and Radiation, Academic Press,
New York and London, 1970, ISBN 0-12-295250-2.
[5] L. D. L ANDAU AND E. M. L IFSHITZ, The Classical Theory of Fields, fourth revised English ed., vol. 2 of Course of Theoretical Physics, Pergamon Press, Ltd., Oxford . . . , 1975,
ISBN 0-08-025072-6.
[6] W. K. H. PANOFSKY AND M. P HILLIPS, Classical Electricity and Magnetism, second ed.,
Addison-Wesley Publishing Company, Inc., Reading, MA . . . , 1962, ISBN 0-201-057026.
167
[7] J. J. S AKURAI, Advanced Quantum Mechanics, Addison-Wesley Publishing Company, Inc., Reading, MA . . . , 1967, ISBN 0-201-06710-2.
[8] D. E. S OPER, Classical Field Theory, John Wiley & Sons, Inc., New York, London, Sydney
and Toronto, 1976, ISBN 0-471-81368-0.
168
Example
8.4 Example
E XAMPLE 8.1
(8.75)
i.e., the difference between the magnetic and electric field energy densities.
From formula (7.79) on page 148 we recall that
0
E x /c Ey /c Ez /c
E x /c
0
Bz
By
(F ) =
Ey /c
Bz
0
Bx
Ez /c
By
Bx
0
(8.76)
0
E x /c Ey /c Ez /c
E x /c
0
Bz
By
F =
Ey /c
Bz
0
Bx
Ez /c By
Bx
0
(8.77)
where denotes the row number and the column number. Then, Einstein summation and
direct substitution yields
F F = F 00 F00 + F 01 F01 + F 02 F02 + F 03 F03
+ F 10 F10 + F 11 F11 + F 12 F12 + F 13 F13
+ F 20 F20 + F 21 F21 + F 22 F22 + F 23 F23
+ F 30 F30 + F 31 F31 + F 32 F32 + F 33 F33
= 0 E 2x /c2 Ey2 /c2 Ez2 /c2
(8.78)
B2
1 2
E
0 c2 0
1
=
2
B2
0 E 2
0
(8.79)
QED
C E ND OF EXAMPLE 8.1
169
F
Formul
(F.1)
B=0
(F.2)
B
t
H=j+ D
t
E=
(F.3)
(F.4)
Constitutive relations
D = E
B
H=
(F.5)
j = E
(F.7)
P = 0 E
(F.8)
(F.6)
(F.9)
171
F. Formul
E =
A
t
(F.10)
1
=0
c2 t
(F.11)
(F.12)
(F.13)
k E
c
(F.14)
d3 x0 eikx j k
40 c |x|
V0
Brad
(x) =
(F.15)
(F.16)
172
0 eik|x|
p k
4 |x|
(F.17)
Electromagnetic radiation
Erad
(x) =
1 eik|x|
(p k) k
40 |x|
(F.18)
40 c |x|
Brad
(x) =
(F.19)
(F.20)
Brad
(x) =
(F.21)
(F.22)
v
s = x x0 (x x0 )
c
v
x x0 = (x x0 ) |x x0 |
c
0
t
|x x0 |
=
t x
s
(F.23)
(F.24)
(F.25)
(F.26)
(F.27)
173
F. Formul
1 0
0
0
0 1 0
0
=
0 0 1 0
0 0
0 1
(F.28)
(F.29)
0 0
0 0
=
0
0
1 0
0
0
0 1
1
= p
1 2
v
=
c
(F.30)
(F.31)
(F.32)
(F.33)
dt
= c d
(F.34)
F.3.5 Four-velocity
u =
174
dx
= (c, v)
d
(F.35)
Vector relations
F.3.6 Four-momentum
p = m0 u =
E
,p
c
(F.36)
(F.37)
F.3.8 Four-potential
A =
,A
c
(F.38)
0
E x /c Ey /c Ez /c
E x /c
0
Bz
By
= A A =
Ey /c
Bz
0
Bx
Ez /c By
Bx
0
(F.39)
x i
i=1
def
def
x i
xi
xi
(F.40)
i = i =
,
,
=
, ,
(F.41)
x1 x2 x3
x y z
175
F. Formul
(F.42b)
= sin x 1 + cos x 2
(F.42c)
(F.43a)
(F.43b)
x 3 = cos r sin
(F.43c)
(F.42a)
(F.44)
(F.45)
(F.46)
Volume element
d3x = dV = dr dS = r2 dr d
(F.47)
176
a b = b a = i j ai b j = ab cos
(F.48)
a b = b a = i jk a j bk x i
(F.49)
a (b c) = (a b) c
(F.50)
a (b c) = b(a c) c(a b) ba c ca b
(F.51)
a (b c) + b (c a) + c (a b) = 0
(F.52)
(F.53)
(a b) (c d) = (a b d)c (a b c)d
(F.54)
Vector relations
(F.55)
(a) = a + a
(F.56)
(a) = a a
(F.57)
(a b) = b ( a) a ( b)
(F.58)
(a b) = a( b) b( a) + (b )a (a )b
(F.59)
(a b) = a ( b) + b ( a) + (b )a + (a )b
(F.60)
(F.61)
= 0
(F.62)
( a) = 0
(F.63)
( a) = ( a) a a a
2
(F.64)
Special identities
In the following x = xi x i and x0 = xi0 x i are radius vectors, k an arbitrary constant
vector, a = a(x) an arbitrary vector field, x i x i , and 0 x 0 x i .
i
x=3
(F.65)
x=0
(F.66)
(k x) = k
x
|x| =
|x|
x x0
|x x0 | =
= 0 |x x0 |
0
|x x |
x
1
= 3
|x|
|x|
1
1
x x0
0
|x x0 |3
|x x0 |
|x x0 |
x
1
= 2
= 4(x)
|x|3
|x|
x x0
1
2
=
= 4(x x0 )
|x x0 |3
|x x0 |
1
kx
k
=k
= 3
|x|
|x|
|x|
x
kx
k
=
if |x| , 0
|x|3
|x|3
(F.67)
(F.68)
(F.69)
(F.70)
(F.71)
(F.72)
(F.73)
(F.74)
(F.75)
177
F. Formul
k
|x|
= k2
1
|x|
= 4k(x)
(k a) = k( a) + k ( a) (k a)
(F.76)
(F.77)
Integral relations
Let V(S ) be the volume bounded by the closed surface S (V). Denote the 3dimensional volume element by d3x( dV) and the surface element, directed along
the outward pointing surface normal unit vector n,
by dS( d2x n).
Then
Z
ZV
V
Z
V
( a) d x =
3
() d3x =
dS a
(F.78)
dS
(F.79)
( a) d3x =
dS a
(F.80)
If S (C) is an open surface bounded by the contour C(S ), whose line element
is dl, then
I
IC
C
dl =
a dl =
dS
(F.81)
ZS
dS ( a)
(F.82)
F.5 Bibliography
[1] G. B. A RFKEN AND H. J. W EBER, Mathematical Methods for Physicists, fourth, international ed., Academic Press, Inc., San Diego, CA . . . , 1995, ISBN 0-12-059816-7.
[2] P. M. M ORSE AND H. F ESHBACH, Methods of Theoretical Physics, Part I. McGraw-Hill
Book Company, Inc., New York, NY . . . , 1953, ISBN 07-043316-8.
[3] W. K. H. PANOFSKY AND M. P HILLIPS, Classical Electricity and Magnetism, second ed.,
Addison-Wesley Publishing Company, Inc., Reading, MA . . . , 1962, ISBN 0-201-057026.
178
M
Mathematical
Methods
M.1.1 Vectors
Radius vector
A vector can be represented mathematically in a number of different ways. One
suitable representation is in terms of an ordered N-tuple, or row vector, of the
coordinates xN where N is the dimensionality of the space under consideration.
179
M. Mathematical Methods
The most basic vector is the radius vector which is the vector from the origin to
the point of interest. Its N-tuple representation simply enumerates the coordinates
which describe this point. In this sense, the radius vector from the origin to a point
is synonymous with the coordinates of the point itself.
In the 3D Euclidean space R3 , we have N = 3 and the radius vector can be
represented by the triplet (x1 , x2 , x3 ) of coordinates xi , i = 1, 2, 3. The coordinates
xi are scalar quantities which describe the position along the unit base vectors x i
which span R3 . Therefore a representation of the radius vector in R3 is
3
def
x = xi x i xi x i
(M.1)
i=1
xi (x1 , x2 , x3 ) (x, y, z)
(M.2)
This component notation is particularly useful in 4D space where we can represent the radius vector either in its contravariant component form
def
x (x0 , x1 , x2 , x3 )
(M.3)
x (x0 , x1 , x2 , x3 )
(M.4)
The relation between the covariant and contravariant forms is determined by the
metric tensor (also known as the fundamental tensor) whose actual form is dictated by the properties of the vector space in question. The dual representation
of vectors in contravariant and covariant forms is most convenient when we work
in a non-Euclidean vector space with an indefinite metric. An example is Lorentz
space L4 which is a 4D Riemannian space utilised to formulate the special theory
of relativity.
We note that for a change of coordinates x x0 = x0 (x0 , x1 , x2 , x3 ), due
to a transformation from a system to another system 0 , the differential radius
vector dx transforms as
x0
dx
(M.5)
dx0 =
x
180
which follows trivially from the rules of differentiation of x0 considered as functions of four variables x .
M.1.2 Fields
A field is a physical entity which depends on one or more continuous parameters.
Such a parameter can be viewed as a continuous index which enumerates the
coordinates of the field. In particular, in a field which depends on the usual
radius vector x of R3 , each point in this space can be considered as one degree of
freedom so that a field is a representation of a physical entity which has an infinite
number of degrees of freedom.
Scalar fields
We denote an arbitrary scalar field in R3 by
def
(M.6)
This field describes how the scalar quantity varies continuously in 3D R3 space.
In 4D, a four-scalar field is denoted
def
(x0 , x1 , x2 , x3 ) (x )
(M.7)
which indicates that the four-scalar depends on all four coordinates spanning
this space. Since a four-scalar has the same value at a given point regardless of
coordinate system, it is also called an invariant.
Analogous to the transformation rule, equation (M.5) on page 180, for the
differential dx , the transformation rule for the differential operator /x under a
transformation x x0 becomes
x
=
x0 x0 x
(M.8)
Vector fields
We can represent an arbitrary vector field a(x) in R3 as follows:
a(x) = ai (x) x i
(M.9)
(M.10)
181
M. Mathematical Methods
(M.11)
(M.12)
where x is the radius four-vector. Again, the relation between a and a is determined by the metric of the physical 4D system under consideration.
Whether an arbitrary N-tuple fulfils the requirement of being an (N-dimensional) contravariant vector or not, depends on its transformation properties during
a change of coordinates. For instance, in 4D an assemblage y = (y0 , y1 , y2 , y3 )
constitutes a contravariant four-vector (or the contravariant components of a fourvector) if and only if, during a transformation from a system with coordinates
x to a system 0 with coordinates x0 , it transforms to the new system according
to the rule
x0
y
(M.13)
y0 =
x
i.e., in the same way as the differential coordinate element dx transforms according to equation (M.5) on page 180.
The analogous requirement for a covariant four-vector is that it transforms,
during the change from to 0 , according to the rule
x
y
(M.14)
x0
i.e., in the same way as the differential operator /x transforms according to
equation (M.8) on page 181.
y0 =
Tensor fields
We denote an arbitrary tensor field in R3 by A(x). This tensor field can be represented in a number of ways, for instance in the following matrix form:
182
1 0 0
(i j ) = 0 1 0
0 0 1
(M.17)
Another common and useful tensor is the fully antisymmetric tensor of rank 3,
also known as the Levi-Civita tensor
(M.19)
In fact, tensors may have any rank n. In this picture a scalar is considered to
be a tensor of rank n = 0 and a vector a tensor of rank n = 1. Consequently, the
notation where a vector (tensor) is represented in its component form is called the
tensor notation. A tensor of rank n = 2 may be represented by a two-dimensional
array or matrix whereas higher rank tensors are best represented in their component forms (tensor notation).
In 4D, we have three forms of four-tensor fields of rank n. We speak of
a contravariant four-tensor field, denoted A1 2 ...n (x ),
a covariant four-tensor field, denoted A1 2 ...n (x ),
2 ...k
a mixed four-tensor field, denoted A1k+1
...n (x ).
a (x ) g a (x )
(M.20)
This rule is often called lowering of index. The raising of index analogue of the
index lowering rule is:
def
a (x ) g a (x )
(M.21)
More generally, the following lowering and raising rules hold for arbitrary
rank n mixed tensor fields:
2 ...k1 k
1 2 ...k1
gk k A1k+1
k+2 ...n (x ) = Ak k+1 ...n (x )
(M.22)
183
M. Mathematical Methods
1 2 ...k1 k
2 ...k1
gk k A1k k+1
...n (x ) = Ak+1 k+2 ...n (x )
(M.23)
Successive lowering and raising of more than one index is achieved by a repeated
application of this rule. For example, a dual application of the lowering operation
on a rank 2 tensor in contravariant form yields
A = g g A
(M.24)
i.e., the same rank 2 tensor in covariant form. This operation is also known as a
tensor contraction.
(M.25)
where we used the fact that the scalar product x i x j is a representation of the
Kronecker delta i j defined in equation (M.16) on page 182. In Russian literature,
the 3D scalar product is often denoted (ab). The scalar product of a in R3 with
itself is
def
(M.26)
(M.27)
(M.28)
where we made use of the index lowering and raising rules (M.20) and (M.21).
The result is a four-scalar, i.e., an invariant which is independent of in which 4D
coordinate system it is measured.
The quadratic differential form
ds2 = g dx dx = dx dx
(M.29)
i.e., the scalar product of the differential radius four-vector with itself, is an invariant called the metric. It is also the square of the line element ds which is the
distance between neighbouring points with coordinates x and x + dx .
184
Dyadic product
The dyadic product field A(x) a(x)b(x) with two juxtaposed vector fields a(x)
and b(x) is the outer product of a and b. Operating on this dyad from the right
and from the left with an inner product of an vector c one obtains
def
def
def
def
A c ab c a(b c)
c A c ab (c a)b
(M.30a)
(M.30b)
x 1
a1 b1 a1 b2 a1 b3
x 2
a2 b1 a2 b2 a2 b3
(M.31)
ab = x 1 x 2 x 3
x 3
a3 b1 a3 b2 a3 b3
which means that we can represent the tensor A(x) in matrix form as
a1 b1 a1 b2 a1 b3
Ai j (xk ) = a2 b1 a2 b2 a2 b3
a3 b1 a3 b2 a3 b3
(M.32)
which we identify with expression (M.15) on page 182, viz. a tensor in matrix
notation.
Vector product
The vector product or cross product of two arbitrary 3D vectors a and b in ordinary R3 space is the vector
c = a b = i jk a j bk x i
(M.33)
Here i jk is the Levi-Civita tensor defined in equation (M.18) on page 183. Sometimes the 3D vector product of a and b is denoted a b or, particularly in the
Russian literature, [ab]. Alternatively,
a b = ab sin e
(M.34)
where is the angle between a and b and e is a unit vector perpendicular to the
plane spanned by a and b.
A spatial reversal of the coordinate system (x10 , x20 , x30 ) = (x1 , x2 , x3 ) changes
sign of the components of the vectors a and b so that in the new coordinate system
a0 = a and b0 = b, which is to say that the direction of an ordinary vector is not
185
M. Mathematical Methods
dependent on the choice of directions of the coordinate axes. On the other hand,
as is seen from equation (M.33) on page 185, the cross product vector c does not
change sign. Therefore a (or b) is an example of a true vector, or polar vector,
whereas c is an example of an axial vector, or pseudovector.
A prototype for a pseudovector is the angular momentum vector L = x p
and hence the attribute axial. Pseudovectors transform as ordinary vectors under
translations and proper rotations, but reverse their sign relative to ordinary vectors
for any coordinate change involving reflection. Tensors (of any rank) which transform analogously to pseudovectors are called pseudotensors. Scalars are tensors
of rank zero, and zero-rank pseudotensors are therefore also called pseudoscalars,
an example being the pseudoscalar x i ( x j x k ). This triple product is a representation of the i jk component of the Levi-Civita tensor i jk which is a rank three
pseudotensor.
x i
def
xi
(M.35)
where x i is the ith unit vector in a Cartesian coordinate system. Since the operator in itself has vectorial properties, we denote it with a boldface nabla. In
component notation we can write
i =
,
,
(M.36)
x1 x2 x3
In 4D, the contravariant component representation of the four-del operator is
defined by
=
,
,
,
(M.37)
x0 x1 x2 x3
whereas the covariant four-del operator is
=
,
,
,
x0 x1 x2 x3
(M.38)
186
and
y0 = 0 x y
(M.40)
respectively.
With the help of the del operator we can define the gradient, divergence and
curl of a tensor (in the generalised sense).
The gradient
The gradient of an R3 scalar field (x), denoted (x), is an R3 vector field a(x):
(x) = (x) = x i i (x) = a(x)
(M.41)
From this we see that the boldface notation for the nabla and del operators is very
handy as it elucidates the 3D vectorial property of the gradient.
In 4D, the four-gradient is a covariant vector, formed as a derivative of a fourscalar field (x ), with the following component form:
(x ) =
(x )
x
(M.42)
The divergence
We define the 3D divergence of a vector field in R3 as
a(x) = x j a j (x) = i j i a j (x) = i ai (x) =
ai (x)
= (x)
xi
(M.43)
a (x )
x
(M.44)
The Laplacian
The 3D Laplace operator or Laplacian can be described as the divergence of the
gradient operator:
2 = = =
2
2
x i x j
= i j i j = 2i = 2 2
xi
x j
xi
i=1 xi
(M.45)
The symbol 2 is sometimes read del squared. If, for a scalar field (x), 2 < 0
at some point in 3D space, it is a sign of concentration of at that point.
187
M. Mathematical Methods
The curl
In R3 the curl of a vector field a(x), denoted a(x), is another R3 vector field
b(x) which can be defined in the following way:
a(x) = i jk x i j ak (x) = i jk x i
ak (x)
= b(x)
x j
(M.46)
where use was made of the Levi-Civita tensor, introduced in equation (M.18) on
page 183.
The covariant 4D generalisation of the curl of a four-vector field a (x ) is the
antisymmetric four-tensor field
G (x ) = a (x ) a (x ) = G (x )
(M.47)
t2
dt L(qi , q i , t)
(M.49)
t1
(M.50)
dt q i
qi
188
(M.51)
Analytical mechanics
L
q i
(M.52)
(M.53)
(M.54)
After differentiating the left and right hand sides of this definition and setting them
equal we obtain
H
H
H
L
L
L
dpi +
dqi +
dt = q i dpi + pi dq i
dqi
dq i
dt
pi
qi
t
qi
q i
t
(M.55)
According to the definition of pi , equation (M.52) above, the second and fourth
terms on the right hand side cancel. Furthermore, noting that according to equation (M.53) the third term on the right hand side of equation (M.55) above is equal
to p i dqi and identifying terms, we obtain the Hamilton equations:
H
dqi
= q i =
pi
dt
dpi
H
= p i =
qi
dt
(M.56a)
(M.56b)
189
M. Mathematical Methods
M.3 Examples
E XAMPLE M.1
BT ENSORS IN 3D SPACE
x3
d2x
x2
V
x1
F IGURE M.1:
T n = T n
(M.57)
Using (M.57) and Newtons second law, we find that the matter of mass m, which at a given
instant is located in V obeys the equation of motion
190
Examples
(M.58)
where Fext is the external force and a is the acceleration of the volume element. In other words
Fext
m
(M.59)
T n = n1 T x 1 + n2 T x 2 + n3 T x 3 + 2 a
dx
m
Since both a and Fext /m remain finite whereas m/d2x 0 as V 0, one finds that in this limit
3
T n = ni T x i ni T x i
(M.60)
i=1
From the above derivation it is clear that equation (M.60) above is valid not only in equilibrium
but also when the matter in V is in motion.
Introducing the notation
T i j = T x i j
(M.61)
for the jth component of the vector T x i , we can write equation (M.60) in component form as
follows
3
T nj
= (T n ) j = ni T i j ni T i j
(M.62)
i=1
Using equation (M.62) above, we find that the component of the vector T n in the direction of
an arbitrary unit vector m
is
T n m = T n m
= T nj
mj =
j=1
j=1
ni Ti j
m j ni T i j m j = n T m
(M.63)
i=1
Hence, the jth component of the vector T x i , here denoted T i j , can be interpreted as the i jth
component of a tensor T. Note that T n m is independent of the particular coordinate system used
in the derivation.
We shall now show how one can use the momentum law (force equation) to derive the
equation of motion for an arbitrary element of mass in the body. To this end we consider a
part V of the body. If the external force density (force per unit volume) is denoted by f and the
velocity for a mass element dm is denoted by v, we obtain
d
dt
Z
V
v dm =
f d3x +
T n d2x
(M.64)
d
v j dm =
dt
f j d3x +
2
T nj
d x =
Z
V
f j d3x +
ni T i j d2x
(M.65)
where, in the last step, equation (M.62) was used. Setting dm = d3x and using the divergence
theorem on the last term, we can rewrite the result as
Z
V
d
v j d3x =
dt
Z
V
f j d3x +
Z
V
T i j 3
dx
xi
(M.66)
Since this formula is valid for any arbitrary volume, we must require that
191
M. Mathematical Methods
T i j
d
vj fj
=0
dt
xi
(M.67)
or, equivalently
v j
T i j
+ v v j f j
=0
t
xi
(M.68)
Note that v j /t is the rate of change with time of the velocity component v j at a fixed point
x = (x1 , x1 , x3 ).
C E ND OF EXAMPLE M.1
E XAMPLE M.2
if = = 0
1
g = 1 if = = i = j = 1, 2, 3
(M.69)
0
if ,
which, in matrix notation, is represented as
1 0
0
0
0 1 0
0
(g ) =
0 0 1 0
0 0
0 1
(M.70)
i.e., a matrix with a main diagonal that has the sign sequence, or signature, {+, , , } or
1 if = = 0
g = 1
(M.71)
if = = i = j = 1, 2, 3
0
if ,
which, in matrix notation, is represented as
1 0 0 0
0 1 0 0
(g ) =
0 0 1 0
0 0 0 1
(M.72)
a (a0 , a1 , a2 , a3 ) = (a0 , a)
(M.73)
According to the index lowering rule, equation (M.20) on page 183, we obtain the covariant
version of this vector as
192
Examples
def
a (a0 , a1 , a2 , a3 ) = g a
(M.74)
a0 = 1 a0 + 0 a1 + 0 a2 + 0 a3 = a0
(M.75)
=1:
a1 = 0 a 1 a + 0 a + 0 a = a
(M.76)
=2:
a2 = 0 a0 + 0 a1 1 a2 + 0 a3 = a2
(M.77)
=3:
a3 = 0 a + 0 a + 0 a + 1 a = a
(M.78)
or
a = (a0 , a1 , a2 , a3 ) = (a0 , a1 , a2 , a3 ) = (a0 , a)
(M.79)
(M.80)
(M.81)
C E ND OF EXAMPLE M.2
E XAMPLE M.3
6
(M.82)
def
(M.83)
(M.84)
Using this in equation (M.82), we see that we can interpret this so that the complex unit vector
is
A
aI
aR
A =
= p 2
a R + i p 2
a I
A
aR a2I + 2iaR aI
aR a2I + 2iaR aI
p
p
aR a2R a2I 2iaR aI
aI a2R a2I 2iaR aI
=
a
+
i
a I C3
R
a2R + a2I
a2R + a2I
(M.85)
193
M. Mathematical Methods
On the other hand, the definition of the scalar product in terms of the inner product of complex
vector with its own complex conjugate yields
def
(M.86)
(M.87)
C E ND OF EXAMPLE M.3
E XAMPLE M.4
(M.88)
The important scalar product of the L4 radius four-vector with itself becomes
x x = (x0 , x) (x0 , x) = (ct, x) (ct, x)
(M.89)
(M.90)
C E ND OF EXAMPLE M.4
194
Examples
E XAMPLE M.5
(M.91)
(M.92)
1 2
2 = 2
c2 t2
(M.93)
which is the dAlembert operator, sometimes denoted , and sometimes defined with an opposite sign convention.
C E ND OF EXAMPLE M.5
E XAMPLE M.6
3
Very often electrodynamic quantities are dependent on the relative distance in R between
two vectors x and x0 , i.e., on |x x0 |. In analogy with equation (M.35) on page 186, we can
define the primed del operator in the following way:
0 = x i
= 0
xi0
(M.94)
Using this, the unprimed version, equation (M.35) on page 186, and elementary rules of differentiation, we obtain the following two very useful results:
(|x x0 |) = x i
x x0
|x x0 |
|x x0 |
=
= x i
0
xi
|x x |
xi0
(M.95)
= 0 (|x x0 |)
and
1
|x x0 |
x x0
=
= 0
|x x0 |3
1
|x x0 |
(M.96)
C E ND OF EXAMPLE M.6
195
M. Mathematical Methods
E XAMPLE M.7
BD IVERGENCE IN 3D
For an arbitrary R3 vector field a(x0 ), the following relation holds:
0 a(x0 )
a(x0 )
1
0
0
=
0
+
a(x
)
|x x0 |
|x x0 |
|x x0 |
(M.97)
which demonstrates how the primed divergence, defined in terms of the primed del operator in
equation (M.94) on page 195, works.
C E ND OF EXAMPLE M.7
E XAMPLE M.8
= 4(x x0 )
|x x0 |
|x x0 |
(M.98)
where (x x0 ) is the 3D Dirac delta function. This formula follows directly from the fact
that
Z
I
Z
x x0
1
x x0
3
2
=
d
x
d3x
=
d
x
n
(M.99)
|x x0 |3
|x x0 |3
|x x0 |
V
V
S
equals 4 if the integration volume V(S ), enclosed by the surface S (V), includes x = x0 , and
equals 0 otherwise.
C E ND OF EXAMPLE M.8
E XAMPLE M.9
BT HE CURL OF A GRADIENT
Using the definition of the R3 curl, equation (M.46) on page 188, and the gradient, equation (M.41) on page 187, we see that
[(x)] = i jk x i j k (x)
(M.100)
(x) x 1
x2 x3 x3 x2
2
2
+
(x) x 2
x3 x1 x1 x3
2
2
+
(x) x 3
x1 x2 x2 x1
(M.101)
196
Examples
(M.102)
(M.103)
BT HE DIVERGENCE OF A CURL
E XAMPLE M.10
With the use of the definitions of the divergence (M.43) and the curl, equation (M.46) on
page 188, we find that
[ a(x)] = i [ a(x)]i = i jk i j ak (x)
(M.104)
Using the definition for the Levi-Civita symbol, defined by equation (M.18) on page 183, we
find that, due to the assumed well-behavedness of a(x),
i jk
ak
xi
x j
2
2
a1 (x)
=
x2 x3 x3 x2
2
2
+
a2 (x)
x3 x1 x1 x3
2
2
+
a3 (x)
x1 x2 x2 x1
i i jk j ak (x) =
(M.105)
0
i.e., that
[ a(x)] 0
(M.106)
(M.107)
C E ND OF EXAMPLE M.10
197
M. Mathematical Methods
M.4 Bibliography
[1] G. B. A RFKEN AND H. J. W EBER, Mathematical Methods for Physicists, fourth, international ed., Academic Press, Inc., San Diego, CA . . . , 1995, ISBN 0-12-059816-7.
[2] R. A. D EAN, Elements of Abstract Algebra, John Wiley & Sons, Inc., New York, NY . . . ,
1967, ISBN 0-471-20452-8.
[3] A. A. E VETT, Permutation symbol approach to elementary vector analysis, American Journal of Physics, 34 (1965), pp. 503507.
[4] P. M. M ORSE AND H. F ESHBACH, Methods of Theoretical Physics, Part I. McGraw-Hill
Book Company, Inc., New York, NY . . . , 1953, ISBN 07-043316-8.
[5] B. S PAIN, Tensor Calculus, third ed., Oliver and Boyd, Ltd., Edinburgh and London, 1965,
ISBN 05-001331-9.
[6] W. E. T HIRRING, Classical Mathematical Physics, Springer-Verlag, New York, Vienna,
1997, ISBN 0-387-94843-0.
198
Index
acceleration field, 100
advanced time, 46
Ampres law, 6
Ampre-turn density, 57
anisotropic, 116
anomalous dispersion, 117
antenna, 77
antenna current, 77
antenna feed point, 78
antisymmetric tensor, 147
associated Legendre polynomial, 87
associative, 138
axial gauge, 49
axial vector, 147, 185
Bessel functions, 84
Biot-Savarts law, 8
birefringent, 116
braking radiation, 107
bremsstrahlung, 107, 113
canonically conjugate four-momentum, 156
canonically conjugate momentum, 156, 188
canonically conjugate momentum density,
163
Cerenkov radiation, 118
characteristic impedance, 29
classical electrodynamics, 1, 9
closed algebraic structure, 138
coherent radiation, 112
collisional interaction, 116
complete -Lorenz gauge, 48
complex field six-vector, 23
complex notation, 33
complex vector, 193
component notation, 180
concentration, 187
conservative field, 12
conservative forces, 160
constitutive relations, 15
contravariant component form, 134, 180
contravariant field tensor, 147
contravariant four-tensor field, 183
contravariant four-vector, 182
contravariant four-vector field, 137
contravariant vector, 134
convection potential, 127
convective derivative, 13
cosine integral, 81
Coulomb gauge, 47
Coulombs law, 2
covariant, 132
covariant component form, 180
covariant field tensor, 148
covariant four-tensor field, 183
covariant four-vector, 182
covariant four-vector field, 137
covariant vector, 134
cross product, 185
curl, 187
cutoff, 129
cyclotron radiation, 109, 113
dAlembert operator, 26, 43, 143, 194
del operator, 186
del squared, 187
differential distance, 136
differential vector operator, 186
dipole antennas, 77
Dirac delta, 196
Diracs symmetrised Maxwell equations, 16
dispersive, 117
displacement current, 11
divergence, 187
dot product, 184
dual electromagnetic tensor, 149
dual vector, 134
duality transformation, 17, 149
199
Index
Euler-Mascheroni constant, 81
event, 138
E1 radiation, 90
E2 radiation, 93
Einsteins summation convention, 180
electric charge conservation law, 10
electric charge density, 4
electric conductivity, 11
electric current density, 8
electric dipole moment, 89
electric dipole moment vector, 54
electric dipole radiation, 90
electric displacement, 15
electric displacement current, 21
electric displacement vector, 53, 55
electric field, 3
electric field energy, 59
electric monopole moment, 53
electric permittivity, 116
electric polarisation, 54
electric quadrupole moment tensor, 54
electric quadrupole radiation, 93
electric quadrupole tensor, 92
electric susceptibility, 55
electric volume force, 60
electricity, 2
electrodynamic potentials, 40
electromagnetic field tensor, 147
electromagnetic scalar potential, 41
electromagnetic vector potential, 40
electromagnetism, 1
electromagnetodynamic equations, 16
electromagnetodynamics, 17
electromotive force (EMF), 12
electrostatic scalar potential, 39
electrostatics, 2
electroweak theory, 1
energy theorem in Maxwells theory, 59
equation of continuity, 10, 144
equations of classical electrostatics, 9
equations of classical magnetostatics, 9
Euclidean space, 139
Euclidean vector space, 135
Euler-Lagrange equation, 162
Euler-Lagrange equations, 163, 188
200
far field, 68
far zone, 71
Faradays law, 12
field, 181
field Lagrange density, 164
field point, 4
field quantum, 129
fine structure constant, 114, 129
four-current, 143
four-del operator, 186
four-dimensional Hamilton equations, 156
four-dimensional vector space, 134
four-divergence, 187
four-gradient, 187
four-Hamiltonian, 156
four-Lagrangian, 154
four-momentum, 142
four-potential, 143
four-scalar, 181
four-tensor fields, 183
four-vector, 137, 181
four-velocity, 141
Fourier integral, 28
Fourier series, 27
Fourier transform, 28, 44
free-free radiation, 107
functional derivative, 162
fundamental tensor, 134, 180, 183
Galileos law, 131
gauge fixing, 49
gauge function, 42
gauge invariant, 42
gauge transformation, 42
Gausss law of electrostatics, 5
general inhomogeneous wave equations, 42
generalised coordinate, 156, 188
generalised four-coordinate, 156
Gibbs notation, 186
gradient, 187
Green function, 44, 87
group theory, 138
group velocity, 117
Hamilton density, 163
201
Index
202
polarisation potential, 86
polarisation vector, 85
positive definite, 139
positive definite norm, 135
potential energy, 160, 188
potential theory, 87
power flux, 59
Poynting vector, 59
Poyntings theorem, 59
Proca Lagrangian, 166
propagator, 44
proper time, 137
pseudo-Riemannian space, 139
pseudoscalar, 179
pseudoscalars, 186
pseudotensor, 179
pseudotensors, 186
pseudovector, 147, 179, 185
quadratic differential form, 136, 184
quantum chromodynamics, 1
quantum electrodynamics, 1, 47
quantum mechanical nonlinearity, 4
radial gauge, 49
radiation field, 68, 71, 100
radiation fields, 71
radiation resistance, 81
radius four-vector, 134
radius vector, 179
raising of index, 183
rank, 182
rapidity, 141
refractive index, 116
relative electric permittivity, 61
relative magnetic permeability, 61
relative permeability, 116
relative permittivity, 116
Relativity principle, 132
relaxation time, 28
rest mass density, 164
retarded Coulomb field, 71
retarded potentials, 46
retarded relative distance, 95
retarded time, 46
Riemann-Silberstein vector, 23
Riemannian metric, 136
vacuum permeability, 6
vacuum permittivity, 2
vacuum polarisation effects, 4
vacuum wave number, 29
variational principle, 188
Vavilov-Cerenkov
cone, 120
Vavilov-Cerenkov
radiation, 118, 120
vector, 179
vector product, 185
velocity field, 100
velocity gauge condition, 48
virtual simultaneous coordinate, 96, 100
wave equations, 25
wave vector, 31, 117
world line, 138
Youngs modulus, 161
Yukawa meson field, 166
203