This action might not be possible to undo. Are you sure you want to continue?

Υ

Bo Thidé

U

P S I L O N

M

E D I A

Bo Thidé

E LECTROMAGNETIC F IELD T HEORY

Also available

**E LECTROMAGNETIC F IELD T HEORY E XERCISES
**

by Tobia Carozzi, Anders Eriksson, Bengt Lundborg, Bo Thidé and Mattias Waldenvik

Sweden Υ UPSILON MEDIA · UPPSALA · SWEDEN .E LECTROMAGNETIC F IELD T HEORY Bo Thidé Swedish Institute of Space Physics and Department of Astronomy and Space Physics Uppsala University.

1998. Sweden All rights reserved.A This book was typeset in L TEX 2ε on an HP9000/700 series workstation and printed on an HP LaserJet 5000GN printer. 1999 and 2000 by Bo Thidé Uppsala. Electromagnetic Field Theory ISBN X-XXX-XXXXX-X . Copyright ©1997.

. .3. . . . . . . . . . . . . .Contents Preface 1 Classical Electrodynamics 1. . . . . .2 Plane waves . . . . . . . . . . . . . . . . .2 The wave equation for B . . . . . . . . . . . . . 1. .1. . . . . . . .3 Electrodynamics . . . . . .1 Equation of continuity . 20 23 24 24 24 25 26 27 . 1. . 2. . .4 Electromagnetic Duality . . . . . . . . . . . . .1 Duality of the electromagnetodynamic equations 16 . . . 1. . 2. . . 1. . . .3. . . . . .2. . . . . . . . . . . . .2 The electrostatic ﬁeld . . . . . . . . . . . .1. . . . .3 The time-independent wave equation for E 2. .5 Maxwell’s microscopic equations 1. . .1. . . . . .2 Magnetostatics . . xi 1 1 1 2 4 4 6 8 9 9 10 11 14 14 15 Example 1. . . . .1. .2 Maxwell from Dirac-Maxwell equations for a ﬁxed mixing angle . . . . i . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6 Maxwell’s macroscopic equations 1. . . . . . . . . . . . . . . . . . . . . . . . . 1. . . . . .1 Telegrapher’s equation . .3. . . . . 1. . . . . . . . . . . 1. . . . . . . . . . . . .4 Duality expressed in the complex ﬁeld six-vector 19 Bibliography . 2 Electromagnetic Waves 2. Example 1. . . . . . . . . . . . . . .4 Faraday’s law of induction . . . . . . . . . . . . .2 Maxwell’s displacement current .1 The wave equation . . 18 Example 1. . . . . . . . . . . . 1. . . . . . .2 The magnetostatic ﬁeld . . . . . . . 1. . .1 Electrostatics . . . . . . . . . . . . . .1. . . . . . . . . . . .1 Ampère’s law . 2. .3 The complex ﬁeld six-vector . . .3. . . . . .3 Electromotive force . . . . . . . . . . .3. . . . . . . . . .1 Coulomb’s law . . . . .2. . . . . . . . . . . . . . . 2. . .1 The wave equation for E . . . . . . . . 17 Example 1. . . . . . . . . . . .3. . . . . . . . . . . . 1. . . . . . .2. . . . 1.

. . . . . . . . . . . . . . . . . . . . . . . . . . Bibliography . . . . . 5. . . . . . . . . Invariant line element and proper time . . . . . . . . Gauge transformations . . . . . . .1 The four-potential . . . .1. . . . . . .3. . . . . . . . . . . . 5. . .3 Covariant classical electrodynamics . . . . . . . . . . Relativistic Electrodynamics 5. . . . . . .3. . . . . . . . . . . The Lorentz transformation matrix . . . . . . . Metric tensor . . . . . . 3. . . . .2 Waves in conductive media . . . . . . . . . . . . . . . . . . . . . 5. . . . Bibliography . . . . . . . . . . . . . . . . . .1 The special theory of relativity . . . . .uu. .2 Solution of the Lorentz equations for the electromagnetic potentials . . . . .2 Lorentz space . .3. . . . . . . . . . . . . . . . . . . . .1 The Lorentz transformation . . . . . . 5. . . . . . .2 The electric ﬁeld . . . . . . . . . . . . . . . . . . . .3 The electromagnetic scalar and vector potentials .2. . . . . . . . . . .3 Minkowski space . 3 Electromagnetic Potentials 3.3. . . . . . . . . . . . . . . . . . . . . 4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 Electromagnetic gauges . . . . . . . . . . . . . . . . . . . . . .ii C ONTENTS 2. . . . . . . . . . . . . . . . . . . . . . Lorentz equations for the electromagnetic potentials . . . . . . . . . 5. .1 The electrostatic scalar potential . . . . . . . . . . . . . . . . The Electromagnetic Fields 4. . . . . . . . . . . . . . . Four-vector ﬁelds . .2 Covariant classical mechanics . . . . . . . . . . . . The Lorentz group . . . . .plasma. . .2 The magnetostatic vector potential . . 5. . . . . Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 Observables and averages . . . . . . . . . . . . . .se/CED/Book Draft version released 13th November 2000 at 22:01. . . . . . . . . . . . . . . .2 The Liénard-Wiechert potentials . .1. . 5. Radius four-vector in contravariant and covariant form Scalar product and norm . . . . . The retarded potentials . . .1 The magnetic ﬁeld . . 3. 5. . . . . . . . . . . . . . . . .3. . . . . . . . . . . . . . . . 29 30 31 33 33 34 34 36 36 36 38 41 41 43 45 47 49 51 51 52 53 54 54 55 56 57 57 58 58 61 62 62 63 65 67 4 5 Downloaded from http://www. . . . . . . . . .3 The electromagnetic ﬁeld tensor . . . 3. . . . . Bibliography . . . . . . . . . . . . . . . . . . 3. . . .1. . . . . 2. . . .

. . . . . . . . . . . Bibliography . . . . . . . . . . . . . .1 Covariant equations of motion . . . . . . 6. . 8. . . .2 Covariant Field Theory . . .1 Lagrange-Hamilton formalism for ﬁelds and interactions The electromagnetic ﬁeld .2 Magnetisation and the magnetising ﬁeld . . . . . . . . . . .1 Monochromatic signals . . . . .3 Energy and momentum .2 The convection potential and the convection force . . . . . . . . . 6. . 8 Electromagnetic Radiation 8. . . . 8. . . . . 8. . . . . . 8. . Bibliography . . . . . .5 Radiation from a localised charge in arbitrary motion 8. . . 7. . . . . . . . . .5. . .1 Field energy difference expressed in the ﬁeld tensor . . . . . . . . Downloaded from http://www. . . . . Lagrange formalism . . . . . .4. . . . . . . .3. . . . . . . . . . . . . . . . . . .1 Charged Particles in an Electromagnetic Field . .2 The momentum theorem in Maxwell’s theory . . . . .4.se/CED/Book . 123 Draft version released 13th November 2000 at 22:01. . . . . . . . . . . . . . . . . . . .1 The energy theorem in Maxwell’s theory . . 8. . . . . . . 8. . . . . .1 Electric polarisation and the electric displacement vector 7. .4. . . . . . . . . . . . . . . . . . . . 7. . . . 69 69 69 69 72 76 77 80 81 84 85 87 87 87 90 91 92 93 95 97 97 99 100 100 102 102 104 104 108 109 110 111 112 114 121 Other ﬁelds . .1 The Hertz potential . . . .2 Electric dipole radiation . . . . . . . . . . . .2. . . . . . . . . .2. . . . . . .uu. . . .3. 8. . . . . . . . . . . . . . .5. 8. . . .3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8. . . . . . .2 Radiated energy . . . . . . . . . . . . . . . . . . . Example 8. . . . . . . . . . . . . . . . . . . . .1. .iii 6 Interactions of Fields and Particles 6. . . . . . . . .4 Electric quadrupole radiation . . . . . . . . . . . . . . . . .4. 8.plasma. 7 Interactions of Fields and Matter 7. . . . Hamiltonian formalism .4 Multipole radiation . Example 8. 7. . . . . . .1 The Liénard-Wiechert potentials . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 Finite bandwidth signals . . . . . . . . . .1 Linear antenna . . . . .2. . . . 8. . . . .3 Radiation from extended sources . . . . . . . . . 6. . . 7. . . . . . . . . . . . . . . . . . . . . . . . . .1. . . .1 The ﬁelds from a uniformly moving charge . . . . . .1 The radiation ﬁelds . . . . . . . . . . . . . . . . . . . . 8. . .1 Electric multipole moments . .3 Magnetic dipole radiation . . . . . Example 6. . . . . .2 Radiation from an accelerated point charge . . . . . . . .

. . . .3 Bremsstrahlung for low speeds and short acceleration times .3 Radiation for small velocities . . .5 Radiation from charges moving in matter ˇ Vavilov-Cerenkov radiation . 127 Example 8. . . . . . . . . . . . . .2. .2 The far ﬁelds from an extended source distribution .4 . . . . . . . . . . . . . . . . F. . . . . . . . . . . . . Radiation in the general case . . . . . . . . . . . . . F. . . . . . . . . .7 The ﬁelds from a point charge in uniform motion . F. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. . . . Maxwell’s stress tensor . . . . . . . . . . . . . F. . . . . . . . . . . . . . . . . . .iv C ONTENTS 8. . . . . . . . . . .7 Four-current density . Poynting’s vector . . . . . . . . . . . . F.3 Lorentz transformation of a four-vector . . . . .1. .3 Special Relativity . Synchrotron radiation . . . . . . . . . . . . 8. F. . . . .1. . . . . . . . .4 The far ﬁelds from a magnetic dipole . . .5. . . . . .1. . . . . . . . . . . . . . . . . . F. . 130 Cyclotron and synchrotron radiation . . . .3 The far ﬁelds from an electric dipole . . . .plasma. . .2. . F. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Vector and scalar potentials . . Cyclotron radiation . . . . . . . . . . . . . . .3. . . . . . . .1 The Electromagnetic Field . . . . . . . . . . . . . . . . . . . . . . F. . . . . F. . . . . . . . . . . . . F.uu. . . . . 8. . . . . . . .se/CED/Book Draft version released 13th November 2000 at 22:01. . . . . .3. . . . . . . . . . . . . . . . . . . . . . . F. . . . . Constitutive relations . . . . . Downloaded from http://www. . .2. . 125 Bremsstrahlung . . . . . . . . . . . . 132 134 134 137 137 139 142 147 149 149 149 149 149 149 150 150 150 150 150 150 150 150 151 151 151 151 152 152 152 152 152 152 153 153 153 F Formulae F. . . . . . . .5 Four-velocity . . . . . .3. . . . . . . . . .5. .2 Covariant and contravariant four-vectors . . . . . . . . . . . . .8 Four-potential . . . . . Bibliography . .3.2 Electromagnetic Radiation . .3 Force and energy . .1 Metric tensor . . . . .3. . . .3.1 Maxwell’s equations . . F.2. .1 Relationship between the ﬁeld vectors in a plane wave F. . . . . . . Lorentz’ gauge condition in vacuum . . . . . . . . .2. . . . . . F. . . . . .6 Four-momentum . . .5 The far ﬁelds from an electric quadrupole . .6 The ﬁelds from a point charge in arbitrary motion . . . . . . . . . . . . . F. . . . . .2. . . . . . . . . . . F. . . . . . F. .4 Invariant line element . . . . F. . . . . . .2 Fields and potentials . . . . . . . . .3. . . . . . .5. . . . . . . F. . . . . . . Virtual photons . . . . .3. .

. The del operator . . . . . . . . . . . . . . . 168 Dyadic product . . . . . . . . Radius vector . . . . Vector ﬁelds . . . . . . . . . .3. . . F. . . . . . . . . . . . . . . . . . . The gradient . . . . . . . . . . . . . F. . . Tensor ﬁelds . . . . . . . . . . . Vectors and Tensors M. . . . M. . . . . . . . . . . . M. . . . .1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Integral relations . . . . . . . . . . . . . . . 168 Example M. . . . . norm and metric in Lorentz space . . . . Scalar ﬁelds . . . . . . . . . . . . . Example M. . . . . . . . . . . . . . . . . . Example M. . . . . . . . Example M. . . . . . . . . . . . . . . . . . .5 The four-del operator in Lorentz space . . . . . . . . . . . . . . . . . . . . . Vector product . .1. . . . . . . . . . Solid angle element . . . . . M. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4 Metric in general relativity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . General relations . . . . . . . . . . . . . . . . . . . . . . . .1.plasma. . . . . Special relations . . . . . . . . . . . . . . .1 Vectors . . . .4 Vector analysis . . . . . . . . . .1 Tensors in 3D space . . . . 153 153 154 154 154 154 154 154 154 154 156 157 157 148 159 159 159 159 161 161 161 162 164 167 167 167 . . . . . .4. . . . .3 Scalar product. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 Scalars. . . . Directed line element . . Vector Relations . . . . . . . . . . . .4 Appendices M Mathematical Methods M. . . . . . . .9 Field tensor . . . . . . . . . . . . . . . . . . . . . . 173 Draft version released 13th November 2000 at 22:01. . . . . Volume element . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .v F. . . . . . . . .2 Vector formulae . . . . . . . . . . . . . 172 The divergence . . . . Base vectors . . . . . . . . . . . . . Directed area element . . Example M. . . . . . . . . . . . . . .se/CED/Book . . .1. . . . .1 Spherical polar coordinates . . . . . . . . . . . .uu. . . .2 Fields . . . . . F. . . .6 Gradients of scalar functions of relative distances in 3D . . . . 169 170 170 170 171 172 Example M. . . . . . .4. . . . . . . . .2 Inner products in complex vector space . . . Scalar product . . . . . Bibliography . . Downloaded from http://www. .3 Vector algebra . . . .

. . . . . . . . . . Bibliography . . . . . . . .2 Analytical Mechanics . . Example M. . . . . . . . . . . . . . . . . . . . . . . . . . The Laplacian .8 The Laplacian and the Dirac delta . . . . .uu. . . . . .1 Lagrange’s equations . . . . . . . . . 173 173 173 174 174 175 176 176 176 177 Downloaded from http://www. . . . . . M. . . . . . . .9 The curl of a gradient . . . . . . . . . . . . Example M.plasma. . . . . . M. . . . . . . . . . . . . . . . . . . . . The curl . . .vi C ONTENTS Example M. . . . . . . . . Example M. . . . . . .7 Divergence in 3D . . . . . . . . . .2 Hamilton’s equations . .se/CED/Book Draft version released 13th November 2000 at 22:01. . . . .2. . . . . . . . . . . . . .2. . . . . . M. . .10 The divergence of a curl . . . . . .

. . . . . . .8 8. . . . . . . . . . . . . . . . . . . . . . Angular distribution of radiation during bremsstrahlung Location of radiation during bremsstrahlung . . . . . . . . . . .2 8. . . . . .1 8. . .3 8. . . . . . . 2 5 12 52 59 59 76 98 112 114 128 129 133 135 138 144 M. . . . . . . . . .1 Surface element of a material body . . .1 5.1 1. . . . . . . . Ampère interaction . . . Radiation from a moving charge in vacuum . . . . . Moving loop in a varying B ﬁeld . . Radiation from a charge in circular motion . Radiation in the far zone .List of Figures 1. . . . . . . . . . . . . . . . . . . . . . . . . Synchrotron radiation lobe width . .5 8. . . . . . . . . . . . . . . . . . . . . . . . .2 1. .3 6. . . Rotation in a 2D Euclidean space . . . . An accelerated charge in vacuum . . . . . Relative motion of two inertial systems . . . .6 8. . . .7 8. The perpendicular ﬁeld of a moving charge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 Tetrahedron-like volume element of matter . . . ˇ Vavilov-Cerenkov cone . . Linear one-dimensional mass chain . . . . . . . 164 M. . . . . . . . . . . . .3 5. . . . . . . . . . . . . . . . . . . .2 5. . . . . .9 Coulomb interaction . . . . . . . . . Minkowski diagram . . . . . . . . . . .1 8. . . . . . . . . . . . . . . . . . 165 vii .4 8. .

.

remarkable physicist and a truly great human.To the memory of L EV M IKHAILOVICH E RUKHIMOV dear friend. .

.

Shortly thereafter. I took my ﬁrst advanced course in electrodynamics at the Theoretical Physics department. professor P ER -O LOF F RÖ MAN . gauge transformations. already at the classical level. both in my studies. momentum and energy of the electromagnetic ﬁeld. radiation and scattering phenomena. force. with the preparation of a new version of his lecture notes on Electricity Theory. and includes such things as electrostatics and magnetostatics and their uniﬁcation into electrodynamics. Uppsala University. It provides a thorough treatment of the theory of electrodynamics. Ever since that time. and I fell in love with it. to become the ﬁve-credit course Classical Electrodynamics in 1997. and the current book is the result of my own teaching of a course in advanced electrodynamics at Uppsala University some twenty odd years after I experienced the ﬁrst encounter with this subject. by B ENGT L UNDBORG who created. mainly from a classical ﬁeld theoretical point of view. covariant formulation of classical electrodynamics. I joined the research group there and took on the task of helping my supervisor. in Swedish. I hope the book may be useful for research workers too. In 1972. research and teaching. developed and taught the earlier. and covariant Lagrangian/Hamiltonian ﬁeld theoretical methods for electromagnetic ﬁelds. To some extent. particles and interactions. In an attempt to encourage participation by other scientists and students in xi . two-credit course Electromagnetic Radiation at our faculty. the electromagnetic potentials. The book is the outgrowth of the lecture notes that I prepared for the four-credit course Electrodynamics that was introduced in the Uppsala University curriculum in 1992. electromagnetic waves and their propagation in vacuum and in media. 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. parts of these notes were based on lecture notes prepared. I have off and on had reason to return to electrodynamics.Preface This book is the result of a twenty-ﬁve year long love affair. Intended primarily as a textbook for physics students at the advanced undergraduate or beginning graduate level. These two things opened up my eyes for the beauty and intricacy of electrodynamics.

my daughter K AROLINA. I dedicate this book to my son M ATTIAS. I have not only received comments on it from fellow Internet physicists around the world. Uppsala Division. but also to C HRISTER WAHLBERG at Uppsala University for interesting discussions on electrodynamics in general and on this book in particular. Sweden November. my high-school physics teacher. I am particularly indebted to Academician professor V ITALIY L. comments and historical footnotes on electromagnetic radiation while cruising on the Volga river during our joint Russian-Swedish summer schools. Lindau. the book serves as a frequently used Internet resource. and who have participated in the teaching and commented on the material covered in the course and in this book. Finally. it was produced within a World-Wide Web (WWW) project. I am grateful not only to Per-Olof Fröman and Bengt Lundborg for providing the inspiration for my writing this book. and to my former graduate students M ATTIAS WALDENVIK and T OBIA C AROZZI as well as A NDERS E RIKSSON. . 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. 2000 B O T HIDÉ Downloaded from http://www. who not only taught me about the practical aspects of the of highpower radio wave transmitters and transmission lines. and to my fellow members of the C APELLA P EDAGOGICA U PSALIENSIS. This way it is my hope 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. By making an electronic version of the book freely down-loadable on the net. from WWW ‘hit’ statistics that at the time of writing this.xii P REFACE the authoring of this book. but know. Thanks are also due to my long-term space physics colleague H ELMUT KOPKA of the Max-Planck-Institut für Aeronomie. Germany. S TAFFAN RÖSBY. This turned out to be a rather successful move. all at the Swedish Institute of Space Physics.uu. G INZBURG for his many fascinating and very elucidating lectures.plasma. and to ensure its quality and scope to make it useful in higher university education anywhere in the world. Uppsala. but also about the more A delicate aspects of typesetting a book in TEX and LTEX.se/CED/Book Draft version released 13th November 2000 at 22:01.

This means that the concepts of localised 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. the limiting processes used will yield results which are correct on small as well as large macroscopic scales. classical electrodynamics. described by one system of coupled dynamic ﬁeld equations. However. Clearly. 1 .1.1 Classical Electrodynamics Classical electrodynamics deals with electric and magnetic ﬁelds and interactions caused by macroscopic distributions of electric charges and currents. this is in contradiction to electromagnetism on a truly microscopic scale. 1. In this Chapter we start with the force interactions in classical electrostatics and classical magnetostatics and introduce the static electric and magnetic ﬁelds and ﬁnd two uncoupled systems of equations for them.1.1.1 Coulomb’s law It has been found experimentally that in classical electrostatics the interaction between two stationary electrically charged bodies can be described in terms of a mechanical force.1 Electrostatics The theory that describes physical phenomena related to the interaction between stationary electric charges or charge distributions in space is called electrostatics. 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 in one theory. Let us consider the simple case described by Figure 1. 1. where charges and currents are known to be spatially extended objects.

1: Coulomb’s law describes how a static electric charge q. as deﬁned in Equation (1.1).se/CED/Book Draft version released 13th November 2000 at 22:01. 1. and the length |x − x | in metres (m). Let F denote the force acting on a charged particle with charge q located at x. In SI units.9979 × 108 m/s is the speed of light in vacuum.uu. the force F is measured in Newton (N). the charge in statcoulomb. According to Coulomb’s law this force is.2 The electrostatic ﬁeld Instead of describing the electrostatic interaction in terms of a “force action at a distance. ¢ ¡ 1 |x − x | (1. given by the expression F(x) = qq qq x − x =− 3 4πε0 |x − x | 4πε0 where we have used results from Example M. located at a point x relative to the origin O.6 on page 172. in vacuum. the charges q and q in Coulomb (C) [= Ampère-seconds (As)].8542 × 10−12 Farad per metre (F/m) is the vacuum permittivity and c ≈ 2. and length in centimetres (cm).1) (1.1. Since the Downloaded from http://www. from a net charge q on the test particle with a small electric net charge q. experiences an electrostatic force from a static electric charge q located at x . The constant ε0 = 107 /(4πc2 ) ≈ 8.2 C LASSICAL E LECTRODYNAMICS q x−x q x x O F IGURE 1.” it turns out that it is often more convenient to introduce the concept of a ﬁeld and to describe the electrostatic interaction in terms of a static vectorial electric ﬁeld Estat deﬁned by the limiting process Estat ≡ lim def q→0 F q where F is the electrostatic force. which we shall use throughout. In CGS units ε0 = 1/(4π) and the force is measured in dyne.plasma. due to the presence of a charge q located at x .2) .

1 In the preface to the ﬁrst edition of the ﬁrst volume of his book A Treatise on Electricity and Magnetism. in the last step. at xi . in his mind’s eye. James Clerk Maxwell describes this in the following. We emphasise that Equation (1.1 Using formulae (1.3) (1.4) 1 d3x (1. . we ﬁnd that the electrostatic ﬁeld Estat at the ﬁeld point x (also known as the observation point).se/CED/Book ¢ ¤ 1 x−x 3 d x =− 4πε0 − x |3 |x ρ(x ) ¡ ££ ££ Estat (x) = ∑ qi x − x i 4πε0 x − x 3 i ¢ ¡ 1 |x − x | (1. 3. 2 In Draft version released 13th November 2000 at 22:01. 2.uu. in which case ρ can be expressed in terms of one or more Dirac delta functions.” fact. This means that we can say that any net electric charge produces an electric ﬁeld in the space that surrounds it. but classically this nonlinearity is negligible. manner: [6] “For instance.5) above is valid for an arbitrary distribution of charges.1. we used formula Equation (M. ﬁrst published in 1873. the assumption of linearity of vacuum2 allows us to superimpose their individual E ﬁelds into a total E ﬁeld i If the discrete charges are small and numerous enough. . almost poetic. vacuum exhibits a quantum mechanical nonlinearity due to vacuum polarisation effects manifesting themselves in the momentary creation and annihilation of electron-positron pairs. Faraday.1 E LECTROSTATICS 3 purpose of the limiting process is to assure that the test charge q does not inﬂuence the ﬁeld. .plasma. respectively. including discrete charges. due to a ﬁeld-producing charge q at the source point x .5) |x − x | ¤ . . they were satisﬁed that they had found it in a power of action at a distance impressed on the electric ﬂuids. regardless of the existence of a second charge anywhere in this space. is given by Estat (x) = q x−x q =− 4πε0 |x − x |3 4πε0 In the presence of several ﬁeld producing discrete charges qi . we introduce the charge density ρ located at x and write the total ﬁeld as Estat (x) = 1 4πε0 ρ(x ) V V where. the expression for Estat does not depend explicitly on q but only on the charge q and the relative radius vector x − x . i = 1. Downloaded from http://www.1) and (1.2).68) on page 172. 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.

2. Equation (1. § ¨¢ ¢ ¢ ¦ ρ(x ) × ¡ V § ¨¢ ¦ ¡ ¡ ¤ × Estat (x) = − ¤ ¤ 1 4πε0 1 =− 4πε0 =0 × ρ(x ) ¡ ¤ ¤ ¤ ¥ 1 |x − x | 1 |x − x | d3x d3x (1.5) on the preceding page.78) on page 175. Taking the divergence of the general Estat expression for an arbitrary charge distribution. due to the presence of a small loop C carrying a current J Downloaded from http://www. and the interaction between these currents. much the same way that charges interact.7) . Let F denote such a force acting on a small loop C carrying a current J located at x. i. 1.. charges moving with constant speeds. we immediately ﬁnd that in electrostatics V I. 1. Equation (M. Estat is an irrotational ﬁeld.e. × [ α(x)] ≡ 0 for any 3D 3 scalar ﬁeld α(x).uu.e. magnetostatics deals with stationary currents.1 Ampère’s law Experiments on the interaction between two small current loops have shown that they interact via a mechanical force.73) on page 174..6) (1. according to formula Equation (M.4 C LASSICAL E LECTRODYNAMICS Since.2 Magnetostatics While electrostatics deals with static charges. we ﬁnd that · Estat (x) = 1 x−x 3 dx ρ(x ) 4πε0 V |x − x |3 1 1 ρ(x ) · =− d3x −x | 4πε0 V |x 1 1 ρ(x ) 2 =− d3x 4πε0 V |x − x | 1 = ρ(x )δ(x − x ) d3x ε0 V ρ(x) = ε0 · which is Gauss’s law in differential form.plasma.se/CED/Book Draft version released 13th November 2000 at 22:01. and using the representation of the Dirac delta function.

At ﬁrst glance. and. However.se/CED/Book dl × dl × ¡ ¦ © © © © F(x) = C C 1 |x − x | (1. According to Ampère’s law this force is. carrying a static electric current J through the tangential line element dl located at x .plasma. dl located at x .8) above appears to be unsymmetric in terms of the loops and therefore to be a force law which is in contradiction with Newton’s third law. Equation (1. carrying a static electric current J through its tangential line element dl located at x.1. in SI units.2: Ampère’s law describes how a small loop C. From the deﬁnition of ε0 and µ0 (in SI units) we observe that 107 1 (F/m) × 4π × 10−7 (H/m) = 2 (s2 /m2 ) 2 4πc c ε0 µ0 = which is a useful relation. in vacuum.2 M AGNETOSTATICS 5 J C dl x−x x C J x O F IGURE 1. by applying the vector triple product “bac-cab” Draft version released 13th November 2000 at 22:01. µ0 = 4π × 10−7 ≈ 1.9) Downloaded from http://www. respectively. The loops can have arbitrary shapes as long as they are simple and closed.uu. experiences a magnetostatic force from a small loop C . given by the expression µ0 JJ 4π µ0 JJ =− 4π dl × dl × (x − x ) |x − x |3 C C Here dl and dl are tangential line elements of the loops C and C .2566 × 10−6 H/m is the vacuum permeability. § ¢ (1.8) .

13).11) C C This clearly exhibits the expected symmetry in terms of loops C and C .plasma. With this deﬁnition of Bstat . we obtain Biot-Savart’s law: Bstat (x) = µ0 x−x 3 j(x ) × dx 4π V |x − x |3 1 µ0 d3x j(x ) × =− 4π V |x − x | Comparing Equation (1.se/CED/Book © F(x) = J C dl × Bstat (x) § ¨¢ ¢ ¡ ¦ © © © © © © ¤ F(x) = − µ0 JJ 4π µ0 JJ − 4π dl · ¡ 1 |x − x | dl (1.8) in the following way C C C C x−x dl · dl |x − x |3 Recognising the fact that the integrand in the ﬁrst integral is an exact differential so that this integral vanishes.12) ¤ (1.14) Draft version released 13th November 2000 at 22:01. The SI unit for the magnetic ﬁeld. Equation (1.12) to an integrated steady state current distribution j(x).5) on page 3 with Equation (1.8) on the previous page may we written Downloaded from http://www. 1. If we generalise expression (1.2. I turns out that Bstat can be deﬁned through dBstat (x) ≡ def µ0 J x−x dl × 4π |x − x |3 which expresses the small element dBstat (x) of the static magnetic ﬁeld set up at the ﬁeld point x by a small line element dl of stationary current J at the source point x . we can rewrite the force expression.10) (1. we see that there exists a close analogy between the expressions for Estat and Bstat but that they differ in their vectorial characteristics. we may attribute the magnetostatic interaction to a vectorial magnetic ﬁeld Bstat .2 The magnetostatic ﬁeld In analogy with the electrostatic case. in the following symmetric way F(x) = − µ0 JJ 4π x−x dl · dl |x − x |3 (1. Equation (1.6 C LASSICAL E LECTRODYNAMICS formula (F.13) (1. is Tesla (T).8) on the previous page.uu.54) on page 155. . sometimes called the magnetic ﬂux density or magnetic induction. we can rewrite (1.

obtained by applying Gauss’s theorem. we determine its divergence and curl. and integrate the second one by parts. and that the second integral vanishes because · j = 0 for Draft version released 13th November 2000 at 22:01.13) on the facing page and utilising formula (F. vanishes when integrated over a large sphere far away from the localised source j(x ).1.uu. Taking the divergence of both sides of Equation (1. by utilising formula (F. the curl of Equation (1. formula (F. and the second term vanishes since. Downloaded from http://www.59) on page 155 as follows: [j(x ) · = xk ˆ = xk ˆ V V S V Then we note that the ﬁrst integral in the result.2 M AGNETOSTATICS 7 In order to assess the properties of Bstat .17) ¢ ¡ ¤ § ¨¢ ¦ · j(x ) ∂ 1 ∂xk |x − x | 1 · dS − |x − x | d3x − · j(x ) ¡ ¡ V ¢ ] 1 |x − x | d3x 1 |x − x | d3x § ¢ ¢ ¢ ¢ ¢ ¡ ¡ ¢ ¡ ¡ ¡ ¦ ¡ ¤ ¤ (1.plasma.15) ¤ ¤ ¤ ¤ (1. according to Equation (M.78) on page 175. Applying the operator “bac-cab” rule. we use the representation of the Dirac delta function Equation (M. × [ α(x)] vanishes for any scalar ﬁeld α(x).se/CED/Book ¢ ¤ ¢ j(x ) ∂ ∂xk · j(x ) 1 |x − x | d3x (1. we obtain · Bstat (x) = − µ0 1 · j(x ) × d3x 4π |x − x | V µ0 1 =− · [ × j(x )]d3x 4π V |x − x | 1 µ0 j(x ) · + × d3x 4π V |x − x | =0 where the ﬁrst term vanishes because j(x ) is independent of x so that × j(x ) ≡ 0.13) on the preceding page can be written × Bstat (x) = − 1 µ0 × j(x ) × d3x 4π |x − x | V µ0 1 =− d3x j(x ) 2 4π V |x − x | µ0 1 d3x + [j(x ) · ] 4π V |x − x | In the ﬁrst of the two integrals on the right hand side.67) on page 155.73) on page 174.61) on page 155.16) ¡ ¡ ¤ ¤ ¤ .

as a consequence. .3 Electrodynamics As we saw in the previous sections. ” Downloaded from http://www. The net result is simply × Bstat (x) = µ0 j(x )δ(x − x ) d3x = µ0 j(x) (1.20a) (1. This uniﬁcation of the theories of electricity and magnetism is motivated by two empirically established facts: 1.8 C LASSICAL E LECTRODYNAMICS stationary currents (no charge accumulation in space). in Maxwell’s displacement current. the laws of electrostatics and magnetostatics can be summarised in two pairs of time-independent. simplicity and order. classical electrodynamics. This fact manifests itself in the equation of continuity and. 2. we must consider classical electrostatics and classical magnetostatics as two independent theories. namely the equations of classical electrostatics 3 ρ(x) ε0 stat × E (x) = 0 · Estat (x) = and the equations of classical magnetostatics · Bstat (x) = 0 stat ×B (x) = µ0 j(x) Since there is nothing a priori which connects Estat directly with Bstat .uu.20b) Draft version released 13th November 2000 at 22:01. .19b) (1. and connected with one another by the rules of strict logic. . Electric charge is a conserved quantity and current is a transport of electric charge. However. these theories are uniﬁed into one theory. when we include time-dependence.19a) (1. 3 The famous physicist and philosopher Pierre Duhem (1861–1916) once wrote: “The whole theory of electrostatics constitutes a group of abstract ideas and general propositions. A change in the magnetic ﬂux through a loop will induce an EMF electric ﬁeld in the loop.plasma.se/CED/Book ¤ (1. formulated in the clear and concise language of geometry and algebra.18) V 1. uncoupled vector differential equations. This is the celebrated Faraday’s law of induction. This whole fully satisﬁes the reason of a French physicist and his taste for clarity.

v) d3v where fα (t. x) = 0 (1. x) = −∂ρ(t. x ) ¡ 1 |x − x | d3x (1. ¢ × B(t. The electric charge conservation law can be formulated in the equation of continuity ∂ρ(t. x) ∂t ¤ V V (1. v) is the (normalised) distribution function for particle species α with electrical charge qα . x) + ∂ ε0 E(t. we would obtain the formal result V V = µ0 j(t. x. x)/∂t. x.23) (1. Doing so. x) = ∑α qα v fα (t. x) + · j(t. x )δ(x − x ) d3x + µ0 ∂ 4π ∂t ρ(t.3. In the case of nonstationary sources and ﬁelds. In general. in accordance with the continuity Equation (1. x) = µ0 j(t. x) ∂t where.24) Downloaded from http://www. j has to be deﬁned in statistical mechanical terms as j(t.1. set · j(t. we will need to consider this formal result further.21) ∂t which states that the time rate of change of electric charge ρ(t. x ) 1 ∂ 1 d3x = − ∂t 4πε0 |x − x | ∂ = E(t. we must. x) ∂t ρ(t. x) + µ0 ∂ ε0 E(t.uu.18) on the preceding page that there we used the fact that in magnetostatics · j(x) = 0. x ) 1 d3x |x − x | ¢ ¡ ¤ ¡ ¤ × B(t. x) = µ0 j(t.3.2 Maxwell’s displacement current We recall from the derivation of Equation (1. we have assumed that a generalisation of Equation (1.plasma.18) on the preceding page.21).5) on page 3 to time-varying ﬁelds allows us to make the identiﬁcation Later.1 Equation of continuity Let j denote the electric current density (A/m2 ). in the last step. The result is Maxwell’s source equation for the B ﬁeld Draft version released 13th November 2000 at 22:01.22) .se/CED/Book § ¢ ¦ ¢ ¡ ¤ 1 ∂ 4πε0 ∂t ρ(t. 1. x). x) is balanced by a divergence in the electric current density j(t.3 E LECTRODYNAMICS 9 1. In the simplest case it can be deﬁned as j = vρ where v is the velocity of the charge density. and formally repeating the steps in the derivation of Equation (1.

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. takes the form j = σ(Estat + EEMF ) (1. In the most general cases. We can view Ohm’s law.25) above.25) where σ is the electric conductivity (S/m). as the ﬁrst term in a Taylor expansion of the law j[E(t.26) Downloaded from http://www. which can be split up in “ordinary” conduction currents. in general. The rate at which this energy is expended is j · E per unit volume. unless the medium is superconducting. x) is applied to a conducting medium. σ is a tensor. x) (1. even in vacuum. this electric ﬁeld will exert work on the charges in the medium and. a current density j(t. This term was introduced.3. x)].se/CED/Book Draft version released 13th November 2000 at 22:01. we still have to use Ohm’s law with care. j will decay away with time.uu. x) will be produced in this medium. by Maxwell in order to make the right hand side of this equation divergence free when j(t. . The conductivity σ is. x) = σE(t. one can sometimes assume a linear relationship between the current density j and E. Examples of media which are highly non-linear are semiconductors and plasma. In the presence of such a ﬁeld EEMF . Stationary currents therefore require that an electric ﬁeld which corresponds to an electromotive force (EMF) is present. This general law incorporates non-linear effects such as frequency mixing. If E is irrotational (conservative). and for certain materials. polarisation currents and magnetisation currents. Ohm’s law. We draw the attention to the fact that even in cases when the linear relation between E and j is a good approximation.3 Electromotive force If an electric ﬁeld E(t.25) is valid for each individual Fourier component of the ﬁeld. Under certain physical conditions. There exist also hydrodynamical and chemical processes which can create currents. called Ohm’s law: j(t. The displacement current is an extra term which behaves like a current density ﬂowing in vacuum. 1. As we shall see later. in a stroke of genius. x) is assumed to represent the density of the total electric current.plasma.25) above.10 C LASSICAL E LECTRODYNAMICS where the last term ∂ε0 E(t. x). time-dependent (temporal dispersive media) but then it is often the case that Equation (1. If the current is caused by an applied electric ﬁeld E(t. for instance in an anisotropic conductor. Equation (1. there will be some energy loss. x)/∂t is the famous displacement current. Equation (1.

x) = − ∂ B(t.2 we derived the differential equations for the electrostatic ﬁeld.28) It has been established experimentally that a nonconservative EMF ﬁeld is produced in a closed circuit C if the magnetic ﬂux through this circuit varies with time. transforms it into the differential equation × E(t. This is formulated in Faraday’s law which.27) above vanishes and that this equation becomes E= EEMF · dl (1.se/CED/Book . In particular. Application of Stokes’ theorem on this integral equation. Let us therefore consider the case.3.uu.30) which is valid for arbitrary variations in the ﬁelds and constitutes the Maxwell equation which explicitly connects electricity with magnetism. on page 4 we derived Equation (1.4 Faraday’s law of induction In Subsection 1.3 E LECTRODYNAMICS 11 The electromotive force is deﬁned as where dl is a tangential line element of the closed loop C.27) C C S dS · ∂ B(t. This implies that the closed line integral of E stat in Equation (1.6) which states that × Estat (x) = 0 and thus that Estat is a conservative ﬁeld (it can be expressed as a gradient of a scalar ﬁeld). The convective derivative is evaluated according to the wellknown operator formula Draft version released 13th November 2000 at 22:01. ¤ ¤ © © © E= C (Estat + EEMF ) · dl (1. x) = E(t.3.4 on the following page.1. x) ∂t (1. x) ∂t (1. Any change of the magnetic ﬂux Φm will induce an EMF. that the “loop” is moved in such a way that it links a magnetic ﬁeld which varies during the movement. reads E(t. x) dt d =− B(t. in Maxwell’s generalised form.1. 1. illustrated if Figure 1. x) · dl = − d Φm (t.29) Downloaded from http://www.plasma. x) · dS = − dt S where Φm 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.

se/CED/Book ¤ ¤ ¤ E(t. x) = − (1. according to Equation (F. x(t)).12 C LASSICAL E LECTRODYNAMICS v dS B(x) v C dl B(x) F IGURE 1.uu.34) Downloaded from http://www. Equation (1. we obtain d ∂ B · dS = − dS · B − (v · )B · dS (1. × (B × v) = (v · )B (1.3: 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.33) Draft version released 13th November 2000 at 22:01.32) dt S ∂t S S During spatial differentiation v is to be considered as constant.31) dt ∂t which follows immediately from the rules of differentiation of an arbitrary differentiable function f (t. . Applying this rule to Faraday’s law.plasma.60) on page 155. and Equation (1.15) on page 7 holds also for time-varying ﬁelds: · B(t.29) on the previous page. d ∂ = +v· (1. x) = 0 (it is one of Maxwell’s equations) so that.

we can conclude that for a stationary observer.39) Hence.32) on the facing page in the following way: E(t. The use of Stokes’ theorem “backwards” on Equation (1. ¤ C ¤ C S S ∂ B · dS ∂t © ¤ ¤ =− S ∂ B · dS − ∂t ¤ C S B · dS © (1. in the moving system.1. rearranging the terms.35) × (B × v) · dS © C (1.. we ﬁnally get E(t. x) = EEMF · dl = − ∂ B · dS − ∂t (B × v) · dl (1.” i.se/CED/Book .36) or. the Maxwell equation ×E = − ∂ B ∂t (1.uu.42) is indeed valid even if the “loop” is moving.40) corresponding to an “effective” electric ﬁeld in the “loop” (moving observer) EEMF = E + (v × B) (1. (EEMF − v × B) · dl = − where EEMF is the ﬁeld which is induced in the “loop.37) above yields × (EEMF − v × B) = − ∂ B ∂t (1.3 E LECTRODYNAMICS 13 allowing us to rewrite Equation (1. x) = EEMF · dl = − d dt S With Stokes’ theorem applied to the last integral. a moving observer measures the following Lorentz force on a charge q qEEMF = qE + q(v × B) (1.e.plasma. an observer measures the electric ﬁeld E = EEMF − v × B (1.41) Hence.37) ¤ Downloaded from http://www.38) In the ﬁxed system. Draft version released 13th November 2000 at 22:01.

and j(t. x). However.e. i.3. they form a system of well-posed partial differential equations which completely determine E and B. .3. in an average sense. including both microscopic and macroscopic scales. E. 1. electric charge. These ﬁelds are the electric displacement D and the magnetising ﬁeld H.43d) ∂B = ∂t ·B = 0 ∂E × B − ε 0 µ0 = µ0 j(t. the material properties of the substances are already included.. x) represents the total.. nonlinear functionals of the primary ﬁelds E and B: D = D[t.e. these derived ﬁelds are complicated nonlocal. As they stand. possibly both time and space dependent.43a) (1. x) and B(t. B] (1. Another name often used for them is the Maxwell-Lorentz equations. electric current.plasma. possibly both time and space dependent.43b) (1. x. x) represents the total. for macroscopic substances it is sometimes convenient to introduce new derived ﬁelds which represent the electric and magnetic ﬁelds in which. The equations are ·E = ×E+ ρ(t. x. x) ∂t In these equations ρ(t. which relate ρ and j to the ﬁelds.se/CED/Book Draft version released 13th November 2000 at 22:01.5 Maxwell’s microscopic equations We are now able to collect the results from the above considerations and formulate the equations of classical electrodynamics valid for arbitrary variations in time and space of the coupled electric and magnetic ﬁelds E(t.43c) (1. conduction currents (motion of free charges) as well as all atomistic (polarisation. the equations therefore incorporate the classical interaction between all electric charges and currents in the system and are called Maxwell’s microscopic equations. free as well as induced (polarisation) charges. E. i.14 C LASSICAL E LECTRODYNAMICS 1.44a) (1. In the most general case. B] H = H[t. Together with the appropriate constitutive relations. magnetisation) currents.43) provide a correct classical picture for arbitrary ﬁeld and source distributions.6 Maxwell’s macroscopic equations The microscopic ﬁeld equations (1. x) ε0 (1.44b) Downloaded from http://www. and the initial and boundary conditions pertinent to the physical situation at hand.uu.

x)B −1 (1. Downloaded from http://www. With these new quantities included in the theory. x) ∂t and are called Maxwell’s macroscopic equations. the Maxwell equations can be written ·E = ×E+ ρe ε0 (1. symmetry. denoted jm = jm (t.45) (1.47c) (1.e. x). denoted ρm = ρm (t.47a) (1. x) by je . x) ∂B =0 ×E+ ∂t ·B = 0 ∂D ×H− = j(t.46) i.47d) 1.se/CED/Book .47b) (1. 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.48a) (1. x) by ρe and the electric current density j = j(t. The ﬁeld equations expressed in terms of the derived ﬁeld quantities D and H are · D = ρ(t.4 Electromagnetic Duality If we look more closely at the microscopic Maxwell equations (1.48d) ∂B = −µ0 jm ∂t · B = µ 0 ρm ∂E × B + ε 0 µ0 = µ0 je ∂t We shall call these equations the Dirac-Maxwell equations or the electromagnetodynamic equations Draft version released 13th November 2000 at 22:01.plasma. x)E H = µ (t.4 E LECTROMAGNETIC D UALITY 15 Under certain conditions..uu.48b) (1. for instance for very low ﬁeld strengths. (1.1. albeit not a complete. Let us for explicitness denote the electric charge density ρ = ρ(t. and a magnetic current density. we may assume that the response of a substance is linear so that D = ε(t. We further make the ad hoc assumption that there exist magnetic monopoles represented by a magnetic charge density.48c) (1. x).48). we see that they exhibit a certain.

51f) cρe → ρm cje → jm cB → −E ρm → −cρe jm → −cje which is a particular case (θ = π/2) of the general duality transformation (depicted by the Hodge star operator) E = E cos θ + cB sin θ c ρe = cρe cos θ + ρm sin θ c je = cje cos θ + jm sin θ ρm = −cρe sin θ + ρm cos θ jm = −cje sin θ + jm cos θ c B = −E sin θ + cB cos θ (1. which behaves as a mixing angle in a two-dimensional “charge space.52c) (1.52b) (1.48b).52d) (1.51e) (1.51c) (1. ρe a (true) scalar.21) on page 9. Using (1.52f) which leaves the Dirac-Maxwell equations.48c) to rewrite this relation.51b) (1.52e) (1. we ﬁnd that · ( × E) = − ∂ ( · B) − µ0 · jm ≡ 0 ∂t (1.49) where we used the fact that.51a) (1.52a) (1. invariant. B a pseudovector (axial vector).se/CED/Book Draft version released 13th November 2000 at 22:01. We notice that the new Equations (1.82) on page 175. then ρm and θ. Since E and j e are (true or polar) vectors.16 C LASSICAL E LECTRODYNAMICS Taking the divergence of (1. we obtain the equation of continuity for magnetic monopoles ∂ρm + · jm = 0 ∂t (1.51d) (1. . according to formula (M. Equation (1.uu. the divergence of a curl always vanishes. and hence the physics they describe (often referred to as electromagnetodynamics).50) which has the same form as that for the electric monopoles (electric charges) and currents. Downloaded from http://www.plasma.48) on the preceding page exhibit the following symmetry (recall that ε0 µ0 = 1/c2 ): E → cB (1.” must be pseudoscalars and jm a pseudovector.

56a) 1 1 2 2 (ρe cos θ + ρe sin θ) = ρe = cos θ cos θ ρm = −cρe sin θ + cρe tan θ cos θ = −cρe sin θ + cρe sin θ = 0 (1.54) QED 1.uu.se/CED/Book ! D UALITY OF THE ELECTROMAGNETODYNAMIC EQUATIONS E XAMPLE 1.53) (1.52) on page 16 (1.55b) E XAMPLE 1.52). Explicit application of the transformation yields ρe cos θ + cµ0 ρm sin θ ε0 1 1 ρe = ρe cos θ + ρm sin θ = ε0 c ε0 1 ∂ B ∂ × E+ − E sin θ + B cos θ = × (E cos θ + cB sin θ) + ∂t ∂t c 1 ∂E ∂B cos θ − µ0 jm cos θ + sin θ + cµ0 je sin θ =− ∂t c ∂t 1 ∂E ∂B − sin θ + cos θ = −µ0 jm cos θ + cµ0 je sin θ c ∂t ∂t = −µ0 (−cje sin θ + jm cos θ) = −µ0 jm · E= · (E cos θ + cB sin θ) = and analogously for the other two Dirac-Maxwell equations.2 Draft version released 13th November 2000 at 22:01. Equations (1. E ND OF EXAMPLE M AXWELL FROM D IRAC -M AXWELL EQUATIONS FOR A FIXED MIXING ANGLE Show that for a ﬁxed mixing angle θ such that ρm = cρe tan θ jm = cje tan θ the Dirac-Maxwell equations reduce to the usual Maxwell equations.55a) (1.plasma. or.1.1 .1 (1.56c) cos θ cos θ jm = −cje sin θ + cje tan θ cos θ = −cje sin θ + cje sin θ = 0 (1.4 E LECTROMAGNETIC D UALITY 17 Show that the symmetric. equivalently. We notice that the inverse of the transformation given by Equation (1.52) on the facing page yields 1 1 ρe = ρe cos θ + ρm sin θ = ρe cos θ + cρe tan θ sin θ c c (1. Explicit application of the ﬁxed mixing angle conditions on the duality transformation (1.56d) Hence. a ﬁxed ratio between the electric and magnetic charges/currents. “hides” the magnetic monopole inﬂuence (ρ m and jm ) on the dynamic equations.56b) 1 1 je = je cos θ + je tan θ sin θ = (je cos2 θ + je sin2 θ) = je (1. a ﬁxed mixing angle. electromagnetodynamic form of Maxwell’s equations (the Dirac-Maxwell equations).48) on page 15 are invariant under the duality transformation (1. Downloaded from http://www.

62) 3.61) QED OF EXAMPLE (1.63) Downloaded from http://www. 1. By varying the mixing angle θ we can change the fraction of magnetic monopoles at will without changing the laws of electrodynamics. x) + icB(t. T HE COMPLEX FIELD SIX . I..e.58) (1. Such particles are referred to as dyons. we ﬁnd that ρe 1 ρe · E= = (1. B ∈ 3 and hence F ∈ has a number of interesting properites: Draft version released 13th November 2000 at 22:01.uu. from the expressions for the transformed charges and currents above.VECTOR The complex ﬁeld six-vector E XAMPLE 1. x) (1. as long as we assume that this fraction is the same for all particles. x) = E(t. E ND The invariance of the Dirac-Maxwell equations under the similarity transformation means that the amount of magnetic monopole density ρm is irrelevant for the physics as long as the ratio ρm /ρe = tan θ is kept constant.59) ε0 cos θ ε0 and · B = µ0 ρm = 0 so that ·E = ρe 1 ρe cos θ − 0 = cos θ ε0 ε0 (1.3 F(t. So whether we assume that the particles are only electrically charged or have also a magnetic charge with a given.plasma.57) Furthermore.18 C LASSICAL E LECTRODYNAMICS yields E = E cos θ − c B sin θ This means that ·E = · E cos θ − c · B sin θ (1. ﬁxed ratio between the two types of charges is a matter of convention.2 . For θ = 0 we recover the usual Maxwell electrodynamics as we know it.60) and so on for the other equations.se/CED/Book # " where E. The inner product of F with itself F · F = (E + icB) · (E + icB) = E 2 − c2 B2 + 2icE · B is conserved. (1. ! 1.

64a) (1. the cross product of F itself vanishes: F × F = (E + icB) × (E + icB) (1. The inner product of F with the complex conjugate of itself F · F∗ = (E + icB) · (E − icB) = E 2 + c2 B2 is proportional to the electromagnetic ﬁeld energy. assuming that θ = θ(t.66) 4.1. we see that the spatial and temporal differenti- Draft version released 13th November 2000 at 22:01.se/CED/Book ! E XAMPLE 1.4 E LECTROMAGNETIC D UALITY 19 E 2 − c2 B2 = Const E · B = Const as we shall see later. 3. introduced in Example 1.uu.plasma.3 . As with any vector. the duality transformation Equations (1. 2. The cross product of F with the complex conjugate of itself F × F∗ = (E + icB) × (E − icB) = E × E + c2 B × B − ic(E × B) + ic(B × E) = 0 + 0 − ic(E × B) − ic(E × B) = −2ic(E × B) (1. EXAMPLE D UALITY EXPRESSED IN THE COMPLEX FIELD SIX .67) is proportional to the electromagnetic power ﬂux. $$ $$ F · F∗ = F = e−iθ F · eiθ F∗ = |F|2 (1.3 on the facing page.65) = 0 + 0 + ic(E × B) − ic(E × B) = 0 = E × E − c2 B × B + ic(E × B) + ic(B × E) (1.4 E ND OF 1.70) Downloaded from http://www.68) = E(cos θ − i sin θ) + icB(cos θ − i sin θ) = e−iθ (E + icB) = e−iθ F 2 while F · F = e2iθ F · F Furthermore. x).69) (1.VECTOR Expressed in the complex ﬁeld vector.52) on page 16 become F = E + ic B = E cos θ + cB sin θ − iE sin θ + icB cos θ from which it is easy to see that (1.64b) (1.

A. MA. L. S CHWINGER . . NY.71b) (1. London. E ND OF EXAMPLE Bibliography [1] R. Reading. Oxford . Downloaded from http://www.. J R . T SAIYANG . [4] J. ISBN 0-73820056-5. K. L ANDAU . G REINER.se/CED/Book Draft version released 13th November 2000 at 22:01. B. third ed.uu. 1982. Ltd. . ISBN 0-201-05702-6. H ALLÉN. NY . 1962. New York. Inc. Perseus Books.20 C LASSICAL E LECTRODYNAMICS ation of F leads to ∂ F ∂θ ∂F = −i e−iθ F + e−iθ ∂t ∂t ∂t ∂ · F ≡ · F = −ie−iθ θ · F + e−iθ · F ∂t F ≡ ∂× F ≡ (1. Dover Publications. ISBN 0-471-30932-X.. ISBN 0-08-025072-6. AND M.plasma. vol. Reading... Electromagnetic Theory. MA . Chapman & Hall. D E R AAD . [10] J. [9] J. [6] J.. Pergamon Press. . Inc. . The Classical Theory of Fields. Classical Electrodynamics. 1999. 1975. . M.4 . M C P HEDRAN. D. 1. Electromagnetic Processes in Dispersive Media. [3] E. New York. vol. S CHWINGER. . A magnetic model of matter. ISBN 0-52141025-8. AND W. P HILLIPS. AND E. ! 1. 1962. Inc. L IFSHITZ. Berlin. JACKSON. New York. Ltd. Heidelberg. ISBN 0-486-64290-9. ISBN 0-486-60636-8.. K. . . 1954. Classical Electrodynamics. C. second ed. [2] W. M AXWELL. [5] L. 1965. B ECKER.. ISBN 0-387-94799-X. Classical Electricity and Magnetism. .. Addison-Wesley Publishing Company. Springer-Verlag. [8] W. [7] D. and that and × F transform as F itself if θ is space-independent. 2 of Course of Theoretical Physics. M ILTON . . Wiley & Sons. AND R.71c) · F × F = −ie−iθ θ × F + e−iθ × F which means that ∂t F transforms as F itself if θ is time-independent. L. . Dover Publications. Inc. A Treatise on Electricity and Magnetism.. NY.71a) (1. D. M ELROSE . fourth revised English ed. Electromagnetic Fields and Interactions. C. Cambridge University Press. . 1991. Classical Electrodynamics.. 1996. third ed.. Cambridge . PANOFSKY. H. Science 165 (1969). New York.

1953. and Singapore.. S TRATTON. Toronto. Draft version released 13th November 2000 at 22:01. New York. NY and London. Classical Electromagnetic Theory. Inc. VANDERLINDE.4 B IBLIOGRAPHY 21 [11] J.plasma. A. ISBN 0-47157269-1. Inc.1. Electromagnetic Theory. New York.uu. McGraw-Hill Book Company. John Wiley & Sons. 1993. [12] J. Brisbane.. Downloaded from http://www.se/CED/Book . ISBN 07-062150-0. Chichester.

se/CED/Book Draft version released 13th November 2000 at 22:01.22 C LASSICAL E LECTRODYNAMICS Downloaded from http://www.uu.plasma. .

However. as we shall see is a homogeneous wave equation. which are usually written in the following form ρ(t. Maxwell’s microscopic equations (1. It turns out that these alternative equations are wave equations.43) on page 14.2 Electromagnetic Waves In this chapter we shall investigate the dynamical properties of the electromagnetic ﬁeld by deriving an set of equations which are alternatives to the Maxwell equations. and then discuss the implications of this equation. these four ﬁrst order. x) and current distributions j(t. x) + ε0 µ0 (2.1a) (2. indicating that electromagnetic waves are natural and common manifestations of electrodynamics.1b) (2. We shall derive the second order equation for E. second order partial equations. In particular. prescribed charge distributions ρ(t. coupled partial differential vector equations can be rewritten as two un-coupled.1d) can be viewed as an axiomatic basis for classical electrodynamics. as is well known from the theory of differential equations.1c) ∂E ∂t (2. 23 . x) of arbitrary time. these equations are well suited for calculating the electric and magnetic ﬁelds E and B from given. one for E and one for B. which.and space-dependent form. x) ε0 ∂B ×E = − ∂t ·B = 0 ·E = × B = µ0 j(t. We shall also show how the B ﬁeld can be easily calculated from the solution of the E equation.

¢ ¡ (2.5) ¡ (2.2) (2. ρ = 0. Ohm’s law.1b) and using (2. also using Equation (1. 2.24 E LECTROMAGNETIC WAVES 2. since ρ = 0.uu.8) ∂t ∂t Downloaded from http://www.1.2) can be rewritten 2 or. and no electromotive force EEMF = 0 and compare the results.67) on page 155 × ( × E) = ( · E) − 2 E Furthermore. to obtain × ( × E) = − ∂ ∂ ∂ ( × B) = −µ0 j + ε0 E ∂t ∂t ∂t According to the operator triple product “bac-cab” rule Equation (F.1a) on the previous page yields ·E = 0 and since EEMF = 0.1 The wave equation Let us derive the wave equation for the electric ﬁeld vector E and the magnetic ﬁeld vector B in a volume with no net charge. yields j = σE we ﬁnd that Equation (2.se/CED/Book ¢ E − µ0 ∂ ∂ σE + ε0 E = 0 ∂t ∂t Draft version released 13th November 2000 at 22:01.plasma. Equation (2.4) (2.9) on page 5 and rearranging.1d). 2 E − µ0 σ ∂ 1 ∂2 E− 2 2E = 0 ∂t c ∂t which is the homogeneous wave equation for E.1 The wave equation for E In order to derive the wave equation for E we take the curl of (2. Equation (1. 2.3) (2.26) on page 10.1d) and use Ohm’s law j = σE to obtain × ( × B) = µ0 × j + ε0 µ0 ∂ ∂ ( × E) = µ0 σ × E + ε0 µ0 ( × E) (2. Take the curl of (2.6) (2.7) .2 The wave equation for B The wave equation for B is derived in much the same way as the wave equation for E.1.

¢ ¢ E+ ω2 c2 1+i σ E= ε0 ω 2 E+ ω2 c2 ¡ ¡ 1+ i E=0 τω (2. it sufﬁces to consider only the E ﬁeld. We notice that it is of exactly the same form as the wave equation for the electric ﬁeld. We therefore make the following Fourier component Ansatz E = E0 (x)e−iωt and insert this into Equation (2. 2.1c) on page 23 can be rewritten ( · B) − 2 B = −µ0 σ ∂2 ∂B − ε0 µ0 2 B ∂t ∂t (2. Equation (2.2. since the results for the B ﬁeld follow trivially.se/CED/Book .1c). with the use of Equation (F.9) Using the fact that.14) Downloaded from http://www. This yields 2 (2.7) on the facing page.12) which we can rewrite as 2 The quantity τ = ε0 /σ is called the relaxation time of the medium in question. we can rewrite this equation as 2 · B = 0 for any medium and rearran(2.uu.13) tends to 2 E+ ω2 E=0 c2 Draft version released 13th November 2000 at 22:01. according to (2.plasma. As is clear from the above.1 T HE WAVE EQUATION 25 which.11) E − µ0 σ ∂ 1 ∂2 E0 (x)e−iωt − 2 2 E0 (x)e−iωt ∂t c ∂t 1 = 2 E − µ0 σ(−iω)E0 (x)e−iωt − 2 (−iω)2 E0 (x)e−iωt c 1 2 2 = E − µ0 σ(−iω)E − 2 (−iω) E = 0 c (2.10) B − µ0 σ ∂ 1 ∂2 B− 2 2B = 0 ∂t c ∂t This is the wave equation for the magnetic ﬁeld.3 The time-independent wave equation for E We now look for a solution to any of the wave equations in the form of a timeharmonic wave. In the limit of long τ.7) on the preceding page.13) (2.67) on page 155 and Equation (2. ging.1. Equation (2.

20d) Downloaded from http://www.18) Draft version released 13th November 2000 at 22:01.15) which is a time-independent diffusion equation for E.9) on page 5. Hence.20c) (2. using the fact that c = 1/ ε0 µ0 according to Equation (1.se/CED/Book % R0 = µ0 ≈ 376.7 Ω ε0 % σ σ 1 σ 1 = = = τω ε0 ω ε0 ck k µ0 σ = R0 ε0 k (2. representing weakly damped propagating waves. and SHF signals. we may say that the displacement current ε 0 ∂E/∂t is negligible relative to the conduction current j = σE. where in the last step we introduced the characteristic impedance for vacuum 2.19) and Maxwell’s equations attain the form ∂E ∂ζ ∂E n× ˆ ∂ζ ∂B n· ˆ ∂ζ ∂B n× ˆ ∂ζ n· ˆ =0 =− =0 = µ0 j(t.26 E LECTROMAGNETIC WAVES which is a time-independent wave equation for E.plasma. . If we introduce the vacuum wave number ω (2. For most metals τ ∼ 10−14 s. Then the del operator becomes ˆ =n ˆ ∂ ∂ζ (2. Alternatively.2 Plane waves Consider now the case where all ﬁelds depend only on the distance ζ to a given plane with unit normal n. In the short τ limit we have instead 2 E + iωµ0 σE = 0 (2.20b) (2.17) (2. x) + ε0 µ0 ∂E ∂E = µ0 σE + ε0 µ0 ∂t ∂t ∂B ∂t (2. the propagation term ∂2 E/c2 ∂t2 is negligible even for VHF. UHF. in metallic conductors.20a) (2.16) k= c √ we can write. which means that the diffusion picture is good for all frequencies lower than optical frequencies.uu.

2.2

P LANE

WAVES

27

Scalar multiplying (2.20d) by n, we ﬁnd that ˆ 0 = n· n× ˆ ˆ ∂B ∂ζ = n· µ0 σ + ε0 µ0 ˆ

which simpliﬁes to the ﬁrst-order ordinary differential equation for the normal component En of the electric ﬁeld dEn σ + En = 0 dt ε0 with the solution En = En0 e−σt/ε0 = En0 e−t/τ (2.23) (2.22)

This, together with (2.20a), shows that the longitudinal component of E, i.e., the component which is perpendicular to the plane surface is independent of ζ and has a time dependence which exhibits an exponential decay, with a decrement given by the relaxation time τ in the medium. Scalar multiplying (2.20b) by n, we similarly ﬁnd that ˆ 0 = n· n× ˆ ˆ or n· ˆ ∂B =0 ∂t

From this, and (2.20c), 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.

2.2.1 Telegrapher’s equation

In analogy with Equation (2.7) on page 24, we can easily derive the equation ∂2 E ∂E 1 ∂2 E − µ0 σ − =0 ∂ζ 2 ∂t c2 ∂t2 (2.26)

This equation, which describes the propagation of plane waves in a conducting medium, is called the telegrapher’s equation. If the medium is an insulator so that σ = 0, then the equation takes the form of the one-dimensional wave equation ∂2 E 1 ∂2 E − =0 ∂ζ 2 c2 ∂t2 (2.27)

Draft version released 13th November 2000 at 22:01.

¢

∂E ∂ζ

= − n· ˆ

∂B ∂t

¢

¡

¢

¡ ¡

∂ E ∂t

(2.21)

(2.24)

(2.25)

Downloaded from http://www.plasma.uu.se/CED/Book

28

E LECTROMAGNETIC WAVES

As is well known, each component of this equation has a solution which can be written Ei = f (ζ − ct) + g(ζ + ct), i = 1, 2, 3 (2.28)

where f and g are arbitrary (non-pathological) functions of their respective arguments. This general solution represents perturbations which propagate along ζ, where the f perturbation propagates in the positive ζ direction and the g perturbation propagates in the negative ζ 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{−iωt}, the solution of Equation (2.27) becomes E = E0 e−i(ωt±kζ) = E0 ei(±kζ−ωt) By introducing the wave vector k = kn = ˆ ωˆ ω n= k ˆ c c (2.30) (2.29)

this solution can be written as E = E0 ei(k·x−ωt) (2.31)

Let us consider the minus sign in the exponent in Equation (2.29) above. This corresponds to a wave which propagates in the direction of increasing ζ. Inserting this solution into Equation (2.20b) on page 26, gives n× ˆ ∂E = iωB = ik n× E ˆ ∂ζ (2.32)

or, solving for B, B= k 1 1ˆ √ n× E = k × E = k × E = ε0 µ0 n× E ˆ ˆ ω ω c (2.33)

Hence, to each transverse component of E, there exists an associated magnetic ﬁeld given by Equation (2.33) above. If E and/or B has a direction in space which is constant in time, we have a plane polarised wave (or linearly polarised wave).

Downloaded from http://www.plasma.uu.se/CED/Book

Draft version released 13th November 2000 at 22:01.

2.2

P LANE

WAVES

29

**2.2.2 Waves in conductive media
**

Assuming that our medium has a ﬁnite conductivity σ, and making the timeharmonic wave Ansatz in Equation (2.26) on page 27, we ﬁnd that the timeindependent telegrapher’s equation can be written ∂2 E ∂2 E + ε0 µ0 ω2 E + iµ0 σωE = 2 + K 2 E = 0 ∂ζ 2 ∂ζ where ω2 c2 (2.34)

where, in the last step, Equation (2.16) on page 26 was used to introduce the wave number k. Taking the square root of this expression, we obtain K=k 1+i σ = α + iβ ε0 ω (2.36)

Squaring, one ﬁnds that

or β2 = α2 − k2 αβ = k2 σ 2ε0 ω

Squaring the latter and combining with the former, one obtains the second order algebraic equation (in α2 ) α2 (α2 − k2 ) = k 4 σ2 4ε2 ω2 0 (2.40)

which can be easily solved and one ﬁnds that

α=k

2

β=k

2

Draft version released 13th November 2000 at 22:01.

0

)

1+

σ ε0 ω

0

)

1+

¢

k2 1 + i

σ ε0 ω

= (α2 − β2 ) + 2iαβ

σ ε0 ω

2

+1 (2.41a)

2

−1

Downloaded from http://www.plasma.uu.se/CED/Book

¢

¢

¢

K 2 = ε 0 µ 0 ω2 1 + i

σ ε0 ω

=

1+i

σ ε0 ω

= k2 1 + i

¡

¡

¡

σ ε0 ω

(2.35)

' &'' % (

% (

&''

'

%

¡

(2.37)

(2.38) (2.39)

(2.41b)

30

E LECTROMAGNETIC WAVES

As a consequence, the solution of the time-independent telegrapher’s equation, Equation (2.34) on the preceding page, can be written E = E0 e−βζ ei(αζ−ωt) (2.42)

With the aid of Equation (2.33) on page 28 we can calculate the associated magnetic ﬁeld, and ﬁnd that it is given by B= 1 ˆ 1 ˆ 1 ˆ K k × E = ( k × E)(α + iβ) = ( k × E) |A| eiγ ω ω ω (2.43)

where we have, in the last step, rewritten α + iβ in the amplitude-phase form |A| exp{iγ}. From the above, we immediately see that E, and consequently also B, is damped, and that E and B in the wave are out of phase. In the case that ε0 ω σ, we can approximate K as follows:

From this analysis we conclude that when the wave impinges perpendicularly upon the medium, the ﬁelds are given, inside this medium, by

Hence, both ﬁelds fall off by a factor 1/e at a distance δ= 2 µ0 σω (2.46)

This distance δ is called the skin depth.

**2.3 Observables and averages
**

In the above we have used complex notation quite extensively. This is for mathematical convenience only. For instance, in this notation differentiations are almost trivial to perform. However, every physical measurable quantity is

Downloaded from http://www.plasma.uu.se/CED/Book

%

B = (1 + i)

µ0 σ ( n× E ) ˆ 2ω

Draft version released 13th November 2000 at 22:01.

1¢

%

%

E = E0 exp −

µ0 σω ζ exp i 2

%

¡

%

√ = ε0 µ0 ω(1 + i)

σ = (1 + i) 2ε0 ω

µ0 σω 2

µ0 σω ζ − ωt 2

%

§0

)

¦

¢

σ K = k 1+i ε0 ω

¡

1 2

σ ε0 ω =k i 1−i ε0 ω σ

1 2

≈ k(1 + i)

σ 2ε0 ω

(2.44)

(2.45a) (2.45b)

.se/CED/Book 3 1 1 F · G ≡ Re {F} · Re {G} = Re F · G∗ = Re F∗ · G 2 2 3 F(t. is F(t. If so. 1999. i.. x) + F0 · G0 e−2iωt 2 Often in physics. Hence. Inc. Bibliography [1] J. x) = Re {F} · Re {G} = Re F0 (x) e−iωt · Re G0 (x) e−iωt 5 2 2 3 3 3 (2.50) ..49) 1 = Re F0 · G∗ + F0 · G0 e−2iωt 0 2 1 = Re F0 e−iωt · G∗ eiωt + F0 · G0 e−2iωt 0 2 1 = Re F(t. NY . representing physical observables. we measure temporal averages ( ) of our physical observables. Downloaded from http://www. x) · G∗ (t. x) = Re F0 (x) e−iωt · Re G0 (x) e−iωt 1 1 = [F0 (x) e−iωt + F∗ (x) eiωt ] · [G0 (x) e−iωt + G∗ (x) eiωt ] 0 0 2 2 1 = F0 · G∗ + F∗ · G0 + F0 · G0 e−2iωt + F∗ · G∗ e2iωt 0 0 0 0 4 (2.. . ISBN 0-471-30932-X. I. x) · G(t. JACKSON. third ed.2. x) · G(t. we can express the real part of the complex vector F as 1 Re {F} = Re F0 (x) e−iωt = [F0 (x) e−iωt + F∗ (x) eiωt ] 0 2 and similarly for G. letting ∗ denotes complex conjugate.47) 3 3 3 2 2 2 3 3 (2.3 B IBLIOGRAPHY 31 always real valued. the physically acceptable interpretation of the scalar product of two complex vectors. we see that the average of the product of the two physical quantities represented by F and G can be expressed as since the temporal average of the oscillating function exp{−2iωt} vanishes.plasma.” It is particularly important to remember this when one works with products of physical quantities. .e.48) 2 2 2 2 4 2 (2. Wiley & Sons. if we have two physical vectors F and G which both are time-harmonic.e. Draft version released 13th November 2000 at 22:01.uu. For instance. Classical Electrodynamics. then we must make the following interpretation Furthermore. “Ephysical = Re {Emathematical }. D. . New York. can be represented by Fourier components proportional to exp{−iωt}.

plasma. 1962.. AND M.32 E LECTROMAGNETIC WAVES [2] W. . Inc. . . Addison-Wesley Publishing Company. Reading. Classical Electricity and Magnetism. MA . PANOFSKY. Downloaded from http://www. ISBN 0-201-05702-6. P HILLIPS. second ed. H. K.se/CED/Book Draft version released 13th November 2000 at 22:01. .uu..

In this Chapter we will introduce and study the properties of such potentials.5) on page 3. it turns out that it is often more convenient to express the theory in terms of potentials. we get Estat (x) = − φstat (x) (3.3) above is called the electrostatic scalar potential. ¤ 33 . The scalar function φstat (x) in Equation (3.7) on page 4.6) on page 4. it may be expressed in terms of the gradient of a scalar ﬁeld. An example of such a quantity is a scalar constant. we obtain Poissons’ equation 2 stat φ (x) = − · Estat (x) = − ρ(x) ε0 (3. after the has been moved out of the integral.1 The electrostatic scalar potential As we saw in Equation (1.3) V where the integration is taken over all source points x at which the charge density ρ(x ) is non-zero and α is an arbitrary quantity which has a vanishing gradient. Hence. namely Equation (1.3 Electromagnetic Potentials Instead of expressing the laws of electrodynamics in terms of electric and magnetic ﬁelds.1) Taking the divergence of this and using Equation (1. If we denote this scalar ﬁeld by −φstat (x).2) A comparison with the deﬁnition of Estat . 3. the electrostatic ﬁeld Estat (x) is irrotational. shows that this equation has the solution φstat (x) = 1 4πε0 ρ(x ) 3 d x +α |x − x | (3.

43) on page 14. we may start from Biot-Savart’s law as expressed by Equation (1. We therefore realise that we can always write Bstat (x) = × Astat (x) (3. From Equation (1. x) = 0 remains valid. 3. In other words. i. the case with temporal and spatial dependent sources ρ(t. x). x) and A(t.e. We saw above that the electrostatic potential (as any scalar potential) is not unique: we may. A similar arbitrariness is true also for the magnetostatic vector potential.se/CED/Book ¢ ¡ ¤ ¤ ¤ (3. .uu.3 The electromagnetic scalar and vector potentials Let us now generalise the static analysis above to the electrodynamic case.5) ¤ Draft version released 13th November 2000 at 22:01.20) on page 8. let us study the electromagnetic potentials φ(t. x) and B(t.plasma. From Equation (M. In the magnetostatic case.15) on page 7 in magnetostatics we found that · Bstat (x) = 0. add to it a quantity whose spatial gradient vanishes. without changing the physics. x).13) on page 6 and “move the out of the integral:” Bstat (x) = x−x 3 µ0 dx j(x ) × 4π V |x − x |3 1 µ0 j(x ) × =− d3x 4π V |x − x | j(x ) 3 µ0 = × dx 4π V |x − x | µ0 4π j(x ) 3 d x + a(x) |x − x | An identiﬁcation of terms allows us to deﬁne the static vector potential as Astat (x) = (3.43c) on page 14 we note that also in electrodynamics the homogeneous equation · B(t. x).2 The magnetostatic vector potential Consider the equations of magnetostatics (1. and corresponding ﬁelds E(t..82) on page 175 we know that any 3D vector a has the property that · ( × a) ≡ 0 and in the derivation of Equation (1.34 E LECTROMAGNETIC P OTENTIALS 3.4) where Astat (x) is called the magnetostatic vector potential. x) and j(t. as described by Maxwell’s equations (1.6) V where a(x) is an arbitrary vector ﬁeld whose curl vanishes. From Equation (M.78) on page 175 we know that such a vector can always be written as the gradient of a scalar ﬁeld. Because of this Downloaded from http://www.

Hence. we obtain ∂ ∂ [ × A(t. Equation (3.7) into Maxwell’s equations (1. x) ∂ + ε0 ∂t ¡ ¢ × E(t. after some simple algebra and the use of Equation (1. However.8) As before we utilise the vanishing curl of a vector expression to write this vector expression as the gradient of a scalar function. x) ∂t ∂t or. x) + ∂ A(t.3 T HE ELECTROMAGNETIC SCALAR AND VECTOR POTENTIALS 35 divergence-free nature of the time.7) above. we can express it as the curl of an electromagnetic vector potential: B(t. On the other hand. x). ¢ A = µ0 j(t. rearranging the terms. we introduce the electromagnetic scalar potential function −φ(t. x).7) and Equation (3.3. or in terms of the ﬁelds E(t.9) becomes equivalent to ∂ A(t. it is a matter of taste whether we want to express the laws of electrodynamics in terms of the potentials φ(t. E(t. x) = − φ(t. x) + ¡ ∂ A(t.12b) Downloaded from http://www. Inserting (3. the general inhomogeneous wave equations E(t. 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. x) − 2 Draft version released 13th November 2000 at 22:01. x). x) (3.9) on page 5. in analogy with the electrostatic case. x) and A(t. × E(t. x) and B(t.and space-dependent magnetic ﬁeld.10) ∂t This means that in electrodynamics. Equation (1.11) and (3. x) = × A(t. x) (3.se/CED/Book . the treatment becomes signiﬁcantly simpler if we use the potentials in our calculations and then. x) = − (3. x) = − φ(t.11) above to calculate the ﬁelds or physical quantities expressed in the ﬁelds.uu. x)] = − × A(t. x) can be calculated from the formula E(t.43) on page 14 we obtain. use Equation (3.9) ·A+ 1 ∂ φ c2 ∂t 1 ∂ ·A+ 2 φ c ∂t (3.11) ∂t and B(t. x) − ¢ ¡ 1 ∂2 φ− c2 ∂t2 1 ∂2 A− c2 ∂t2 2 φ= ρ(t. If.7) Inserting this expression into the other homogeneous Maxwell equation.30) on page 11. x) from Equation (3. at the ﬁnal stage. x) = 0 ∂t (3.12a) (3. x) (3.plasma.

which represent eight scalar ﬁrst-order. Lorentz introduced ·A+ 1 ∂ φ=0 c2 ∂t (3.5 on page 171.14) the Lorentz equations for the electromagnetic potentials. partial differential equations. x) Draft version released 13th November 2000 at 22:01.se/CED/Book 6 6 6 (3. . x) we can consider this as a speciﬁcation of × A. because this choice simpliﬁes the system of coupled equations (3. we are free to choose · A in whatever way we like and still obtain the same physical results! This is somewhat analogous to the freedom of adding an arbitrary scalar constant (whose grad vanishes) to the potential energy in classical mechanics and still get the same force. coupled.2 on page 34 that in electrostatics and magnetostatics we have a certain mathematical degree of freedom.1 on page 33 and in Section 3. partial differential equations. x) ε0 where 2 is the d’Alembert operator discussed in Example M.plasma. Since this divergence does not enter the derivation above. 3. we must also specify · A. x) = B(t. We shall call (3. representing in all four scalar equations (one for φ and one each for the three components A 1 . if we write Equation (3. and A3 of A) are completely equivalent to the formulation of electrodynamics in terms of Maxwell’s equations. With a judicious choice of · A. Equations (3. But we know from Helmholtz’ theorem that in order to determine the (spatial behaviour) of A completely. the calculations can be simpliﬁed considerably. However.36 E LECTROMAGNETIC P OTENTIALS These two second order.uu.13) which is called the Lorentz gauge condition. Downloaded from http://www. Gauge transformations We saw in Section 3.3.14a) (3.1 Electromagnetic gauges Lorentz equations for the electromagnetic potentials As they stand.7) on the preceding page in the form × A(t.14b) 2 A = µ0 j(t.12) look complicated and may seem to be of limited use. A2 . coupled.12) on the previous page into the following set of uncoupled partial differential equations which we call the Lorentz inhomogeneous wave equations: 1 ∂2 φ− c2 ∂t2 2 def 1 ∂ 2 A ≡ 2 2A− c ∂t 2 φ ≡ def 2 φ= ρ(t.

A transformation of the potentials φ and A which leaves the ﬁelds. we obtain the transformed ﬁelds ∂ ∂ ∂ ∂ ∂ Γ− A+ Γ = − φ − A (3.plasma. ¢ A− 2 Γ = µ0 j(t. x) and A (t. x) according to the mapping scheme ∂ Γ(t.15). If we transform φ(t. invariant is called a gauge transformation. in particular. x) simultaneously into new ones φ (t. x) → A (t. If. the ﬁelds themselves are.15) above. x) ε0 (3. once again Equation (M. x) and A(t.12) on page 35. By deﬁnition. x) is an arbitrary.se/CED/Book . can be further gauge transformed according to (3.15) on the resulting Lorentz equations (3. these equations will be transformed into 2 We notice that if we require that the gauge function Γ(t.78) on page 175 was used.16a) E = − φ − A = − φ− ∂t ∂t ∂t ∂t ∂t B = ×A = ×A− × Γ = ×A (3.3. differentiable scalar function called the gauge function. gauge invariant. with an arbitrary choice of · A. x) → φ (t.3 T HE ELECTROMAGNETIC SCALAR AND VECTOR POTENTIALS 37 up to terms of vanishing gradients and curls. x) and A(t.16b) where. and apply the gauge transformation (3. to pick suitable forms for the potentials and still get the same physical result. x) ∂t A(t.13) on the facing page. and insert the transformed potentials into Equation (3. x) calculated from (3. x) − Γ(t. x) φ(t. The potentials φ(t. x) = φ(t. of course. Comparing these expressions with (3.11) and (3. Equation (3.7) on page 35 for the magnetic ﬁeld.17a) (3. and hence Maxwell’s equations. x) = A(t. A physical law which does not change under a gauge transformation is said to be gauge invariant. x) itself be restricted to fulﬁl the wave equation 1 ∂2 Γ− c2 ∂t2 2 Γ=0 Draft version released 13th November 2000 at 22:01. In fact.11) on page 35 for the electric ﬁeld and into Equation (3. x) and the vector potential A(t.15b) where Γ(t. x) + (3.15a) (3. we choose · A according to the Lorentz condition.14) on the preceding page.uu. x) are related to the physically observables gives leeway for similar “manipulation” of them also in electrodynamics. x) ¢ ¡ 1 ∂2 φ− c2 ∂t2 1 ∂2 A− c2 ∂t2 2 φ+ ∂ ∂t ¡ 1 ∂2 Γ− c2 ∂t2 1 ∂2 Γ− c2 ∂t2 2 Γ = ρ(t.18) Downloaded from http://www.17b) (3.7) we see that the ﬁelds are unaffected by the gauge transformation (3. the way the electromagnetic scalar potential φ(t.

· A = 0. We assume that our sources are well-behaved enough in time t so that the Fourier transform pair for the generic source function F −1 [ fω (x)] ≡ f (t.plasma. x) eiωt dt (3. .20a) (3. Each of these four scalar equations is an inhomogeneous wave equation of the following generic form: where Ψ is a shorthand for either φ or one of the vector components of A. • The temporal gauge. comprises the Lorentz gauge. The set of potentials which have been gauge transformed according to Equation (3.15) on the preceding page with a gauge function Γ(t.18) on the previous page. • The axial gauge. also known as the Hamilton gauge.e.uu. deﬁned by · A = 0.19) −∞ ∞ fω (x) e−iωt dω f (t..20b) −∞ Draft version released 13th November 2000 at 22:01. deﬁned by φ = 0. also known as the transverse gauge. x) which is restricted to fulﬁl Equation (3.14) are invariant. deﬁned by φ = 0. Other useful gauges are • The radiation gauge. x) = F [ f (t.3.38 E LECTROMAGNETIC P OTENTIALS these transformed Lorentz equations will keep their original form.se/CED/Book ¤ 1 2π ¤ 6 2 Ψ(t. deﬁned by A3 = 0. i. x) and A(t. x) = f (t.2 Solution of the Lorentz equations for the electromagnetic potentials As we see. x)] ≡ fω (x) = def def ∞ Downloaded from http://www. and f is the pertinent generic source component. x) represent a set of uncoupled equations involving four scalar equations (one equation for φ and one equation for each of the three components of A).14) on page 36 for φ(t. the Lorentz equations (3. 3. • The Coulomb gauge. those gauge transformed potentials for which the Lorentz equations (3. The process of choosing a particular gauge condition is referred to as gauge ﬁxing. x) (3.

x ) d3x (3. x) = Ψω (x) = ∞ Inserting the Fourier representations (3.26) Draft version released 13th November 2000 at 22:01.se/CED/Book . and that the same is true for the generic potential component: Ψ(t.3 T HE ELECTROMAGNETIC SCALAR AND VECTOR POTENTIALS 39 exists.21a) (3. Hence. x ) + k2G(x. each point on a wave front acts as a point source for spherical waves which form a new wave from a superposition of the individual waves from each of the point sources on the old wave front. ¤ 1 2π −∞ Ψω (x) e−iωt dω ∞ −∞ (3. depends only on x − x and there is no angular dependence in the equation. x) eiωt dt (3.23) (3.23) above which corresponds to the frequency ω is given by the superposition Ψω = fω (x )G(x.19) on the facing page turns into 2 Ψω (x) + k2 Ψω (x) = − fω (x) which is a 3D inhomogeneous time-independent wave equation. In Equation (3. As postulated by Huygen’s principle. often called the 3D inhomogeneous Helmholtz equation.24). x ) = −δ(x − x ) and the solution of Equation (3.uu.plasma.24) ¤ Downloaded from http://www.22) the generic 3D inhomogeneous wave equation Equation (3.19) on the preceding page. which represents the point source. and using the vacuum dispersion relation for electromagnetic waves ω = ck (3.3.23) can therefore be expressed as a superposition of solutions of an equation where the source term has been replaced by a point source: 2 G(x. the Dirac generalised function δ(x − x ). If we interpret r as the radial coordinate in a spherically polar coordinate system. The solution of (3. the “spherically symmetric” G(r) is given by the solution of d2 (rG) + k2 (rG) = −rδ(r) dr2 (3. the solution can only be dependent on r = |x − x |.25) The function G(x.20a) and (3.21a) into Equation (3. x ) is called the Green’s function or the propagator.21b) ¤ Ψ(t.

27) (3.28) where C ± are constants.31) 4π Insertion of the general solution Equation (3. In order to evaluate the constants C ± . 9 A0 exp −iω t + k|x−x | ω |x − x | 9 @0 exp −iω t − k|x−x | ω ¤ ) ) 8 8 ¤ 7¤ ¤ =− δ( x − x ) d x 3 dω d3x (3.se/CED/Book Draft version released 13th November 2000 at 22:01. away from the source point x .33) dω d3x ¤ 4 ££ ¢ C+ + C− 2 5 1 d3x + k2 C + + C − |x − x | ££ ££ ££ ¡ ¤ 7¤ ¤ ¤ 5 ££ ££ 1 d3x |x − x | 4 (3. we ﬁnd that the integrand in the ﬁrst integral can be written as −4πδ(|x − x |) and.40 E LECTROMAGNETIC P OTENTIALS Away from r = |x − x | = 0.uu. from Equation (M.25) on the previous page gives C+ + C− = Ψω (x) = C + fω (x ) In virtue of the fact that the volume element d3x in spherical polar coordinates is proportional to |x − x |2 . x) = C + fω (x ) + C− fω (x ) |x − x | Downloaded from http://www.32) |x − x | |x − x | The Fourier transform to ordinary t domain of this is obtained by inserting the above expression for Ψω (x) into Equation (3.plasma.30) .73) on page 174.28) above.e. Equation (3. hence.24) on the previous page can under this assumption be approximated by 1 (3. x−x → 0 G( x − x ) ∼ C + |x − x | |x − x | Equation (3.24) on the preceding page and integrate over a small volume around r = |x − x | = 0. this equation takes the form d2 (rG) + k2 (rG) = 0 dr2 with the well-known general solution G = C+ eikr e−ik|x−x | e−ikr eik|x−x | + C− + C− ≡ C+ r r |x − x | |x − x | (3.28) into Equation (3. we insert the general solution. Since 1 1 (3. Furthermore.. i. into Equation (3.29) + C− . that eik|x−x | 3 e−ik|x−x | 3 d x + C − fω (x ) dx (3. the second integral vanishes when |x − x | → 0.21a) on the preceding page: Ψ(t.

A.uu. FADEEV..20a) on page 38.22) on page 39]: k |x − x | |x − x | = t− ω c k |x − x | |x − x | = t+ tadv = tadv (t. x ) 3 dx |x − x | (3. Ben- Draft version released 13th November 2000 at 22:01. x ) 3 A(t. i.19) on page 38. according to Equation (3. if we assume that causality requires that the potential at (t.31) on the preceding page. x) = C + f (tret . according to Equation (3. we obtain Ψ(t. As they stand. Bibliography [1] L. x − x ) = t + ω c tret = tret (t.e. The retarded potentials From the above discussion on the solution of the inhomogeneous wave equation we conclude that under the assumption of causality the electromagnetic potentials in vacuum can be written ρ(tret . Gauge Fields: Introduction to Quantum Theory. at (tret . x − x ) = t − and use Equation (3. we shall use them frequently in the following. We note that the solution at time t at the ﬁeld point x is dependent on the behaviour at other times t of the source at x and that both retarded and advanced t are mathematically acceptable solutions.1 on page 36.34a) (3. we must in Equation (3. x) = (3. x) is set up by the source at an earlier time. x) = dx 4π |x − x | φ(t.3 B IBLIOGRAPHY 41 If we introduce the retarded time tret and the advanced time tadv in the following way [using the fact that in vacuum k/ω = 1/c. C + = 1/(4π). 50 in Frontiers in Physics: A Lecture Note and Reprint Series.3.35) This is a solution to the generic inhomogeneous wave equation for the potential components Equation (3. D. x ) 3 1 dx 4πε0 |x − x | µ0 j(tret . No.plasma.14) on page 36 they are valid in the Lorentz gauge but may be gauge transformed according to the scheme described in Subsection 3. x ) 3 d x + C− |x − x | f (tadv .se/CED/Book . x ). ¤ ££ ££ ¤ ¤ ¤ ££ ££ (3.36b) Since these retarded potentials were obtained as solutions to the Lorentz equations (3.35) set C − = 0 and therefore. AND A.34b) Downloaded from http://www.36a) (3. However. S LAVNOV.3.

. D. [2] M. Gauge Field Theories: An Introduction with Applications. . .42 E LECTROMAGNETIC P OTENTIALS jamin/Cummings Publishing Company. . K. Inc. Reading.. Inc. 1999. NY . . AND M.. . [4] W. . . New York. ISBN 0-471-63117-5.plasma.. 1980. MA . [3] J. Inc. ISBN 0-471-30932-X. . ISBN 0-201-05702-6. MA . Wiley & Sons. New York.uu.. Addison-Wesley Publishing Company. Downloaded from http://www. G UIDRY.. Classical Electricity and Magnetism. H.se/CED/Book Draft version released 13th November 2000 at 22:01. . Wiley & Sons. . Reading. 1991.. Classical Electrodynamics. third ed. 1962. NY . JACKSON. ISBN 08053-9016-2. . PANOFSKY. P HILLIPS. . second ed. Inc.

the existence of a single Fourier component assumes a monochromatic source (i. In this Chapter we will derive the electric and magnetic ﬁelds from the potentials.e. it is often physically more lucid to calculate them from the electromagnetic potentials derived in Chapter 3.1b) ¤ Ψ(t. in principle. or even from the wave equations in Chapter 2.4 The Electromagnetic Fields While. x) and current densities j(t. we can still use this approach even for sources and ﬁelds of ﬁnite duration. Strictly speaking. For charges and currents varying in time we can therefore. x). the electric and magnetic ﬁelds can be calculated from the Maxwell equations in Chapter 1. a source containing only one single frequency component). which in turn requires that the electric and magnetic ﬁelds exist for inﬁnitely long times. x) = Ψω (x) = ∞ That such transform pairs exists is true for most physical variables which are not strictly monotonically increasing and decreasing with time. without loss of generality. work with individual Fourier components.35) for the generic inhomogeneous wave equation (3.. We recall that in order to ﬁnd the solution (3.20a) on page 38 for the generic source term Ψ(t.1a) (4. by taking the proper limits. 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 ρ(t. x) eiωt dt 43 . However. deﬁned by the temporal Fourier transform ¤ 1 2π −∞ Ψω (x) e−iωt dω ∞ −∞ (4.19) on page 38 we presupposed the existence of a Fourier transform pair (3.

Similarly.44 E LECTROMAGNETIC P OTENTIALS pairs ρ(t. we must require that Aω = A∗ . .plasma. x) = φω (x) = ∞ where in the last step. x) = jω (x) = ∞ ∞ under the assumption that only retarded potentials produce physically acceptable solutions. we made use of the explicit expression for the temporal Fourier transform of the generic potential component Ψω (x).1 The temporal Fourier transform pair for the retarded vector potential can then be written φ(t. x) eiωt dt = (4. Similar transform pairs and requirements of real-valuedness exist for the ﬁelds themselves.2a) (4.4b) φ(t.2b) ¤ ¤ ¤ ¤ ρ(t.se/CED/Book ¤ ¤ 1 2π −∞ Aω (x) e−iωt dω ∞ −∞ A(t. x) = ρω (x) = and j(t. See also [1]. John A.4a) 1 4πε0 ρω (x ) eik|x−x | 3 dx |x − x | (4. the following Fourier transform pair for the vector potential must exist: A(t.3b) j(t.5b) (4.6) Draft version released 13th November 2000 at 22:01.32) on page 40. fact.3a) (4. Feynman derived in 1945 a fully self-consistent electrodynamics using both the retarded and the advanced potentials [6].5a) jω (x ) eik|x−x | 3 dx |x − x | (4. −ω φω = φ∗ −ω in order that all physical quantities be real. Wheeler and Richard P. 1 In Downloaded from http://www. x) eiωt dt = µ0 4π ¤ ¤ 1 2π −∞ ¤ 1 2π −∞ ¤ 1 2π −∞ ρω (x) e−iωt dω ∞ −∞ (4. x) = Aω (x) = ∞ Clearly. x) eiωt dt φω (x) e−iωt dω ∞ −∞ (4.uu. Equation (3. x) eiωt dt jω (x) e−iωt dω ∞ −∞ (4.

We note that in this context..7d) φ(t. deﬁned by Equation (4.7a) (4. Fourier transform back to ordinary t space.5a) and Equation (4.1 The magnetic ﬁeld Let us now compute the magnetic ﬁeld from the vector potential.uu. at the ﬁnal stage. we can make the formal identiﬁcation ρω = ρ0 δ(ω − ω0 ). we obtain Bω (x) = × Aω (x) = µ0 × 4π jω (x ) eik|x−x | 3 dx |x − x | (4.4. we can safely assume that all formulae derived for a general temporal Fourier representation of the source (general distribution of frequencies in the source) are valid for these simple limiting cases. x) (4. let ρ0 mean the same as the Fourier amplitude ρω and so on. We are working in the Lorentz gauge and note that in ω space the Lorentz condition.7c) (4.plasma. Using the Fourier transformed version of Equation (4. without any loss of stringence.1 T HE MAGNETIC FIELD 45 In the limit that the sources can be considered monochromatic containing only one single frequency ω0 .se/CED/Book . x) = j0 (x)e −iω0 t (4.13) on page 36.5b) on the preceding page. ¤ Downloaded from http://www.10) V Draft version released 13th November 2000 at 22:01. Equation (3.7) on page 35: B(t. x) = φ0 (x)e−iω0 t A(t. and formula (3. x) = ρ0 (x)e−iω0 t j(t.7b) (4.8) The calculations are much simpliﬁed if we work in ω space and. x) = A0 (x)e −iω0 t where again the real-valuedness of all these quantities is implied. and that we therefore.9) which provides a relation between (the Fourier transforms of) the vector and scalar potentials. takes the form k · Aω − i φω = 0 c (4. 4.8) and Equation (4. x) = × A(t. As discussed above. we have the much simpler expressions ρ(t.5b) on the preceding page. jω = j0 δ(ω − ω0 ) etc.

46 T HE E LECTROMAGNETIC F IELDS Using formula (F. the radiation ﬁeld or the far ﬁeld. F D GC ¤ µ0 + 4πc V ˙ ret .60) on page 155. the induction ﬁeld.uu.13) .plasma. x) = = ∞ −∞ Bω (x) e−iωt dω jω (x )ei(k|x−x |−ωt) dω × (x − x ) d3x |x − x |3 (−iω)jω (x )ei(k|x−x |−ωt) dω × (x − x ) |x − x |2 V + = µ0 4π V Induction ﬁeld Radiation ﬁeld The ﬁrst term. x ) × (x − x ) j(t d3x |x − x |2 F D EC V j(tret . we obtain the magnetic ﬁeld in the temporal (t) domain by taking the inverse Fourier transform (using the identity −ik = −iω/c): B(t. is the electrodynamic version of the BiotSavart law in electrostatics.se/CED/Book ¢ ˙ ret . dominates at large distances and represents energy that is transported out to inﬁnity. The second term. Note how the spatial derivatives ( ) gave rise to a time derivative (˙)! Downloaded from http://www.12) (4. we can rewrite this as From this expression for the magnetic ﬁeld in the frequency (ω) domain.11) ¤ B ¤ d3x (4. dominates near the current source but falls off rapidly with distance from it. x ) × (x − x ) 3 dx |x − x |3 ¤ 1 c ¤ ¤ B µ0 4π § ¤ ¤ ¦ = µ0 4π jω (x )eik|x−x | × (x − x ) 3 dx V |x − x |3 (−ik)jω(x )eik|x−x | × (x − x ) 3 dx + |x − x |2 V § ¢ § ¨¢ ¢ ¡ ¦ ¡ ¤ ¦ ¤ Bω (x) = − µ0 4π µ0 =− 4π eik|x−x | d3x |x − x | V x−x eik|x−x | d3x jω (x ) × − V |x − x |3 x − x ik|x−x | 1 + jω (x ) × ik e d3x |x − x | |x − x | V jω (x ) × ¡ (4. x ) def j(t ≡ ¡ where ∂j ∂t t=tret Draft version released 13th November 2000 at 22:01. formula (1.13) on page 6.

4.2

T HE

ELECTRIC FIELD

47

**4.2 The electric ﬁeld
**

In order to calculate the electric ﬁeld, we use the temporally Fourier transformed version of formula (3.11) on page 35, inserting Equations (4.4b) and (4.5b) as the explicit expressions for the Fourier transforms of φ and A: Eω (x) = − φω (x) + iωAω (x) 1 4πε0 1 = 4πε0 eik|x−x | 3 eik|x−x | 3 iµ0 ω jω (x ) dx + dx 4π V |x − x | |x − x | V (4.14) ρω (x )eik|x−x | (x − x ) 3 dx 3 V |x − x | ρω (x )(x − x ) jω (x ) eik|x−x | 3 − ik − dx c |x − x | |x − x | V ρω (x )

Using the Fourier transform of the continuity equation (1.21) on page 9 · jω (x ) − iωρω (x ) = 0 we see that we can express ρω in terms of jω as follows i · jω (x ) (4.16) ω Doing so in the last term of Equation (4.14), and also using the fact that k = ω/c, we can rewrite this Equation as ρω (x ) = − (4.15)

**The last integral can be further rewritten in the following way:
**

V

= But, since ∂ ∂xm

V

Draft version released 13th November 2000 at 22:01.

Downloaded from http://www.plasma.uu.se/CED/Book

¢

¡

¢

¢

jωm

xl − xl ik|x−x | e |x − x |2

=

xl − xl ik|x−x | e |x − x |2 xl − xl ik|x−x | ∂ + jωm e ∂xm |x − x |2 ∂ jωm ∂xm

¢

Iω =

· jω (x )](x − x ) eik|x−x | 3 − ikjω (x ) dx |x − x | |x − x | eik|x−x | 3 ∂ jωm xl − xl ˆ − ik jωl (x ) xl dx ∂xm |x − x | |x − x | [

D EC

V

Iω

§ F

¢

¢

¡

¤B

−

1 c

¡ ¤

¦

Eω (x) =

1 4πε0

V

ρω (x )eik|x−x | (x − x ) 3 dx |x − x |3 [ · jω (x )](x − x ) eik|x−x | 3 − ikjω (x ) dx |x − x | |x − x |

§

¢

¤

¡

¤

¤

=−

¤

¦

(4.17)

¡ ¡ ¤ ¡ ¤

(4.18)

(4.19)

48

T HE E LECTROMAGNETIC F IELDS

we can rewrite Iω as ∂ ∂xm xl − xl eik|x−x | 3 x eik|x−x | + ikjω ˆ dx 2 l |x − x | |x − x | xl − xl jωm xl eik|x−x | d3x ˆ |x − x |2

+

V

where, according to Gauss’s theorem, the last term vanishes if jω is assumed to be limited and tends to zero at large distances. Further evaluation of the derivative in the ﬁrst term makes it possible to write

V

V

Using the triple product “bac-cab” formula (F.54) on page 155 backwards, and inserting the resulting expression for Iω into Equation (4.17) on the preceding page, we arrive at the following ﬁnal expression for the Fourier transform of the total E-ﬁeld: 1 4πε0 1 = 4πε0 + + − iµ0 ω eik|x−x | 3 eik|x−x | 3 dx + dx jω (x ) 4π V |x − x | |x − x | V ρω (x )eik|x−x | (x − x ) 3 dx V |x − x |3 [jω (x )eik|x−x | · (x − x )](x − x ) 3 (4.22) dx V |x − x |4 [jω (x )eik|x−x | × (x − x )] × (x − x ) 3 dx |x − x |4 V [jω (x )eik|x−x | × (x − x )] × (x − x ) 3 dx |x − x |3 V ρω (x )

Taking the inverse Fourier transform of Equation (4.22) above, once again using the vacuum relation ω = kc, we ﬁnd, at last, the expression in time do-

Downloaded from http://www.plasma.uu.se/CED/Book

Draft version released 13th November 2000 at 22:01.

§

¤

ik c

¤

1 c

¤

1 c

¤

¤

Eω (x) = −

I

¤

¦

H ¤

− ik

−

jω · (x − x ) (x − x ) |x − x |3

e

ik|x−x |

eik|x−x | + jω |x − x |

dx

3

¢

¤

Iω = −

−jω

eik|x−x | 2 jω · (x − x ) (x − x )eik|x−x | + 2 |x − x | |x − x |4

¢

∂ ∂xm

¡

V

§

¢

¦

¤ ¡

¤

Iω = −

jωm

¡

(4.20)

d3x (4.21)

4.2

B IBLIOGRAPHY

49

**main for the total electric ﬁeld: E(t, x) = =
**

∞ −∞

Eω (x) e−iωt dω

V

Retarded Coulomb ﬁeld

+

Intermediate ﬁeld

+

Intermediate ﬁeld

V

+

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 (3.35) on page 41 in Chapter 3.

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., 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.

Draft version released 13th November 2000 at 22:01.

Downloaded from http://www.plasma.uu.se/CED/Book

F

D GC

¤

1 4πε0 c2

[˙ ret , x ) × (x − x )] × (x − x ) 3 j(t dx |x − x |3

F

D EC

¤

1 4πε0 c

V

[j(tret , x ) × (x − x )] × (x − x ) 3 dx |x − x |4

F

D GC

¤

1 4πε0 c

V

[j(tret , x ) · (x − x )](x − x ) 3 dx |x − x |4

F

D EC

¤

B

B

B

¤ B

1 4πε0

ρ(tret , x )(x − x ) 3 dx |x − x |3

(4.23)

50

T HE E LECTROMAGNETIC F IELDS

[4] W. K. H. PANOFSKY, AND M. P HILLIPS, Classical Electricity and Magnetism, second ed., Addison-Wesley Publishing Company, Inc., Reading, MA . . . , 1962, ISBN 0-201-05702-6. [5] J. A. S TRATTON, Electromagnetic Theory, McGraw-Hill Book Company, Inc., New York, NY and London, 1953, ISBN 07-062150-0. [6] J. A. W HEELER , AND R. P. F EYNMAN, Interaction with the absorber as a mechanism for radiation, Reviews of Modern Physics 17 (1945).

Downloaded from http://www.plasma.uu.se/CED/Book

Draft version released 13th November 2000 at 22:01.

to the introduction of a characteristic. and to ensure that our laws of physics be independent of any speciﬁc coordinate frame. 1905) The velocity of light is independent of the motion of the source. rectilinear motion relative to each other and is based on two postulates: Postulate 1 (Relativity principle. Postulate 2 (Einstein. To take this ﬁnite speed of propagation of information into account. an inertial system is a system in which free bodies move uniformly and do not experience any acceleration. or rigid coordinate system. in which the law of inertia (Galileo’s law. ﬁnite speed of √ propagation that equals the speed of light c = 1/ ε0 µ0 and which can be considered as a constant of nature..e. i. in the same way. 1905) All laws of physics (except the laws of gravitation) are independent of the uniform translational motion of the system on which they operate.1 The special theory of relativity An inertial system. This is the object of the current chapter. Poincaré. Newton’s ﬁrst law) holds. requires a treatment of electrodynamics in a relativistically covariant (coordinate independent) form.5 Relativistic Electrodynamics We saw in Chapter 3 how the derivation of the electrodynamic potentials led. A consequence of the ﬁrst postulate is that all geometrical objects (vectors. 51 . In other words. is a system of reference. in a most natural way. The special theory of relativity describes how physical processes are interrelated when observed in different inertial systems in uniform. or inertial reference frame. tensors) in an equation describing a physical process must transform in a covariant manner. 5.

For convenience. and t and (x . y.1. .uu.plasma. z) in Σ is represented by P(t . Downloaded from http://www. z). we shall make frequent use of these shorthand notations. y . y . x. y. At time t = t = 0 the origins O and O and the x and x axes of the two inertial systems coincide and at a later time t they have the relative location as depicted in Figure 5. x . x . An event represented by P(t. respectively. At time t. the two postulates of special relativity require that the spatial coordinates and times as measured by an observer in Σ and Σ . z ). z ) O z x z O x F IGURE 5.1 The Lorentz transformation Let us consider two three-dimensional inertial systems Σ and Σ in vacuum which are in rectilinear motion relative to each other in such a way that Σ moves with constant velocity v along the x axis of the Σ system. y .se/CED/Book P Draft version released 13th November 2000 at 22:01. the inertial system Σ has been translated a distance vt along the x axis in Σ. 5.1. y.52 R ELATIVISTIC E LECTRODYNAMICS vt y Σ Σ y v P(t. The times and the spatial coordinates as measured in the two systems are t and (x. In the following. z ) in Σ . x. At time t = t = 0 the origin O of Σ coincided with the origin O of Σ. z) P(t . As shown by Einstein.1: Two inertial systems Σ and Σ in relative motion with velocity v along the x = x axis. let us introduce the two quantities v c 1 β= γ= (5.2) where v = |v|.1) 1 − β2 (5.

x. z ) at time t in Σ in such a way that both observers conclude that the speed (spatial distance divided by time) of light in vacuum is c. x1 .6) Downloaded from http://www.1 T HE SPECIAL THEORY OF RELATIVITY 53 respectively. the speed of light in Σ and Σ is the same. are connected by the following transformation: ct = γ(ct − xβ) x = γ(x − vt) y =y z =z (5.2 Lorentz space Let us introduce an ordered quadruple of real numbers.4) (5.3d) Taking the difference between the square of (5. linear.1.5.5) which means that if a light wave is transmitted from the coinciding origins O and O at time t = t = 0 it will arrive at an observer at (x. enumerated with the help of upper indices µ = 0. Using this fact. A linear coordinate transformation which has this property is called a (homogeneous) Lorentz transformation. § ¨¢ ¢ v2 c2 2 2 = c t − x2 ¦ = 1 c2 t 2 1 − − x2 1 − 5 (5. 3. y. x3 ) = (ct.3b) (5. and the remaining components are the components of the ordinary 3 radius vector x deﬁned in Equation (M.3c) (5. 2. where the zeroth component is ct (c is the speed of light and t is time).uu. 1. Draft version released 13th November 2000 at 22:01. x) We then interpret this quadruple xµ as (the component form of) a radius fourvector in a real.3b) we ﬁnd that c2 t 2 − x 2 = γ2 c2 t2 − 2xcβt + x2 β2 − x2 + 2xvt − v2 t2 1− v2 c2 v2 c2 From Equations (5.1) on page 160: xµ = (x0 . four-dimensional vector space.4) above to c2 t 2 − x 2 − y2 − z2 = c2 t 2 − x 2 − y 2 − z 2 (5. z) ≡ (ct. x2 . y. Hence. 5. we ﬁnd that we can generalise the result in Equation (5.se/CED/Book ¡ ¡ 4 ¥ .3a) and the square of (5.3a) (5.3) we see that the y and z coordinates are unaffected by the translational motion of the inertial system Σ along the x axis of system Σ. z) at time t in Σ and an observer at (x . y .plasma.

9) QRRS . or signature. +..uu. +}. x3 ) = (ct. −. a space where a distance and a scalar product are deﬁned. x2 . with a main diagonal with sign sequence. Downloaded from http://www. We therefore need to deﬁne in this space a metric tensor. which we shall denote gµν and choose as (in matrix notation): 1 0 0 0 0 −1 0 0 0 0 −1 0 0 0 0 −1 i. −} seems to be the most commonly used one. +. In current physics literature. {+. as deﬁned in Equation (5. a vector in contravariant component form). one can alternatively choose a signature {−.e.8) This process is an example of contraction and is often called the “lowering” of index. −..54 R ELATIVISTIC E LECTRODYNAMICS Metric tensor We want our space to be a Riemannian space. The simple diagonal form of gµν .7) UU ET V (5. also known as the fundamental tensor.se/CED/Book Draft version released 13th November 2000 at 22:01. −. The latter has the advantage that the transition from 3D to 4D becomes smooth. Equation (5.e. is. more accurately. x).6) on the previous page.plasma. Index lowering of the contravariant radius four-vector x µ amounts to multiplying the column vector representation of xµ from the left by the matrix representation. while it will introduce some annoying minus signs in the theory. i. the signature {+. means that the index lowering operation in our chosen ﬂat 4D space is nearly trivial: x0 x1 x2 x3 1 0 0 0 0 −1 0 0 0 0 −1 0 0 0 0 −1 x0 x1 x2 x3 x0 −x1 −x2 −x3 1 Without changing the physics. −}. V QRRS V QRRS V QRRS = = UU ET UU ET UU ET V QRRS (gµν ) = UU ET (5. −. by deﬁnition. the prototype of a contravariant vector (or. of gµν to obtain the column vector representation of xµ .1 Radius four-vector in contravariant and covariant form The radius four-vector xµ = (x0 . The corresponding covariant vector xµ is obtained as (the upper index µ in xµ is summed over and is therefore a dummy index and may be replaced by another dummy index ν): xµ = gµν xν (5. Equation (5.7) above.7). x1 .

−z) = c2 t2 − x2 − y2 − z2 (5. −y. x1 . that this norm is conserved (invariant) during a Lorentz transformation. x.uu. x1 .11) and Equation (5.12b) δµ ν δν µ µ gµ ν gν µ = = (5. we ﬁnd. y.13) . the trace equals the space dimensionality. that the following relations between the contravariant.5. x3 ) · (x0 .se/CED/Book 5 4 ¥ W ¥ W (5. a mixed four-tensor of rank 2 which fulﬁls Draft version released 13th November 2000 at 22:01. z) · (ct. We see.11) which acts as a norm or distance in our 4D space.11) that our space has an indeﬁnite norm which means that we deal with a non-Euclidean space.. x2 . Scalar product and norm Taking the scalar product of xµ with itself. x3 ) = (ct. −x) · (ct.e. we get. covariant and mixed forms of the metric tensor hold: gµν = gνµ g = gµν gνκ g = g gκµ = νκ κµ µν (5. The 4 metric tensor Equation (5.12a) (5. x2 . the zeroth component is referred to as the time component. x) = (ct.1 T HE SPECIAL THEORY OF RELATIVITY 55 which we can describe as xµ = gµν xν = (ct. −x) (5. We notice further from Equation (5. the covariant radius four-vector xµ is obtained from the contravariant radius four-vector xµ simply by changing the sign of the last three components. δµ = δν = ν µ 0 if µ = ν 1 if µ = ν Downloaded from http://www. These components are referred to as the space components. gµν xν xµ = xµ xµ = (x0 .12d) Here we have introduced the 4D version of the Kronecker delta δν .10) i. linear 4D space with a positive deﬁnite norm which is conserved during ordinary rotations is a Euclidean vector space which we denote 4 . by deﬁnition.5) on page 53. after trivial algebra.12c) (5. The four-dimensional space (or space-time) with these properties is called Lorentz space and is denoted 4 .7) on the preceding page has a number of interesting properties: Firstly. as in any vector space with deﬁnite norm. The corresponding real.plasma. −x. we see that this tensor has a trace Tr gµν = −2 whereas in 4 . by comparing Equation (5. Secondly.

16) Since dτ measures the time when no spatial changes are present. As we see.56 R ELATIVISTIC E LECTRODYNAMICS Invariant line element and proper time The differential distance ds between the two points xµ and xµ + dxµ in 4 can be calculated from the Riemannian metric.18) Downloaded from http://www. it is called the proper time.14) Y ¡ ¡ ¢ ¡ X &'' ( P . given by the quadratic differential form ds2 = gµν dxν dxµ = dxµ dxµ = (dx0 )2 − (dx1 )2 − (dx2 )2 − (dx3 )2 where the metric tensor is as in Equation (5.uu. The square root of this expression is the invariant line element ds = c 1 1− 2 c 1− dx 1 dt 2 =c where we introduced dτ = dt/γ (5. but if dx2 + dy2 + dz2 > c2 dt2 ds is a space-like interval. whereas dx2 + dy2 + dz2 = c2 dt2 (5. If in some inertial system dx2 + dy2 + dz2 < c2 dt2 ds is a time-like interval.5) on page 53 in terms of the differential interval ds and comparing with Equation (5.se/CED/Book Draft version released 13th November 2000 at 22:01.17) is invariant during a Lorentz transformation. this form is indeﬁnite as expected for a non-Euclidian space.7) on page 54.plasma. % % =c v2 1 (v x )2 + (vy )2 + (vz )2 dt = c 1 − 2 dt 2 c c c 1 − β2 dt = dt = c dτ γ ¢ ¢ dx 2 + dt 2 dx 3 + dt 2 dt (5.20) (5.14). we may say that every coordinate transformation which preserves this differential interval is a Lorentz transformation. Conversely.19) (5.15) W (5. Expressing Equation (5. we ﬁnd that ds2 = c2 dt2 − dx2 − dy2 − dz2 (5.

as described by xκ = (ct.25) (5. A vector which has a light-like interval is called a null vector. can be written x µ = Λµν xν The inverse transform then takes the form xν = (Λµν )−1 x µ (5. −a(xκ )) (5.. i. b) = a0 b0 − a · b (5. The Lorentz transformation matrix Introducing the transformation matrix γ −βγ 0 0 −βγ γ 0 0 0 0 1 0 0 0 0 1 the linear Lorentz transformation (5. x.21) (5.3) on page 53. a) for a general contravariant four-vector ﬁeld in 4 and ﬁnd that the “lowering of index” rule. 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 µ does.1 T HE SPECIAL THEORY OF RELATIVITY 57 is a light-like interval.23) Downloaded from http://www. from one inertial system Σ to another inertial system Σ . an invariant scalar quantity α(x κ ) which depends on time and space. x1 . x3 )..22) which is a scalar ﬁeld. is called a four-vector. x2 .20) on page 163. In analogy with the notation for the radius four-vector we introduce the notation aµ = (a0 . i.5. −a) · (b0 . the coordinate transformation xµ → x µ = x µ (x0 . for such an arbitrary four-vector yields the dual covariant four-vector ﬁeld The scalar product between this four-vector ﬁeld and another one bµ (xκ ) is gµν aν (xκ )bµ (xκ ) = (a0 .e. The time-like. Equation (M.plasma.uu.se/CED/Book . space-like or light-like aspects of an interval ds is invariant under a Lorentz transformation. z).e. V QRRS Λµν = UU ET 5 4 W aµ (xκ ) = gµν aν (xκ ) = (a0 (xκ ). y. we may also say that in this case we are on the light cone.24) Draft version released 13th November 2000 at 22:01.

. x2 . and deﬁned by the speed parameters β1 and β2 .27b) (5.14) on page 56 transforms into dS 2 = (dX 0 )2 + (dX 1 )2 + (dX 2 )2 + (dX 3 )2 (5.25) on the previous page. The trajectory for an event as a function of time and space is called a world line. the world line for a light ray which propagates in vacuum is the trajectory x0 = x1 .1. we shall not make any further use of group theory. We shall call the 4D Euclidean space constructed in this way the Minkowski space 4 .uu.” Such a point is therefore called an event.27d) (5. x2 .se/CED/Book ` ¥ Draft version released 13th November 2000 at 22:01. 5.plasma. For instance. z) = (x1 . x1 .26) This means that the nonempty set of Lorentz transformations constitute a closed algebraic structure with a binary operation which is associative. y. In other words. Furthermore. However tempting. into a 4D differential form which is positive deﬁnite just as is ordinary 3D Euclidean space 3 .27e) dS = ids √ where i = −1. this set of Lorentz transformations constitute a mathematical group. one can show that this set possesses at least one identity element and at least one inverse element. . x3 ).27c) (5.58 R ELATIVISTIC E LECTRODYNAMICS The Lorentz group It is easy to show. x3 ) in 4D space-time is a way of saying that “something takes place at a certain time t = x0 /c and at a certain place (x. Downloaded from http://www. by means of direct algebra.28) i. that two successive Lorentz transformations of the type in Equation (5. respectively.3 Minkowski space Specifying a point xµ = (x0 . Introducing X 0 = ix0 = ict X =x X =x X =x 3 2 1 1 2 3 (5. correspond to a single transformation with speed parameter β= β1 + β2 1 + β1 β2 (5.27a) (5.e. we see that Equation (5.

uu. including O = O. X 0 = ict) to (x 1 .3: Minkowski diagram depicting geometrically the transformation (5. X 0 = ict ) corresponds to an ordinary rotation through an angle θ. is not simultaneous with O in the unprimed system but occurs there at time |P − P | /c. Draft version released 13th November 2000 at 22:01. This rotation 2 2 leaves the Euclidean distance x1 + X 0 = x2 − c2 t2 invariant.se/CED/Book . b a b a P ct x1 x1 = x Downloaded from http://www. which is also simultaneous with all points on the x axis.31) from the unprimed system to the primed system. w x0 = ct x0 x0 = x 1 ϕ ϕ O=O P F IGURE 5.2: Minkowski space can be considered an ordinary Euclidean space where a Lorentz transformation from (x1 . Here w denotes the world line for an event and the line x0 = x1 ⇔ x = ct the world line for a light ray in vacuum. Note that the event P is simultaneous with all points on the x1 axis (t = 0). including the origin O while the event P .5.1 T HE SPECIAL THEORY OF RELATIVITY 59 X0 X 0 θ x1 θ x1 F IGURE 5. to an observer at rest in the primed system.plasma.

it sufﬁces to consider the simpliﬁed case where the relative motion between Σ and Σ is along the x axes. at best.32a) (5. if we leave the ﬂat 4 space and enter the curved space of general relativity.se/CED/Book Draft version released 13th November 2000 at 22:01.27e) on page 58 backwards. and transform back to the original space and time variables by using Equation (5.3) on page 53 if we let sinh ϕ = γβ cosh ϕ = γ tanh ϕ = β (5. but is not very physical. If we introduce the angle ϕ = −iθ.26) on page 58 for successive Lorentz transformation then corresponds to the tanh addition formula tanh(ϕ1 + ϕ2 ) = tanh ϕ1 + tanh ϕ2 1 + tanh ϕ1 tanh ϕ2 The use of ict and 4 . Let us therefore immediately return to 4 where all components are real valued.32b) (5. Then dS 2 = (dX 0 )2 + (dx1 )2 (5.32c) It is therefore possible to envisage the Lorentz transformation as an “ordinary” rotation in the 4D Euclidean space 4 This rotation i 4 corrsponds to a coordinate change in 4 as depicted in Figure 5. Besides. Downloaded from http://www. be illustrative. Equation (5.plasma.uu.30a) (5. ` W ` W ` W (5.31b) which are identical to the transformation equations (5.60 R ELATIVISTIC E LECTRODYNAMICS As before. which leads to the interpretation of the Lorentz transformation as an “ordinary” rotation.3 on the preceding page. often called the rapidity or the Lorentz boost parameter. every interval is invariant under a rotation of the X 0 x1 plane through an angle θ into X 0 x 1 : X 0 = −x1 sin θ + X 0 cos θ x = x cos θ + X sin θ 1 1 0 (5. we obtain ct = −x sinh ϕ + ct cosh ϕ x = x cosh ϕ − ct sinh ϕ (5. As in all Euclidean spaces.31a) (5.29) and we consider X 0 and x1 as orthogonal axes in an Euclidean space.30b) See Figure 5. the “ict” trick will turn out to be an impasse. may.2 on the previous page.33) .

35) (5.plasma.5.38) (5.34) (5. we obtain pµ pµ = gµν pν pµ = (p0 )2 − (p1 )2 − (p2 )2 − (p3 )2 = (E. 2 V c SQ 0 ) uµ = dx µ v = γ 1. v) = ds m0 cv T c .40) Downloaded from http://www.36) i.e. particularly in the frame where p = 0 where we thus have E = m 0 c2 which is probably the most well-known physics formula ever.41) Draft version released 13th November 2000 at 22:01.se/CED/Book . this equation holds in any inertial frame. (5. cp) Scalar multiplying this four-momentum with itself..37) (5. = ds c 1 v T W The measure ds of the differential “distance” in four-velocity 4 allows us to deﬁne the (5. 1 − v2 c 2 V c SQ pµ = m0 c2 dx µ = γm0 c (c. −cp) · (E. The zeroth (time) component of the four-momentum pµ is given by We interpret this as the energy E. ¢ ¡ c p0 = γm0 c2 = c m = γm0 = 1 − v2 c 2 m0 c2 1− v2 c2 = mc2 v2 1− 2 c = (m0 c ) 2 2 c . that Lorentz covariance implies that the mass-like term in the ordinary 3D linear momentum is not invariant.uu. we can write pµ = (E. i.e..2 C OVARIANT CLASSICAL MECHANICS 61 5. when multiplied with the scalar invariant m0 c2 yields the four-momentum From this we see that we can write p = mv where m0 (5.39) (5.2 Covariant classical mechanics 1 − v2 c 1 − v2 c c m0 c2 1− v2 c2 2 which. cp) = E 2 − c2 |p|2 = (m0 c2 )2 v 1 − c2 2 Since this is an invariant.

the d’Alembert operator is invariant and.42) (5.plasma. can be written ∂µ Aµ ≡ (5. 5.47) and the Lorentz gauge condition. Equation (1. As is shown in Example M.45) where φ is the scalar potential and A the vector potential.46) Draft version released 13th November 2000 at 22:01.14) on page 36 in the following compact (and covariant) way: 2 µ A = jµ ε0 With the help of the above. cA) (5.se/CED/Book 0 2 ) j µ = ρ0 u µ = ρ0 dxµ v = ρ.21) on page 9 is ∂µ jµ ≡ ∂ jµ =0 ∂xµ ∂Aµ =0 ∂xµ (5. we can write the inhomogeneous wave equations (Lorentz equations) Equation (3. the d’Alembert operator is the scalar product of the four-del with itself: 2 = ∂µ ∂µ = ∂µ ∂µ = 1 ∂2 − c2 ∂t2 Since it has the characteristics of a four-scalar.13) on page 36.43) 6 6 (5.62 R ELATIVISTIC E LECTRODYNAMICS 5.3. we can formulate our electrodynamic equations covariantly. the covariant form of the equation of continuity.1 The four-potential If we introduce the four-potential Aµ = (φ.3 on page 34. For instance.5 on page 171. Equation (3. deﬁned in Section 3.48) Downloaded from http://www. the homogeneous wave equation is Lorentz covariant. The four-vector (in contravariant component form) where we introduced ρ = γρ0 is called the four-current.3 Covariant classical electrodynamics In the rest inertial system the charge density is ρ0 .44) (5. ρ ds c (5.uu. . hence.

We can therefore conclude that the electron charge is a universal constant.50) ¡ ££ 5 (5.e.. x3 ).43) on the facing page we see that ρdV = ρ0 dV0 (5. x1 . a 3D spatial volume transforms according to 1 dV = d3x = dV0 = dV0 γ then from Equation (5. we obtain Rµ Rµ = (c(t − t ). evaluated in the rest system (signiﬁed by the index “0”). x 1 .5. x 3 ). x2 .plasma.14) on page 36 in vacuum for the case of a well-localised charge q at a source point deﬁned by the radius four-vector x µ = (x 0 = ct .2 The Liénard-Wiechert potentials Let us now solve the Lorentz equation (the inhomogeneous wave equation) (3.3 C OVARIANT CLASSICAL ELECTRODYNAMICS 63 The gauge transformations (3. due to relative motion along the x direction). In the rest system we know that the solution is simply Aµ 0 = where |x − x |0 is the usual distance from the source point to the ﬁeld point.53) We know that in vacuum the signal (ﬁeld) from the charge q at x µ propagates to xµ with the speed of light c so that x − x = c(t − t ) Draft version released 13th November 2000 at 22:01. −(x − x )) = c2 (t − t )2 − x − x (5.55) . x − x ) · (c(t − t ).uu.51) i. The ﬁeld (observation) point is denoted by the radius four-vector x µ = (x0 = ct. the charge in a given volume is conserved.54) ££ (5.52) 4 2 (5. Let us introduce the relative radius four-vector between the source point and the ﬁeld point: Rµ = xµ − x µ = (c(t − t ).se/CED/Book ££ ££ ¢ q 1 . Downloaded from http://www.0 4πε0 |x − x |0 % P 1 − β2 = dV0 1− v2 c2 (5. x 2 .49) If only one dimension Lorentz contracts (for instance. x − x ) Scalar multiplying this relative four-vector with itself.3. 5.15) on page 37 in covariant form are simply A µ = Aµ − ∂µ [cΓ(xν )] = Aµ − ∂ cΓ(xν ) ∂xµ (5.

we see that Rµ Rµ = 0 or that Equation (5. valid in any inertial system.56) Now we want to ﬁnd the correspondence to the rest system solution. According to Equation (5. If we evaluate it in the rest system the result is: = (1. in the rest system.se/CED/Book ¢ =γ ££ 0 uν Rν = γ 1.34) on page 61 that in the rest system 0 1− v2 c2 c and (Rµ )0 = ( x − x . −(x − x )) c (x − x ) · v x−x − c Draft version released 13th November 2000 at 22:01.uu. −(x − x )0 ) = x − x We therefore see that the expression Aµ = (5.34) on page 61 and Equation (5. v 1 − c2 2 V uµ 1 v T ££ ££ ) ££ 4 ££ (5.53) on the previous page can be written Rµ = ( x − x . ££ ££ ££ ££ ££ ££ uµ Rµ = uµ Rµ 0 = (uµ )0 (Rµ )0 ££ ££ c ££ 5 ££ ¡ c SQ = .60) 0 (5. x − x ) (5. x − x )0 = ( x − x Like all scalar products. so we can evaluate it in any inertial system. uµ Rµ is invariant. (x − x )0 ) (5.58) 4 0 . It is therefore the correct solution.54) on the preceding page. in an arbitrary inertial system.plasma.61) subject to the condition Rµ Rµ = 0 has the proper transformation properties (proper tensor form) and reduces. to the solution Equation (5. We note from Equation (5.52) on the preceding page.57) Introducing s = x−x − (x − x ) · v c Downloaded from http://www.63) . 0) (v=0) (5.52) on the preceding page. 0) · ( x − x uµ q 4πε0 uν Rν 0 . Equation (5.57) ££ 5 = (1.64 R ELATIVISTIC E LECTRODYNAMICS Inserting this into Equation (5.62) (5. v · ( x − x .59) (5.

s cs = (φ.64) from which we see that the solution (5. the scalar and vector potentials are given by the expressions q 1 q 1 = 4πε0 s 4πε0 |x − x | − (x−x )·v c v v q q A(t. (5.3. ¢ ¡ ¢ 1 v . i.68) + (a1 b2 − a2 b1 ) x3 ˆ + (a3 b1 − a1 b3 ) x2 ˆ = (a2 b3 − a3 b2 ) x1 ˆ We notice that the kth component of the vector c can be represented as ck = ai b j − a j bi = ci j = −c ji .67a) (5. was used.se/CED/Book . x) = = 4πε0 c2 s 4πε0 c2 |x − x | − (x−x )·v c φ(t.61) can be written Aµ (xκ ) = q 4πε0 where in the last step the deﬁnition of the four-potential.66) Downloaded from http://www. cA) ¡ 1 v .3 C OVARIANT CLASSICAL ELECTRODYNAMICS 65 we can write uν Rν = γs and uµ = uν Rν (5.69) In other words. moving relative an observer with a velocity v.plasma. x) = These potentials are called the Liénard-Wiechert potentials. Equation (5. j = k (5.67b) 5.65) (5.3 The electromagnetic ﬁeld tensor Consider a vectorial (cross) product c between two ordinary vectors a and b: c = a×b = ˆ i jk ai b j xk (5.uu. s cs (5. Writing the solution in the ordinary 3D-way.5.45) on page 62. the pseudovector c = a × b can be considered as an antisymmetric tensor of rank two! Draft version released 13th November 2000 at 22:01. we conclude that for a very localised charge volume.

45) on page 62: Aµ = (φ.72a) ∂A ∂t (5. In matrix representation. Equation (5. the Maxwell equation ×E = − ∂B ∂t (5. this can be written Bi j = ∂A j ∂Ai − = ∂i A j − ∂ j Ai ∂xi ∂x j ∂φ ∂Ai Ei = − i − = −∂i φ − ∂t Ai ∂x ∂t (5.74) and Equation (5. covariant four-tensor of rank 2 is called the electromagnetic ﬁeld tensor.66 R ELATIVISTIC E LECTRODYNAMICS The same is true for the curl operator ×.72b) E = − φ− In component form.76) .plasma.se/CED/Book Draft version released 13th November 2000 at 22:01.73b) From this. For instance. 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.71) We know from Chapter 3 that the ﬁelds can be derived from the electromagnetic potentials in the following way: B= ×A (5.74) Inspection of (5.75) This anti-symmetric (skew-symmetric).uu. we notice the clear difference between the axial vector (pseudovector) B and the polar vector (“ordinary vector”) E.70) can in this tensor notation be written ∂Bi j ∂E j ∂Ei − j =− i ∂x ∂x ∂t (5. V QRRS Fµν = ∂µ Aν − ∂ν Aµ = UU ET (5.73) above makes it natural to deﬁne the covariant four-tensor Fµν = ∂Aν ∂Aµ − = ∂µ Aν − ∂ν Aµ ∂xµ ∂xν (5. the covariant ﬁeld tensor can be written 0 Ex Ey Ez −E x 0 −cBz cBy −Ey cBz 0 −cB x −Ez −cBy cB x 0 Downloaded from http://www. −cA) (5.73a) (5.

A HARONI. Equation (5. Draft version released 13th November 2000 at 22:01. ISBN 0-486-64870-2.5. Bibliography [1] J. New York. One can show that the two Maxwell source equations ·E = ρ ε0 (5. ¢ ¢ × B = µ 0 j + ε0 ∂E ∂t V ¡ QRRS F µν = ∂µ Aν − ∂ν Aµ = UU ET (5..83) ∂B ∂t (5.81) (5.se/CED/Book . 1980. Inc. 1985.83) above constitute Maxwell’s equations in four-dimensional formalism. Dover Publications.. Dynamics and Classical Theory of Fields and Particles. second. The Special Theory of Relativity.79) Downloaded from http://www. New York. O.77) ¡ (5. ISBN 0-486-64038-8. revised ed. NY. Inc. [2] A.plasma. Dover Publications.80) and Equation (5.80) Hence. BARUT.3 B IBLIOGRAPHY 67 The matrix representation for the contravariant ﬁeld tensor is 0 −E x −Ey −Ez Ex 0 −cBz cBy Ey cBz 0 −cB x Ez −cBy cB x 0 It is perhaps interesting to note that the ﬁeld tensor is a sort of four-dimensional curl of the four-potential vector Aµ .uu.78) ∂E ∂t = µ0 ρv + ε0 correspond to ∂F νµ jµ = ∂xν ε0 and that the two Maxwell “ﬁeld” equations ×E = − ·B = 0 correspond to ∂Fµν ∂Fνκ ∂Fκµ + µ + =0 ∂xκ ∂x ∂xν (5..82) (5.

. Reading. [6] W. . [8] B. ISBN 0-12-295250-2. 1972. K. second ed. ISBN 0-201-05702-6. Addison-Wesley Publishing Company. S PAIN. Academic Press. 1975. Addison-Wesley Publishing Company. Oliver and Boyd.. . MA . Ltd.. . ISBN 0-201-06710-2. third ed.. . Oxford . [5] C. . 1965.uu. AND M. D. . . . G RANDY.. ISBN 05-001331-9. vol. M ØLLER. ISBN 0-08-025072-6. J.. T.se/CED/Book Draft version released 13th November 2000 at 22:01. S AKURAI. New York and London. Reading. Oxford University Press. The Theory of Relativity. 2 of Course of Theoretical Physics. . 1962. P HILLIPS. fourth revised English ed.. AND E. . L ANDAU . Pergamon Press. Classical Electricity and Magnetism. PANOFSKY. [4] L. 1970. 1967. Tensor Calculus. Inc. . Glasgow . Inc. H. Ltd. Advanced Quantum Mechanics.plasma. L IFSHITZ.. [7] J. M. Edinburgh and London. Downloaded from http://www.68 R ELATIVISTIC E LECTRODYNAMICS [3] W. Introduction to Electrodynamics and Radiation.. MA . . The Classical Theory of Fields. second ed.

15) on page 56. 6.1 Charged Particles in an Electromagnetic Field We ﬁrst establish a relativistically correct theory describing the motion of charged particles in prescribed electric and magnetic ﬁelds. uµ ) ds = 0 (6. We shall do both. From these equations we may then calculate the charged particle dynamics in the most general case. The analysis is based on Lagrangian and Hamiltonian methods. Lagrange formalism Call the 4D Lagrange function L(4) and assume that it fulﬁls the variational principle δ x1 µ x0 µ L(4) (xµ . ¤ 69 . is fully covariant.1) where ds is the invariant line element given by Equation (5. and the endpoints are ﬁxed.1 Covariant equations of motion We can obtain an equation of motion if we ﬁnd a Lagrange function in 4D for our problem and then apply a variational principle or if we ﬁnd a Hamiltonian in 4D and solve the corresponding Hamilton’s equations.6 Interactions of Fields and Particles In this Chapter we study the interaction between electric and magnetic ﬁelds and electrically charged particles.1. 6. and yields results which are relativistically correct.

74) on page 66 in the following way L(4) = m0 c2 µ u uµ + quµ Aµ (xν ) 2 (6. The Lagrange function must be invariant. A free particle has only kinetic energy. If the particle mass is m0 then in 3D the kinetic energy is m0 v2 /2. According to formula (M. This suggests that in 4D the Lagrangian for a free particle should be free L(4) = m0 c2 µ u uµ 2 (6. This implies that L(4) must not contain higher than the second power of the fourvelocity uµ . leads to According to Equation (5.34) on page 61. u ) ds = δ µ µ x1 µ µ ¡ m0 c2 µ u uµ + quµ Aµ ds 2 ds (6.plasma.6) . The variation principle (6.3) inserted.1) with the 4D Lagrangian (6.uu.3) We call this the four-Lagrangian.2) For an interaction with the electromagnetic ﬁeld we can introduce the interaction with the help of the four-potential given by Equation (5. § ¨¢ µ x0 ¦ = x1 µ mo c2 uµ δuµ + q Aµ δuµ + uµ ¡ x0 ∂Aµ ν δx ∂xν ds = 0 § ¨¢ 4 µ ¦ = 5 µ x1 ∂Aµ m0 c2 ∂ uµ uµ δuµ + q Aµ δuµ + uµ ν δxν 2 ∂uµ ∂x ¡ ¤ x0 x0 ¢ δ x1 µ µ L(4) (x . This implies that L(4) must be a scalar.84) on page 176 the ordinary 3D Lagrangian is the difference between the kinetic and potential energies. the four-velocity is uµ = dx µ ds (6.se/CED/Book 5 4 ¢ dx µ ds d δxµ ds Draft version released 13th November 2000 at 22:01.4) ¡ ¤ ¤ ¤ (6.70 I NTERACTIONS OF F IELDS AND PARTICLES We must require that L(4) fulﬁls the following conditions: 1.5) which means that we can write the variation of uµ as a total derivative with respect to s : δuµ = δ = Downloaded from http://www. 2. The Lagrange function must yield linear equations of motion.

7) gives δ x1 x0 µ µ L(4) (xµ . A change of irrelevant summation index from µ to ν in the ﬁrst two terms of the right hand member of (6. uµ )ds x1 x0 µ µ Draft version released 13th November 2000 at 22:01.1 C HARGED PARTICLES IN AN E LECTROMAGNETIC F IELD 71 Inserting this into the ﬁrst two terms in the last integral in Equation (6.8) yields. § ¢ ¦ = −m0 c 2 duν ds + qu µ ∂Aµ ∂Aν − ∂xν ∂xµ ¢ ∂Aµ dAν −m0 c −q + quµ ν ds ds ∂x 2 duν δx ds ν δx ds ν Downloaded from http://www.9) above. we obtain δ x1 µ x0 µ L(4) (xµ .11) . we can reform the expression for dAν /ds as follows: dAν ∂Aν dxµ ∂Aν µ = µ = µu ds ∂x ds ∂x (6. uµ ) ds x1 x0 µ µ = Applying well-known rules of differentiation and the expression (5. the following expression: δ x1 x0 µ µ L(4) (xµ .se/CED/Book ¢ ¢ 4 ¡ 4 µ 5 5 x1 µ ¡ ¡ ¡ ¤ ¤ ¤ ¤ ¤ ¤ ¤ ¤ ∂Aµ d d mo c uµ δxµ + qAµ δxµ + quµ ν δxν ds ds ds ∂x (6. uµ ) ds 2 = x0 Partial integration in the two ﬁrst terms in the right hand member of (6. we obtain the ﬁnal variational principle expression δ x1 x0 µ µ L(4) (xµ .34) for the four-velocity. after moving the common factor δxν outside the partenthesis.10) By inserting this expression (6.10) into the second term in right-hand member of Equation (6.uu. uµ ) ds x1 x0 µ µ = dAµ µ ∂Aµ −m0 c δx − q δx + quµ ν δxν ds ds ds ∂x 2 duµ µ where the integrated parts do not contribute since the variations at the endpoints vanish.plasma.9) (6.8) (6.4) on the facing page.6.7) (6. and noting the common factor quµ of the resulting term and the last term.

t) is the ordinary 3D Hamiltonian. we have found an equation of motion for a charged particle in a prescribed electromagnetic ﬁeld: m0 c2 = quµ ∂ν Aµ − ∂µ Aν With the help of Equation (5.75) on page 66 we can express this equation in terms of the electromagnetic ﬁeld tensor in the following way: m0 c2 duν = quµ Fνµ ds (6. qi .16) Downloaded from http://www.34) on page 61.14b) where H(pi . qi .plasma. as given by Equation (5. To this end we introduce a canonically conjugate four-momentum pµ in an analogous way as the ordinary 3D conjugate momentum: pµ = ∂L(4) ∂uµ (6. this expression shall vanish and µ µ δxν is arbitrary along the world-line between the ﬁxed end points x0 and x1 . to deﬁne the four-Hamiltonian H(4) = pµ uµ − L(4) (6. according to the variational principle.72 I NTERACTIONS OF F IELDS AND PARTICLES Since.89) on page 176 in Chapter M. qi is a ˙ ˙ generalised coordinate and pi is its canonically conjugate momentum. the spatial part of this 4-vector equation is the covariant (relativistically correct) expression for the Newton-Lorentz force equation. the expression inside in the integrand in the right hand member of (6. t) = pi qi − L(qi .13) This is sought-for version of the covariant equation of motion for a particle in an electromagnetic ﬁeld. Hamiltonian formalism The usual Hamilton equations for a 3D space are given by Equation (M.11) must vanish. We seek a similar set of equations in 4D space.uu. It is often referred to as the Minkowski equation.14a) (6. As the reader can easily verify.15) and utilise the four-velocity uµ . In other words. These six ﬁrst-order partial differential equations are ∂H dqi = ∂pi dt ∂H dpi =− ∂qi dt (6.12) .se/CED/Book Draft version released 13th November 2000 at 22:01. 5 4 ¢ duν = quµ ds ¡ ∂Aµ ∂Aν − ∂xν ∂xµ (6.

17b) µ ∂x ds which form the four-dimensional Hamilton equations.16) on the preceding page. deﬁned in Equation (5.20) above that H(4) is indeed a scalar invariant.3) on page 70.15) on page 56.uu. Equation (6.1 C HARGED PARTICLES IN AN E LECTROMAGNETIC F IELD 73 With the help of these.6.39) on page 61. we obtain m0 c2 µ u uµ − quµ Aµ (xν ) (6. we clearly see from Equation (6.18). from the deﬁnition (6. we see that H(4) = pµ uµ − L(4) = pµ uµ − = m0 c u + qA 2 µ µ Inserting this into (6.19) Downloaded from http://www.15) on the facing page and Equations (6. This later consistency check is left as an exercise to the reader. ¢ pµ = ∂L(4) ∂ = ∂uµ ∂uµ ¡ m0 c2 µ u uµ + quµ Aµ (xν ) 2 (6. we obtain H(4) = m0 c2 uµ uµ + qAµ uµ − m0 c2 µ u uµ − quµ Aµ (xν ) 2 (6. considered as the generalised four-coordinate.18) 2 Furthermore.14) on the facing page.plasma.21) Draft version released 13th November 2000 at 22:01.15) on the preceding page be identical to the ordinary 3D Hamiltonian H and hence that this component solves the Hamilton equations Equations (6.20) m0 c2 µ u uµ = 2 Since the four-velocity scalar-multiplied by itself is uµ uµ = 1. whose value is simply H(4) = m0 c2 2 (6. According to Equation (5. Our strategy now is to use Equation (6. Equation (6.se/CED/Book . and the expression for L(4) . Using the deﬁnition of H(4) . Hence we require that the component p0 of the conjugate fourmomentum vector deﬁned according to Equation (6. and the invariant line element ds. a four-momentum has a zeroth (time) component which we can identify with the total energy.17) above to derive an explicit algebraic expression for the canonically conjugate momentum four-vector. we introduce the following eight partial differential equations: ∂H(4) dxµ = (6. the radius four-vector xµ .17a) ∂pµ ds dpµ ∂H(4) =− (6.15) of the conjugate four-momentum p µ .

26b) (6. Using the fact that expression (6.26e) .74 I NTERACTIONS OF F IELDS AND PARTICLES However.19) on the preceding page was used.23) above for H(4) is equal to the scalar value m0 c2 /2.26d) (6. as derived in Equation (6.17) and using the relation (6.26a) (6.23) (6.13) on page 72.24) ∂Aν ν = −qu ∂xµ duµ ∂Aµ dpµ = −mo c2 − q ν uν =− ds ds ∂x where in the last step Equation (6. and using Equation (5. the fact that pµ = (p0 .se/CED/Book ¢ m0 c2 duµ = quν ds ∂Aν ∂Aµ − ∂xµ ∂xν = quν Fµν Draft version released 13th November 2000 at 22:01.plasma. one gets That this four-Hamiltonian yields the correct covariant equation of motion can be seen by inserting it into the four-dimensional Hamilton’s equations (6.77) on page 67.25) (6. we obtain which is identical to the covariant equation of motion Equation (6.uu. cA) A pµ = φp − c (p · A) pµ pµ = (p0 )2 − c2 (p)2 µ 0 2 µ Aµ Aµ = φ2 − c2 (A)2 Downloaded from http://www. cp) A = (φ. Rearranging terms.19) provides the algebraic relationship and if this is used in (6.21) on the preceding page. at the same time (6.20) to eliminate uµ .26c) (6. ¢ 5 4 5 4 5 4 H(4) = 5 1 1 m0 c2 pµ − qAµ pµ − qAµ 2 2 m0 c m0 c2 1 = pµ − qAµ pµ − qAµ 2m0 c2 1 = pµ pµ − 2qAµ pµ + q2 Aµ Aµ 2m0 c2 5 ¡ 4 4 ¡ 4 uµ = 5 1 pµ − qAµ m0 c2 (6.22): q ∂Aν ∂H(4) =− (pν − qAν ) µ µ 2 ∂x m0 c ∂x q ∂Aν =− m0 c2 uν µ m0 c2 ∂x (6. We can then safely conclude that the Hamiltonian in question is correct.22) (6.

The ordinary Lagrange and Hamilton functions L and H are related to each other by the 3D transformation [cf.16) between L (4) and H(4) ] L = p·v−H (6.30) Since the fourth component (time component) p0 of the four-momentum pµ is the total energy.6. the 4D transformation (6.uu.30) must be identiﬁed with the ordinary Hamilton function H.32) above).31) is the ordinary 3D Hamilton function for a charged particle moving in scalar and vector potentials associated with prescribed electric and magnetic ﬁelds.33) and if we make the identiﬁcation Draft version released 13th November 2000 at 22:01.se/CED/Book c c c .plasma. This means that H ≡ p0 = qφ + c (p − qA)2 + m2 c2 0 (6. c p − qA = 1− v2 c2 = mv (6.31)) and (Equation (6. we obtain the equation m0 c2 1 = (p0 )2 − c2 (p)2 − 2qφp0 + 2qc2 (p · A) + q2 φ2 − q2 c2 (A)2 2 2m0 c2 (6. the positive solution in (6.29) (p − qA)2 + m2 c2 0 (6.27) which is a second order algebraic equation in p0 which can be written (m0 c2 )2 = (p0 )2 − 2qφp0 − c2 (p)2 + 2qc2 (p · A) − q2 c2 (A)2 + q2 φ2 (6. we obtain the explicit expression for the ordinary 3D Lagrange function L = p · v − qφ − c m0 v (p − qA)2 + m2 c2 0 (6.28) or (p0 )2 − 2qφp0 − c2 (p)2 − 2qp · A + q2 (A)2 + q2 φ2 − m2 c4 0 = (p0 )2 − 2qφp0 − c2 (p − qA)2 + q2 φ2 − m2 c4 = 0 0 with two possible solutions p0 = qφ ± c (6.1 C HARGED PARTICLES IN AN E LECTROMAGNETIC F IELD 75 and Equation (6.19) on page 73.32) Using the explicit expressions (Equation (6.34) Downloaded from http://www.

se/CED/Book % = −qφ + qA · v − m0 c 2 c v2 1− 2 c (6. drawing inferences from this model problem. and (i + 1)th mass point. The equilibrium distance between the neighbouring mass points is a and ηi−1 (t).2 Covariant Field Theory So far. the particles. we have considered two classes of problems. ηi+1 (t) are the instantaneous deviations. ideal springs with spring constants k. ηi (t).35) Draft version released 13th November 2000 at 22:01.uu.plasma. or we have derived the equations of motion for charged particles in given. respectively. and their interactions in a uniﬁed way. prescribed distributions of charges and currents. ith. This involves transition to a ﬁeld picture with an inﬁnite number of degrees of freedom. Let us now put the ﬁelds and the particles on an equal footing and present a theoretical description which treats the ﬁelds. Downloaded from http://www. identical mass points m. where the quantity mv is called the kinetic momentum.1: A one-dimensional chain consisting of N discrete. along the x axis. . 6. Then. We shall ﬁrst consider a simple mechanical problem whose solution is well known. of positions of the (i − 1)th. connected to their neighbours with identical. we apply a similar view on the electromagnetic problem. we can rewrite this expression for the ordinary Lagrangian as follows: L = qA · v + mv2 − qφ − c m2 v2 − m2 c2 0 What we have obtained is the correct expression for the Lagrangian describing the motion of a charged particle in scalar and vector potentials associated with prescribed electric and magnetic ﬁelds.76 I NTERACTIONS OF F IELDS AND PARTICLES ηi−1 m m m ηi ηi+1 m m x k a k a k a k a F IGURE 6. prescribed ﬁelds. Either we have calculated the ﬁelds from given.

If we now let N → ∞ and. At equilibrium the mass points are at rest. which we choose to be the x axis. is the so called linear Lagrange density. ˆ The solution to this mechanical problem can be obtained if we can ﬁnd a Lagrangian (Lagrange function) L which satisﬁes the variational equation ˙ δ L(ηi . let the springs become inﬁnitesimally short according to the following scheme: a → dx m dm → =µ a dx ka → Y ηi+1 − ηi ∂η → a ∂x (6.6. ηi .39) d Downloaded from http://www.38) 2 (6. each with mass m and connected to its neighbour along a one-dimensional straight line. Let us denote the magnitude of the deviation for ˆ mass point i from its equilibrium position by ηi (t) x. by identical ideal springs with spring constants k. After perturbation. § 0 i ) ¦ = ηi+1 − ηi 1 m 2 ˙ ηi − ka 2 a a d L = ∑a N i L= ¤ 1 N m˙ 2 − k(ηi+1 − ηi )2 ηi 2∑ i=1 (6.40d) Draft version released 13th November 2000 at 22:01. as given by Equation (6.40b) (6.plasma. in the following way: i=1 Here.37) (6.2 C OVARIANT F IELD T HEORY 77 6.37).40c) (6. the motion of mass point i will be a one-dimensional oscillatory motion along x. at the same time. t) dt = 0 (6.84) on page 176.se/CED/Book .2.36) According to Equation (M. the Lagrangian is L = T − V where T denotes the kinetic energy and V the potential energy of a classical mechanical system with conservative forces.40a) linear mass density Young’s modulus (6. distributed evenly with a distance a to their two nearest neighbours.1 Lagrange-Hamilton formalism for ﬁelds and interactions Consider N identical mass points.uu. In our case the Lagrangian is Let us write the Lagrangian.

. . .44) (6. ∂η ∂η . .45) .36) on the previous page yields: δ L dt = =0 The last integral can be integrated by parts. the variation principle (6. 2. Under the assumption of time independence and ﬁxed endpoints. or wave functions. be it classical scalar or vectorial ﬁelds. ∂t ∂x dx dt (6.41) fh ¤ ge¤ ¡ ¤ d d d (6.se/CED/Book 0 ) ∂ ∂η ∂x V d SQ δ ∂ ∂ = − δη ∂η ∂x ∂ 0 ) T 0 ) ∂ ∂ Draft version released 13th November 2000 at 22:01. as we shall see. This results in the expression where the variation is arbitrary (and the endpoints ﬁxed). ∂η ∂η 1 ∂η .42) (6.43) dx dt ¢ ¢ X ¢ η.t = µ ∂t ∂x 2 ∂t 2 −Y ∂η ∂x 2 Y ¡ ¡ ¡ fh ¤ ge¤ d ¤ e¤ d d ¤ L= dx (6. If we introduce the functional derivative Downloaded from http://www. to a continuous description. ˆ 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 has now become dependent also on the partial derivative with respect to x of the “ﬁeld coordinate” η. N. . in which the mass points were identiﬁed by a discrete integer variable i = 1.uu. namely their position along x. spinors and other ﬁelds that appear in quantum physics. where the inﬁnitesimal mass points were instead identiﬁed by a continuous real parameter x. But. on an equal footing. ip ∂η ∂t SQ − ∂η ∂x V d V d SQ ∂ ∂ − ∂η ∂t ∂ ∂ ∂x ∂ δη dx dt = 0 ip ¨¢ ∂ ∂ ∂η ∂ ∂η δη + δ + δ ∂η ∂η ∂η ∂t ∂x ∂ ∂t ∂ ∂x ¡ 0 ) T d ¢ ¡ 0 ) T ¢ d d d =δ η. This means that the integrand itself must vanish.plasma.78 I NTERACTIONS OF F IELDS AND PARTICLES we obtain where Notice how we made a transition from a discrete description. the transition is well worth the price because it allows us to treat all ﬁelds.

Inserting the linear mass point chain Lagrangian density. ∂η d4x ∂xµ ¢ ¡ 0 ) ∂ ∂η ∂t V d ¤ q¤ ¤ SQ ∂ δ − δη ∂t ∂ T d ¤ d d d =0 (6.uu. Introducing the three-dimensional functional derivative Draft version released 13th November 2000 at 22:01.47) (6. 0 ) ∂ ∂η ∂xi V d SQ δ ∂ ∂ = − δη ∂η ∂xi ∂ T 0 ) ∂ ∂η ∂xµ V d SQ ∂ ∂ − µ ∂η ∂x ∂ =0 0 ) ∂ ip SQ = ∂η ∂xµ V d ∂ ∂ − ∂η ∂xµ ∂ T ¢ T d d =δ ¡ fh e¤ ¤ d δ L dt = δ d3x dt η. we obtain the equation of motion for our one-dimensional linear mechanical structure.2 C OVARIANT F IELD T HEORY 79 we can express this as which is the one-dimensional Euler-Lagrange equation..42) on the facing page. Equation (6.46) above.se/CED/Book .e.50) Downloaded from http://www.plasma. For this 3D case we get the variational principle =0 where the variation δη is arbitrary and the endpoints are ﬁxed. It is: µ ∂2 ∂2 η−Y 2η = ∂t2 ∂x i. This means that the integrand itself must vanish: This constitutes the three-dimensional Euler-Lagrange equations.46) ∂2 µ ∂2 − 2 Y ∂t2 ∂x η=0 (6. A generalisation of the above 1D results to a three-dimensional continuum is straightforward.49) (6. into Equation (6. the one-dimensional wave equation for compression waves which propag√ ate with phase speed vφ = Y/µ along the linear structure.6.48) δη d x 4 (6.

t = π − i ∂x ∂t ¡ 0 ) d 0 ) ∂ ∂η ∂t ∂ V d SQ ∂ δ − δη ∂t ∂ T ¡ r r d d =0 (6. x). the four-potential A The entire system of particles and ﬁelds consists of a mechanical part. x) = ∂ and deﬁne the Hamilton density If. ∂t ∂xi r d d (6. If we want to describe the electromagnetic ﬁeld in terms of a Lagrange density and Euler-Lagrange equations.80 I NTERACTIONS OF F IELDS AND PARTICLES we can express this as I analogy with particle mechanics (ﬁnite number of degrees of freedom). The electromagnetic ﬁeld Above. it comes natural to express in terms of µ (xκ ). a ﬁeld part and an interaction part. we differentiate this expression and identify terms.se/CED/Book d d d tot = mech + inter + ﬁeld Draft version released 13th November 2000 at 22:01. η. as usual.55) .52) η. we may introduce the canonically conjugate momentum density π(xµ ) = π(t. we used a scalar ﬁeld η(t. when we described the mechanical ﬁeld. We therefore assume that the total Lagrange density tot for this system can be expressed as where the mechanical part has to do with the particle motion (kinetic energy).plasma.uu.2) on page 70 and V is Downloaded from http://www. ∂η ∂η . ∂η ∂η . we obtain the following Hamilton density equations ∂ ∂η = ∂π ∂t δ ∂π =− δη ∂t The Hamilton density functions are in many ways similar to the ordinary Hamilton functions and lead to similar results. It is given by L(4) /V where L(4) is given by Equation (6.54b) (6.54a) (6.51) ∂η ∂t (6.53) (6. d ¢ d ¢ π.

61) (6. Using the ﬁeld tensor. Einstein summation Draft version released 13th November 2000 at 22:01. by explicit calculation. A convenient expression for this interaction Lagrange density is For the ﬁeld part 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). Expressed in the rest mass density ρ0 .58) (6.62) Downloaded from http://www. y sttu F µν = ∂µ Aν − ∂ν Aµ = ww xv y sttu F = ∂ A −∂ A = 0 Ex Ey Ez −E x 0 cBz −cBy −E y −cBz 0 cB x −E z cBy −cB x 0 ww xv d d d d d d inter = jµ Aµ (6.1 Show.77) on page 67 we recall that µν µ ν ν µ and from formula (5.60) (6. that F µν F µν = 2(c2 B2 − E 2 ) From formula (5. the mechanical Lagrange m density can be written mech 1 = ρ0 c2 uµ uµ 2 m (6.56) The inter part which describes the interaction between the charged particles and the external electromagnetic ﬁeld. F IELD ENERGY DIFFERENCE EXPRESSED IN THE FIELD TENSOR E XAMPLE 6.uu.76) on page 66 that 0 −E x −E y −E z Ex 0 cBz −cBy Ey −cBz 0 cB x Ez cBy −cB x 0 where µ denotes the row number and ν the column number.6.57) (6. we express this ﬁeld Lagrange density as ﬁeld 1 = ε0 F µν Fµν 4 so that the total Lagrangian density can be written tot 1 1 = ρ0 c2 uµ uµ + jµ Aµ + ε0 F µν Fµν m 2 4 From this we can calculate all physical quantities.59) (6.plasma.se/CED/Book .2 C OVARIANT F IELD T HEORY 81 the volume. Then.

Equation (6. expression (6.se/CED/Book Draft version released 13th November 2000 at 22:01.49) on page 79 Using (with η replaced by Aν ).uu.82 I NTERACTIONS OF F IELDS AND PARTICLES and direct substitution yields F µν F µν = F 00 F 00 + F 01 F 01 + F 02 F 02 + F 03 F 03 + F 10 F 10 + F 11 F 11 + F 12 F 12 + F 13 F 13 + F 20 F 20 + F 21 F 21 + F 22 F 22 + F 23 F 23 + F 30 F 30 + F 31 F 31 + F 32 F 32 + F 33 F 33 2 2 = 0 − E 2 − Ey − Ez x 2 − E y + c 2 B2 + 0 + c 2 B2 z x − E 2 + 0 + c 2 B2 + c 2 B2 x z y 2 − E z + c 2 B2 + c 2 B2 + 0 y x (6.63) = −2E 2 + 2c2 B2 = 2(c2 B2 − E 2 ) 2 2 = −2E 2 − 2E y − 2E z + c2 B2 + c2 B2 + c2 B2 x x y z QED OF EXAMPLE in the 3D Euler-Lagrange equations. we can derive the dynamics for the whole system. Downloaded from http://www. For instance.65) (6. the electromagnetic part of the Lagrangian density 1 = jν Aν + ε0 F µν Fµν (6.plasma. we note from Equation (6.1 d d .1 that tot Furthermore. To see this. § 4 ¦ = 5 ε0 ∂ ∂µ ∂κ Aλ ∂κ Aλ − ∂κ Aλ ∂λ Aκ 2 ∂(∂µ Aν ) § ¦ = ε0 ∂ ∂µ ∂κ Aλ ∂κ Aλ − ∂κ Aλ ∂λ Aκ 4 ∂(∂µ Aν ) − ∂λ Aκ ∂κ Aλ + ∂λ Aκ ∂λ Aκ § 4 ¦ § ∂µ 5 d ¦ d ∂ EM ∂Aν = jν ∂ EM ∂ ε0 = ∂µ F κλ Fκλ ∂(∂µ Aν ) 4 ∂(∂µ Aν ) ε0 ∂ = ∂µ (∂κ Aλ − ∂λ Aκ )(∂κ Aλ − ∂λ Aκ ) 4 ∂(∂µ Aν ) d d EM = inter + ﬁeld (6.66) ! E ND 6. yields two of Maxwell’s equations.64) above and the results in Example 6.64) 4 inserted into the Euler-Lagrange equations.49) on page 79.

77) on page 67 for the covariant component form of the electromagnetic ﬁeld tensor F µν . expression (6.68) (6.49) on page 79.69) (6. § d ¦ d ∂ EM ∂F µν ∂ EM = jν − ε0 µ = 0 ∂(∂µ Aν ) ∂x 5 4 d 5 5 § 4 d ¦ 4 = 2∂µ Aν (6.plasma.72) ε0 which is the Maxwell source equation for the electric ﬁeld.67) ∂κ Aλ + gκα gλβ ∂κ Aλ ∂α Aβ = ∂κ Aλ ∂(∂µ Aν ) ∂(∂µ Aν ) ∂ ∂ = ∂κ Aλ ∂κ Aλ + ∂α Aβ ∂α Aβ ∂(∂µ Aν ) ∂(∂µ Aν ) Similarly. For ν = 1 we get ∂By ρu x 1 ∂E x ∂Bz ∂F 01 ∂F 11 ∂F 21 ∂F 31 + + + =− +0−c +c = (6.71) Downloaded from http://www. ∂ ∂κ Aλ ∂λ Aκ = 2∂ν Aµ ∂(∂µ Aν ) so that ∂µ ∂F µν ∂ EM = ε0 ∂µ ∂µ Aν − ∂ν Aµ = ε0 µ ∂(∂µ Aν ) ∂x This means that the Euler-Lagrange equations. Equation (1.43a) on page 14.70) (6. setting ν = 0 in this covariant equation and using the matrix representation formula (5. we obtain ∂F 00 ∂F 10 ∂F 20 ∂F 30 ∂E x ∂Ey ∂Ez + + = + + + = 0+ 0 1 2 3 ∂x ∂x ∂x ∂x ∂x ∂y ∂z ·E = ρ (6.uu.se/CED/Book .73) 0 1 2 3 ∂x ∂x ∂x ∂x c ∂t ∂y ∂z ε0 Draft version released 13th November 2000 at 22:01. for the Lagrangian density EM and with Aν as the ﬁeld quantity become ∂Aν − ∂µ or ∂F µν jν = ∂xµ ε0 Explicitly.6.2 C OVARIANT F IELD T HEORY 83 But ∂ ∂ ∂ ∂κ Aλ ∂κ Aλ = ∂κ Aλ ∂κ Aλ + ∂κ Aλ ∂κ Aλ ∂(∂µ Aν ) ∂(∂µ Aν ) ∂(∂µ Aν ) ∂ ∂ = ∂κ Aλ ∂κ Aλ + ∂κ Aλ gκα ∂α gλβ Aβ ∂(∂µ Aν ) ∂(∂µ Aν ) ∂ ∂ (6.

can be derived from a Lagrangian density together with a variational principle (the Euler-Lagrange equations). Downloaded from http://www. the dynamic equations for most any ﬁelds.84 I NTERACTIONS OF F IELDS AND PARTICLES or. consider a real. scalar ﬁeld η which has the following Lagrange density: = 1 1 ∂µ η∂µ η − m2 η2 = 2 2 Insertion into the 1D Euler-Lagrange equation. Equation (6. In summary.78) which describes the Yukawa meson ﬁeld for a scalar meson with mass m. we can write the result as × B − ε 0 µ0 ∂E = µ0 j(t.77) (6. With π= 1 ∂η c2 ∂t (6. 3.75) which is the Maxwell source equation for the magnetic ﬁeld. and not only electromagnetic ones. ¢ ¡ ∂η ∂η − m2 η2 ∂xµ ∂xµ 5 (6. ∂By ∂Bz ∂E x − − ε0 µ0 = µ0 j x ∂z ∂y ∂t (6. Both linear and non-linear ﬁelds are studied with this technique. Equation (1.74) and similarly for ν = 2. x) ∂t (6. using ε0 µ0 = 1/c2 and identifying ρu x = j x .se/CED/Book 9 4 8 = η 5 1 2 2 c π + 2 2 + m2 η2 Draft version released 13th November 2000 at 22:01.80) .43d) on page 14. Other ﬁelds In general.uu. in three-vector form.plasma. yields the dynamic equation ( 2 − m2 )η = 0 e−m|x| |x| with the solution η = ei(k·x−ωt) (6. As a simple example.79) we obtain the Hamilton density which is positive deﬁnite.76) 4 6 r d (6.46) on page 79.

Space experiments of this kind onboard satellites have led to stringent upper bounds on the photon mass. ISBN 0-486-64038-8. L IFSHITZ.. 1967. vol. G RANDY. H. .se/CED/Book . or. If massive photons would exist. Dover Publications. Draft version released 13th November 2000 at 22:01. Reading. NY. Reading. D. .82) This equation describes an electromagnetic ﬁeld with a mass. ISBN 0-12-295250-2. . L. ISBN 0-201-05702-6. . [7] J. AND E. ISBN 0-201-02918-9.plasma. [3] H. fourth revised English ed. it will have an impact on electrodynamics as well as on cosmology and astrophysics.. PANOFSKY.2 B IBLIOGRAPHY 85 Another Lagrangian density which has attracted quite some interest is the Proca Lagrangian which leads to the dynamic equation ∂F µν jν + m2 Aν = µ ∂x ε0 (6. 1980. Pergamon Press. d d EM = inter + ﬁeld 1 = jν Aν + ε0 F µν Fµν + m2 Aµ Aµ 4 (6. 2 of Course of Theoretical Physics. Paris.6. Inc. P HILLIPS. ISBN 288124-719-9. ISBN 0-08-025072-6. 1970. including those of the earth and galactic spiral arms. Inc. Addison-Wesley Publishing Company.. . 1989. K. New York. AND M. [5] L. 1962.. MA . . [6] W. Ltd. Advanced Quantum Mechanics. Tokyo and Melbourne. . New York and London. G INZBURG. Classical Electricity and Magnetism. Montreux. MA . . Addison-Wesley Publishing Company.. Gordon and Breach Science Publishers.. massive photons.81) d Downloaded from http://www. The Classical Theory of Fields. BARUT. . New York. If the photon really has a mass. Bibliography [1] A. 1975. .uu. Academic Press.. ISBN 0-201-06710-2. Dynamics and Classical Theory of Fields and Particles. Classical Mechanics. large-scale magnetic ﬁelds. second ed. 1981. Inc. Applications of Electrodynamics in Theoretical Physics and Astrophysics. L ANDAU . Oxford . [4] W. . Addison-Wesley Publishing Company. in other words. S AKURAI. [2] V. J. M. second ed.. . Revised third ed.. Inc. T. Introduction to Electrodynamics and Radiation. G OLDSTEIN. London. Reading. MA . O. would be signiﬁcantly modiﬁed to yield measurable discrepances from their usual form.

se/CED/Book Draft version released 13th November 2000 at 22:01.86 I NTERACTIONS OF F IELDS AND PARTICLES Downloaded from http://www. .plasma.uu.

2) (x − x0 )(x − x0 )ρ(x ) d3x (7. derived ﬁelds are introduced in order to incorporate effects of macroscopic matter when this is immersed fully or partially in an electromagnetic ﬁeld.1) where the ρ is the charge density introduced in Equation (1. when macroscopic matter is present.1 Electric multipole moments The electrostatic properties of a spatial volume containing electric charges and located near a point x0 can be characterized in terms of the total charge or electric monopole moment q= V ρ(x ) d3x (7. 7. the electric quadrupole moment tensor Q= V ¤ ¤ ¤ (x − x0 )ρ(x ) d3x (7.7 Interactions of Fields and Matter The microscopic Maxwell equations (1. i = 1.43) derived in Chapter 1 are valid on all scales where a classical description is good.3) 87 .7) on page 4. it is sometimes convenient to use the corresponding macroscopic Maxwell equations (in a statistical sense) in which auxiliary. However.1. the electric dipole moment vector p= V with components pi . 2.1 Electric polarisation and the electric displacement vector 7. 3.

88 I NTERACTIONS OF F IELDS AND M ATTER with components Qi j . 2. As can be seen from this expression. . unless there is Downloaded from http://www.se/CED/Book Draft version released 13th November 2000 at 22:01.uu.4) on page 88. § § ¤ ¤ ¢ ¤ ©¦ ¦ φp (x) = 1 4πε0 1 = 4πε0 P(x ) · P(x ) 3 d3x − dx |x − x | V V |x − x | ˆ P(x ) · n 2 · P(x ) 3 d x− dx |x − x | |x − x | S V · § (7. in some medium. the electric potential from this volume distribution P(x ) of electric dipole moments p at the source point x can be written φp (x) = x−x 3 1 P(x ) · dx 4πε0 V |x − x |3 1 1 d3x P(x ) · =− 4πε0 V |x − x | 1 1 = d3x P(x ) · 4πε0 V |x − x | Using the expression Equation (M.4) ¤ ¤ ¤ .plasma. the electrostatic potential Equation (3. only the ﬁrst few terms are important if the ﬁeld point (observation point) is far away from x0 . unit: C/m2 ). we can rewrite this expression for the potential as follows: where the ﬁrst term. Particularly important is the dipole moment.5) (7.6) ¢ 1 Qi j + |x − x0 |3 ¢ ¢ ¡ ¡ ¡ ¡ ¦ φstat (x) = q (x − x0 )i 1 1 p + 2 i |x − x | 4πε0 |x − x0 | |x − x0 | 0 3 (x − x0 )i (x − x0 ) j 1 − δi j + . Let P denote the electric dipole moment density (electric dipole moment per unit volume. j = 1.71) on page 173 and applying the divergence theorem. In particular. In analogy with the second term in the expansion Equation (7. . i. 3.3) on page 33 from a charge distribution located near x0 can be Taylor expanded in the following way: where Einstein’s summation convention over i and j is implied. also known as the electric polarisation. which describes the effects of the induced. non-cancelling dipole moment on the surface of the volume. can be neglected. the major contributions to the electrostatic interactions come from the net charge and the lowest order electric multipole moments induced by the polarisation due to an applied electric ﬁeld. and higher order electric moments. 2 |x − x0 | |x − x0 | 2 (7. For a normal medium.

and at the same time introducing the electric displacement vector (C/m2 ) D = ε0 E + P we obtain · (ε0 E + P) = · D = ρtrue (x) (7. for low enough ﬁeld strengths |E|. The version of Equation (1.3) on page 33 for the electrostatic potential from a static charge distribution ρ.12) is a good approximation.9) where ρtrue is the “true” charge density in the medium. In general.plasma.11) Often. ¤ Downloaded from http://www.se/CED/Book . the relationship is not of a simple linear form as in Equation (7.uu.1 E LECTRIC POLARISATION AND THE ELECTRIC DISPLACEMENT VECTOR 89 a discontinuity in P · n at the surface. we see that − · P(x) has the characteristics of a charge density and that. Doing so. we ﬁnd that the contribution ˆ from the electric dipole moments to the potential is given by φp = 1 4πε0 − V · P(x ) 3 dx |x − x | (7. and ρtotal = ρtrue + ρpol for the “total” charge density. By introducing the notation ρpol = − · P for the “polarised” charge density in the medium. the effective charge density becomes ρ(x) − · P(x). to the lowest order. This is one of Maxwell’s equations and is valid also for time varying ﬁelds.7.7) Comparing this expression with expression Equation (3.10) (7. we can write down the following alternative version of Maxwell’s equation (7. For electromagnetically anisotropic media such as a magnetised plasma or a birefringent crystal.8) Rewriting this equation. in which the second term is a polarisation term. the susceptibility is a tensor. the linear and isotropic relationship between P and E P = ε0 χE (7.23a) on page 91 ·E = ρtotal (x) ε0 (7.7) on page 4 where “true” and polarisation charges are separated thus becomes ·E = ρ(x) − · P(x) ε0 (7.12) Draft version released 13th November 2000 at 22:01. The quantity χ is the electric susceptibility which is material dependent.

14) (7. but which may produce net effects at discontinuities and boundaries. ε = ε0 (1 + χ) (7. nature that are inaccessible to direct observation. The lowest order magnetic moment. We shall call such currents magnetisation currents and denote them jm . approximately. Inserting the approximation (7. In analogy with “true” charges for the electric case.13) 7. No magnetic monopoles have been observed yet. a physical transport of true charges. is the magnetic dipole moment m= 1 2 (x − x0 ) × j(x ) d3x (7. In such a situation the principle of superposition is no longer valid and non-linear effects such as frequency conversion and mixing can be expected.9) on the previous page. often atomic.e. .90 I NTERACTIONS OF F IELDS AND M ATTER on the preceding page but non-linear terms are important.15) V For a distribution of magnetic dipole moments in a volume. We identify them with ∂P/∂t.. we can write the latter D = εE where. we may describe this volume in terms of the magnetisation.plasma.12) into Equation (7. In analogy with electric polarisation P there may be a form of charge transport associated with the changes of the polarisation with time. we may have “true” currents jtrue . We call such currents induced by an external ﬁeld polarisation currents. There may also be intrinsic currents of a microscopic. 2. or magnetic dipole moment per unit Downloaded from http://www.2 Magnetisation and the magnetising ﬁeld An analysis of the properties of stationary magnetic media and the associated currents shows that three such types of currents exist: 1.se/CED/Book ¤ Draft version released 13th November 2000 at 22:01. i.2). So there is no correspondence in the magnetic case to the electric monopole moment (7.1).uu. 3. corresponding to the electric dipole moment (7.

3 E NERGY AND MOMENTUM 91 volume. in analogy with the electric case.23c) (7. we can write B H= (7.3 Energy and momentum As mentioned in Chapter 1.uu.9) on page 89. introduce a magnetic susceptibility for the medium. Denoting it χm .se/CED/Book . ¢ × B = µ0 jtrue + ¡ ∂P + ×M ∂t (7.23a) (7.16) In a stationary medium we therefore have a total current which is (approximately) the sum of the three currents enumerated above: ∂P jtotal = jtrue + + ×M (7. approximately. Maxwell’s equations expressed in terms of the derived ﬁeld quantities D and H can be written · D = ρ(t.7. we can write this Maxwell equation in the following form ∂D ∂E ∂P = jtrue + − ε0 (7.23d) × H = j(t. Equation (7.21) µ where.23b) (7. x) + Draft version released 13th November 2000 at 22:01.plasma.17) ∂t We then obtain the Maxwell equation Moving the term × M to the left hand side and introducing the magnetising ﬁeld (magnetic ﬁeld intensity.18) Downloaded from http://www. x) ·B = 0 ×E = − ∂B ∂t ∂ D ∂t (7.22) 7.19) µ0 and using the deﬁnition for D. Ampère-turn density) as B H= −M (7. Via the deﬁnition of the vector potential one can show that the magnetisation current and the magnetisation is simply related: jm = ×M (7. M. µ = µ0 (1 + χm ) (7.20) × H = jtrue + ∂t ∂t ∂t We may.

uu.23c) by H. we obtain H · ( × E) − E · ( × H) = · (E × H) ∂D 1∂ ∂B −E·j−E· =− (H · B + E · D) − j · E = −H · ∂t ∂t 2 ∂t (7. F D GC F ¤B D EC F D EC ¤B ¤B F D EC V V ¤ ¤ V V V V ¤ ¤ V V S ¤ ¤B ¤ (7. These equations are convenient to use in certain simple cases. (7.plasma.28) .26) Inserting this into Equation (7. Downloaded from http://www. they describe uniquely (but only approximately!) the properties of the electric and magnetic ﬁelds in matter.se/CED/Book Draft version released 13th November 2000 at 22:01. 7.24) Integration over the entire volume V and using Gauss’s theorem (the divergence theorem).92 I NTERACTIONS OF F IELDS AND M ATTER and are called Maxwell’s macroscopic equations.25) But. Equation (1.25) j · EEMF d3x = ∂ j2 3 dx + σ ∂t 1 (E · D + H · B) d3x 2 Field energy Applied electric power Joule heat + S (E × H) · dS Radiated power which is the energy theorem in Maxwell’s theory also known as Poynting’s theorem.1 The energy theorem in Maxwell’s theory Scalar multiplying (7. We shall use them in the following considerations on the energy and momentum of the electromagnetic ﬁeld and its interaction with matter.27) (7.3. Together with the boundary conditions and the constitutive relations. we obtain − ∂ ∂t 1 (H · B + E · D) d3x = 2 j · E d3x + (E × H) · dS (7. according to Ohm’s law in the presence of an electromotive force ﬁeld.26) on page 10: j = σ(E + EEMF ) which means that j · E d3x = j2 3 dx − σ j · EEMF d3x (7.23d) by E and subtracting.

29) (7.3 E NERGY AND MOMENTUM 93 It is convenient to introduce the following quantities: 1 E · D d3x 2 V 1 Um = H · B d3x 2 V S = E×H Ue = (7.33) .31) where U e is the electric ﬁeld energy.23) and symmetrising.se/CED/Book ¢ ¡ F EC B D ¢ ¢ ¡ ¡ ¤ ¤ ×H− (7.32) can be expressed as [E( · D) − D × ( × E)]i = ∂D ∂E 1 E· −D· 2 ∂xi ∂xi + ∂ ∂x j and Draft version released 13th November 2000 at 22:01. Using Maxwell’s equations (7. 7. and S is the Poynting vector (power ﬂux). U m is the magnetic ﬁeld energy.32) =0 1 E i D j − E · D δi j 2 (7. For this purpose we consider the force density given by the Lorentz force per unit volume ρE+j×B.uu. both measured in J. measured in W/m2 .plasma.30) (7. we obtain ρE + j × B = ( · D)E + ∂D ×B ∂t ∂D ×B = E( · D) + ( × H) × B − ∂t = E( · D) − B × ( × H) ∂ ∂B − (D × B) + D × ∂t ∂t = E( · D) − B × ( × H) ∂ − (D × B) − D × ( × E) + H( · B) ∂t = [E( · D) − D × ( × E)] + [H( · B) − B × ( × H)] ∂ − (D × B) ∂t One veriﬁes easily that the ith vector components of the two terms in square brackets in the right hand member of (7.3.2 The momentum theorem in Maxwell’s theory Let us now investigate the momentum balance (force actions) in the case that a ﬁeld interacts with matter in a non-relativistic way. Downloaded from http://www.7.

36) (7. § ¢ ¢ ¢ ¦ (Fev )i = (ρE + j × B)i − 1 2 E· ∂D ∂E ∂B ∂H −D· + H· −B· ∂xi ∂xi ∂xi ∂xi ¢ 1 1 E i D j − E · D δ i j + Hi B j − H · B δ i j 2 2 § 1¢ ∂B ∂H −B· ∂xi ∂xi + ∂ (D × B)i ∂t (7.se/CED/Book Draft version released 13th November 2000 at 22:01.41) ¢ ¡ ¢ [H( · B) − B × ( × H)]i = 1 ∂B ∂ ∂H H· + −B· 2 ∂xi ∂xi ∂x j ¡ ¡ ¢ ¡ 1 Hi B j − B · H δ i j 2 (7.39) (7. Using these two expressions in the ith component of Equation (7.uu.35) (7.38) (7.40) i Downloaded from http://www.plasma.94 I NTERACTIONS OF F IELDS AND M ATTER respectively. we get (ρE + j × B)i − = ∂ ∂x j 1 2 E· ∂D ∂E −D· ∂xi ∂xi + H· Introducing the electric volume force Fev via its ith component and the Maxwell stress tensor 1 1 T i j = E i D j − E · D δ i j + Hi B j − H · B δ i j 2 2 we ﬁnally obtain the force equation Fev + ∂T i j ∂ (D × B) = ∂t ∂x j i (7.37) If we introduce the relative electric permittivity κ and the relative magnetic permeability κm as D = κε0 E = εE B = κm µ0 H = µH we can rewrite (7.38) as ∂T i j κκm ∂S = Fev + 2 ∂x j c ∂t (7.34) ¡ § ¡ ¦ ¡ ¡ ¦ .32) on the previous page and re-shufﬂing terms.

In vacuum (7.. Inc.43) or Bibliography [1] W. Addison-Wesley Publishing Company. Integration over the entire volume V yields Fev d3x + d dt κκm 3 Sd x = c2 T n d2x ˆ (7. Reading.42) Force on the matter Field momentum Maxwell stress which expresses the balance between the force on the matter. . ISBN 0-201-05702-6. P HILLIPS. Draft version released 13th November 2000 at 22:01. the rate of change of the electromagnetic ﬁeld momentum and the Maxwell stress. ¤ d mech d ﬁeld p + p = dt dt S T n d2x ˆ ¤ ¤ V V S F D EC ¤B F D EC ¤B F D EC V V S ¤B ¤ (7. This equation is called the momentum theorem in Maxwell’s theory. 1962. . K. .se/CED/Book . Classical Electricity and Magnetism..3 B IBLIOGRAPHY 95 where S is the Poynting vector deﬁned in Equation (7. AND M.42) becomes ρ(E + v × B) d3x + 1 d c2 dt S d3x = T n d2x ˆ (7.uu. MA . PANOFSKY.plasma.29) on page 93. H.7.44) Downloaded from http://www. second ed.

plasma.96 I NTERACTIONS OF F IELDS AND M ATTER Downloaded from http://www.uu.se/CED/Book Draft version released 13th November 2000 at 22:01. .

which give the total electric and magnetic ﬁelds.2) . x ) def j(t ≡ ∂j ∂t t=tret ¤ 1 4πε0 c2 −∞ Erad (x) e−iωt dω ω V [˙ ret . we obtain the radiation ﬁelds Brad (t.1b) (8.8 Electromagnetic Radiation In this chapter we will develop the theory of electromagnetic radiation. In Chapter 3 we were able to derive general expressions for the scalar and vector potentials from which we then. We shall now study these ﬁelds further under the assumption that the observer is located in the far zone. We therefore study the radiation ﬁelds which are dominating in this zone.e.12) on page 46 and Equation (4. x) = = where ∞ −∞ ∞ Instead of studying the ﬁelds in the time domain. x ) × (x − x ) j(t d3x |x − x |2 (8.1a) ¡ ¤ ¤ (8. x ) × (x − x )] × (x − x ) 3 j(t dx |x − x |3 97 ¤ Brad (x) e−iωt dω = ω µ0 4πc V ˙ ret . 8.23) on page 49. in Chapter 4 calculated the total electric and magnetic ﬁelds from arbitrary distributions of charge and current sources.1 The radiation ﬁelds From Equation (4. x) = Erad (t.. we can often make a spectrum analysis into the frequency domain and study each Fourier compon- ¢ ˙ ret . The only limitation in the calculation of the ﬁelds was that the advanced potentials were discarded. very far away from the source region(s). and therefore study electric and magnetic ﬁelds which are capable of carrying energy and momentum over large distances. i.

3a) (8. . centred on x0 . x) eiωt dt Erad (x) = ω 2π −∞ k [jω (x ) × (x − x )] × (x − x ) ik|x−x | 3 = −i e dx 4πε0 c V |x − x |3 1 [jω (x ) × k] × (x − x ) ik|x−x | 3 = −i dx e 4πε0 c V |x − x |2 If the source is located inside a volume V near x0 and has such a limited spatial extent that max |x − x0 | |x − x |.98 E LECTROMAGNETIC R ADIATION S dS = nd2x ˆ ˆ k x−x x − x0 x x x0 x − x0 V O F IGURE 8. has a large enough radius |x − x0 | max |x − x0 |.22) on page 48. At distances much larger than the extent of V.11) on page 46 and Equation (4.3b) Draft version released 13th November 2000 at 22:01. and the integration surface S .uu.1a) (8. The Fourier representation of the radiation ﬁelds (8. respectively and are explicitly given by Brad (x) = ω 1 ∞ rad B (t. ent separately. and the unit vector ˆ ˆ k of the radiation k vector from x are nearly coincident.1b) were included in Equation (4.plasma. A superposition of all these components and a transformation back to the time domain will then yield the complete solution. normal to the surface S which has its centre at x0 .1 Downloaded from http://www.1: Relation between the surface normal and the k vector for radiation generated at source points x near the point x0 in the source volume V.se/CED/Book ¤ ¤ ¤ ¤ ¤ ¤ (8. we see from Figure 8. the unit vector n. x) eiωt dt 2π −∞ jω (x ) × (x − x ) ik|x−x | 3 kµ0 dx = −i e 4π V |x − x |2 µ0 jω (x ) × k ik|x−x | 3 = −i dx e 4π V |x − x | 1 ∞ rad E (t.

Within approximation (8.2 R ADIATED ENERGY 99 on the preceding page that we can approximate k x − x ≡ k · (x − x ) ≡ k · (x − x0 ) − k · (x − x0 ) ≈ k |x − x0 | − k · (x − x0 ) Recalling from Formula (F.3b) for the radiation ﬁelds can be approximated as µ0 ik|x−x0 | jω (x ) × k −ik·(x −x0 ) 3 e e dx 4π V |x − x | (8.se/CED/Book .49) on page 154 that dS = |x − x0 |2 dΩ = |x − x0 |2 sin θ dθ dϕ ˆ and noting from Figure 8. Equation (8.1 on the preceding page that k and n are nearly ˆ parallel. with integrals over points in the source volume only.8. We have to treat signals with limited lifetime and hence ﬁnite frequency bandwidth differently from monochromatic signals. we see that we can approximate.4) Downloaded from http://www.5) Both these approximations will be used in the following.plasma. and Erad .1a).1b) on page 97. Brad (x) ≈ −i ω 8.4) the expressions (8.e..6a) µ0 eik|x−x0 | −ik·(x −x0 ) 3 [jω (x ) × k] e ≈ −i dx 4π |x − x0 | V 1 [jω (x ) × k] × (x − x ) −ik·(x −x0 ) 3 Erad (x) ≈ −i eik|x−x0 | dx e ω 4πε0 c |x − x |2 V 1 eik|x−x0 | (x − x0 ) ≈i × [jω (x ) × k] e−ik·(x −x0 ) d3x 4πε0 c |x − x0 | |x − x0 | V (8.2 Radiated energy Let us consider the energy that carried in the radiation ﬁelds Brad . Equation (8.6b) I.uu.48) and formula (F. if max |x − x0 | |x − x |.3a) and (8. ˆ ˆ ˆ k · dS k· n = dS ≈ dΩ |x − x0 |2 |x − x0 |2 (8. then the ﬁelds can be approximated as spherical waves multiplied by dimensional and angular factors. ¤ ¤ ¤ ¤ ££ ££ (8. Draft version released 13th November 2000 at 22:01.

100 E LECTROMAGNETIC R ADIATION 8.1 Monochromatic signals If the source is strictly monochromatic.18) on page 26.5) on the preceding page.6b) and the fact that 1/c = √ √ ε0 µ0 and R0 = µ0 /ε0 according to the deﬁnition (2. V which is the radiated power per unit solid angle. we obtain S = V or. 8. ££ ¤£ £ 1 dP = R0 dΩ 32π2 2 (jω × k)e−ik·(x −x0 ) d3x −i(ω+ω )t ££ ££ ¤£ £ 1 1 R 2 0 32π |x − x0 |2 2 (jω × k)e−ik·(x −x0 ) d3x x − x0 |x − x0 | 3 ££ 2 2 3 ££ 2 S = E×H = dt 3 1 1 Re E × B∗ = Re Eω e−iωt × (Bω e−iωt )∗ 2µ0 2µ0 1 1 = Re Eω × B∗ e−iωt eiωt = Re Eω × B∗ ω ω 2µ0 2µ0 3 (8.se/CED/Book ¤ ¤ −∞ (Eω × Hω )e Draft version released 13th November 2000 at 22:01. making use of (8. we can obtain the temporal average of the radiated power P directly.plasma. The total energy transmitted through a unit area is the time integral of the Poynting vector: ∞ −∞ S(t) dt = = ∞ −∞ ∞ (E × H) dt dω ∞ −∞ dω ∞ −∞ If we carry out the temporal integration ﬁrst and use the fact that ∞ −∞ e−i(ω+ω )t dt = 2πδ(ω + ω ) Downloaded from http://www. simply by averaging over one period so that Using the far-ﬁeld approximations (8.2.2.11) .2 Finite bandwidth signals A signal with ﬁnite pulse width in time (t) domain has a certain spread in frequency (ω) domain.uu.6a) and (8.8) (8.7) ££ ¤ ¤ 2 (8. To calculate the total radiated energy we need to integrate over the whole bandwidth.9) ¤ ¤ (8.10) (8.

13) (8.15) includes only source coordinates.18) on page 26 for the vacuum resistance R 0 we obtain V which.uu. Draft version released 13th November 2000 at 22:01.3 R ADIATION FROM EXTENDED SOURCES 101 Equation (8. i. Parseval’s identity] ∞ −∞ S(t) dt = 2π = 2π = 2π = 2π 2π = µ0 2π = µ0 ∞ −∞ 0 (Eω × H−ω ) dω ∞ 0 0 −∞ −∞ 0 (Eω × B−ω + E−ω × Bω ) dω (Eω × B∗ + E∗ × Bω ) dω ω ω ∞ 0 where the last step follows from the real-valuedness of Eω and Bω .e.. is U= Inserting the approximations (8. We insert the Fourier transforms of the ﬁeld components which dominate at large distances.5) into Equation (8.plasma. ££ ¤£ £ dUω 1 dω ≈ R0 dΩ 4π (jω × k)e−ik·(x −x0 ) d3x ££ ¤ ££ ¤ ¤ S 0 V ££ ££ % 1 4π µ0 ε0 ∞ jω × k ik|x−x | 3 e dx |x − x | ££ ¤ 0 ∞ 0 2 ˆ ˆ k · ndS dω 2 dω Downloaded from http://www. is a good approximation to the energy that is radiated per unit solid angle dΩ in a frequency band dω.14) (8.13) and also introducing U= 0 ∞ Uω dω and recalling the deﬁnition (2.3b). It is important to notice that Formula (8. the radiation ﬁelds (8.15) ¤ ££ ¤ ¡ ¤ ¡ ¤ ¤ ¤ ¡¤ ¤ ¤ (Eω × H−ω ) dω + (Eω × H−ω ) dω . The result.10) on the preceding page can be written [cf. at large distances. after integration over the area S of a large sphere which encloses the source.4) and (8. 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).12) (8.3a) and (8.se/CED/Book ¢ (Eω × H−ω ) dω + ¤ ∞ ∞ 0 ∞ (E−ω × Hω ) dω ¢ (Eω × H−ω ) dω − (Eω × H−ω ) dω ¢ (8.8.

1 Linear antenna Let us apply Equation (8. we need Downloaded from http://www. in practice. symmetric or in any other way simple enough that a direct calculation of the radiated ﬁelds and energy is possible. at least in principle.3. transmitting antenna. Certain radiation systems have a geometry which is one-dimensional. In the general case.102 E LECTROMAGNETIC R ADIATION 8. the current must vanish at the end points −L/2 and L/2.se/CED/Book ££ ££ (8. one can.9) on page 100 for calculating the power from a linear. This is for instance the case when the current ﬂows in one direction in space only and is limited in extent. thin conductor of length L which carries a one-dimensional time-varying current so that it produces electromagnetic radiation. Poynting ﬂux and energy for an arbitrary current density Fourier component. However. The current therefore forms a standing wave with wave number k = ω/c and can be written j0 (x ) = I0 δ(x1 )δ(x2 ) sin[k(L/2 − x3 )] x3 ˆ sin(kL/2) where the amplitude I0 is constant. An example of this is a linear antenna. Since we can assume that the antenna wire is inﬁnitely thin. one has to resort to approximations. We assume that the conductor resistance and the energy loss due to the electromagnetic radiation are negligible. We shall consider both these situations. The antenna is a straight.plasma. 8. it is often difﬁcult to evaluate the integrals unless the current has a simple distribution in space. calculate the radiated ﬁelds.uu. fed across a small gap at its centre with a monochromatic source.3 Radiation from extended sources As shown above.16) inserted.16) Draft version released 13th November 2000 at 22:01. In order to evaluate formula (8.9) on page 100 with the explicit monochromatic current (8. .

For the technologically important case of a half-wave antenna.18) above reduces to 0 The integral in (8.21) Downloaded from http://www. formula (8.17) 2 sin θ dθ (8.19) Ω (8.9) on page 100 and integrating over θ.21) can be evaluated numerically.8.plasma..18) (8.e. 4 ¤ 2 P(λ/2) = R0 I0 1 4π π cos2 cos θ dθ sin θ π 2 ¢ ≈ 197 ¡ ¢ 5 ¡ ¢ kL→0 L λ 2 2 R0 J0 L λ 2 ¢ ¡ ¤ 1 4π π cos[(kL/2) cosθ] − cos(kL/2) sin θ sin(kL/2) ¢ 2 = 4I0 cos[(kL/2) cosθ] − cos(kL/2) sin θ sin(kL/2) (8.3 R ADIATION FROM EXTENDED SOURCES 103 the expression (j0 × k)e−ik·(x −x0 ) d3x L/2 V = = 2 0 sin[k(L/2 − x3 )] cos(kx3 cos θ) dx3 2 inserting this expression and dΩ = 2π sin θ dθ into formula (8.20) (8. It can also be evaluated Draft version released 13th November 2000 at 22:01.uu.se/CED/Book ££ 2 I0 k2 sin2 θ eikx0 cos θ 2 sin (kL/2) 2 L/2 ££ ££ £ sin[k(L/2 − x3 )] I0 k sin θe−ikx3 cosθ e−ikx0 cosθ dx3 sin(kL/2) −L/2 ££ ¤ £ £ £ ££ £ ££ ££ ££ ££ ¡ ££ ¡ ¤£ ££ ££ ¤£ ££ £ 2 2 2 . The quantity Rrad (L) = π P(L) P(L) = 1 2 = R0 2 6 Ieff 2 I0 L λ 2 is called the radiation resistance. we ﬁnd that the total radiated power from the antenna is 2 P(L) = R0 I0 0 One can show that lim P(L) = π 12 where λ is the vacuum wavelength. i. for L = λ/2 or kL = π.

24a) (8.21) on page 9. can be written ∂ρ(t. 8.21) on the preceding page we obtain the value Rrad (λ/2) ≈ 73 Ω. and the cosine integral Ci(x) were introduced.se/CED/Book Draft version released 13th November 2000 at 22:01.1 The Hertz potential Let us consider the continuity equation. ¤ 4 ¤ 0 ) ¤ ¤ ¤ ¤ 4 cos2 1 cos2 π v cos θ 2 dθ = cos θ → v = dv = 2 sin θ −1 1 − v 1 + cos(πu) π cos2 v = 2 2 1 1 + cos(πu) 1 = du 2 −1 (1 + u)(1 − u) (8. We shall use Hertz’ method to obtain this expansion. x) + · j(t.5772 .24b) (8. which.plasma. according to expression (1. . 8. and when we are interested in evaluating the radiation far from the source volume.22) 1 1 1 + cos(πu) 1 1 1 + cos(πu) du + du = 4 −1 (1 + u) 4 −1 (1 − u) 1 1 1 + cos(πu) v = du = 1 + u → 2 −1 (1 + u) π 2π 1 − cos v 1 1 = dv = [γ + ln 2π − Ci(2π)] 2 0 v 2 ≈ 1.4.104 E LECTROMAGNETIC R ADIATION analytically as follows: π 0 where in the last step the Euler-Mascheroni constant γ = 0. we can introduce an approximation which leads to a multipole expansion where individual terms can be evaluated analytically. x) such that · π = −ρtrue ∂π = jtrue ∂t (8. Inserting this into the expression Equation (8.22 π 2 5 5 ¤ .uu.4 Multipole radiation In the general case. x) = 0 ∂t If we introduce a vector ﬁeld π(t.23) Downloaded from http://www. .

25a) (8. Therefore. Furthermore. Draft version released 13th November 2000 at 22:01.30b) Clearly.27) Downloaded from http://www. Πe acts as a “super-potential” in the sense that it is a potential from which we can obtain other potentials. as can be seen from (8. It is called the Hertz’ vector or polarisation potential and. x) = ¤ 6 (8.9) on page 89]. ¤ Πe (t. As we see.8.30a) (8. this is no essential limitation.29) we see that we can calculate the magnetic and electric ﬁelds. Equation (7.26) 1 4πε0 π(tret .4 M ULTIPOLE RADIATION 105 and compare with Equation (8.19) on page 38. it satisﬁes the inhomogeneous wave equation 2 Πe = 1 ∂2 e Π − c2 ∂t2 2 Πe = π ε0 This equation is of the same type as Equation (3. and has therefore the retarded solution with Fourier components Πe (x) = ω 1 4πε0 πω (x )eik|x−x | 3 dx |x − x | (8.24) and (8.23) on the facing page. we see that the quantity π is related to the “true” charges in the same way as P is related to polarised charge.25). We introduce a further potential Πe with the following property · Πe = −φ 1 ∂Πe =A c2 ∂t (8.se/CED/Book .uu.25b) where φ and A are the electromagnetic scalar and vector potentials. x) satisﬁes this continuity equation. a long distance from the source region. as follows B= E= 1 ∂C c2 ∂t ×C (8.plasma. we see that π(t. where · E = 0. Since we are mainly interested in the ﬁelds in the far zone. π is referred to as the polarisation vector. x ) 3 dx |x − x | (8. respectively.28) If we introduce the help vector C such that C= × Πe (8. respectively. the last equation is valid only outside the source volume. if we compare with the electric polarisation [cf.

uu.32) (8.32) above.31) where eik|x−x | is a Green´s function |x − x | Θ is the angle between x − x0 and x − x0 Pn (cos Θ) is the Legendre polynomial of order n jn (k x − x0 ) is the spherical Bessel function of the ﬁrst kind of order n h(1) (k |x − x0 |) is the spherical Hankel function of the ﬁrst kind of order n n According to the addition theorem for Legendre polynomials.27) on the previous page the Hertz’ vector.plasma. we can in a formally exact way expand the Fourier component of the Hertz’ vector as Πe = ω × ik ∞ n ∑ ∑ (2n + 1)(−1)mh(1) (k |x − x0 |) Pm (cos θ) eimϕ n n 4πε0 n=0 m=−n V πω (x ) jn (k x − x0 ) P−m (cos θ n )e −imϕ We notice that there is no dependence on x − x0 inside the integral. into Equation (8. in a formal series.33a) (8.33b) (8. x ) in the vicinity of x0 . together with Equation (8. the integrand is only dependent on the relative source vector x − x0 .31). Under these assumptions. φ) Inserting Equation (8. θ. ££ ££ ££ ££ ££ ££ ££ ¤ ££ (8. x − x0 = ( x − x0 . θ . For this purpose we recall from potential theory that eik|x−x | eik|(x−x0 )−(x −x0 )| ≡ |x − x | |(x − x0 ) − (x − x0 )| ∞ n=0 = ik ∑ (2n + 1)Pn (cos Θ) jn (k x − x0 )h(1) (k |x − x0 |) n (8.34) dx 3 . in spherical polar con ordinates. φ ) x − x0 = (|x − x0 | .106 E LECTROMAGNETIC R ADIATION Assume that the source region is a limited volume around some central point x0 far away from the ﬁeld (observation) point x.28) on the preceding page. we can write Pn (cos Θ) = n m=−n ∑ (−1)m Pm (cos θ)P−m(cos θ )eim(ϕ−ϕ ) n n where Pm is an associated Legendre polynomial and. due to the presence of non-vanishing π(tret . we can expand expression (8.se/CED/Book Draft version released 13th November 2000 at 22:01. Downloaded from http://www.

Draft version released 13th November 2000 at 22:01.8. Then we may to a good approximation replace h(1) with the ﬁrst term in n its asymptotic expansion: h(1) (k |x − x0 |) ≈ (−i)n+1 n eik|x−x0 | k |x − x0 | (8. i. Taking the inverse Fourier transform of Πe will yield the Hertz’ vector in time domain.30) on page 105 in order to obtain the radiation ﬁelds. the expansions of the ω spherical Hankel and Bessel functions used above may not be consistent and must be replaced by more accurate expressions.uu.32) on the preceding page.36) and replace jn with the ﬁrst term in its power series expansion: jn (k x − x0 ) ≈ Inserting these expansions into Equation (8.38a) Downloaded from http://www. if many Πe (n) terms are needed for an accurate result. however. Let us now study the lowest order contributions to the expansion of Hertz’ vector.34) on the preceding page.e.37) (8.. In the general case. ££ ££ ¤ 1 eik|x−x0 | 2n n! 4πε0 |x − x0 | (2n)! 5 ££ ££ 4 ££ ££ ££ ££ 1 k |x − x0 | (8.38b) n=0 ∑ Πe (n) ω ∞ V This expression is approximately correct only if certain care is exercised. we obtain the multipole expansion of the Fourier component of the Hertz’ vector Πe ≈ ω where Πe (n) = (−i)n ω πω (x ) (k x − x0 )n Pn (cos Θ) d3x (8. and Pn (cos θ) gives the angular distribution of the radiation.se/CED/Book . when the following inequalities k x − x0 hold.29) on page 105 will yield C. the angular distribution must be computed with the help of formula (8.35) 2n n! k x − x0 (2n + 1)! n (8. which inserted ω into Equation (8.4 M ULTIPOLE RADIATION 107 We are interested in the case where the ﬁeld point is many wavelengths away from the well-localised sources.38b).plasma. Θ = θ in expression (8. For a linear source distribution along the polar axis. The resulting expression can then in turn be inserted into Equation (8.

If a spherical coordinate system is chosen with its polar axis along pω .41) (8.108 E LECTROMAGNETIC R ADIATION 8.40a) (8. we can now write down the Fourier components of the radiation parts of the magnetic and electric ﬁelds from the dipole: ωµ0 eik|x−x0 | (pω × k) 4π |x − x0 | 1 eik|x−x0 | Erad = − [(pω × k) × k] ω 4πε0 |x − x0 | Brad = − ω (8.42b) ¡ Draft version released 13th November 2000 at 22:01. Equation (7.30) on page 105.38b) on the preceding page. with the spherically polar components (8. we obtain Πe (0) = ω eik|x−x0 | 4πε0 |x − x0 | πω (x ) d3x = 1 eik|x−x0 | pω 4πε0 |x − x0 | (8.2 Electric dipole radiation Choosing n = 0 in expression (8.se/CED/Book § ¢ (0) ˆ Cω = Cω. the components of Πe (0) are ω 1 eik|x−x0 | pω cos θ 4πε0 |x − x0 | 1 eik|x−x0 | Πθ = −Π(0) sin θ = − pω sin θ ω 4πε0 |x − x0 | Πϕ = 0 Πr = Π(0) cos θ = ω Evaluating formula (8.42a) (8. .43a) (8.2) on page 87 which describes the static dipole moment.40c) 1 eik|x−x0 | ˆ − ik pω sin θ φ |x − x0 | |x − x0 | (8. cf.4.39) V where pω = V πω (x ) d3x is the Fourier component of the electric dipole moment.40b) (8. we obtain ω Applying this to Equation (8.uu. we obtain directly the Fourier components of the ﬁelds Bω = −i ωµ0 1 eik|x−x0 | ˆ − ik pω sin θ φ 4π |x − x0 | |x − x0 | 1 ik 1 x − x0 − Eω = 2 cos θ 2 4πε0 |x − x0 | |x − x0 | |x − x0 | ik|x−x0 | 1 ik ˆ e − k2 sin θ θ pω + − 2 |x − x0 | |x − x0 | |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 = ω/c.ϕ φ = 1 4πε0 ¢ ¢ ¢ ¡ ¤ ¡ ¡ ¦ (8.40) of Πe (0) inserted.29) on page 105 for the help vector C.plasma.43b) Downloaded from http://www.

se/CED/Book .e.i )(xi − x0.i ) πω (x ) and introducing ηi = xi − x0.i the jth component of the integrand in Πe (1) can be broken up into ω 1 {[(x − x0 ) · (x − x0 )] πω (x )} j = ηi πω.48) j Downloaded from http://www.8.24) on page 104.4.i η j = [πω. the ﬁrst being symmetric and the second antisymmetric in the indices i. We note that the antisymmetric part can be written as 1 1 ηi πω.i η j 2 1 + ηi πω. π(t.44) (8. also known as E1 radiation.i (8. the term [(x − x0 ) · (x − x0 )] πω (x ) can be rewritten [(x − x0 ) · (x − x0 )] πω (x ) = (xi − x0.3 Magnetic dipole radiation The next term in the expression (8.46b) (8. as the sum of two parts.38b) on page 107 for the expansion of the Fourier transform of the Hertz’ vector is for n = 1: Πe (1) = −i ω eik|x−x0 | k x − x0 πω (x ) cos Θ d3x 4πε0 |x − x0 | V 1 eik|x−x0 | [(x − x0 ) · (x − x0 )] πω (x ) d3x = −ik 4πε0 |x − x0 |2 V Here.plasma.46a) (8.. 5 3 5 4 ££ 4 ££ ¤ 2 ¤ (8. j ηi − πω. 8.4 M ULTIPOLE RADIATION 109 These ﬁelds constitute the electric dipole radiation. j )] 2 2 1 = [πω (η · η ) − η (η · πω )] j 2 1 (x − x0 ) × [πω × (x − x0 )] = 2 Equations (8.49) Draft version released 13th November 2000 at 22:01. j (ηi ηi ) − η j (ηi πω.uu. j ηi − πω. j. j ηi + πω. and the fact that we are considering a single Fourier component.i η j 2 i. x) = πω e−iωt (8.45) ηi = xi − x0.47) 5 4 (8.

we can write the antisymmetric part of the integral in formula (8. we insert this expression into Equation (8.38b) on page 107 for the expansion of the Hertz’ vector can be expressed in terms e. all terms belonging to the near ﬁelds and transition ﬁelds and keeping only the terms that dominate at large distances.antisym ω (1) =− k eik|x−x0 | (x − x0 ) × mω 4πε0 ω |x − x0 |2 In analogy with the electric dipole case.54b) which are the ﬁelds of the magnetic dipole radiation (M1 radiation).plasma. 8.4. as before.53) Draft version released 13th November 2000 at 22:01.54a) (8.4 Electric quadrupole radiation The symmetric part Πω of the n = 1 contribution in the Equation (8.30) on page 105 then gives the B and E ﬁelds.uu.se/CED/Book ¤ ¤ (8. we obtain Brad (x) = − ω µ0 eik|x−x0 | (mω × k) × k 4π |x − x0 | k eik|x−x0 | mω × k Erad (x) = ω 4πε0 c |x − x0 | (8. part of Π e (1) ω can be written Πe. Discarding.51) ¤ (8. with which Equations (8.44) on the preceding page as 1 (x − x0 ) × πω (x ) × (x − x0 ) dV 2 V 1 = i (x − x0 ) × jω (x ) × (x − x0 ) d3x 2ω V 1 = −i (x − x0 ) × mω ω where we introduced the Fourier transform of the magnetic dipole moment mω = 1 2 (x − x0 ) × jω (x ) d3x (8.52) V The ﬁnal result is that the antisymmetric.110 E LECTROMAGNETIC R ADIATION allow us to express πω in jω as πω = i jω ω (8. magnetic dipole.29) on page 105 to evaluate C.50) Hence. .sym (1) Downloaded from http://www.

56a) (8. 8. x) = (8. 5 5 4 4 ¤ ¤ ¤ (8. Tedious. unstructured and rigid “charge distribution” with a small. x ) 3 1 dx 4πε0 V |x − x | µ0 j(tret .3) on page 87: Q(t) = V (x − x0 )(x − x0 )ρ(t.8. The part of this “charge distribution” dq which we are considering is located in dV = d3x in the sphere in Figure 8. respectively.56b) Downloaded from http://www.se/CED/Book .5 Radiation from a localised charge in arbitrary motion The derivation of the radiation ﬁelds for the case of the source moving relative to the observer is considerably more complicated than the stationary cases studied above.57) is V ρ(tret . we consider a charge q .30) on page 105.5 R ADIATION FROM A LOCALISED CHARGE IN ARBITRARY MOTION 111 of the electric quadrupole tensor. Since we assume that the electron (or any other other similar electric charge) is moving with a velocity v whose direction is arbitrary and whose magnitude can be almost comparable to the speed of light. The radiation components of the ﬁelds in the far ﬁeld zone (wave zone) are given by iµ0 ω eik|x−x0 | k · Qω × k 8π |x − x0 | i eik|x−x0 | k · Qω × k × k Erad (x) = ω 8πε0 |x − x0 | Brad (x) = ω This type of radiation is called electric quadrupole radiation or E2 radiation. In order to handle this non-stationary situation. yields the resulting ﬁelds. can be thought of as a localised. we cannot say that the charge and current to be used in (8.55) Again we use this expression in Equation (8. x) = dx 4π V |x − x | φ(t. which is deﬁned in accordance with Equation (7.57b) and consider a source region with such a limited spatial extent that the charges and currents are well localised. x ) 3 A(t. Draft version released 13th November 2000 at 22:01. x ) d3x .29) on page 105 to calculate the ﬁelds via Equations (8.uu. for instance an electron.plasma. ﬁnite radius.36) on page 41 in Chapter 3 ρ(tret . classically. part of the charge distribution will “leak” out of the volume element d3x . we use the retarded potentials (3. x ) d3x and V vρ(tret . because in the ﬁnite time interval during which the observed signal is generated. Speciﬁcally. but fairly straightforward algebra (which we will not present here). x ) d3x (8. which.5 on the following page.57a) (8.

Since dt = dr/c and dS dr = d3x we can rewrite this expression for the net charge as dq = ρ(tret . x ) (x − x ) · v dS dt |x − x | (8.se/CED/Book v x−x x (t ) dS (8. as time increases. x ) d3x − ρ(tret .5. centred on x and expanding. x ) 1 − d3x c |x − x | Downloaded from http://www. t + dt) is (r.uu. .1 The Liénard-Wiechert potentials The charge distribution in Figure 8.2: Signals which are observed at the ﬁeld point x at time t were generated at source points x (t ) on a sphere.58) where the last term represents the amount of “source leakage” due to the fact that the charge distribution moves with velocity v. The radius interval of this sphere from which radiation is received at the ﬁeld point x during the time interval (t. The source charge element moves with an arbitrary velocity v and gives rise to a source “leakage” out of the source volume dV = d3x . x ) dS dr − ρ(tret . with the velocity c outward from the centre.plasma.5 on page 112 which contributes to the ﬁeld at x(t) is located at x (t ) on a sphere with radius r = |x − x | = c(t − t ). r + dr) and the net amount of charge in this radial interval is dq = ρ(tret . x ) (x − x ) · v 3 dx c |x − x | (x − x ) · v = ρ(tret .112 E LECTROMAGNETIC R ADIATION x(t) dr dV q F IGURE 8.59) ¢ ¡ Draft version released 13th November 2000 at 22:01. 8.

64a) (8. The potentials (8.5 R ADIATION FROM A LOCALISED CHARGE IN ARBITRARY MOTION 113 or ρ(tret . x ) d3x = dq 1 − (x−x )·v c|x−x | (8. x) = (x−x )·v 2 2 s 4πε0 c |x − x | − c 4πε0 c c φ(t.8.64c) Downloaded from http://www.63) are precisely the LiénardWiechert potentials which we derived in Section 5. The result is (recall that j = ρv) φ(t. It is important to realise that in the complicated derivation presented here. assuming that the integrands do not change sign in the integration volume.62a) (8.57) on page 111 for the retarded potentials.61) This is the expression to be used in the formulae (8.63a) (8.60) which leads to the expression ρ(tret . the observer is in a coordinate system which has an “absolute” meaning and the Draft version released 13th November 2000 at 22:01. x) = = 2 φ(t. x) = where s = x−x − (x − x ) · v c x−x v · = x−x 1− |x − x | c x−x v = (x − x ) · − |x − x | c is the retarded relative distance. x) = 1 4πε0 dq |x − x | − (x−x )·v c v dq |x − x | − (x−x )·v c (8. ¢ ¢ ¡ ¤ µ0 4π ¤ (8.uu. x) = A(t.63b) £ ¡£ ££ ££ ££ (8. use the mean value theorem and the fact that V dq = q to evaluate these expressions to become 1 q q 1 = (x−x )·v 4πε0 |x − x | − c 4πε0 s q v q v v A(t.2 on page 63 by using a covariant formalism. x ) 3 dq dx = |x − x | |x − x | − (x−x )·v c (8.plasma.3.se/CED/Book .64b) (8.62b) For a sufﬁciently small and well localised charge distribution we can.

45) on page 62.se/CED/Book ¢ Aµ (xκ ) = q 4πε0 ¡ 1 v . After time t the particle.66a) (8.66b) ∂A(t. x) = − φ(t. s cs = (φ. whereas in the covariant derivation two frames of equal standing were moving relative to each other with v. Equation (5.plasma. velocity v is that of the particle.2 Radiation from an accelerated point charge Consider a localised charge q and assume that its trajectory is known experimentally as a function of retarded time x = x (t ) (8.67) Downloaded from http://www.uu. has followed a yet unknown trajectory. x) (8. . deﬁnes the virtual simultaneous coordinate x0 (t).5. x) E(t. x) = × A(t. The electric and magnetic ﬁelds are calculated from the potentials in the usual way: B(t. Extrapolating tangentially the trajectory from x (t ). x) − ∂t 8. which moves with nonuniform velocity. cA) (8. Expressed in the four-potential.3: Signals which are observed at the ﬁeld point x at time t were generated at the source point x (t ). the Liénard-Wiechert potentials become The Liénard-Wiechert potentials are applicable to all problems where a spatially localised charge emits electromagnetic radiation.114 E LECTROMAGNETIC R ADIATION ? q x (t ) θ |x − x | v c x0 (t) θ0 v(t ) x − x0 x−x x(t) F IGURE 8. and we shall now study such emission problems.65) Draft version released 13th November 2000 at 22:01. based on the velocity v(t ).

8. If we choose the ﬁeld point x as ﬁxed. In these formulae the unprimed . we have in general no knowledge of the velocity and acceleration at times later than t . which in turn is expressed implicitly in t via Equation (8. and the unprimed time derivative operator ∂/∂t means that we differentiate with respect to t while keeping x ﬁxed.68) to the relative vector x − x yields d (x − x (t )) = −v(t ) dt d2 (x − x (t )) = −˙ (t ) v dt 2 (8. the trajectory at times later than t is not (yet) known. But the Liénard-Wiechert potentials φ and A. we drop the subscript “ret” on t from now on). be calculated from the implicit relation t = t (t. i. According to formulae (8.5 R ADIATION FROM A LOCALISED CHARGE IN ARBITRARY MOTION 115 (in the interest of simplifying our notation. Despite this complication it is Draft version released 13th November 2000 at 22:01.e. x2 .68a) and the retarded relative distance s(t . at least in principle.uu. The retarded velocity and acceleration at time t are given by v(t ) = dx dt dv d2 x = dt dt 2 (8.68a) (8. x) given by Equation (8. This means that the expressions for the potentials φ and A contain terms which are expressed explicitly in t .69a) (8.69b) The retarded time t can. application of (8. This means that we know the trajectory of the charge q .68b) ˙ a(t ) = v(t ) = As for the charge coordinate x itself.e. Because of the ﬁnite speed of propagation of the ﬁelds. x3 ) while keeping t ﬁxed. i. x) = t − |x − x (t )| c (8.. for all times up to the time t at which a signal was emitted in order to precisely arrive at the ﬁeld point x at time t. Downloaded from http://www.64) on page 113.66) on the facing page 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 particular not at the time of observation t. the spatial derivative differentiation operator = xi ∂/∂xi means that ˆ we differentiate with respect to the coordinates x = (x1 ..70) and we shall see later how this relation can be taken into account in the calculations.70).se/CED/Book .63) on page 113. Equations (8. x .plasma. are expressed in the charge velocity v(t ) given by Equation (8.

respectively.70) on the previous page we ﬁnd that ∂t ∂t |x − x (t (t.64) on page 113. we obtain the following useful operator identity ∂ ∂t = ∂t ∂t ∂ ∂t = |x − x | s x x x Likewise. as we shall see below. x))| =− · ( )t (x − x ) c c |x − x | x−x (x − x ) · v =− + ( )t t c |x − x | c |x − x | Downloaded from http://www. by applying the operator (∂/∂t)x to Equation (8. With this convention. Making use of Equation (8.116 E LECTROMAGNETIC R ADIATION possible.plasma. we ﬁnd that x − x (t ) = x − x (t ) = − x x Furthermore. by applying ( )t to Equation (8.70) on the preceding page we obtain ( )t t = −( )t x−x |x − x (t (t.73) above. ¢ ¡ ¢ ¡ ¢ ¡ ¢ ∂t ∂t = x |x − x | |x − x | = s |x − x | − (x − x ) · v/c ¢ ¡ § ¢ ¡ ¢ ¢ ¡ ∂ ∂t x ∂ ∂t x 5 4 ¢ ¦ ¡ ¢ ¢ ¢ ∂ ∂t ∂ x−x · |x − x | ∂t ¡ ££ ££ ¡ ¡ ¡ ¡ (x − x ) · v |x − x | (8. The differential operator method We introduce the convention that a differential operator embraced by parentheses with an index x or t means that the operator in question is applied at constant x and t. To this end.uu. x) is the retarded relative distance given by Equation (8.73) (8.71) (8.72) (8. to determine the electric and magnetic ﬁelds and associated quantities at the time of observation t. x))| c x x ∂ ∂t |x − x | = 1− ∂t x c ∂t (x − x ) · v ∂t = 1+ c |x − x | ∂t x = 1− This is an algebraic equation in (∂t /∂t)x which we can solve to obtain where s = s(t .75) .se/CED/Book Draft version released 13th November 2000 at 22:01. we need to investigate carefully the action of differentiation on the potentials.74) (8.

Downloaded from http://www.76) which gives the following operator relation when ( )t is acting on an arbitrary function of t and x: x With the help of the rules (8. We ﬁnd.5 R ADIATION FROM A LOCALISED CHARGE IN ARBITRARY MOTION 117 This is an algebraic equation in ( )t with the solution ( )t t = − x−x cs (8.64a) on page 113 for the retarded relative distance s(t .80) . we see that we can evaluate (∂s/∂t )x in the following way ∂s ∂t = ∂ ∂t x−x − x x =− ˙ (x − x ) · v v2 (x − x ) · v + − c c |x − x | Draft version released 13th November 2000 at 22:01.77) ¢ ¡ (8.78a) ¡ (8.74) we are now able to replace t by t in the operations which we need to perform.77) and (8.78b) (8.79) − ˙ |x − x | v 2 c ¢ ¡ (8.se/CED/Book I 4 = H ∂ 1 x − x (t ) − ∂t c ∂ x − x (t ) ∂v(t ) · v − (x − x ) · ∂t ∂t 5 ¢ (x − x ) · v c § ¢ − (x − x ) − |x − x | v/c cs ¡ q (x − x ) − |x − x | v/c 2 4πε0 s |x − x | § ¢ x ∂s ∂t ¢ § ¢ £ ¡ £ ¡ ££ ¢ ¢ ££ ££ ££ ¡ ££ ¡ ¢ ¦ ££ ¡ ¦ ¦ ¢ ££ ( )t = ( )t t ∂ ∂t + ( )t = − x−x cs ¡ ¡ ∂ ∂t + ( )t x (8. Equations (8.63) on page 113.8. we obtain E(t.plasma. x). for instance. that 1 q 4πε0 s x−x q v x − x ∂s =− − − 2 |x − x | 4πε0 s c cs ∂t x ∂A ∂A ∂ µ0 q v ≡ = ∂t ∂t x ∂t 4π s q ∂s = x − x s˙ − x − x v v 4πε0 c2 s3 ∂t φ ≡ ( φ)t = Utilising these relations in the calculation of the E ﬁeld from the LiénardWiechert potentials.uu. x) = − φ − = ∂A ∂t x Starting from expression (8.

83) (8. the electric ﬁeld generated by an arbitrarily moving charged particle at x (t ) is given by the expression Radiation ﬁeld The ﬁrst part of the ﬁeld.68) on page 115. the virtual simultaneous coordinate.81) (8.71) on page 116 and Equations (8. During the arbitrary motion.81) in the following way x − x0 = (x − x ) − (8. this relative vector is given by |x − x | v c This allows us to rewrite Equation (8. we interpret x − x0 as the coordinate of the ﬁeld point x relative to the virtual simultaneous coordinate x0 (t). Hence. x) = q v2 (x − x0 ) 1 − 2 4πε0 s3 c + (x − x ) × ˙ (x − x0 ) × v 2 c F§ ¢ D GC ¦ x−x + 2 × c Coulomb ﬁeld when v → 0 |x − x | v ˙ (x − x ) − ×v c F¢ ¢ D EC ¡ ¡ B B E(t. respectively. were used. is radiated into the far zone and is therefore also called the radiation ﬁeld. the velocity ﬁeld.118 E LECTROMAGNETIC R ADIATION where Equation (8.63) on page 113 and formula (8. The second part of the ﬁeld. But. according to (8.85) Downloaded from http://www. x−x x−x q × ( )t φ = ×v 2 s2 |x − x | c |x − x | 4πε0 c (8. x) = q 4πε0 s3 (x − x ) − |x − x | v c ¡ ¡ 1− v2 c2 (8. From Figure 8. the acceleration ﬁeld. Since the time it takes from a signal to propagate (in the assumed vacuum) from x (t ) to x is |x − x | /c.74) on page 116.uu. ¢ B(t.82) In a similar manner we can compute the magnetic ﬁeld: where we made use of Equation (8.2 on page 114 we see that the position the charged particle would have had if at t 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 x (t ) is x0 (t). x) = x−x ∂ × cs ∂t q x−x ∂A x−x =− ×v− × 2 s2 |x − x | 4πε0 c c |x − x | ∂t x ¢ ¡ × A ≡ ( ) t × A = ( )t × A − A x § ¡ ¢ ¦ E(t.se/CED/Book Draft version released 13th November 2000 at 22:01.plasma.5.78a). tends to the ordinary Coulomb ﬁeld when v → 0 and does not contribute to the radiation.84) .

Using this formula and formula (8.86) (8.se/CED/Book ¢ (x − x ) · v (x − x ) · v + −2 x−x c c 2 § q |x − x | v ˙ (x − x ) × (x − x ) − ×v 2 s3 4πε0 c c q ˙ = (x − x ) × [(x − x0 ) × v] 4πε0 c2 s3 ¢ ¡ § ¢ x ¡ ¦ ££ ££ ¦ B(t. x) where in the last step we again used formula (8. x) c |x − x | ¡ (8.5 R ADIATION FROM A LOCALISED CHARGE IN ARBITRARY MOTION 119 so that The radiation part of the electric ﬁeld is obtained from the acceleration ﬁeld in formula (8. x) = x−x ∂A × −( φ)t − c |x − x | ∂t x−x = × E(t. the square of this retarded relative can be written s (t . According to Equation (8.plasma.64) on page 113. x) = = |x−x |→∞ lim E(t.86) above. An example of such a quantity is the retarded relative distance s(t . ¢ (x − x ) × v c 2 Downloaded from http://www.90) .89) ¡ (8. x) c |x − x | (8.uu.8.81) on the preceding page as Erad (t. the radiation part of the magnetic ﬁeld can be written Brad (t. x) = x − x 2 2 If we use the following handy identity (x − x ) · v c = 2 + |x − x |2 v2 |x − x |2 v2 2 cos2 θ + sin θ c2 c2 |x − x |2 v2 |x − x |2 v2 = (cos2 θ + sin2 θ ) = c2 c2 Draft version released 13th November 2000 at 22:01.82) on the facing page. x).87) ¡ ££ ¢ ££ (8.88) The direct method An alternative to the differential operator transformation technique just described is to try to express all quantities in the potentials directly in t and x. x) = x−x × Erad (t.

in other words. this expression becomes s2 = x − x 2 where in the penultimate step we used Equation (8. from Equation (8. the retarded distance s in the LiénardWiechert potentials (8. viz.uu.e. in the case the particle velocity at time t can be calculated or projected..94) .e.63) can be expressed in terms of the virtual simultaneous coordinate x0 (t).plasma. when we obtain the ﬁrst knowledge of its existence at the source point x at the retarded time t .89) on the previous page for s2 . ¢ ¢ = (x − x ) − |x − x | v c 2 − (x − x0 ) × v c 2 ¢ −2 x−x (x − x ) · v |x − x |2 v2 (x − x0 ) × v + − 2 c c c ¡ ¢ (x − x0 ) × v |x − x |2 v2 − c2 c ¢ 2 2 (x − x ) × v |x − x |2 v2 = − c2 c ¡ ¡ ¡ ££ ¡ ££ ¡ ¢ ¢ ££ ¡ ££ ¡ ¡ 2 (8. can. yields the relation (x − x ) · v c 2 = Inserting the above into expression (8.se/CED/Book ¢ ≡ |x − x0 |2 − (x − x0 ) × v c 2 ¢ (x − x0 (t)) × v = (x − x0 (t)) − c 2 2 Draft version released 13th November 2000 at 22:01. when inserted into Equation (8. when the motion of the charge is somehow known. and in the ﬁeld coordinate x(t).. What we have just demonstrated is that.91) (8. we obtain the following identity: (x − x ) × v = (x − x0 ) × v which. where we make our observations..82) on page 118.120 E LECTROMAGNETIC R ADIATION we ﬁnd that (x − x ) · v c 2 Furthermore. in this special case we are able to express the retarded relative distance as s = s(t.91). We have. x) and we do not have to involve the retarded time t or transformed differential operators in our calculations. Downloaded from http://www.93) (8. and hence s itself. shown that all quantities in the deﬁnition of s. be expressed in terms of the time t alone.82) on page 118. the point at which the particle will have arrived at time t. i. I.92) (8.

5.uu.95a) (8. x) = ¡ (x − x0 ) × v c 2 (8. Since the particle is not accelerated.8. we obtain s2 = |x − x0 |2 1 − v2 c2 − (x − x0 ) · v c 2 which we shall use in the following example of a uniformly.1 Downloaded from http://www.. v ≡ 0.e. the angle between x − x0 and v is θ0 while then angle between x − x and v is θ .5 R ADIATION FROM A LOCALISED CHARGE IN ARBITRARY MOTION 121 Taking the square root of both sides of Equation (8.95c) and standard vector analytic formulae. As depicted in Figure 8. from its posi˙ tion at the retarded time. It will therefore move uniformly in a straight line with the constant velocity v. 9 A0 v v × (x − x0 ) = 2 (x − x0 ) + × c c § ¢ = 2 (x − x0 ) 1 − v2 c2 + vv · (x − x0 ) c2 Y ¢ ¢ ¡ ¢ |x − x0 |2 1 − v2 c2 − ¡ ¢ ) ¡ ¡ ¡ % = |x − x0 | v2 2 sin θ0 c2 (x − x0 ) · v c ¢ 2 s(t.95c) X ¦ 8 (8. we obtain the following alternative ﬁnal expressions for the retarded relative distance s in terms of the charge’s virtual simultaneous coordinate x 0 (t): |x − x0 |2 − 1− = Using Equation (8. x (t ). T HE FIELDS FROM A UNIFORMLY MOVING CHARGE In the special case of uniform motion. time and space derivatives are closely related in the following way when they operate on functions of x(t) : ∂ → −v · ∂t (8. We note that in the case of uniform velocity v. x (t).96) E XAMPLE 8.94) on the preceding page. the virtual simultaneous coordinate x0 will be identical to the actual simultaneous coordinate of the particle at time t. isolated space and we know that it will not be affected by any external forces. This gives us the possibility to extrapolate its position at the observation time.95b) (8. i.66) on page 114. the localised charge moves in a ﬁeld-free. with the Draft version released 13th November 2000 at 22:01.97) Hence. unaccelerated motion of the charge.se/CED/Book .2 on page 114.plasma. x0 (t) = x (t). the E and B ﬁelds can be obtained from formulae (8.

In a similar way.63) on page 113 as follows: ∂A 1 ∂vφ v ∂φ E = − φ− = − φ− 2 = − φ− 2 ∂t c ∂t c ∂t v v vv (8. the magnetic ﬁeld can be calculated and one ﬁnds that q v2 (x − x0 ) 1 − 2 4πε0 s3 c follows.122 E LECTROMAGNETIC R ADIATION potentials given by Equations (8. x) = 2 − 1 · φ = − − 1 · s2 c 8πε0 s3 c2 v v q (x − x0 ) + × = × (x − x0 ) 4πε0 s3 c c From Equation (8. From Equation (8. the same result also follows from Equation (8.plasma.99) q v v × (x − x0 ) =− (x − x0 ) + × 4πε0 s3 c c When this expression for φ is inserted into Equation (8. Downloaded from http://www.101) .98a).83) on page 118 ˙ with v ≡ 0 inserted.uu.98a) = − φ+ · φ = − 1− 2 · φ c c c vv = 2 −1 · φ c v v v B = ×A = × 2φ = φ× 2 = − 2 × φ c c c v v vv v v · φ − φ = 2 × 2 −1 · φ = 2× (8. x) = µ0 q 4πs3 1− v2 c2 v × (x − x0 ) = = − v v · (x − x0 ) c c 1 v×E c2 = q v v v2 (x − x0 ) + · (x − x0 ) − (x − x0 ) 2 4πε0 s3 c c c Draft version released 13th November 2000 at 22:01.100) (8.se/CED/Book B(t. What remains is just ˆ ˆ to express φ in quantities evaluated at t and x.100) we conclude that E is directed along the vector from the simultaneous coordinate x0 (t) to the ﬁeld (observation) coordinate x(t). the following result vv vv q E(t.96) on the previous page we ﬁnd that 1 q q =− φ= s2 4πε0 s 8πε0 s3 (8.98b) c c c c c v = 2 ×E c Here 1 = xi xi is the unit dyad and we used the fact that v × v ≡ 0.63a) on page 113 and Equation (8. − v v vv v v · (x − x0 ) − 2 · × × (x − x0 ) c c c c c (8. Of course.

2 E ND OF 8. x) = 1 ∂2 − c2 ∂t2 2 Ψ(t. x) and a generic source component f (t. in a time-independent form. v → 0 ⇒ E goes over into the Coulomb ﬁeld ECoulomb 2.104b) (8. sin θ0 ≈ 0 ⇒ E → (1 − v2 /c2 )ECoulomb 5.104a) (8. for a generic potential component Ψ(t. e ξ1 = 1 − v2 /c2 1− v2 c2 ∂2 Ψ ∂2 Ψ ∂2 Ψ + + = − f (x) ∂x2 ∂x2 ∂x2 1 2 3 x1 2 Ψ(t. ∂/∂ξ2 . v → c ⇒ E becomes dependent on θ0 4.1 d . sin θ0 ≈ 1 ⇒ E → (1 − v2 /c2 )−1/2 ECoulomb EXAMPLE T HE CONVECTION POTENTIAL AND THE CONVECTION FORCE Let us consider in more detail the treatment of the radiation from a uniformly moving rigid charge distribution.19) on page 38.103) (8. x) (8. v → c. formula (3. v → c.104c) def ξ ≡ (∂/∂ξ1 . Draft version released 13th November 2000 at 22:01.e. Downloaded from http://www.uu.102) (8. this equation can be written ˆ i. x). ∂/∂ξ3 ).se/CED/Book ! E XAMPLE 8. Transforming ξ2 = x 2 ξ3 = x 3 and introducing the vectorial nabla operator in ξ space.5 R ADIATION FROM A LOCALISED CHARGE IN ARBITRARY MOTION 123 From these explicit formulae for the E and B ﬁelds we can discern the following cases: 1. If we return to the original deﬁnition of the potentials and the inhomogeneous wave equation.8.. x) = f (t.plasma. we ﬁnd that under the assumption that v = v x1 . v → 0 ⇒ B goes over into the Biot-Savart ﬁeld 3.

111) as v v v v v F = q E+v× 2 ×E = q (8.63) on page 113.108) (8. ξ2 .plasma. Let us now consider the action of the ﬁelds produced from a moving. This equation has the well-known Coulomb potential solution Ψ(ξ) = V After inverse transformation back to the original coordinates. formula (F.109a) φ(t.uu. x) = which we recognise as the Liénard-Wiechert potentials. x) = 4πε0 c2 s φ(t. ξ3 ) ≡ − f (ξ) f (ξ ) 3 dξ ξ−ξ (8. on the last term yields v v v v v2 × × φ = · φ − φ c c c c c2 Downloaded from http://www.110a) (8.103) reduces to an ordinary Poisson equation 2 ξ Ψ(ξ) = −f( 1 4π in this space. we obtain ρ(x ) 3 1 dx (8. x) = 4πε0 V s 1 v vρ(x ) 3 A(t. in the denominator.105) (8. is of more general validity than the one leading to Equations (8.63) on page 113. We notice. on a charged particle q.124 E LECTROMAGNETIC R ADIATION the time-independent equation (8. x) (8. v2 s = (x1 − x1 ) + 1 − 2 c 2 g $$ f f $$ f e 1 − v2 /c2 ξ1 .106) f [(x2 − x2 ) + (x3 − x3 ) ] 2 2 1 2 (8.110b) (8.54) on page 155. these expressions reduce to q 4πε0 s qv A(t. realising that for a rigid charge distribution ρ moving with velocity v the current is given by j = ρv. also moving with velocity v. x) = d x = 2 φ(t. This force is given by the Lorentz force F = q(E + v × B) (8. Ψ(x) = (8.111) This means that we can rewrite expression (8.112) · φ − φ− × × φ c c c c c Applying the “bac-cab” rule.107) Applying this to the explicit scalar and vector potential components. that the derivation here. Equations (8. this becomes f (x ) 3 1 dx 4π V s where. however.113) . cf.se/CED/Book Draft version released 13th November 2000 at 22:01. based on a mathematical technique which in fact is a Lorentz transformation.109b) 4πε0 c2 V s c For a localised charge where ρ d3x = q . rigid charge distribution represented by q moving with velocity v.

116) ££ e h 2 + (x2 − x2 )2 + (x3 − x3 )2 (8.se/CED/Book ! ££ ££ i x1 − x 1 ψ = 1− v2 c2 ££ q 4πε0 s (8.5 R ADIATION FROM A LOCALISED CHARGE IN ARBITRARY MOTION 125 which means that we can write F = −q ψ where ψ = 1− v2 c2 φ (8.plasma. formula (8. This is the idea behind the Trouton-Noble experiment. c v c (8.108) on the facing page as = γ2 (x1 − x1 )2 + (x2 − x2 )2 + (x3 − x3 )2 = Const These Heaviside ellipsoids are equipotential surfaces. 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.uu.118) and formula (8. EXAMPLE 1 − v2 /c2 Radiation for small velocities If the charge moves at such low speeds that v/c simpliﬁes to s = x−x − (x − x ) · v ≈ x−x . the force between the two charges is in general not directed along the line which connects the charges.82) on page 118 x − x0 = (x − x ) − |x − x | v ≈ x−x . When the rigid charge distribution is well localised so that we can use the potentials (8. aimed at measuring the absolute speed of the earth or the galaxy.64) on page 113 v c (8. so that a compensating torque appears due to mechanical stresses within the charge-bar system. Downloaded from http://www.8. deﬁned through Equation (8. c 1. and since the force is proportional to the gradient of ψ. A consequence of this is that a system consisting of two co-moving charges connected with a rigid bar. will experience a torque.119) Draft version released 13th November 2000 at 22:01.114) (8.117) E ND OF 8. which means that it is perpendicular to the ellipsoid surface.110) the convection potential becomes The convection potential from a point charge is constant on ﬂattened ellipsoids of revolution.2 .115) The scalar function ψ is called the convection potential or the Heaviside potential.

x) = q 4πc3 |x − x |2 [˙ × (x − x )]. with the use of formula (8. v v c (8. The electric and magnetic ﬁelds.124) ˙ where θ is the angle between v and x − x0 .123a) (8.126 E LECTROMAGNETIC R ADIATION so that the radiation ﬁeld Equation (8.87) on page 119 can be approximated by Erad (t. respectively. dependent on the instantaneous position of the charge at x (t ).uu. and the expressions for the Poynting ﬂux and power derived from them.121) and the electric dipole ﬁeld Equations (8. Downloaded from http://www.43) on page 108 if we introduce p = q x (t ) and at the same time make the transitions ˙ ¨ q v = p → −ω2 pω (8. x) = q ˙ (x − x ) × [(x − x ) × v]. The total radiated power (integrated over a closed spherical surface) becomes P= v µ0 q 2 (˙ )2 ˙ q 2 v2 = 6πc 6πε0 c3 (8.120) and Equation (8.120) from which we obtain.123b) (8. are here instantaneous values. The angular distribution is that which is “frozen” to the point from which the energy is radiated.plasma.121) It is interesting to note the close correspondence which exists between the nonrelativistic ﬁelds (8. . 4πε0 c2 |x − x |3 v c (8. the magnetic ﬁeld Brad (t.86) on page 119.122) x − x = x − x0 The power ﬂux in the far zone is described by the Poynting vector as a function of Erad and Brad . Equation (8.120) and (8. 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. We use the close correspondence with the dipole case to ﬁnd that it becomes S= x−x v µ0 q 2 (˙ )2 sin2 θ 2 c |x − x |2 |x − x | 16π (8.se/CED/Book Draft version released 13th November 2000 at 22:01.121) above.125) which is the Larmor formula for radiated power from an accelerated charge.

the magnetic ﬁeld Brad (t. This condition (for an arbitrary magnitude of v) inserted into expression (8. x) = q |x − x | [˙ × (x − x )]. The Poynting vector is related to the time t when it is measured and to a ﬁxed surface in space. yields Erad (t. ¢ ∂t ∂t dΩ x Downloaded from http://www.131) Draft version released 13th November 2000 at 22:01.126) from which we obtain.plasma.127) The difference between this case and the previous case of v c is that the approximate expression (8.130) .uu. The radiated power into a solid angle element dΩ.3 Bremsstrahlung An important special case of radiation is when the velocity v and the acceler˙ ˙ ation v are collinear (parallel or anti-parallel) so that v × v = 0.8. x) = q ˙ (x − x ) × [(x − x ) × v]. with the use of formula (8.87) on page 119 for the radiation ﬁeld.se/CED/Book 5 4 ££ 5 ££ ¡ 4 6 dΩ (8.5.86) on page 119. we must instead use the correct expression (8. is given by the formula µ0 q 2 v 2 sin2 θ ˙ dU rad (θ) dΩ = S · (x − x ) x − x dΩ = dt 16π2 c 1 − v cos θ c On the other hand.74) on page 116.128) 6 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. The angular distribution of the power ﬂux (Poynting vector) therefore becomes S= sin2 θ ˙ µ0 q 2 v 2 16π2 c |x − x |2 1 − v cos θ c x−x |x − x | (8. the radiation loss due to radiation from the charge at retarded time t : dU rad dU rad dΩ = dt dt Using formula (8. We must be careful when we compute the energy (S integrated over time).129) (8.64) on page 113. 4πε0 c2 s3 ˙ v v (8.118) on page 125 for s is no longer valid. we obtain dU rad dU rad s dΩ = dΩ = S · (x − x )s dΩ dt dt |x − x | (8.5 R ADIATION FROM A LOCALISED CHARGE IN ARBITRARY MOTION 127 8. measured relative to the particle’s retarded position. v 4πε0 c3 s3 ˙ v v (8.

From Figure 8.plasma.129) on the previous page with expression (8.3. Let us explain this in geometrical terms.5.134) Draft version released 13th November 2000 at 22:01. Inserting Equation (8. . As shown in 8. t + dt ) and within the solid angle element dΩ the particle radiates an energy [dU rad (θ)/dt ] dt dΩ.uu.3 we see that the volume element subtending the solid angle element x − x2 is d3x = dS dr = x − x2 Downloaded from http://www. and v = 0. are plotted in Figure 8.128 E LECTROMAGNETIC R ADIATION v = 0.se/CED/Book ££ ££ ££ ££ dΩ = dS 2 2 dΩ dr 5 dU rad (θ) sin2 θ ˙ µ0 q 2 v 2 dΩ = dt 16π2 c 1 − v cos θ c 5 dΩ (8. one outer with its origin in x1 (t ) and one inner with its origin in x1 (t + dt ) = x1 (t ) + v dt and radius c[t − (t + dt )] = c(t − t − dt ).5. v = 0.5. for three different particle speeds.5c.131). we see that they differ by a factor 1 − v cos θ/c which comes from the extra factor s/ |x − x | introduced in (8. Comparing expression (8.5c v = 0.132).128) on the previous page for S into (8. During the interval (t .25c v=0 v F IGURE 8.3 this energy is at time t located between two spheres.133) (8.131).25c.4: Polar diagram of the energy loss angular distribution factor sin2 θ/(1 − v cos θ/c)5 during bremsstrahlung for particle speeds v = 0.132) 4 (8. we obtain the explicit expression for the energy loss due to radiation evaluated at the retarded time The angular factors of this expression.

but relatively straightforward integration of formula (8. ˜ Let the radiated energy. dr denotes the differential distance between the two spheres and can be evaluated in the following way v cos θ where formula (8.132) on the preceding page. t + dt ).129) on page 127 and expression (8. Draft version released 13th November 2000 at 22:01.se/CED/Book . Hence.8.plasma. The observation point (ﬁeld point) is at the ﬁxed location x.uu.135) (8. integrated over Ω. t + dt ) is located in a volume element whose size is θ dependent.5: Location of radiation between two spheres as the charge moves with velocity v from x1 to x2 during the time interval (t . be denoted U rad . After tedious. ££ ££ d3x = dS dr = s dS c dt x − x2 ££ ££ I £ £ ££ = c− x − x2 ·v x − x2 dt = cs dt x − x2 F ££ D GC ££ B dr = x − x2 + c dt − x − x2 · v dt − x − x2 x − x2 ££ $$ $$ ££ ££ H ££ (8.5 R ADIATION FROM A LOCALISED CHARGE IN ARBITRARY MOTION 129 dS dr x dΩ θ q x2 vdt x1 x − x2 + c dt F IGURE 8. This explains the difference between expression (8.136) Downloaded from http://www. Here.64) on page 113 was used in the last step.132) on the facing page. the volume element under consideration is We see that the energy which is radiated per unit solid angle during the time interval (t .

se/CED/Book $$ $$ $$ $$ $$ s ≈ x−x 0 v 1 − c2 2 ) f $$ SHORT ACCELERATION TIMES (8. x − x0 ≈ x − x (8.3 Calculate the bremsstrahlung when a charged particle.138) (8.140) (8. This is a consequence of the inﬁnitely short “impulsive step” δ(t −t0 ) in the time domain which produces an inﬁnite spectrum in the frequency domain. according to formula (8. Eω = The total radiation energy is given by the expression Downloaded from http://www. According to the “bremsstrahlung expression” for Erad .130 E LECTROMAGNETIC R ADIATION one obtains ˜ ˙ dU rad µ0 q 2 v2 = dt 6πc 1 3 (8. yielding q sin θ ∆v eiωt0 (8.141) From the general expression (8.144) 8π2 ε0 c2 |x − x | We note that the magnitude of this Fourier component is independent of ω. This way we obtain a classical picture of bremsstrahlung (braking radiation). Often.82) on page 118. we assume v/c −∞ and.86) on page 119 we conclude that E ⊥ B and that it sufﬁces to consider E ≡ Erad . We approximate the velocity change at time t = t0 by a delta function: ˙ v(t ) = ∆v δ(t − t0 ) which means that ∆v = ∞ ˙ v dt 1 so that.143) Draft version released 13th November 2000 at 22:01. according to formula (8. .142) (8. is accelerated or decelerated during an inﬁnitely short time interval. we can integrate this expression over t and obtain the total energy radiated during the acceleration or deceleration of the particle.64) on page 113.plasma.uu. Also. moving at a non-relativistic speed.137) If we know v(t ). E= q sin θ ∆v δ(t − t0 ) 4πε0 c2 |x − x | In this simple case B ≡ Brad is given by B= E c Fourier transforming expression (8. B REMSSTRAHLUNG FOR LOW SPEEDS AND E XAMPLE 8. an atomistic treatment is required for an acceptable result.139) (8. Equation (8.126) on page 127.142) above for E is trivial.

for an electron where q = − |e|. ω + dω) is S For our inﬁnite spectrum.149) 2 which expresses that the highest possible frequency in the spectrum is that for which all kinetic energy difference has gone into one single ﬁeld quantum (photon) with energy hω. ∞]. Equation (8. dω 2π ˜ rad We see that the energy spectrum Uω is independent of frequency ω.uu. where we used the value of the ﬁne structure constant e2 /(4πε0 hc) ≈ 1/137.8. where e is the elementary charge.plasma.144) on the preceding page.146) f (8.12) on page 101] the following equality holds: ∞ −∞ which means that the radiated energy in the frequency interval (ω. This means that if we integrate it over all frequencies ω ∈ [0. with an upper cutoff limit set by the quantum condition 1 hω = m(∆v)2 ¯ (8. we obtain 0 2 In reality.147) (8. ¯ Even if the number of photons becomes inﬁnite when ω → 0. all spectra have ﬁnite widths. we ﬁnd that ˜ rad Uω dω = dNω (8. Equation (8.3 .se/CED/Book ! dNω = e2 2 4πε0 hc 3π ¯ ∆v c = q2 3πε0 c ∆v c 2 dω 1 2 ≈ ω 137 3π f f q 2 (∆v)2 = 16π3 ε0 c3 f q 2 (∆v)2 ˜ rad Uω dω = 16π3 ε0 c3 S 2π sin2 θ 2 d x dω |x − x |2 π dϕ 0 f ˜ rad Uω dω = 4πε0 c f E 2 dt = 4π f f = ε0 c −∞ ∞ −∞ E 2 dt d2x S ∞ 0 |E ω |2 dω |E ω |2 d2x dω sin2 θ sin θ dθ dω f f S f f 1 µ0 ∞ −∞ S EB dt d2x = 1 µ0 c f f f ˜ U rad = ˜ dU rad dt = dt ∞ E× B µ0 · dS dt ∞ −∞ E 2 dt d2x (8. these photons have negligible energies so that the total radiated energy is still ﬁnite.150) hω ¯ or.5 R ADIATION FROM A LOCALISED CHARGE IN ARBITRARY MOTION 131 = S According to Parseval’s identity [cf. EXAMPLE Draft version released 13th November 2000 at 22:01. If we adopt the picture that the total energy is quantised in terms of N ω ¯ photons radiated during the process.148) ∆v c 2 dω ω (8.145) (8. a divergent integral would result. Downloaded from http://www.151) E ND OF 8.

uu. The power ﬂux is given by the Poynting vector.156) Downloaded from http://www.152d) (8.86) on page 119.5.153) (8.87) on page 119 for the magnetic ﬁeld and the radiation part of the electric ﬁeld are general.e..152b) (8.e. ˆ ˆ x − x = x − x ( x2 sin α + x3 cos α) where α is the angle between x − x and the normal to the plane of the particle orbit (see Figure 8.152f) Because of the rotational symmetry we can. without loss of generality. With the charged particle orbiting in the x1 x2 plane as in Figure 8.5.131) on page 127. A very important special case is circular motion. which.152c) (8.155) Inserting this into Equation (8.154b) .. can be written S= 1 1 x−x (E × B) = |E|2 µ0 cµ0 |x − x | (8. an orbit radius a. rotate our coordinate system in around the x3 axis so the relative vector x − x from the source point to an arbitrary ﬁeld point always lies in the x2 x3 plane. the ˙ case v ⊥ v. ϕ) |x − x | s 2 = |E| dt cµ0 (8.152a) (8.4). we obtain ϕ(t ) = ω0 t x (t ) = a[ x1 cos ϕ(t ) + x2 sin ϕ(t )] ˆ ˆ ˙ v(t ) = x (t ) = aω0 [− x1 sin ϕ(t ) + x2 cos ϕ(t )] ˆ ˆ ˙ ¨ v(t ) = x (t ) = −aω2 [ x1 cos ϕ(t ) + x2 sin ϕ(t )] ˆ 0 ˆ v = |˙ | = ˙ v aω2 0 v = |v| = aω0 (8.plasma. i.86) and formula (8. and an angular frequency ω0 .4 Cyclotron and synchrotron radiation Formula (8.4 on the next page. valid for any kind of motion of the localised charge.5. From the above expressions we obtain ˙ (x − x ) · v = − x − x v sin α sin ϕ = x − x v cos θ ˙ ˙ where in the last step we simply used the deﬁnition of a scalar product and the ˙ fact that the angle between v and x − x is θ.se/CED/Book Draft version released 13th November 2000 at 22:01. we obtain dU rad (α.152e) (8.154a) (8. ££ ££ (x − x ) · v = x − x v sin α cos ϕ ££ ££ ££ ££ ££ ££ (8. with the help of formula (8.132 E LECTROMAGNETIC R ADIATION 8. i.

expression (8. after some cumbersome algebra. inserted. With the radiation part of the electric ﬁeld. one ﬁnds. Integration of Equation (8.se/CED/Book . The radius of the orbit is a.plasma. where the retarded distance s is given by expression (8.6: Coordinate system for the radiation from a charged particle at x (t ) in circular motion with velocity v(t ) along the tangent and con˙ stant acceleration v(t ) toward the origin. and using (8.8. 0 1 − v2 c 2 2 0 5 ) 2 2 5 4 4 ) Downloaded from http://www. But we can parametrise the solid angle dΩ in the angle ϕ and the (ﬁxed) angle α so that dΩ = sin α dα dϕ. x ) x−x x3 F IGURE 8. after some algebra. x) x v q a α 0 θ ˙ v ϕ(t ) x1 (t . The x1 x2 axes are chosen so that the relative ﬁeld point vector x − x makes an angle α with the x3 axis which is normal to the plane of the orbital motion.5 R ADIATION FROM A LOCALISED CHARGE IN ARBITRARY MOTION 133 x2 (t.158) Draft version released 13th November 2000 at 22:01.157) = 5 dt 16π2 c 1 − v sin α cos ϕ c The angles θ and ϕ vary in time during the rotation. ϕ) µ0 q 2 v2 1 − c sin α cos ϕ − 1 − c2 sin α sin ϕ ˙ (8. the angular integrated expression ˜ dU rad µ0 q 2 v2 ˙ = dt 6πc 1 (8.87) on page 119.154b) on the facing page.64) on page 113. that 2 2 v v dU rad (α.157) above over this dΩ gives.154a) and (8.uu. so that θ refers to a moving coordinate system.

134

E LECTROMAGNETIC R ADIATION

In Equation (8.157) on the preceding page, two limits are particularly interesting: 1. v/c 1 which corresponds to cyclotron radiation. 1 which corresponds to synchrotron radiation.

Cyclotron radiation For a non-relativistic speed v duces to c, Equation (8.157) on the previous page re-

dU rad (α, ϕ) µ0 q 2 v2 ˙ = (1 − sin2 α sin2 ϕ) dt 16π2 c But, according to Equation (8.154b) on page 132 sin2 α sin2 ϕ = cos2 θ

where θ is deﬁned in Figure 8.5.4 on the previous page. This means that we can write ˙ µ0 q 2 v 2 ˙ dU rad (θ) µ0 q 2 v2 = (1 − cos2 θ) = sin2 θ 2c 2c dt 16π 16π (8.161)

Consequently, a ﬁxed observer near the orbit plane 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 (8.157) on the preceding page becomes very small if sin α cos ϕ ≈ 1, which deﬁnes the forward direction of the particle motion (α ≈ π/2, ϕ ≈ 0). Equation (8.157) on the previous page then becomes dU rad (π/2, 0) µ0 q 2 v2 1 ˙ = 2c dt 16π 1− v c

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 (8.161) and Equation (8.162) above 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.

Downloaded from http://www.plasma.uu.se/CED/Book

5

j

4

j

2. v/c

(8.159)

(8.160)

3

(8.162)

Draft version released 13th November 2000 at 22:01.

8.5

R ADIATION

FROM A LOCALISED CHARGE IN ARBITRARY MOTION

135

(t, x) x−x ∆θ

x2

v q a 0 ∆θ ˙ v ϕ(t ) x1 (t , x )

x3 F IGURE 8.7: When the observation point is in the plane of the particle orbit, i.e., α = π/2 the lobe width is given by ∆θ.

**In the orbit plane (α = π/2), Equation (8.157) on page 133 gives
**

2 v v dU rad (π/2, ϕ) µ0 q 2 v2 1 − c cos ϕ − 1 − c2 sin ϕ ˙ = 5 dt 16π2 c 1 − v cos ϕ c

which vanishes for angles ϕ0 such that cos ϕ0 = v c 1− v2 c2

Hence, the angle ϕ0 is a measure of the synchrotron radiation lobe width ∆θ; see Figure 8.5.4. For ultra-relativistic particles, deﬁned by

1−

v2 c2

one can approximate ϕ0 ≈ sin ϕ0 = 1− v2 1 = c2 γ (8.166)

Draft version released 13th November 2000 at 22:01.

%

%

c

γ=

1

%

sin ϕ0 =

1,

1−

v2 c2

1,

0

5

)

2

2

5

4

4

(8.163)

(8.164a) (8.164b)

(8.165)

Downloaded from http://www.plasma.uu.se/CED/Book

136

E LECTROMAGNETIC R ADIATION

Hence, synchrotron radiation from ultra-relativistic charges is characterized by a radiation lobe width which is approximately ∆θ ≈ 1 γ (8.167)

This angular interval is swept by the charge during the time interval ∆t = ∆θ ω0 (8.168)

during which the particle moves a length interval ∆l = v∆t = v ∆θ ω0 (8.169)

in the direction toward the observer who therefore measures a pulse width of length

As a general rule, the spectral width of a pulse of length ∆t is ∆ω 1/∆t. In the ultra-relativistic synchrotron case one can therefore expect frequency components up to ωmax ≈ 1 = 2γ3 ω0 ∆t

(8.171)

A spectral analysis of the radiation pulse will exhibit Fourier components nω 0 from n = 1 up to n ≈ 2γ3 . When N electrons are contributing to the radiation, we can discern between three situations: 1. All electrons are very close to each other so that the individual phase differences are negligible. The power will be multiplied by N 2 relative to a single electron and we talk about coherent radiation.

Downloaded from http://www.plasma.uu.se/CED/Book

Draft version released 13th November 2000 at 22:01.

0

j

)

0

)

0

F EC B D

)

∆t = ∆t −

∆l v∆t v v ∆θ v 1 = ∆t − = 1− ∆t = 1 − ≈ 1− c c c c ω0 c γω0 2 1− v 1+ v 1 1 v 1 1 c c ≈ 1− 2 = = v γω0 c 2γω0 2γ3 ω0 1+ c 1/γ2 ≈2 (8.170)

5

F GC B D

4

5

4

8.5

R ADIATION

FROM A LOCALISED CHARGE IN ARBITRARY MOTION

137

2. The electrons are perfectly evenly distributed in the orbit. This is the case, for instance, for electrons in a circular current in a conductor. In this case the radiation ﬁelds cancel completely and no far ﬁelds are generated. 3. The electrons are unevenly distributed in the orbit. This happens for an √ open ring current which is subject to ﬂuctuations of order N as for all open systems. As a result we get incoherent radiation. Examples of this can be found both in earthly laboratories and under cosmic conditions. Radiation in the general case We recall that the general expression for the radiation E ﬁeld from a moving charge concentration is given by expression (8.87) on page 119. This expression in Equation (8.156) on page 132 yields the general formula

**Integration over the solid angle Ω gives the totally radiated power as
**

2 v ˜ dU rad µ0 q 2 v2 1 − c2 sin θ ˙ = 2 3 dt 6πc 1 − v2 c

2

˙ where θ is the angle between v and v. ˙ ˙ In the limit v v, sin θ = 0, which corresponds to bremsstrahlung. For v ⊥ v, sin θ = 1, which corresponds to cyclotron radiation or synchrotron radiation. Virtual photons According to formula (8.98a) on page 122 and Figure 8.5.4 on the next page E⊥ = Ez = v2 q 1 − 2 (x − x0 ) · x3 ˆ 4πε0 s3 c b q = 4πε0 γ2 (vt)2 + b2 /γ2 3/2

which represents a contracted 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

−∞

Draft version released 13th November 2000 at 22:01.

Downloaded from http://www.plasma.uu.se/CED/Book

§ ¨¢

¢

¦

Eω,⊥ =

¤

1 2π

∞

E⊥ (t) eiωt dt =

q 4π2 ε0 bv

bω K1 vγ

bω vγ

§ ¢

¡

¡

¦

dU rad (θ) µ0 q 2 |x − x | = (x − x ) × dt 16π2 cs5

¡

(x − x ) −

|x − x | v c

2

˙ ×v

(8.172)

¢

0

(8.173)

¡

)

(8.174)

(8.175)

138

E LECTROMAGNETIC R ADIATION

vt q b v = vx ˆ

B E⊥ z ˆ F IGURE 8.8: The perpendicular ﬁeld of a charge q moving with velocity v = v x is E ⊥ z. ˆ ˆ

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 q , bω vγ (8.176a) Eω,⊥ ∼ 2 4π ε0 bv Eω,⊥ ∼ 0, bω vγ (8.176b) showing that the “pulse” length is of the order b/vγ. Due to the equipartition of the ﬁeld energy into the electric and magnetic ﬁelds, the total ﬁeld energy can be written U = ε0

2 E⊥ d3x = ε0 bmax ∞ −∞ 2 E⊥ vdt 2πb db

(8.177)

where the volume integration is over the plane perpendicular to v. With the use of Parseval’s identity for Fourier transforms, formula (8.12) on page 101, we can rewrite this as U=

0

U dω = 4πε0 v

∞ vγ/ω bmin

from which we conclude that Uω ≈ q2 vγ ln 2π2 ε0 v bmin ω

Downloaded from http://www.plasma.uu.se/CED/Book

¢

¤

¡

¤

q2 ≈ 2 2π ε0 v

0

db dω b

¤

bmin

0

Eω,⊥

££

££

∞

bmax

¤

¤

¤

V

bmin

¤

∞

2

dω 2πb db (8.178)

¤

(8.179)

Draft version released 13th November 2000 at 22:01.

8.5

R ADIATION

FROM A LOCALISED CHARGE IN ARBITRARY MOTION

139

where an explicit value of bmin can be calculated in quantum theory only. As in the case of bremsstrahlung, it is intriguing to quantise the energy into photons [cf. Equation (8.150) on page 131]. Then we ﬁnd that

where α = e2 /(4πε0 hc) ≈ 1/137 is the ﬁne structure constant. ¯ Let us consider the interaction of two electrons, 1 and 2. The result of this interaction is that they change their linear momenta from p1 to p1 and p2 to ¯ p2 , respectively. Heisenberg’s uncertainty principle gives bmin ∼ h/ p1 − p1 so that the number of photons exchanged in the process is of the order cγ dω 2α ln p1 − p1 (8.181) π hω ¯ ω Since this change in momentum corresponds to a change in energy hω = E 1 − ¯ E1 and E1 = m0 γc, we see that

a formula which gives a reasonable account of electron- and photon-induced processes.

**8.5.5 Radiation from charges moving in matter
**

When electromagnetic radiation is propagating through matter, new phenomena may appear which are (at least classically) not present in vacuum. As mentioned earlier, one can under certain simplifying assumptions include, to some extent, the inﬂuence from matter on the electromagnetic ﬁelds by introducing new, derived ﬁeld quantities D and H according to D = ε(t, x)E = κε0 E B = µ(t, x)H = κm µ0 H (8.183) (8.184)

Expressed in terms of these derived ﬁeld quantities, the Maxwell equations, often called macroscopic Maxwell equations, take the form · D = ρ(t, x) ∂ ×E+ B = 0 ∂t ·B = 0 ∂ × H − D = j(t, x) ∂t (8.185a) (8.185b) (8.185c) (8.185d)

Draft version released 13th November 2000 at 22:01.

I

H

Nω dω ≈

2α ln π

E1 cp1 − cp1 m0 c2 E 1 − E 1

££

££

)

Nω dω ≈

dω ω

(8.182)

Downloaded from http://www.plasma.uu.se/CED/Book

££

££

0£

£

¢

Nω dω ≈

2α cγ ln π bmin ω

££

¡

dω ω

(8.180)

140

E LECTROMAGNETIC R ADIATION

Assuming for simplicity that the electric permittivity ε and the magnetic permeability µ, and hence the relative permittivity κ and the relative permeability κm all have ﬁxed values, independent on time and space, for each type of material we consider, we can derive the general telegrapher’s equation [cf. Equation (2.26) on page 27] ∂2 E ∂E ∂2 E − σµ − εµ 2 = 0 ∂ζ 2 ∂t ∂t (8.186)

describing (1D) wave propagation in a material medium. In Chapter 2 we concluded that the existence of a ﬁnite conductivity, manifesting itself in a collisional interaction between the charge carriers, causes the waves to decay exponentially with time and space. Let us therefore assume that in our medium σ = 0 so that the wave equation simpliﬁes to ∂2 E ∂2 E − εµ 2 = 0 ∂ζ 2 ∂t If we introduce the phase velocity in the medium as 1 1 c vϕ = √ = √ =√ εµ κε0 κm µ0 κκm (8.188) (8.187)

√ where, according to Equation (1.9) on page 5, c = 1/ ε0 µ0 is the speed of light, i.e., the phase speed of electromagnetic waves in vacuum, then the general solution to each component of Equation (8.187) Ei = f (ζ − vϕ t) + g(ζ + vϕ t), c √ √ def = κκm = c εµ ≡ n vϕ i = 1, 2, 3 (8.189)

The ratio of the phase speed in vacuum and in the medium (8.190)

is called the refractive index of the medium. In general n is a function of both time and space as are the quantities ε, µ, κ, and κm themselves. If, in addition, the medium is anisotropic or birefringent, all these quantities are rank-two tensor ﬁelds. Under our simplifying assumptions, in each medium we consider n = Const for each frequency component of the ﬁelds. Associated with the phase speed of a medium for a wave of a given frequency ω we have a wave vector, deﬁned as ˆ k ≡ k k = kˆ ϕ = v

def

ω vϕ vϕ vϕ

(8.191)

Downloaded from http://www.plasma.uu.se/CED/Book

Draft version released 13th November 2000 at 22:01.

8.5

R ADIATION

FROM A LOCALISED CHARGE IN ARBITRARY MOTION

141

As in the vacuum case discussed in Chapter 2, assuming that E is time-harmonic, i.e., can be represented by a Fourier component proportional to exp{−iωt}, the solution of Equation (8.187) can be written E = E0 ei(k·x−ωt) (8.192)

where now k is the wave vector in the medium given by Equation (8.191) on the preceding page. With these deﬁnitions, the vacuum formula for the associated magnetic ﬁeld, Equation (2.33) on page 28, 1 1 ˆ √ ˆ k×E = k×E B = εµ k × E = vϕ ω (8.193)

is valid also in a material medium (assuming, as mentioned, that n has a ﬁxed constant scalar value). A consequence of a κ = 1 is that the electric ﬁeld will, in general, have a longitudinal component. It is important to notice that depending on the electric and magnetic properties of a medium, and, hence, on the value of the refractive index n, the phase speed in the medium can be smaller or larger than the speed of light: vϕ = c ω = n k (8.194)

where, in the last step, we used Equation (8.191) on the preceding page. If the medium has a refractive index which, as is usually the case, dependent on frequency ω, we say that the medium is dispersive. Because in this case also k(ω) and ω(k), so that the group velocity vg = ∂ω ∂k (8.195)

has a unique value for each frequency component, and is different from v ϕ . Except in regions of anomalous dispersion, vϕ is always smaller than c. In a gas of free charges, such as a plasma, the refractive index is given by the expression n2 (ω) = 1 − where ω2 = ∑ p

σ

ω2 p 2 ω

(8.196)

Nσ q2 σ ε0 mσ

(8.197)

Draft version released 13th November 2000 at 22:01.

Downloaded from http://www.plasma.uu.se/CED/Book

uu. can experience that the particle speeds are higher than the phase ˇ speed in the medium. In an inhomogeneous plasma. Equations (8.198b) Hence. for each given frequency. a charge in uniform. which we know is an absolute limit for the motion of anything. N σ = Nσ (x) so that the refractive index and also the phase and group velocities are space dependent. the phase and group velocities in a plasma are different from each other. of the retarded (and advanced) potentials in Chapter 3 and the Liénard-Wiechert potentials. consider a medium where ε = Const > ε0 µ = µ0 This implies that in this medium the phase speed is vϕ = 1 c =√ <c n εµ0 (8.64) on page 113 for s. which. . then at that point vϕ → ∞ while vg → 0 and the wave Fourier component at ω is reﬂected there.se/CED/Book Draft version released 13th November 2000 at 22:01. Let us now consider a charge in uniform. of course. entering into the medium at high speeds |v|. respectively. of charged particle species σ. A medium of this kind has the interesting property that particles. we realise that we obtain the latter in the medium by a simple formal replacement c → c/n in the expression (8. are below the phase speed in vacuum.98a) on page 122.198a) (8. the speed of propagation of (the phase planes of) electromagnetic waves is less than the speed of light in vacuum. This is the basis for the Vavilov-Cerenkov radiation that we shall now study. rectilinear motion in a medium with electric properties which are different from those of a (classical) vacuum. in the vacuum case. Speciﬁcally. see in particular Equation (8. Here mσ and Nσ denote the mass and number density.63) on page 113. rectilinear motion in vacuum does not give rise to any radiation. the Liénard-Wiechert potentials in a medium characterized by a Downloaded from http://www. If the frequency ω is such that it coincides with ωp at some point in the medium.142 E LECTROMAGNETIC R ADIATION is the plasma frequency.1.199) (8. in this particular medium. Hence. As can be easily seen. If we recall the general derivation. ˇ Vavilov-Cerenkov radiation As we saw in Subsection 8. including particles.plasma.

if v/c 1/n we recover the well-known vacuum case but with modiﬁed phase speed. Equation (3. the retarded distance s.200a) (8.201) (8. the potentials become singular.5 R ADIATION FROM A LOCALISED CHARGE IN ARBITRARY MOTION 143 refractive index n.200) vanish when n(x − x ) · nv v = x−x cos θc = x − x c c or. the ﬁeld exists within a sphere of radius |x − x | around the particle while the particle moves a distance l = v(t − t ) (8.se/CED/Book . when cos θc = c nv (8.plasma. x) = 1 1 qv qv = 2 4πε0 c |x − x | − n (x−x )·v 4πε0 c2 s c ££ £ ££ ££ £ φ(t. x) = 1 q 1 q = (x−x )·v 4πε0 |x − x | − n 4πε0 s c (8. x − x ) = t − so that the usual time interval t − t between the time measured at the point of observation and the retarded time in a medium becomes t−t = |x − x | n c (8. and therefore the denominators in Equations (8. equivalently.202b) (8. We also note that the retarded and advanced times in the medium are [cf.34) on page 41] k |x − x | |x − x | n = t− ω c k |x − x | |x − x | n tadv = tadv (t.200b) ££ ££ ££ ££ ££ £££ £ (8. x − x ) = t + = t+ ω c tret = tret (t.203) For v/c ≥ 1/n.8. ££ ££ ££ (x − x ) · v c ££ £ ££ ££ ££ ££ £ A(t.uu.203) above. are where now s = x−x −n 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. During the time interval t − t given by expression (8.205) In the direction deﬁned by this angle θc .206) Draft version released 13th November 2000 at 22:01.202a) (8.204) Downloaded from http://www.

ˇ ﬁrst observation of this radiation was made by P. . M. put Cerenkov as the author.207) This cone of potential singularities and ﬁeld sphere circumferences propagates ˇ with speed c/n in the form of a shock front. On May 8. after some controversy with the American editors who claimed the results to be wrong. eventually published in 1937. private communication]. Ginzburg.144 E LECTROMAGNETIC R ADIATION θc αc q v F IGURE 8.se/CED/Book Draft version released 13th November 2000 at 22:01. Sommerfeld sent a letter to Tamm via Austria.9: Instantaneous picture of the expanding ﬁeld spheres from a point charge moving with constant speed v/c > 1/n in a medium where ˇ n > 1. Cerenkov in 1934.plasma. L. saying that he was surprised that his old 1904 ideas were now becoming interesting.uu. In the direction θc where the potentials are singular. and submitted it to Nature. along the direction of v. Tamm and I. ˇ Frank and Cerenkov received the Nobel Prize in 1958 “for the discovery and the interpretation ˇ of the Cerenkov effect” [V. predictions of a similar effect had been made as early as 1888 by Heaviside. A. all ﬁeld spheres are tangent to a straight cone with its apex at the instantaneous position of the particle and with the apex half angle αc deﬁned according to sin αc = cos θc = c nv (8. Vavilov’s research group at the Lebedev Institute in Moscow. 1 The Downloaded from http://www.1 ˇ The Vavilov-Cerenkov cone is similar in nature to the Mach cone in acoustics. In the same year. ˇ Vavilov wrote a manuscript with the experimental ﬁndings. This generates a Vavilov-Cerenkov shock wave in the form of a cone. I. Frank published the theory for the effect (“the singing electron”). but the paper was rejected. who was then a post-graduate student in S. I . called Vavilov-Cerenkov radiation. 1937. Vavilov explained the results in terms of radioactive particles creating Compton electrons which gave rise to the radiation (which was the correct interpretation). and by Sommerfeld in his 1904 paper “Radiating body moving with velocity of light”. Tamm. In the manuscript.E. The paper was then sent to Physical Review and was. In fact.

210a) (8. The integral in (8. respectively. Consequently.211) sin θ dΩ 2 (8. x1 .plasma.5 R ADIATION FROM A LOCALISED CHARGE IN ARBITRARY MOTION 145 In order to make some quantitative estimates of this radiation.3b) on page 98. For a Fourier component one obtains [cf. the resulting Fourier transˇ forms of the Vavilov-Cerenkov magnetic and electric radiation ﬁelds can be calculated from the expressions (8.211) is singular of a “Dirac delta type.209).210b) In this manner. and the direction to ˆ ˆ the observer. The integration of (8.uu.209) (8. 5 4 rad Uω dΩ = q 2 nω2 sin2 θ sin2 1 − nv cos θ c 4π3 ε0 c3 1 − nv cos θ ω c v Xω v 2 dΩ ££ § ¨¢ 5 4 ¦ q 2 nω2 = 16π3 ε0 c3 ωx − kx cos θ exp i v −∞ ∞ ££ ££ ¡ ¤£ £ ££ ¤£ £ rad Uω dΩ ≈ 1 4πε0 nc (jω × k)e−ik·x d3x ££ ££ 2 dΩ 2 (8.1 if only we make the replacements ε0 → ε = n 2 ε0 nω k→ c (8. The magnitude of this maximum grows and its width narrows as X → ∞. X] on the x axis we can evaluate the integral to obtain which has a maximum in the direction θc as expected.208) This Fourier component can be used in the formulae derived for a linear current in Subsection 8.3a) and (8.se/CED/Book . Equation (8.8. using jω from Equation (8.212) Downloaded from http://www. we note that we can describe the motion of each charged particle q as a current density: j = q v δ(x − vt ) = q v δ(x − vt )δ(y )δ(z ) x1 ˆ which has the trivial Fourier transform jω = q iωx /v e δ(y )δ(z ) x1 ˆ 2π (8.212) over Ω therefore picks up the main contributions from θ ≈ θc .3. The total energy content is then obtained from Equation (8.” If we limit the spatial extent of the motion of the particle to the closed interval [−X.15) on page 101] V dx where θ is the angle between the direction of motion. Draft version released 13th November 2000 at 22:01. k.12) on page 101 (integrated over a closed sphere at large distances).

¢ cXπ 1− 2 2 ωn n v ¡ 9 0 ω v c2 ¤ ¡ ¡ )8 −1 1 + nvξ c ¢ sin θc 2 9 0 1 )8 sin2 Xω v 2 dξ ≈ 1 − ¡ 1 + nvξ c 9 0 )8 −1 1 + nvξ c ω v c2 n2 v2 9 0 1 )8 ¤ q 2 nω2 sin2 θc ≈ 2π2 ε0 c3 sin2 ¤ 0 0 1 + nvξ c F EC B D ξ Xω v 2 ¤ ˜ rad Uω = 2π π rad Uω sin θ dθ = 2π π rad Uω d (− cos θ) (8.213) is strongly in peaked near ξ = −c/(nv).213) dξ cX ωn sin2 x dx 2 −∞ x ∞ ¤ (8. i. equivalently. near cos θc = c/(nv) so we can extend the integration limits to ±∞ without introducing too much error.uu. we ﬁnd that the radiation energy per frequency unit and per unit length is ˜ rad Uω dω q 2ω = 2X 4πε0 c2 This result was derived under the assumption that v/c > 1/n(ω). under the condition that the expression inside the parentheses in the right hand side is positive. or.214) (8. ω + dω) q 2X ˜ rad Uω dω = 2πε0 c2 As mentioned earlier.215) (8.se/CED/Book ¢ 1− c2 n2 (ω)v2 ¢ 1− c2 n2 v2 ω dω dω Draft version released 13th November 2000 at 22:01. Realising this.146 E LECTROMAGNETIC R ADIATION we can set sin2 θ ≈ sin2 θc and the result of the integration is The integral (8. For all media it is true that n(ω) → 1 when ω → ∞.plasma.216) . Our derivation above for a ﬁxed value of n is valid for each individual Fourier component. the refractive index is usually frequency dependent. so there exist alˇ ways a highest frequency for which we can obtain Vavilov-Cerenkov radiation from a fast charge in a medium. Downloaded from http://www.. Via yet another variable substitution we can therefore approximate = leading to the ﬁnal approximate result for the total energy loss in the frequency interval (ω.e.

VANDERLINDE. Inc. London.plasma. Classical Electromagnetic Radiation. . . (London) Ltd.. ISBN 012-472257-1.uu. Revised third ed. Wiley & Sons. Toronto. and Singapore. G INZBURG. John Wiley & Sons. [3] J.8. MA . D. NY. Paris. ISBN 0-47157269-1. Inc. Electromagnetic Fields and Interactions. ISBN 0-471-30932-X. Classical Electrodynamics. ... . Classical Electromagnetic Theory. second ed. New York.. Applications of Electrodynamics in Theoretical Physics and Astrophysics. ISBN 0-201-05702-6. H EALD. NY . . New York.. B ECKER. H. [6] J. . Montreux.. New York. K.. Addison-Wesley Publishing Company. Draft version released 13th November 2000 at 22:01. 1980. Academic Press. JACKSON. ISBN 288124-719-9. New York. Tokyo and Melbourne. AND M. Orlando. A. [5] W.. Brisbane. [4] J. Chichester. P HILLIPS. ISBN 0-486-64290-9. M ARION . Inc.5 B IBLIOGRAPHY 147 Bibliography [1] R. second ed. PANOFSKY. [2] V. 1989. . Reading. 1962. Dover Publications. 1999. Inc. L. 1982. 1993. AND M. Classical Electricity and Magnetism. Gordon and Breach Science Publishers. . third ed. .. Downloaded from http://www. . Inc. B.se/CED/Book .

se/CED/Book Draft version released 13th November 2000 at 22:01.uu.148 E LECTROMAGNETIC R ADIATION Downloaded from http://www.plasma. .

5) (F.3) (F.2) (F.10) E = − φ− 149 .F Formulae F.1 Maxwell’s equations ·D = ρ ∂ B ∂t ∂ ×H = j+ D ∂t ×E = − Constitutive relations D = εE B H= µ (F.1 The Electromagnetic Field F.2 Fields and potentials Vector and scalar potentials B= ×A (F.1.1) (F.8) F.7) (F.6) ·B = 0 (F.1.4) j = σE P = ε0 χE (F.9) ∂ A ∂t (F.

15) (F.2 The far ﬁelds from an extended source distribution −iµ0 eik|x| d3 x e−ik·x jω × k 4π |x| V i eik|x| x× ˆ Erad (x) = d3 x e−ik·x jω × k ω 4πε0 c |x| V F.uu.2.14) F.17) (F.12) F.3 Force and energy Poynting’s vector S = E×H Maxwell’s stress tensor 1 T i j = Ei D j + Hi B j − δi j (Ek Dk + Hk Bk ) 2 (F.1 Relationship between the ﬁeld vectors in a plane wave B= ˆ k×E c (F.plasma.18) Draft version released 13th November 2000 at 22:01.se/CED/Book ¤ ¤ Brad (x) = ω (F. .13) (F.2.16) (F.2.3 The far ﬁelds from an electric dipole ωµ0 eik|x| pω × k 4π |x| 1 eik|x| (pω × k) × k Erad (x) = − ω 4πε0 |x| Brad (x) = − ω Downloaded from http://www.1.150 F ORMULAE Lorentz’ gauge condition in vacuum ·A+ 1 ∂ φ=0 c2 ∂t (F.11) F.2 Electromagnetic Radiation F.

3 S PECIAL R ELATIVITY 151 F.2.28) (F.22) ¡ 4 4 + (x − x ) × ˙ (x − x0 ) × v 2 c (F.24) ££ ££ (F.plasma. x) B(t.se/CED/Book § ¢ ¦ E(t.27) ¡ (F.F.2.25) (F.23) (F.29) (F.5 The far ﬁelds from an electric quadrupole Brad (x) = ω iµ0 ω eik|x| k · Qω × k 8π |x| i eik|x| Erad (x) = k · Qω × k × k ω 8πε0 |x| F. x) B(t.6 The ﬁelds from a point charge in arbitrary motion s = x − x − (x − x ) · v c x − x0 = (x − x ) − |x − x | v c F. x) = c2 (x − x0 ) × v c 2 Draft version released 13th November 2000 at 22:01.20) F.7 The ﬁelds from a point charge in uniform motion E(t. x) = q v2 (x − x0 ) 1 − 2 4πε0 s3 c E(t.uu.2.19) (F.21) (F. x) = (x − x ) × c|x − x | 5 5 (F.4 The far ﬁelds from a magnetic dipole Brad (x) = − ω µ0 eik|x| (mω × k) × k 4π |x| k eik|x| mω × k Erad (x) = ω 4πε0 c |x| (F.26) (F.2.30) . x) = q v2 (x − x0 ) 1 − 2 3 4πε0 s c v × E(t. ¢ s= |x − x0 |2 − ¢ ¡ ¡ ¢ ∂t ∂t = x |x − x | s Downloaded from http://www.

3 Special Relativity F.34) (F.32) F.36) (F.3.uu.1 Metric tensor 1 0 0 0 0 −1 0 0 0 0 −1 0 0 0 0 −1 F.31) (F.35) (F.4 Invariant line element ds = c dt = c dτ γ (F.3.2 Covariant and contravariant four-vectors vµ = gµν vν (F.se/CED/Book 0 ) uµ = P γ= 1 1 − β2 dx µ v = γ.152 F ORMULAE F. γ ds c V QRRS Λµν = γ −γβ 0 0 −γβ γ 0 0 0 0 1 0 0 0 0 1 UU ET V QRRS gµν = UU ET (F.37) F.3 Lorentz transformation of a four-vector x µ = Λµν xν (F.33) β= v c F.3.plasma.38) Draft version released 13th November 2000 at 22:01. .3.3.5 Four-velocity Downloaded from http://www.

ρ v c (F. 2.9 Field tensor ∂Aν ∂Aµ F µν = − = ∂xµ ∂xν 0 −E x −Ey −Ez Ex 0 −cBz cBy Ey cBz 0 −cB x Ez −cBy cB x 0 F. .se/CED/Book .3. and x3 ≡ z. The differential vector operator is in Cartesian coordinates given by ≡ ∑ xi ˆ i=1 3 ∂ def ∂ def ≡ xi ˆ ≡ ∂ ∂xi ∂xi where xi . x2 ≡ y.F.41) F. . cp) (F. z) and let |x| denote the magnitude (“length”) of x. x3 ) ≡ (x. . .3.3.uu. cA) (F. x2 . arbitrary vector ﬁelds. 3 is the ith unit vector and x1 ≡ x. ∂x1 ∂x2 ∂x3 = Draft version released 13th November 2000 at 22:01.7 Four-current density F.plasma. i = 1. β(x).42) (F. In ˆ ˆ ˆ ˆ ˆ ˆ ˆ component (tensor) notation can be written = ∂i = ∂ ∂ ∂ . ¢ ¢ i ∂ ∂ ∂ . .4 V ECTOR R ELATIONS 153 F.8 Four-potential Aµ = (φ.40) ) (F.39) F.6 Four-momentum pµ = m0 c2 uµ = (E.44) Downloaded from http://www. ∂x ∂y ∂z UU ET V ¡ QRRS 0 jµ = ρ0 uµ = ρ. c(x). . .4 Vector Relations Let x be the radius vector (coordinate vector) from the origin to the point (x1 . .43) ¡ (F. b(x).3. d(x). be arbitrary scalar ﬁelds and a(x). y. Let further α(x).

52) a · (b × c) = (a × b) · c (F.1 Spherical polar coordinates Base vectors r = sin θ cos ϕ x1 + sin θ sinϕ x2 + cos θ x3 ˆ ˆ ˆ ˆ ˆ θ = cos θ cos ϕ x1 + cos θ sin ϕ x2 − sin θ x3 ˆ ˆ ˆ (F.50) F.53) Downloaded from http://www.se/CED/Book Draft version released 13th November 2000 at 22:01.47) (F.45a) (F.45c) (F.49) (F.48) (F.154 F ORMULAE F.4.51) (F.2 Vector formulae General relations a · b = b · a = δi j ai b j = ab cos θ a × b = −b × a = ˆ i jk a j bk xi (F.45b) (F.46c) ϕ = − sin ϕ x1 + cos ϕ x2 ˆ ˆ ˆ ˆ x1 = sin θ cos ϕˆ + cos θ cos ϕθ − sin ϕϕ ˆ r ˆ ˆ x2 = sin θ sin ϕˆ + cos θ sin ϕθ + cos ϕϕ ˆ r ˆ ˆ x3 = cos θˆ − sin θθ ˆ r Directed line element ˆ dx x = dl = drˆ + r dθθ + r2 sin θ dϕϕ ˆ r ˆ Solid angle element dΩ = sin θ dθ dϕ Directed area element d2x n = dS = dS r = r2 dΩˆ ˆ ˆ r Volume element d3x = dV = drdS = r2 dr dΩ (F.46b) (F.plasma. .uu.46a) (F.4.

56) (F.F.55) (F.66) a (F. Downloaded from http://www.65) (F.54) (F.58) · (αa) = a · α + α · a × (αa) = α × a − a × α · (a × b) = b · ( × a) − a · ( × b) × (a × b) = a( · b) − b( · a) + (b · )a − (a · )b (a · b) = a × ( × b) + b × ( × a) + (b · )a + (a · )b · α= 2 (F.62) (F.60) (F.67) Draft version released 13th November 2000 at 22:01.64) × α=0 · ( × a) = 0 × ( × a) = ( · a) − 2 (F.plasma.57) (F.uu.se/CED/Book .4 V ECTOR R ELATIONS 155 a × (b × c) = b(a · c) − c(a · b) a × (b × c) + b × (c × a) + c × (a × b) = 0 (a × b) · (c × d) = a · [b × (c × d)] = (a · c)(b · d) − (a · d)(b · c) (a × b) × (c × d) = (a × b · d)c − (a × b · c)d (αβ) = α β + β α (F.59) (F.63) α (F.61) (F.

73) k·x |x|3 if |x| = 0 (F.76) Draft version released 13th November 2000 at 22:01.71) 1 |x| = 4πδ(x) (F.68) ×x = 0 (F.72) =− k·x |x|3 (F. ·x = 3 (F.uu.plasma.74) (F.75) (F.69) |x| = x |x| x |x|3 (F. .70) 2 × (k × a) = k( · a) + k × ( × a) − (k · a) Downloaded from http://www.se/CED/Book ¢ ¢ k |x| =k 2 1 |x| = −4πkδ(x) ¢ ¡ § ¢ × k× x |x|3 =− ¡ § ¢ ¦ ¡ ¢ k |x| = k· 1 |x| ¢ ¡ ¢ · x |x|3 =− 2 ¡ ¢ ¦ ¡ ¡ ¡ ¡ 1 |x| =− (F.156 F ORMULAE Special relations In the following k is an arbitrary constant vector.

1995. Arfken and Hans J. ISBN 0-12-059816-7. whose line element is dl. Denote the 3dimensional volume element by d3x(≡ dV) and the surface element. directed along the outward pointing surface normal unit vector n.uu.4 B IBLIOGRAPHY 157 Integral relations Let V(S ) be the volume bounded by the closed surface S (V). Panofsky and Melba Phillips. [3] Wolfgang K.. . Part I. .se/CED/Book . MA . . NY . Inc.79) (F. Inc. New York. H. 1953. Morse and Herman Feshbach.80) (F. international edition. by dS(≡ d2x n). CA . ISBN 0-201-05702-6. Reading. San Diego. Weber. ISBN 07-043316-8. [2] Philip M. . second edition. ¤ C a · dl = ¤ C S dS × α S dS · ( × a) © ( × a) d3x = © ( α) d3x = © S ¤ ¤ ¤ © © dS α (F. 1962.78) S dS × a (F. Draft version released 13th November 2000 at 22:01.81) Downloaded from http://www. Inc. Addison-Wesley Publishing Company. Classical Electricity and Magnetism. .. . Mathematical Methods for Physicists. ˆ ˆ ( · a) d3x = V S dS · a (F. .. .F.plasma. Methods of Theoretical Physics. Academic Press. McGraw-Hill Book Company. fourth. then α dl = Bibliography [1] George B.77) V V If S (C) is an open surface bounded by the contour C(S ). .

se/CED/Book Draft version released 13th November 2000 at 22:01. .158 F ORMULAE Downloaded from http://www.plasma.uu.

One suitable representation is in terms of an ordered N-tuple. The difference between a scalar.1. A scalar describes a scalar quantity which may or may not be constant in time and/or space. . M.M Mathematical Methods M. l. . The most basic vector is radius vector which is the vector from the origin to the point of interest. which are used in four-dimensional (4D) space. However. 3. ν. 2. Throughout we adopt the convention that Latin indices i. Vectors and Tensors Every physical observable can be described by a geometric object. tensors or pseudotensors and will not exploit differential forms to any signiﬁcant degree. pseudovector and a pseudotensor is that the latter behave differently under such coordinate transformations which cannot be reduced to pure rotations. vector and tensor and a pseudoscalar. . vectors. of the coordinates xN where N is the dimensionality of the space under consideration. generalisations to more abstract notions of these quantities are commonplace. k. Its N-tuple representation simply enumerates the coordinates 159 ¥ . run over the range 0. pseudovectors.1 Vectors Radius vector Any vector can be represented mathematically in several different ways. . and Greek indices µ. A vector describes some kind of physical motion due to vection and a tensor describes the motion or deformation due to some form of tension. . pseudoscalars. . We will describe the observables in classical electrodynamics mathematically in terms of scalars. 2. run over the range 1. 1. j. 3 to denote vector or tensor components in the real Euclidean three-dimensional (3D) conﬁguration space 3 .1 Scalars. κ. . or row vector. λ.

The coordinates xi are scalar quantities which describe the position along the unit base vectors x i ˆ 3 .3) (M. x2 . x1 . x3 ) ≡ (x. Therefore a representation of the radius vector in 3 is which span ˆ ˆ x = ∑ x i xi ≡ x i xi def i=1 3 where we have introduced Einstein’s summation convention (EΣ) which states that a repeated index in a term implies summation over the range of the index in question. In the 3D space 3 .plasma. Typographically. the radius vector from the origin to a point is synonymous with the coordinates of the point itself. x2 .5) dx µ = ν dxν ∂x which follows trivially from the rules of differentiation of x µ considered as functions of four variables xν . In this sense. the differential radius vector dxµ transforms as ∂x µ (M. due to a transformation from a system Σ to another system Σ . Whenever possible and convenient we shall in the following always assume EΣ and suppress explicit summation in our formulae.uu. x3 ). we represent a 3D vector by a boldface letter or symbol in a Roman font. we have N = 3 and the radius vector can be represented by the triplet (x1 . 3. 2. z) def This component notation is particularly useful in 4D space where we can represent the radius vector either in its contravariant component form xµ ≡ (x0 . x2 . x1 .1) (M. i = 1. y.se/CED/Book Draft version released 13th November 2000 at 22:01. Downloaded from http://www. 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. x3 ) of coordinates xi . we may describe the radius vector in component notation as follows: xi ≡ (x1 .160 M ATHEMATICAL M ETHODS which describe this point. 4 is often utilised to formulate the special theory of relativity. x2 . x3 ) or its covariant component form xµ ≡ (x0 .4) . ¥ W W ¥ ¥ (M.2) (M. x2 . x3 ) def def 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 physics. An example is Lorentz space 4 which is a 4D Riemannian space. We note that for a change of coordinates xµ → x µ = x µ (x0 . Alternatively. x1 .

x3 ) ≡ α(xi ) def α(x0 . a four-scalar ﬁeld is denoted ¥ We denote an arbitrary scalar ﬁeld in ¥ 3 by (M. In particular.8) a(x) = xi ai (x) ˆ In component notation this same vector can be represented as ai (x) = (a1 (x). Since a four-scalar has the same value at a given point regardless of coordinate system.M.plasma. V ECTORS AND T ENSORS 161 M. x2 .1. a3 (x)) = ai (x j ) Draft version released 13th November 2000 at 22:01.uu. Scalar ﬁelds α(x) = α(x1 . Such a parameter can be viewed as a “continuous index” which enumerates the “coordinates” of the ﬁeld. for the differential dxµ .9) (M.2 Fields A ﬁeld is a physical entity which depends on one or more continuous parameters. again. Vector ﬁelds (M. a2 (x).10) Downloaded from http://www.7) which indicates that the four-scalar α depends on all four coordinates spanning this space.se/CED/Book ¥ This ﬁeld describes how the scalar quantity α varies continuously in 3D space. follows trivially from the rules of differentiation. the transformation rule for the differential operator ∂/∂xµ under a transformation xµ → x µ becomes ∂ ∂xν ∂ = ∂x µ ∂x µ ∂xν which. Analogous to the transformation rule. Equation (M. x1 . In 4D. it is also called an invariant. in a ﬁeld which depends on the usual radius vector x of 3 . ¥ We can represent an arbitrary vector ﬁeld a(x) in 3 as follows: (M.5) on the facing page.6) 3 .1 S CALARS . x3 ) ≡ α(xµ ) def (M. x2 . 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.

the tensor ﬁeld described here is a tensor of rank two. depends on its transformation properties during a change of coordinates. a3 (xν )) (M. according to the rule yµ = ∂xν yν ∂x µ (M.162 M ATHEMATICAL M ETHODS In 4D. in the last member. Again. for instance in the following matrix form: where. a2 (xν ). a1 (xν ).11) where xν is the radius four-vector.. Strictly speaking.15) . y3 ) constitutes a contravariant four-vector (or the contravariant components of a four-vector) if and only if.uu. as aµ (xν ) = (a0 (xν ). 4 SQ A(x) = ≡ Ai j (xk ) 5 V A11 (x) A12 (x) A13 (x) A21 (x) A22 (x) A23 (x) A31 (x) A32 (x) A33 (x) T ¥ def (M. during the change from Σ to Σ . a3 (xν )) or. For instance.e. in covariant component form. the relation between aµ and aµ is determined by the metric of the physical 4D system under consideration. during a transformation from a system Σ with coordinates xµ to a system Σ with coordinates x µ . it transforms to the new system according to the rule yµ= ∂x µ ν y ∂xν (M. in 4D an assemblage yµ = (y0 .5) on page 160. in the same way as the differential coordinate element dx µ transforms according to Equation (M.. a2 (xν ).8) on the preceding page. we have again used the more compact component notation. Tensor ﬁelds We denote an arbitrary tensor ﬁeld in 3 by A(x). a1 (xν ). y2 .12) (M. an arbitrary four-vector ﬁeld in contravariant component form can be represented as aµ (xν ) = (a0 (xν ).e.se/CED/Book Draft version released 13th November 2000 at 22:01.13) i. Downloaded from http://www. Whether an arbitrary N-tuple fulﬁls the requirement of being an (N-dimensional) contravariant vector or not. in the same way as the differential operator ∂/∂xµ transforms according to Equation (M.14) i. The analogous requirement for a covariant four-vector is that it transforms. y1 .plasma. This tensor ﬁeld can be represented in a number of ways.

M. We speak of • a contravariant four-tensor ﬁeld.. denoted Aµ1 µ2 . j.. In 4D. lm = = δ jl δkm − δ jm δkl V 1 0 0 0 1 0 0 0 1 T SQ mo n k δi j = 0 if i = j 1 if i = j ¥ A particularly simple rank-two tensor in bol δi j .se/CED/Book .. • a covariant four-tensor ﬁeld.18) (M. This metric tensor determines the relation between an arbitrary contravariant four-vector aµ and its covariant counterpart aµ according to the following rule: aµ (xκ ) ≡ gµν aν (xκ ) def µ µ . we have three forms of four-tensor ﬁelds of rank n.µ Draft version released 13th November 2000 at 22:01. j.plasma.µn (xν ).2. k is an odd permutation of 1. A tensor of rank n = 2 may be represented by a twodimensional array or matrix whereas higher rank tensors are best represented in their component forms (tensor notation). k are equal if i.20) Downloaded from http://www.µnk (xν ). k is an even permutation of 1. j. k+1 The 4D metric tensor (fundamental tensor) mentioned above is a particularly important four-tensor of rank 2. • a mixed four-tensor ﬁeld. In covariant component form we shall denote it gµν . In this picture a scalar is considered to be a tensor of rank n = 0 and a vector a tensor of rank n = 1.uu. tensors may have any rank n..19) (M. denoted Aµ1 2. Consequently. V ECTORS AND T ENSORS 163 The 3D Kronecker delta has the following matrix representation (δi j ) = Another common and useful tensor is the fully antisymmetric tensor of rank 3. also known as the Levi-Civita tensor 1 0 −1 if i.3 if at least two of i.3 i jk with the following further property i jk ilm In fact..17) (M..16) (M...1 S CALARS . denoted Aµ1 µ2 .2. the notation where a vector (tensor) is represented in its component form is called the tensor notation.µn (xν ). with the following properties: 3 is the 3D Kronecker delta sym- (M.

.uu.νk−1 (xκ ) µk νk+1 .1: Surface element d2x and the unit normal vector n... For example.νn k+1 ν ν2 .νk−1 ... According to Newton’s third law.se/CED/Book T ENSORS IN 3D SPACE n ˆ d2x ˆ F IGURE M.1 Consider the tetrahedron-like volume element V indicated in Figure M. ﬂuid..24) i.23) Successive lowering and raising of more than one index is achieved by a repeated application of this rule.1 on the facing page be a directed surface ˆ element of this volume element and let the vector T n d2x be the force that matter.22) (M. a dual application of the lowering operation on a rank 2 tensor in contravariant form yields Aµν = gµκ gλν Aκλ (M. .. whose atomistic structure is irrelevant for the present analysis. E XAMPLE M..1 on the next page of a solid.νn (xκ ) = Aν1 ν2νk+2 .. acts on matter which ˆ lies on the opposite side of d2x. which at a Downloaded from http://www.... the following lowering and raising rules hold for arbitrary rank n mixed tensor ﬁelds: ν .νnk (xκ ) νk+1 k (M.plasma. we ﬁnd that the matter of mass m.25) and Newton’s second law.e. lying ˆ on the side of d2x toward which the unit normal vector n points..νn (xκ ) = Aν1 ν2 ..νk−1 νk gµk νk Aν1 ν2νk+2 . this surface force fulﬁls T− n = −T n ˆ ˆ (M. Let dS = d2x n in Figure M. This force concept is meaningful only if the forces are short-range enough that they can be assumed to act only in the surface proper.νk−1 µ gµk νk Aν1νk+1 . the same rank 2 tensor in covariant form..25) Using (M.. Draft version released 13th November 2000 at 22:01..21) More generally. or gaseous body. The raising of index analogue of the index lowering rule is: aµ (xκ ) ≡ gµν aν (xκ ) def (M.164 M ATHEMATICAL M ETHODS This rule is often called lowering of index. This operation is also known as a tensor contraction.

Introducing the notation T i j = T xi ˆ j for the jth component of the vector T xi .1 S CALARS .uu. (M.27) ˆ ˆ ˆ ˆ dx m Since both a and Fext /m remain ﬁnite whereas m/d2x → 0 as V → 0.26) where Fext is the external force and a is the acceleration of the volume element. one ﬁnds that in this limit T n = ∑ ni T x i ≡ n i T xi ˆ ˆ ˆ i=1 3 From the above derivation it is clear that Equation (M.28) on the precedˆ Draft version released 13th November 2000 at 22:01.28) b (M. In other words m Fext T n = n 1 T x 1 + n 2 T x2 + n 3 T x3 + 2 a − (M.M.se/CED/Book .28) above is valid not only in equilibrium but also when the matter in V is in motion. we can write Equation (M. given instant is located in V obeys the equation of motion T n d2x − T x1 d2x − T x2 d2x − T x3 d2x + Fext = ma ˆ ˆ ˆ ˆ (M. V ECTORS AND T ENSORS 165 x3 n ˆ d2x x2 V x1 F IGURE M.29) a Downloaded from http://www.2: Terahedron-like volume element V containing matter.plasma.

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.se/CED/Book Draft version released 13th November 2000 at 22:01. x3 ).166 M ATHEMATICAL M ETHODS ing page in component form as follows T n j = (T n) j = ∑ ni T i j ≡ ni T i j ˆ ˆ i=1 3 (M.1 .31) (M. EXAMPLE Downloaded from http://www. Equation (M. equivalently ρ ∂v j ∂T i j + ρv · v j − f j − =0 ∂t ∂xi (M. Note that T nm is independent of the particular coordinate ˆˆ system used in the derivation.uu.35) or. 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. the jth component of the vector T xi . we must require that ρ ∂T i j d vj − fj − =0 dt ∂xi (M. can be interpreted as the ˆ i jth component of a tensor T.30) was used.plasma. ! f f ρ d v j d3x = dt f j d3x + V ∂T i j 3 dx ∂xi f f f f d v j dm = dt f j d3x + f f f d dt v dm = f d3x + S T n d2x ˆ T nj d2x = ˆ i = ∑ T n jm j = ∑ ˆ 3 3 h T nm = T n · m ˆˆ ˆ ˆ ∑ ni T i j 3 m j ≡ ni T i j m j = n· T · m ˆ ˆ (M. here denoted T i j . we obtain V V The jth component of this equation can be written V V S V where.36) Note that ∂v j /∂t is the rate of change with time of the velocity component v j at a ﬁxed point x = (x1 .33) f f (M.30) Using Equation (M.32) f j d3x + S ni T i j d2x (M. Setting dm = ρ d 3x and using the divergence theorem on the last term.34) E ND OF M. To this end we consider a part V of the body. we can rewrite the result as V V Since this formula is valid for any arbitrary volume. in the last step.30) above. x1 . we ﬁnd that the component of the vector T n in the ˆ direction of an arbitrary unit vector m is ˆ j=1 j=1 i=1 Hence.

in two real vectors aR and aI in 3 in the following way # " ¥ (M.39) i. V ECTORS AND T ENSORS 167 M.40) (M.41) (M. is an invariant called the metric.3 Vector algebra Scalar product The scalar product (dot product. if we like.1.21). I NNER PRODUCTS IN The inner product of A with itself may be deﬁned as from which we ﬁnd that Using this in Equation (M.uu..20) and (M. i. The result is a four-scalar.e.e.40).38) where we made use of the index lowering and raising rules (M. the scalar product of the differential radius four-vector with itself.16) on page 163. inner product) of two arbitrary 3D vectors a and b in ordinary 3 space is the scalar number a · b = x i ai · x j b j = x i · x j ai b j = δi j ai b j = ai bi ˆ ˆ ˆ ˆ where we used the fact that the scalar product xi · x j is a representation of ˆ ˆ the Kronecker delta δi j deﬁned in Equation (M. The quadratic differential form ds2 = gµν dxν dxµ = dxµ dxµ (M. # p A= a2 − a2 + 2iaR · aI ∈ R I # A2 ≡ A · A = a2 − a2 + 2iaR · aI ≡ A2 ∈ R I def def # ˆ A ≡ aR + iaI = aR aR + iaI aI ≡ A A ∈ ˆ ˆ def def 3 " A 3D complex vector A is a vector in 3 (or.1 S CALARS ..M. the scalar product is often denoted (ab). expressed in terms of E XAMPLE M.2 (M. we see that we can interpret this so that the complex Draft version released 13th November 2000 at 22:01. In 4D space we deﬁne the scalar product of two arbitrary four-vectors aµ and bµ in the following way aµ bµ = gνµ aν bµ = aν bν = gµν aµ bν (M.42) Downloaded from http://www.plasma. an invariant which is independent on in which inertial system it is measured. In Russian literature.37) COMPLEX VECTOR SPACE 6 ). It is also the square of the line element ds which is the distance between neighbouring points with coordinates x µ and xµ + dxµ .se/CED/Book .

−a) · (b0 . hence.43) (M.47) 4 The metric is the quadratic differential (M. ! q q which is the indeﬁnite.48) M. x) = (ct. −x) · (ct. −x) · (x0 .plasma. real norm of form q The important scalar product of the 4 radius four-vector with itself becomes (M. the deﬁnition of the scalar product in terms of the inner product of complex vector with its own complex conjugate yields |A|2 ≡ A · A∗ = a2 + a2 = |A|2 R I with the help of which we can deﬁne the unit vector as ˆ A = A= |A| = aR a2 + a 2 R I aR + i ˆ aI a2 + a 2 R I aI ˆ (M.168 M ATHEMATICAL M ETHODS unit vector is ˆ A A= = A aR a2 − a2 + 2iaR · aI R I aR + i ˆ aI a2 − a2 + 2iaR · aI R I aI ˆ a2 + a 2 R I a2 + a 2 R I On the other hand. the scalar product in Equation (M. x) = (ct)2 − (x1 )2 − (x2 )2 − (x3 )2 = s2 4.44) M.46) ds2 = dxµ dxµ = c2 (dt)2 − (dx1 )2 − (dx2 )2 − (dx3 )2 E ND OF EXAMPLE Downloaded from http://www.3 ! # aR + i ˆ p p aR a2 + a 2 R I aI a2 + a 2 R I aI ∈ ˆ 3 E ND OF # = aR + i ˆ p p p aR a2 − a2 − 2iaR · aI R I p p p aI a2 − a2 − 2iaR · aI R I aI ∈ ˆ 3 (M.2 q .se/CED/Book Draft version released 13th November 2000 at 22:01.7) on page 54 for an example] and. NORM AND METRIC IN L ORENTZ SPACE E XAMPLE M. (M.uu.45) def a2 + a 2 R I a2 + a 2 R I EXAMPLE S CALAR PRODUCT.38) on the preceding page can be evaluated almost trivially and becomes aµ bµ = (a0 . b) = a0 b0 − a · b xµ xµ = (x0 .3 In 4 the metric tensor attains a simple form [see Equation (5.

r. a dyadic product is often called tensor product and is frequently denoted a ⊗ b. In such a space. V ECTORS AND T ENSORS 169 In the general theory of relativity. φ).e. the metric takes the form ds2 = c2 eκ (dt)2 − eλ (dr)2 − r2 (dθ)2 − r2 sin2 θ(dφ)2 (M.50) In general relativity the metric tensor is not given a priori but is determined by the Einstein equations.plasma. In matrix notation the outer product of a and b is written ˆ ab = x1 x2 ˆ x3 ˆ which means that we can represent the tensor A(x) in matrix form as Ai j (xk ) = which we identify with expression (M.. Operating on this dyad from the right and from the left with an inner product of an vector c one obtains A · c ≡ ab · c ≡ a(b · c) c · A ≡ c · ab ≡ (c · a)b def def def def (M. respectively. several important problems are treated in a 4D spherical polar coordinate system.49) E ND OF M. θ. Draft version released 13th November 2000 at 22:01. φ) and the metric tensor is eκ 0 0 0 0 e−λ 0 0 0 0 −r2 0 0 0 0 −r2 sin2 θ where κ = κ(ct. new vectors.se/CED/Book ! T T T y SQ 5 SQ 5 sttu gµν = ww xv 4 4 M ETRIC IN GENERAL RELATIVITY E XAMPLE M. a tensor in matrix notation. viz. EXAMPLE 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.53) Downloaded from http://www.52) (M. θ. proportional to a and b. r. r.M.51a) (M.4 . θ. Then the radius four-vector can be given as xµ = (ct.uu. In mathematics. V a1 b1 a1 b2 a1 b3 a1 b2 a2 b2 a2 b3 a1 b3 a3 b2 a3 b3 V SQ V a1 b1 a1 b2 a1 b3 a1 b2 a2 b2 a2 b3 a1 b3 a3 b2 a3 b3 x1 ˆ x2 ˆ x3 ˆ (M.51b) i.4 (M.15) on page 162.1 S CALARS . φ) and λ = λ(ct.

18) on page 163. M. On the other hand.” Pseudovectors transform as ordinary vectors under translations and proper rotations. [ab]. ∂x1 ∂x2 ∂x3 (M. . 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.4 Vector analysis The del operator In 3 the del operator is a differential vector operator. x3 ) = (−x1 . as is seen from Equation (M. . Since the ˆ operator in itself has vectorial properties. −x3 ) changes sign of the components of the vectors a and b so that in the new coordinate system a = −a and b = −b.55) ∂ ∂ ∂ .54) ¥ ≡ xi ˆ ∂ def ≡ ∂ ∂xi (M. or pseudovector. Sometimes the vector product of a and b is denoted a ∧ b or.54) above.uu. In “component” notation we can write Downloaded from http://www. or polar vector. and zero-rank pseudotensors are therefore also called pseudoscalars.170 M ATHEMATICAL M ETHODS Vector product The vector product or cross product of two arbitrary 3D vectors a and b in ordinary 3 space is the vector c = a×b = Here i jk is the Levi-Civita tensor deﬁned in Equation (M. −x2 . which is to say that the direction of an ordinary vector is not dependent on the choice of directions of the coordinate axes. Therefore a (or b) is an example of a “true” vector. particularly in the Russian literature. Scalars are tensors of rank zero. denoted in Gibbs’ notation by and deﬁned as def where xi is the ith unit vector in a Cartesian coordinate system. A prototype for a pseudovector is the angular momentum vector and hence the attribute “axial.plasma.se/CED/Book ¢ ∂i = ¡ ¥ ˆ i jk a j bk xi (M. whereas c is an example of an axial vector.56) Draft version released 13th November 2000 at 22:01.1. we denote it with a boldface nabla. the cross product vector c does not change sign. A spatial reversal of the coordinate system (x1 . This triple ˆ ˆ ˆ product is a representation of the i jk component of the Levi-Civita tensor i jk which is a rank three pseudotensor. x2 . an example being the pseudoscalar xi · ( x j × xk ).

14) on page 162 as y µ = ∂ν x µ yν and yµ = ∂µ x ν yν respectively. sometimes denoted an opposite sign convention.∂ = c ∂t 1∂ . c ∂t q In the contravariant form of the four-del operator can be represented as ∂ = µ 1∂ . V ECTORS AND T ENSORS 171 In 4D. the contravariant component representation of the four-del operator is deﬁned by ∂µ = whereas the covariant four-del operator is ∂µ = We can use this four-del operator to express the transformation properties (M. −∂ = c ∂t 1∂ . and sometimes deﬁned with M. ∂x0 ∂x1 ∂x2 ∂x3 ¢ E ND OF 5 5 4 ¡ ¡ 4 ∂ ∂ ∂ ∂ .13) and (M.59) (M. one obtains ∂µ ∂µ = 2 = which is the d’Alembert operator.60) (M.uu.62) (M.5 .M.1 S CALARS .plasma.57) (M.− c ∂t ¢ ∂ ∂ ∂ ∂ .63) . EXAMPLE Draft version released 13th November 2000 at 22:01. Downloaded from http://www. .5 (M.58) (M.DEL OPERATOR IN L ORENTZ SPACE 4 and the covariant form as Taking the scalar product of these two.61) E XAMPLE M. T HE FOUR . .se/CED/Book ! d d 1 ∂2 − c2 ∂t2 2 ∂µ = 1∂ . ∂x0 ∂x1 ∂x2 ∂x3 (M. . .

6 " ¥ b ¥ The gradient of an a(x): 3 scalar ﬁeld α(x). the “unprimed” version.64) b a ..65) G RADIENTS OF SCALAR FUNCTIONS OF RELATIVE DISTANCES IN 3D E XAMPLE M. In 4D. on |x − x |. is an 3 vector ﬁeld (M.66) Using this.se/CED/Book Draft version released 13th November 2000 at 22:01. we obtain the following two very useful results: |x − x | = xi ˆ ∂|x − x | x−x ∂|x − x | = = − xi ˆ ∂xi |x − x | ∂xi |x − x | (M. denoted α(x). Equation (M.e. In analogy with Equation (M. i.67) and EXAMPLE Downloaded from http://www. divergence and curl of a tensor (in the generalised sense).6 Very often electrodynamic quantities are dependent on the relative distance in 3 between two vectors x and x . with the following component form: ∂µ α(xν ) = ∂α(xν ) ∂xµ (M.68) E ND OF M. ! 1 |x − x | =− x−x =− |x − x |3 a =− 1 |x − x | (M.plasma. the four-gradient is a covariant vector.uu. and elementary rules of differentiation.55) on page 170.55) on page 170. formed as a derivative of a four-scalar ﬁeld α(xµ ).172 M ATHEMATICAL M ETHODS With the help of the del operator we can deﬁne the gradient. we can deﬁne the “primed” del operator in the following way: = xi ˆ ∂ =∂ ∂xi (M. The gradient α(x) = ∂α(x) = xi ∂i α(x) = a(x) ˆ 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.

V ECTORS AND T ENSORS 173 The divergence · a(x) = ∂ · x j a j (x) = δi j ∂i a j (x) = ∂i ai (x) = ˆ which. EXAMPLE The Laplacian The 3D Laplace operator or Laplacian can be described as the divergence of the gradient operator: 2 =∆= · = 3 ∂ ∂2 ∂2 ∂ xi · x j ˆ ˆ = δi j ∂i ∂ j = ∂2 = 2 ≡ ∑ 2 i ∂xi ∂x j ∂xi i=1 ∂xi (M.1 S CALARS . works.M.70) M. Downloaded from http://www.7 E ND OF ¥ ¥ We deﬁne the 3D divergence of a vector ﬁeld in 3 as (M. We may think of the divergence as a scalar product between a vectorial operator and a vector. is a scalar ﬁeld in 3 .69) ∂ai (x) = α(x) ∂xi (M.7 .72) The symbol 2 is sometimes read del squared.71) E XAMPLE M. Again we see that the boldface notation for the 3D del operator is very convenient.se/CED/Book ! a(x ) · |x − x | " For an arbitrary vector ﬁeld a(x ). it is a sign of concentration of α at that point. as indicated by the notation α(x).66) on the preceding page.uu. If. for a scalar ﬁeld α(x). the result of a divergence operation is a scalar. Draft version released 13th November 2000 at 22:01. The four-divergence of a four-vector aµ is the following four-scalar: ∂µ aµ (xν ) = ∂µ aµ (xν ) = ∂aµ (xν ) ∂xµ D IVERGENCE IN 3D 3 which demonstrates how the “primed” divergence.plasma. deﬁned in terms of the “primed” del operator in Equation (M. As is the case for any scalar product. 2 α < 0 at some point in 3D space. the following relation holds: · a(x ) = + a(x ) · |x − x | 1 |x − x | (M.

” EXAMPLE The curl × a(x) = ˆ i jk xi ∂ j ak (x) = ˆ i jk xi ∂ak (x) = b(x) ∂x j where use was made of the Levi-Civita tensor. introduced in Equation (M.76) (M.73) 1 |x − x | E ND OF M.64) on page 172. ¥ In 3 the curl of a vector ﬁeld a(x).77) ! " E XAMPLE M.75) T HE CURL OF A GRADIENT E XAMPLE M.se/CED/Book ∂2 ∂2 − ∂x1 ∂x2 ∂x2 ∂x1 ∂2 ∂2 − ∂x3 ∂x1 ∂x1 ∂x3 ∂2 ∂2 − ∂x2 ∂x3 ∂x3 ∂x2 α(x) x2 ˆ α(x) x3 ˆ Draft version released 13th November 2000 at 22:01. denoted × a(x). vanishes: = = + + ≡0 i jk Downloaded from http://www.18) on page 163.8 · 1 |x − x | = 2 " A very useful formula in 3D ¥ T HE L APLACIAN AND THE D IRAC DELTA 3 is = −4πδ(x − x ) (M. we see that × [ α(x)] = ˆ i jk xi ∂ j ∂k α(x) ˆ i jk xi ∂ j ∂k α(x) ∂ ∂ α(x) xi ˆ ∂x j ∂xk α(x) x1 ˆ which.8 .74) (M.uu. and the gradient.plasma. due to the assumed well-behavedness of α(x). (M. Equation (M.174 M ATHEMATICAL M ETHODS where δ(x − x ) is the 3D Dirac delta “function.74) above. Equation (M. is another b(x) which can be deﬁned in the following way: 3 vector ﬁeld (M. The covariant 4D generalisation of the curl of a four-vector ﬁeld aµ (xν ) is the antisymmetric four-tensor ﬁeld Gµν (xκ ) = ∂µ aν (xκ ) − ∂ν aµ (xκ ) = −Gνµ (xκ ) A vector with vanishing curl is said to be irrotational.9 Using the deﬁnition of the 3 curl.

uu.M.83) E ND OF EXAMPLE M.e.10 . the four-divergence of the four-curl is not zero. we ﬁnd that · [ × a(x)] = ∂i [ × a(x)]i = i jk ∂i ∂ j ak (x) (M.9 " so that the four-curl of a four-gradient vanishes just as does a curl of a gradient in " for any arbitrary. (M. ∂i i jk ∂ j ak (x) = = ∂ ∂xi i jk ∂ ak ∂x j a1 (x) ≡0 i. Downloaded from http://www. due to the assumed well-behavedness of a(x). d E XAMPLE M.18) on page 163.10 ! M.se/CED/Book ! " for any arbitrary. well-behaved scalar ﬁeld α(x). we ﬁnd that..1 S CALARS . a (xκ ) = 0 + ∂2 ∂2 − ∂x1 ∂x2 ∂x2 ∂x1 + ∂2 ∂2 − ∂x3 ∂x1 ∂x1 ∂x3 ∂2 ∂2 − ∂x2 ∂x3 ∂x3 ∂x2 a2 (x) a3 (x) (M. well-behaved 3 vector ﬁeld a(x). E ND OF EXAMPLE (M.plasma. Equation (M.79) 3. V ECTORS AND T ENSORS 175 We thus ﬁnd that × [ α(x)] ≡ 0 In 4D we note that for any well-behaved four-scalar ﬁeld α(xκ ) (∂µ ∂ν − ∂ν ∂µ )α(xκ ) ≡ 0 Hence.82) (M.80) Using the deﬁnition for the Levi-Civita symbol.69) and the curl.78) 3 T HE DIVERGENCE OF A CURL With the use of the deﬁnitions of the divergence (M.81) (M.74) on the preceding page. that · [ × a(x)] ≡ 0 In 4D. a gradient is always irrotational. deﬁned by Equation (M. for ∂νGµν = ∂µ ∂ν aν (xκ ) − 2 µ Draft version released 13th November 2000 at 22:01.

M.plasma.84) ¡ (M.2. qi . the Hamiltonian (Hamilton function) H can be deﬁned via the Legendre transformation H(pi .85) Draft version released 13th November 2000 at 22:01. the Lagrange function or Lagrangian L is given by where qi is the generalised coordinate. Evett [3]. t) = L qi .1 Lagrange’s equations As is well known from elementary analytical mechanics.se/CED/Book ¢ ∂ ∂t ∂L ∂qi ˙ − ∂L =0 ∂qi ¢ L(qi .176 M ATHEMATICAL M ETHODS Numerous vector algebra and vector analysis formulae are given in Chapter F. Those which are not found there can often be easily derived by using the component forms of the vectors and tensors. The Lagrangian satisﬁes the Lagrange equations We deﬁne the to the generalised coordinate qi canonically conjugate momentum pi according to pi = ∂L ∂qi ˙ (M. A short but very useful reference in this respect is the article by A.2 Hamilton’s equations From L.89) ˙ ∂pi ∂qi ∂t ∂qi ∂qi ˙ ∂t Downloaded from http://www.86) and note from Equation (M. qi .uu. t) = pi qi − L(qi .87) M. t) ˙ ˙ (M. ˙ ¡ dqi . together with the Kronecker and Levi-Civita tensors and their generalisations to higher ranks.2 Analytical Mechanics M. T the kinetic energy and V the potential energy of a mechanical system.88) 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 = qi dpi + pi dqi − ˙ ˙ dqi − dqi − dt (M. .2.85) above that ∂L = pi ˙ ∂qi (M. qi .t = T −V dt (M.

1965. Dean. .. ISBN 07-043316-8. Arfken and Hans J. Tensor Calculus. noting that according to Equation (M. . 1995. [3] Arthur A. . . Evett. Inc. New York. ISBN 05-001331-9. . American Journal of Physics. Draft version released 13th November 2000 at 22:01. Morse and Herman Feshbach. the second and fourth terms on the right hand side cancel.plasma.M.uu.se/CED/Book . CA . ISBN 0-471-20452-8. international edition. [4] Philip M. . NY . 1953. McGraw-Hill Book Company.90b) Bibliography [1] George B. Furthermore. [5] Barry Spain. Inc. fourth..86) on the facing page. Edinburgh and London.2 B IBLIOGRAPHY 177 According to the deﬁnition of pi . 1967. Oliver and Boyd. 34. A. Ltd. Wiley & Sons.90a) (M. third edition.89) on the facing page is equal to − pi dqi and ˙ identifying terms. NY . Methods of Theoretical Physics. Academic Press. San Diego. Downloaded from http://www. Weber. we obtain the Hamilton equations: dqi ∂H = qi = ˙ ∂pi dt ∂H dpi = − pi = − ˙ ∂qi dt (M. Inc. . Part I. Equation (M.. Elements of Abstract Algebra.. Permutation symbol approach to elementary vector analysis. . . New York. 1965. [2] R. Mathematical Methods for Physicists. ISBN 0-12-059816-7.87) on the preceding page the third term on the right hand side of Equation (M.

uu. .se/CED/Book Draft version released 13th November 2000 at 22:01.178 M ATHEMATICAL M ETHODS Downloaded from http://www.plasma.

5 Ampère-turn density. 160 covariant ﬁeld tensor. 26 charge density. 41 Ampère’s law. 51 covariant component form. 131 cyclotron radiation. 137 d’Alembert operator. 38 axial vector. 160 concentration. 160 contravariant ﬁeld tensor.Index acceleration ﬁeld. 67 contravariant four-tensor ﬁeld. 72. 66. 170 del squared. 130. 72 canonically conjugate momentum. 173 differential distance. 170 179 . 80 characteristic impedance. 65 associated Legendre polynomial. 54 convection potential. 62. 56 differential vector operator. 106 associative. 174 cutoff. 14 contraction. 141 antisymmetric tensor. 163 contravariant four-vector. 140 anomalous dispersion. 125 convective derivative. 104 Coulomb gauge. 171 del operator. 58 axial gauge. 30 complex vector. 66 covariant four-tensor ﬁeld. 54. 176 canonically conjugate momentum density. 134. 91 anisotropic. 8 closed algebraic structure. 6 birefringent. 162 covariant four-vector ﬁeld. 163 covariant four-vector. 57 contravariant vector. 58 coherent radiation. 2 covariant. 170 curl. 11 conservative forces. 130 bremsstrahlung. 173 conservative ﬁeld. 54 contravariant component form. 36. 167 component notation. 11 cosine integral. 118 advanced time. 3 classical electrodynamics. 57 covariant vector. 140 braking radiation. 170 Biot-Savart’s law. 140 complex ﬁeld six-vector. 77 constitutive relations. 137 canonically conjugate four-momentum. 162 contravariant four-vector ﬁeld. 18 complex notation. 54 cross product. 38 Coulomb’s law. 136 collisional interaction.

173 dot product. 161 ﬁeld Lagrange density. 163 Downloaded from http://www. 8 equations of classical magnetostatics. 18 E1 radiation. 108 electric dipole moment vector. 9 electric conductivity. 97 Faraday’s law. 3 ﬁeld quantum. 141 displacement current. 58 far ﬁeld. 10 electric dipole moment. 72 four-Lagrangian. 111 electric susceptibility. 79 Euler-Lagrange equations. 66 electromagnetic potentials. 172 four-Hamiltonian. 53 four-divergence. 173 four-gradient. 2 electric ﬁeld energy. 87 electric quadrupole radiation. 55 Euler-Lagrange equation. 89 electric ﬁeld. 34 electromagnetic scalar potential. 16 equations of classical electrostatics. 81 ﬁeld point. 46 far zone. 169 Einstein’s summation convention. 73 four-dimensional vector space. 89 electric volume force. 35 electromagnetic vector potential. 161 four-tensor ﬁelds. 109 electric displacement.se/CED/Book Draft version released 13th November 2000 at 22:01. 174 Dirac-Maxwell equations. 167 duality transformation. 88 electric quadrupole moment tensor. 58 Euclidean vector space. 79 Euler-Mascheroni constant. 87 electric permittivity. 62 four-scalar. 33 electrostatics. 169 dyons. 11 ﬁeld. 16 electromotive force (EMF). 62 equation of continuity for magnetic monopoles. 61 four-potential. 171 four-dimensional Hamilton equations. 54 dyadic product. 139 four-current.180 M ATHEMATICAL M ETHODS Dirac delta. 70 four-momentum. 93 electric monopole moment. 92 equation of continuity. 111 electric quadrupole tensor. 14 electric displacement vector. 1 energy theorem in Maxwell’s theory. 9. 131 ﬁne structure constant. 160 electric charge conservation law. 62 four-del operator.plasma. 35 electromagnetodynamic equations. . 111 Einstein equations. 131. 15 dispersive. 16 dummy index. 15 electromagnetodynamics. 109 E2 radiation. 87 electric dipole radiation.uu. 10 electrostatic scalar potential. 8 Euclidean space. 140 electric polarisation. 104 event. 10 divergence. 94 electromagnetic ﬁeld tensor.

se/CED/Book . 76 Kronecker delta. 4 general inhomogeneous wave equations. 51 inhomogeneous Helmholtz equation. 49 invariant. 80 Hamilton density equations. 77. 167 instantaneous. 81 intermediate ﬁeld. 176 Heaviside potential. 173 Laplacian. 161 invariant line element. 163 Lagrange density. 58 irrotational. 143 incoherent radiation. 37 Gauss’s law. 39 inhomogeneous time-independent wave equation. 141 Hamilton density. 176 kinetic momentum.uu. 38 gauge function. Downloaded from http://www. 160. 176 generalised four-coordinate. 176 Hamilton gauge. 38 functional derivative. 37 gauge transformation. 137 indeﬁnite norm. 124 light cone. 51 gauge ﬁxing. 61 Fourier component. 51 inertial system. 126 law of inertia. 57 Draft version released 13th November 2000 at 22:01. 176 Levi-Civita tensor. 38 inner product. 176 Lagrangian. 65. 38 Hamiltonian.181 four-vector. 174 Kelvin function. 25 Fourier transform. 51 Legendre polynomial. 77 Lagrange equations. 176 Lagrange function. 73 Gibbs’ notation. 39 inhomogeneous wave equation. 170 gradient. 39 Green´s function. 163 Galileo’s law. 77. 37 gauge invariant. 39 identity element. 78 fundamental tensor. 162 four-velocity. 4. 35 generalised coordinate. 176 Laplace operator. 57. 72. 56 inverse element. 104 Hertz’ vector. 106 Legendre transformation. 54. 177 Hamilton function. 125 Helmholtz’ theorem. 58 group velocity. 163 Liénard-Wiechert potentials.plasma. 24 Huygen’s principle. 113. 23. 138 kinetic energy. 77. 57 light-like interval. 58 in a medium. 36 help vector. 55 induction ﬁeld. 105 Hodge star operator. 46 inertial reference frame. 126 interaction Lagrange density. 106 group theory. 80 Hamilton equations. 16 homogeneous wave equation. 72. 172 Green’s function. 105 Hertz’ method. 173 Larmor formula for radiated power.

6 magnetic monopoles. 93 magnetic ﬁeld intensity. 142 Poisson equation. 4 massive photons. 124 Poissons’ equation. 14 mechanical Lagrange density. 60 Lorentz equations. 14. 90. 110 magnetic ﬁeld. 54. 58 matrix form. 167 metric tensor. 164 M1 radiation. 38 Lorentz gauge condition. 34 magnetostatics.182 M ATHEMATICAL M ETHODS line element. 14 Maxwell-Lorentz equations. 92 Maxwell’s microscopic equations. 15 magnetic dipole moment. 140 magnetic susceptibility. 124 lowering of index. 167 linear mass density. 131 physical measurable. 95 monochromatic. 93. 6 magnetic ﬁeld energy. 160. 62 Lorentz inhomogeneous wave equations.se/CED/Book Draft version released 13th November 2000 at 22:01. 110 Mach cone. 51 Newton-Lorentz force equation. 55 non-linear effects. 58 mixed four-tensor ﬁeld.uu. 27 outer product. 53. 28 longitudinal component. 168 null vector.plasma. 15 magnetic current density. 101. 163 Minkowski equation. 72 non-Euclidean space. 141 plasma frequency. 10 one-dimensional wave equation. 170 polarisation charges. 91 magnetostatic vector potential. 33 polar vector. 90 magnetisation currents. 160. 13. 89 Downloaded from http://www. 163 mixing angle. 140 photon. 36. 160 Lorentz transformation. 10 norm. 139 magnetic charge density. 49 Newton’s ﬁrst law. 57 observation point. 30 plane polarised wave. 138 phase velocity. 169 Parseval’s identity. 3 Ohm’s law. 124 Lorentz gauge. 6 magnetic induction. 28 plasma. 36 Lorentz space. 144 macroscopic Maxwell equations. 72 Minkowski space. 162 Maxwell stress tensor. 43 multipole expansion. 15. 90 magnetising ﬁeld. 66. 36 Lorentz force. 16 momentum theorem in Maxwell’s theory. 94 Maxwell’s macroscopic equations. 131. 55. 85 mathematical group. 15 magnetic permeability. 91 magnetic ﬂux. 77 linearly polarised wave. 107 near zone. . 55. 104. 27 Lorentz boost parameter. 91 magnetisation. 110 magnetic dipole radiation. 11 magnetic ﬂux density. 81 metric.

167 quantum mechanical nonlinearity. 121 skew-symmetric. 41 retarded relative distance. 54 simultaneous coordinate. 163 tensor product. 99 super-potential. 140 temporal dispersive media. 118 radiation ﬁelds. 55 special theory of relativity. 159 pseudoscalars. 79 time component. 38 radiation resistance. 93 Poynting vector. 3 space components. 97 radiation gauge. 169 three-dimensional functional derivative. 94 relative magnetic permeability.183 polarisation currents. 90 polarisation potential. 159. 164 rank. 159 tensor contraction. 159 scalar. 162 tensor notation.plasma. 164 tensor ﬁeld. 56 pseudoscalar. 170 pseudotensor. 51 spherical Bessel function of the ﬁrst kind. 27. 161 scalar product. 38 tensor. 144 signature. 55 potential energy. 167 shock front. 170 quadratic differential form. 173 scalar ﬁeld. 85 propagator. 137 synchrotron radiation lobe width. 57.uu. 106 power ﬂux. 103 radius four-vector. 51 relaxation time. 66 skin depth. 140 Relativity principle.se/CED/Book . 56 Riemannian space. 25 rest mass density. 30 source point. 106 spherical Hankel function of the ﬁrst kind. 93 Poynting’s theorem. 81 retarded Coulomb ﬁeld. 105 synchrotron radiation. 159 pseudotensors. 135 telegrapher’s equation. 55 Draft version released 13th November 2000 at 22:01. 176 potential theory. Downloaded from http://www. 46. 105 polarisation vector. 3 radiation ﬁeld. 106 spherical waves. 94 relative permeability. 53 radius vector. 160 row vector. 162 rapidity. 54. 55 space-like interval. 41 Riemannian metric. 49. 140 relative electric permittivity. 56 space-time. 60 refractive index. 10 temporal gauge. 140 relative permittivity. 134. 39 proper time. 170 pseudovector. 159. 77. 65. 92 Proca Lagrangian. 113 retarded time. 58 positive deﬁnite norm. 56. 159 raising of index. 105 positive deﬁnite. 49 retarded potentials.

140 world line. . 28.plasma. 38 vacuum permeability. 26 time-like interval. 144 vector. 84 Downloaded from http://www. 142.uu.se/CED/Book Draft version released 13th November 2000 at 22:01. 26 time-independent telegrapher’s equation. 87 transverse components.184 M ATHEMATICAL M ETHODS time-harmonic wave. 58 Young’s modulus. 170 velocity ﬁeld. 26 ˇ Vavilov-Cerenkov radiation. 77 Yukawa meson ﬁeld. 3 vacuum wave number. 114. 159 vector product. 25 time-independent diffusion equation. 56 total charge. 29 time-independent wave equation. 5 vacuum permittivity. 118 wave vector. 2 vacuum polarisation effects. 118 virtual simultaneous coordinate. 27 transverse gauge.

- Austria
- IEEC2016-23(Reviewer Comments) (1)
- Proposal Format
- Advt_PMA138.pdf
- Types of Research
- PPolice_SB_Ad (1)
- Tariq2015 - Copy (3)
- Cold War Notes 1
- ;lm;m';;''Higher Education
- Antonyms
- References
- PP Muzaffargarh
- Dubai Telecom Companies
- Basic Computer
- Packages E Form
- List of CPD Courses 2013
- EntryTest List ComED FullTime 2013
- Ecology
- Chemistry
- PP Muzaffargarh
- adreq
- Quiz
- ScholarshipApplication_3
- SAPPK Multiple Jobs Advertisement Regioanl Office Gawader
- 13 19 Visa Assistant Isb

Sign up to vote on this title

UsefulNot useful- 1999 2000.Bombardier.traxter.service.manual
- 8314125 Electromagnetic Field Theory Bo Thide
- Electromagnetic Field Theory - Bo Thidé.pdf
- EMFT_Book
- Electro Magnetic Field Theory
- Electromagnetic Field Theory
- Thide Bo - Electromagnetic Field Theory
- EMFT_Book
- Physics
- Electrodynamics
- Electrodynamics of Superconductors Hirsch
- First Order Linear Equations
- Difference Equations
- Coulomb's Law Electro Statics
- CH 3 Statics Problems 2-D and 3-D
- Solving Equations
- Mechanics Lagrangian and Hamilitonian
- Fisico
- Statics Lecture 05
- preface.pdf
- 2009_farina_1
- Equilibrium of Rigid Body(2)
- Differentiation Problems
- Nonlinear_Equations
- Solution Of Nonlinear Equation
- Ordinary Differential Equations Notes
- page32
- 2-Chapters Chapter One Introduction 2
- nonlinear equation
- constant-coeff-diff-eq rolando.pdf
- Electromagnetic Field Theory - Bo Thide.pdf