Relativity Made Relatively Easy: Andrew Steane

Relativity made relatively easy
Andrew Steane
October 6, 2011
2
Contents
1 Introduction 15
2 Basic ideas 17
2.1 Classical physics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.2 Special relativity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.2.1 The postulates of Special Relativity . . . . . . . . . . . . . . . . . . . . . 20
2.2.2 Central ideas about spacetime . . . . . . . . . . . . . . . . . . . . . . . . 21
2.3 Matrix methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3 The Lorentz transformation 29
3.1 Introducing the Lorentz transformation . . . . . . . . . . . . . . . . . . . . . . . 29
3.2 Derivation of Lorentz transformation . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.3 Velocities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.4 Lorentz invariance and four-vectors . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.4.1 Rapidity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.4.2 Lorentz invariant quantities . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3.5 Basic 4-vectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3
4 CONTENTS
3.5.1 Proper time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.5.2 Velocity, acceleration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.5.3 Momentum, energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.5.4 The direction change of a 4-vector under a boost . . . . . . . . . . . . . . 47
3.5.5 Force . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
3.5.6 Wave vector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.6 The joy of invariants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.7 Moving light sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3.7.1 The Doppler effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3.7.2 Aberration and the headlight effect . . . . . . . . . . . . . . . . . . . . . . 55
3.7.3 Stellar aberration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
3.7.4 Visual appearances* . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
3.8 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
4 Dynamics 63
4.1 Force . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
4.1.1 Transformation of force . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
4.2 Motion under a pure force . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
4.2.1 Constant force (the ‘relativistic rocket’) . . . . . . . . . . . . . . . . . . . 69
4.2.2 4-vector treatment of hyperbolic motion . . . . . . . . . . . . . . . . . . . 74
4.2.3 Circular motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
4.2.4 Motion under a central force . . . . . . . . . . . . . . . . . . . . . . . . . 77
4.2.5 (An)harmonic motion* . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
4.3 The conservation of energy-momentum . . . . . . . . . . . . . . . . . . . . . . . . 82

CONTENTS 5
4.3.1 Elastic collision, following Lewis and Tolman . . . . . . . . . . . . . . . . 83
4.3.2 Energy-momentum conservation using 4-vectors . . . . . . . . . . . . . . . 83
4.3.3 Mass-energy equivalence . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
4.4 Collisions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
4.4.1 Elastic collisions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
4.5 Composite systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
4.6 Energy flux, momentum density, and force . . . . . . . . . . . . . . . . . . . . . . 98
4.7 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
5 Further kinematics 101
5.1 The Principle of Most Proper Time . . . . . . . . . . . . . . . . . . . . . . . . . . 101
5.2 4-dimensional gradient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
5.3 Current density, continuity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
5.4 Wave motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
5.4.1 Wave equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
5.4.2 Particles and waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
5.5 Acceleration and rigidity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
5.5.1 The great train disaster1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
5.5.2 Lorentz contraction and internal stress . . . . . . . . . . . . . . . . . . . . 116
5.6 General Lorentz boost . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
5.7 Lorentz boosts and rotations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
5.7.1 Two boosts at right angles . . . . . . . . . . . . . . . . . . . . . . . . . . 121
5.7.2 The Thomas precession . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123

1 This is loosely based on the discussion by Fayngold.
6 CONTENTS
5.7.3 Analysis of circular motion . . . . . . . . . . . . . . . . . . . . . . . . . . 126
5.8 The Lorentz group* . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
5.8.1 Further group terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
5.9 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
6 Relativity and electromagnetism 135
6.1 Maxwell’s equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
6.1.1 Moving capacitor plates . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
6.2 The fields due to a moving point charge . . . . . . . . . . . . . . . . . . . . . . . 143
6.3 Covariance of Maxwell’s equations . . . . . . . . . . . . . . . . . . . . . . . . . . 147
6.3.1 Transformation of the fields: 4-vector method . . . . . . . . . . . . . . . . 151
6.4 Electromagnetic waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
6.5 Solution of Maxwell’s equations for a given charge distribution . . . . . . . . . . 154
6.5.1 The four-vector potential of a uniformly moving point charge . . . . . . . 154
6.5.2 The general solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
6.5.3 The Liénard-Wiechart potentials . . . . . . . . . . . . . . . . . . . . . . . 161
6.5.4 The field of an arbitrarily moving charge . . . . . . . . . . . . . . . . . . . 167
6.5.5 Two example fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
6.6 Radiated power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
6.6.1 Linear and circular motion . . . . . . . . . . . . . . . . . . . . . . . . . . 180
6.6.2 Angular distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
6.7 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181

CONTENTS 7
II An introduction to General Relativity 187
7 An introduction to General Relativity 189
7.1 The Principle of Equivalence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
7.1.1 Free fall or free float? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
7.1.2 Weak Principle of Equivalence . . . . . . . . . . . . . . . . . . . . . . . . 194
7.1.3 The Eötvös-Pekár-Fekete experiment . . . . . . . . . . . . . . . . . . . . . 195
7.1.4 The Strong Equivalence Principle . . . . . . . . . . . . . . . . . . . . . . . 198
7.1.5 Falling light and gravitational time dilation . . . . . . . . . . . . . . . . . 200
7.2 The uniformly accelerating reference frame . . . . . . . . . . . . . . . . . . . . . 206
7.2.1 Accelerated rigid motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
7.2.2 Rigid constantly accelerating frame . . . . . . . . . . . . . . . . . . . . . . 209
7.3 Newtonian gravity from Principle of Most Proper Time . . . . . . . . . . . . . . 219
7.4 Gravitational red shift and energy conservation . . . . . . . . . . . . . . . . . . . 221
7.5 Warped spacetime . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
7.5.1 Two-dimensional spatial surfaces . . . . . . . . . . . . . . . . . . . . . . . 225
7.5.2 Three spatial dimensions . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
7.5.3 Time and space together . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
7.5.4 Gravity and curved spacetime . . . . . . . . . . . . . . . . . . . . . . . . . 233
7.6 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
8 Physics from the metric 239
8.1 Example exact solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
8.1.1 The acceleration due to gravity . . . . . . . . . . . . . . . . . . . . . . . . 240

8 CONTENTS
8.2 Schwarzschild metric: basic properties . . . . . . . . . . . . . . . . . . . . . . . . 243
8.3 Geometry of Schwarzschild solution . . . . . . . . . . . . . . . . . . . . . . . . . . 245
8.3.1 Shapiro time delay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
8.3.2 Circular orbits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
8.4 Gravitational lensing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
8.5 Black holes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
8.5.1 Horizon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
8.5.2 Energy considerations near a horizon . . . . . . . . . . . . . . . . . . . . . 257
8.6 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
9 Tensors and index notation 261
9.1 Introducing tensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
9.1.1 Outer product . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
9.1.2 The vector product . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
9.1.3 Differentiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
9.2 Contravariant and covariant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
9.3 Index notation and tensor algebra . . . . . . . . . . . . . . . . . . . . . . . . . . 270
9.3.1 Rules for tensor algebra . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
9.3.2 Index notation for derivatives . . . . . . . . . . . . . . . . . . . . . . . . . 279
9.4 Some basic results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280
9.4.1 Antisymmetric tensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284
10 Angular momentum 287
10.1 Conservation of angular momentum . . . . . . . . . . . . . . . . . . . . . . . . . 287

CONTENTS 9
10.2 Spin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
10.2.1 Introducing spin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290
10.2.2 Pauli-Lubanski vector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
10.2.3 Thomas precession revisited . . . . . . . . . . . . . . . . . . . . . . . . . . 295
11 Lagrangian mechanics 299
11.1 Classical Lagrangian mechanics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
11.2 Relativistic motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
11.2.1 From classical Euler-Lagrange . . . . . . . . . . . . . . . . . . . . . . . . . 302
11.2.2 Manifestly covariant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
12 Further electromagnetism 309
12.1 Fundamental equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
12.1.1 The dual field and invariants . . . . . . . . . . . . . . . . . . . . . . . . . 315
12.1.2 Motion of particles in a static uniform field . . . . . . . . . . . . . . . . . 315
12.1.3 Precession of the spin of a charged particle . . . . . . . . . . . . . . . . . 315
12.2 Electromagnetic energy and momentum . . . . . . . . . . . . . . . . . . . . . . . 315
12.2.1 Examples of energy density and energy flow . . . . . . . . . . . . . . . . . 319
12.2.2 Field momentum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
12.2.3 Stress-energy tensor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328
12.2.4 Resolution of the “4/3 problem” and the origin of mass . . . . . . . . . . 338
12.3 Self-force and radiation reaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
13 Fluids and stressed bodies 341
13.1 Stress-energy tensor for an arbitrary system . . . . . . . . . . . . . . . . . . . . . 341

10 CONTENTS
13.1.1 Interpreting the terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 342
13.2 Conservation of energy and momentum again . . . . . . . . . . . . . . . . . . . . 345
13.3 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346
14 General relativity 347
14.1 Gravitational field theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
14.2 A simple example: the uniform static field . . . . . . . . . . . . . . . . . . . . . . 351
14.3 Some more geometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
14.4 Equation of motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358
14.5 Schwarzschild metric . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
14.5.1 Derivation of the metric . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
14.5.2 Orbits and the perihelion of Mercury . . . . . . . . . . . . . . . . . . . . . 363
14.5.3 Photon orbits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367
14.6 Interior Schwarzschild solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368
14.7 Black holes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371
14.7.1 Kruskal spacetime . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371
14.7.2 Reissner-Nordström and Kerr metrics . . . . . . . . . . . . . . . . . . . . 374
14.7.3 Black hole thermodynamics . . . . . . . . . . . . . . . . . . . . . . . . . . 376
14.8 Rotation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378
14.8.1 Canonical form of the metric . . . . . . . . . . . . . . . . . . . . . . . . . 383
14.8.2 de Sitter and Lense-Thirring precession . . . . . . . . . . . . . . . . . . . 384
14.8.3 Frame dragging? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388
14.9 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389

CONTENTS 11
15 General covariance 391
15.0.1 Geodesic equation and Christoffel symbols . . . . . . . . . . . . . . . . . . 391
15.0.2 Local flatness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395
15.0.3 Covariant differentiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396
15.0.4 Parallel transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
15.1 Riemann curvature and Field Equations . . . . . . . . . . . . . . . . . . . . . . . 400
15.2 General covariance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 404
15.2.1 Maxwell’s equations in curved spacetime . . . . . . . . . . . . . . . . . . . 405
15.2.2 Energy and momentum . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405
15.2.3 Derivation of the interior Schwarzschild solution . . . . . . . . . . . . . . 406
15.2.4 On the connection between gravity and general covariance . . . . . . . . . 406
15.3 Linearized General Relativity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409
15.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409
16 Cosmology 411
16.0.1 Topology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411
16.1 Cosmology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412
16.1.1 FRW metric . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412
16.1.2 Friedmann equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412
17 Spinors 415
17.1 Introducing spinors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
17.2 The rotation group and SU(2)* . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421
17.2.1 Rotation of rank 1 spinors . . . . . . . . . . . . . . . . . . . . . . . . . . . 425

12 CONTENTS
17.3 Lorentz transformation of spinors . . . . . . . . . . . . . . . . . . . . . . . . . . . 427
17.3.1 Obtaining 4-vectors from spinors . . . . . . . . . . . . . . . . . . . . . . . 430
17.4 Chirality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432
17.4.1 Chirality, spin and parity violation . . . . . . . . . . . . . . . . . . . . . . 434
17.4.2 Index notation* . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440
17.4.3 Invariants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 442
17.5 Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 444
17.6 Dirac spinor and particle physics . . . . . . . . . . . . . . . . . . . . . . . . . . . 445
17.6.1 Moving particles and classical Dirac equation . . . . . . . . . . . . . . . . 448
17.7 Spin matrix algebra (Lie algebra)* . . . . . . . . . . . . . . . . . . . . . . . . . . 451
17.7.1 Dirac spinors from group theory* . . . . . . . . . . . . . . . . . . . . . . . 453
18 Classical field theory 455
18.1 Wave equation and Klein-Gordan equation . . . . . . . . . . . . . . . . . . . . . 456
18.1.1 The wave equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 456
18.1.2 Klein-Gordan equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457
18.2 The Dirac equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 459
18.2.1 Massive Dirac equation in 4 dimensions . . . . . . . . . . . . . . . . . . . 463
18.3 Lagrangian mechanics for fields* . . . . . . . . . . . . . . . . . . . . . . . . . . . 465
18.4 Conserved quantities and Noether’s theorem . . . . . . . . . . . . . . . . . . . . . 465
18.5 Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466
19 Relativistic quantum mechanics 467
19.1 A false start . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467

Copyright A. Steane, Oxford University 2010, 2011; not for redistribution. 13
19.2 An outline of quantum field theory . . . . . . . . . . . . . . . . . . . . . . . . . . 468
19.2.1 Basic concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469
19.2.2 Free field theories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 470
19.2.3 Particles, spin and exchange symmetry . . . . . . . . . . . . . . . . . . . . 473
19.2.4 Vacuum energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475
19.2.5 Antiparticles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 477
19.2.6 Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 481
19.3 Single-particle Dirac theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 481
19.3.1 Spin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483
19.3.2 Particle and antiparticle solutions . . . . . . . . . . . . . . . . . . . . . . 484
19.3.3 Low velocity limit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 486
19.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 487
20 The dance of the universe 489
21 Appendix 1. Some basic arguments 491
21.1 Simultaneity and radar coordinates . . . . . . . . . . . . . . . . . . . . . . . . . . 491
21.2 Proper time and time dilation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 493
21.3 Lorentz contraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 494
21.4 Doppler effect, addition of velocities . . . . . . . . . . . . . . . . . . . . . . . . . 496
22 Appendix 2. The field of an arbitrarily moving charge 499
23 Appendix 3. Fundamental constants 503

14 Copyright A. Steane, Oxford University 2010, 2011; not for redistribution.
Chapter 1
Introduction
This book presents an extensive study of Special Relativity, aimed at an undergraduate level.
It is not intended to be the first introduction to the subject for most students, although for a
bright student it could function as that. Therefore basic ideas such as time dilation and space
contraction are recalled but not discussed at length. However, I think it is also beneficial to have
a thorough discussion of those concepts at as simple a level as possible, so I have provided one
in another book called The Wonderful World of Relativity. The present book is self-contained
and does not require knowledge of the first one, but a more basic text such as The Wonderful
World or something similar is recommended as a preparation for this book.
The later chapters of the book go further than most undergraduate courses will want to go; they
are intended to fill the gap between undergraduate and graduate study, and to offer general
reading for the professional physicist.
Acknowledgements
I have, of course, learned relativity mostly from other people. All writers in this area have
learned from the pioneers of the subject, especially Einstein, Lorentz, Maxwell, Minkowski
and Poincaré. I am indebted also to tutors such N. Stone and W. S. C. Williams at Oxford
University, and to authors who have preceded me, especially texts by (in alphabetical order)
A. Einstein, R. Feynman, A. P. French, J.D. Jackson, H. Muirhead, W. Rindler, F. Rohrlich,
R. Shankar, E. F. Taylor and J. A. Wheeler, and W.S.C. Williams.
Einstein emphasized the need to think of a reference frame in physical not abstract terms, as a
physical entity made of rods and clocks. As a student I resisted this idea, feeling that a more
abstract idea, liberated from mere matter, must be superior. I was wrong. The whole point of
Relativity is to see that abstract notions of space and time are superfluous and misleading.
15
Feynman offered very useful guidance on how to approach things simply while retaining rigour.
Readers familiar with his, Leighton and Sands’ ‘Lectures on Physics’ will recognize that my
treatment of Poynting’s argument follows his quite closely, because I felt there was little room
for improvement. I am happy to acknowledge this.
I learned a significant number of detailed points from Rindler’s work; my contribution has been
to clarify where possible. Appendix 2, for example, re-presents an argument I found in his book,
with more comfort and explanation for the reader. The material in chapter 12 on radiation
reaction and self-force is heavily indebted to Rohrlich. The presentation of General Relativity
in chapter 14 owes much to J. Binney of Oxford University, as well as to W. Rindler once again.
Chapter 2
Basic ideas
The primary purpose of this chapter is to offer a way in for readers completely unfamiliar
with Special Relativity, and to recall the main ideas for readers who have some preliminary
knowledge of the subject. For the former category, Appendix 1 contains some of the basic
arguments that will not be repeated in the main text (and that can be found in introductory
texts such as The Wonderful World of Relativity). The right moment to turn to that appendix,
if you need to, is after completing section 2.2.2 of this chapter.
In order to discuss space and time without being vague, it is extremely helpful to introduce
the notion of a reference body. This is usually called an “inertial frame of reference”, but that
phrase is in some respects unfortunate. The phrase “frame of reference” is used in an abstract
way in everyday language, but in physics we mean something more concrete: a large rigid
physical object which could, in principle, exist in the vicinity of any system whose evolution
we wish to discuss. Such a “reference body” clarifies what we mean when we talk of distance
and time. By ‘distance’ we mean the number of particles or rods of the reference body between
two places. The reference body keeps track of time as well, since the particles making it can be
imagined to be tiny regular clocks (think of an atom with an internal vibration, for example).
By ‘time’ at any given place we mean the number of repetitions of some such regularly repeating
process (‘clock’) at that place.
“Frame of reference” and “reference body” are synonyms in physics. Most people like to think
of a frame of reference as having the form of a scaffolding of ideally thin and rigid rods, with
clocks attached. I sometimes like to think of it as a large brick (but one with the unusual
property that others things can move through it unimpeded). It is a mistake to try to be too
abstract here. Although the scaffolding or rigid body is not necessarily present, our reasoning
about distance and time must be consistent with the fact that such a body might in principle
be present in any region of spacetime.
An observer is a reasoning being who could in principle be situated at rest in some given frame
17
y y
v
z z
S S
x x
Figure 2.1: Two reference frames (=reference bodies) in standard configuration. S0 moves in the
x direction relative to S, with its axes aligned with those of S. The picture shows the situation
at the moment (defined in S) when the axes of S0 have just swept past those of S. The whole
reference frame of S0 is in motion together at the same velocity v relative to S. Equally, the
frame of S is in motion at velocity −v relative to S0 .
of reference. We use the word ‘observe’ to mean not what the observer directly sees, but what
he or she can deduce to be the case at each time and place in his/her reference frame. For
example, suppose two explosions occur, and an observer is located closer to one than to the other
in his own reference frame. If such an observer receives light flashes from the two explosions
simultaneously, then he ‘observes’ (i.e. deduces) that the explosions were not simultaneous in
his reference frame.
2.1 Classical physics
Let us briefly survey the connection between inertial reference frames according to classical
physics, as developed by Galileo and Newton and others.
A crucial idea, first presented at length by Galileo, is the idea that the behaviour of physical
systems is the same in any given inertial reference frame, irrespective of whether that frame
may be in uniform motion with respect to others. For example, it is possible to play table
tennis in a carriage of a moving railway train without noticing the motion of the train (as long
as the rails are smooth and the train has constant velocity). There is no need to adjust one’s
calculations of the trajectory of the ball or the choice of force to apply using the bat: all the
behaviour is the same as it would be in a motionless train. This idea, which we shall state more
carefully in a moment, is called the Principle of Relativity; it is obeyed by both classical and
relativistic physics.
When we analyze the motions of bodies, it is useful to introduce a coordinate system (in both
space and time), which means we measure distances and times relative to a reference body (=
inertial frame of reference). An event is a point in space and time. It is useful to know, for
any given event, how the coordinates of the event relative to one reference body relate to the
coordinates of the same event relative to another reference body. If reference frames F and
F0 have all their axes aligned, but frame F0 moves along the positive x direction relative to F
at speed v, then we say the reference frames are in standard configuration (figure 2.1). The
coordinates of any given event, as determined in two reference frames in standard configuration,
are related, according to classical physics, by
t0 = t,
x0 = x − vt,
y0 = y,
z0 = z. (2.1)
This set of equations is called the Galilean transformation. It can also be written in matrix
notation as
    
t0 1 0 0 0 t
 x0   −v 1 0 0 x 
 0  =   , (2.2)
y   0 0 1 0  y 
z0 0 0 0 1 z
or
µ ¶ µ ¶
t0 t
= G , (2.3)
r0 r
where
 
1 0 0 0
 −v 1 0 0
G≡
 0
. (2.4)
0 1 0
0 0 0 1
The inverse Galilean transformation is
    0 
t 1 0 0 0 t
x v 1 0 0  x0 
  =   . (2.5)
y 0 0 1 0   y0 
z 0 0 0 1 z0
which can also be written
µ ¶ µ ¶
t −1 t0
= G . (2.6)
r r0
The reader is invited to verify this, i.e. check that the matrix given in (2.5) is indeed the inverse
of G.
Matrix notation makes it easy to check things like the effect of transforming from one reference
frame to another and then to a third. For example, the net effect of transforming to another
frame and then back to the first is given by G −1 G which is, of course, the identity matrix.
2.2 Special relativity
2.2.1 The postulates of Special Relativity
Turning now to Special Relativity, we shall find that the Principle of Relativity is still obeyed,
but the Galilean transformation fails.
The Main Postulates of Special Relativity are
Postulate 1, “Principle of Relativity”: The motions of bodies included in a given

space are the same among themselves, whether that space is at rest or moves uni-
formly forward in a straight line.
Postulate 2, “Light speed postulate”:
Version A:There is a finite maximum speed for signals.
Version B:There is an inertial reference frame in which the speed of light in vacuum
is independent of the motion of the source.
The Principle of Relativity (Postulate 1) is obeyed by classical physics; the Light Speed Postu-
late is not. The Principle of Relativity can also be stated
The laws of physics take the same mathematical form in all inertial frames of refer-
ence.
In Postulate 2, either version A or version B is sufficient on its own to allow Special Relativity
to be developed. Version A does not mention light; this makes it clear that Special Relativ-
ity underlies all theories in physics, not just electromagnetism. For this reason version A is
preferred. However we will preserve the practice of calling this postulate the “Light speed pos-
tulate” because in vacuum, far from material objects, light waves move at the maximum speed
for signals. With this piece of information about light, one can use either version to derive the
other.
Einstein used version B of the Light Speed Postulate. It is often stated as “the speed of light
is independent of the motion of the source.” In this statement the fact that motion can only
ever be relative motion is taken for granted, and it is a statement about what is observed in
any reference frame. In our version B we chose to make a slightly more restricted statement
(picking just one reference frame), merely because it is interesting to hone ones assumptions
down to the smallest possible set. By combining this with Postulate 1 it immediately follows
that all reference frames will have this property.
In order to make clear what is assumed and what is derived, it is useful to add two further
postulates to the list:
Postulate 0, “Euclidean geometry”: The rules of Euclidean geometry apply to all

spatial measurements within any given inertial reference frame.
Postulate 3, “Conservation of momentum”: Internal interactions among the parts
of an isolated system cannot change the system’s total momentum, where momentum
is a vector function of rest mass and velocity.
Postulate 0 (Euclidean geometry) is obeyed by Special Relativity but not by General Relativity.
Postulate 3 (conservation of momentum) allows the central elements of dynamics to be deduced,
including the famous formula “E = mc2 ” (that formula cannot be derived from the Main
Postulates alone).
2.2.2 Central ideas about spacetime
Recall that a ‘point in spacetime’ is called an event. This is something happening at an instant
of time at a point in space, with infinitesimal time duration and spatial extension. For an
example, tap the tip of a pencil once on a table top, or click your fingers.
A particle is a physical object of infinitesimal spatial extent, which can exist for some some
extended period of time. The line of events which gives the location of the particle as a
function of time is called its worldline, see figure 2.2.
If two events have coordinates (t1 , x1 , y1 , z1 ) and (t2 , x2 , y2 , z2 ) in some reference frame, then
the quantity
s2 = −c2 (t2 − t1 )2 + (x2 − x1 )2 + (y2 − y1 )2 + (z2 − z1 )2 (2.7)

Figure 2.2: A spacetime diagram showing a worldline and a light cone (past and future
branches). The cross (×) marks an example event. The apex of the cone is another event.
is called the squared spacetime interval between them. Note the crucial minus sign in front of
the first term. We emphasize it by writing (2.7) as
s2 = −c2 ∆t2 + ∆x2 + ∆y 2 + ∆z 2 . (2.8)
If s2 < 0 then the time between the events is sufficiently long that a particle or other signal
(moving at speeds less than c) could move from one event to the other. Such a pair of events
is said to be separated by a timelike interval. If s2 > 0 then the time between the events is too
short for any physical influence to move between them. This is called a spacelike interval. If
s2 = 0 then we have a null interval, it means that a light pulse or other light-speed signal could
move directly from one event to the other.
Although the parts ti , xi , yi , zi needed to calculate an interval will vary from one reference frame
to another, we will find in chapter 3 that the net result, s2 , is independent of reference
frame: all reference frames agree on the value of this quantity. This is similar to the fact
that the length of a vector is unchanged by rotations of the vector. A quantity whose value
is the same in all reference frames is called a Lorentz invariant (or Lorentz scalar). Lorenz
invariants play a central role Special Relativity.
The set of events with a null spacetime interval from any given event lie on a cone called the
light cone. The part (or ‘branch’) of this cone extending into the past is made of the worldlines
of photons that form a spherical pulse of light collapsing onto the event, the part extending into
the future is made of the worldlines of photons that form a spherical pulse of light emitted by
the event. The cone is an abstraction: the incoming and outgoing light pulses don’t have to be
there. The past part of the light cone surface of any event A is called the past light cone of A,
the future part of the surface is called the future light cone of A. The whole of the future cone
(i.e. the body of the cone as well as the surface) is called the absolute future of A, it consists
of all events which could possibly be influenced by A (in view of the Light Speed Postulate).
The whole of the past cone is called the absolute past of A, it consists of all events which could
possibly influence A. The rest of spacetime, outside either branch of the light cone, can neither
Einstein’s train.
Suppose a fast-moving train is moving past a platform, and suppose that at the moment
when the front of the train reaches the far end of the platform (where it is about to
leave the station), a firecracker explodes there, leaving scorch marks on the train and
platform. Similarly, when the back of the train arrives at the start of the platform (at
the other end of the station), a firecracker explodes there, leaving scorch marks on the
train and platform. We consider a train whose length is such that, in the reference
frame of the platform, the lengths of train and platform are the same. In this case the
two explosions are simultaneous in the reference frame of the platform. The flashes of
light emitted by the explosions therefore arrive at the centre of the platform at the same
time. However, the flashes of light do not arrive at the centre of the train together. An
observer standing on the platform finds that when the flashes arrive at him, the train
has moved on, so that the flash from the front of the train has already moved past the
centre of the train, and the flash from the back has not yet arrived at the centre of the
train. It follows that the flash from the front of the train arrives at a passenger seated
in the middle of the train before the flash from the back does.
Observers in all reference frames must agree with this fact, i.e. one flash arrives at the
passenger before the other, because the flashes could be arranged to trigger events at
the passenger. Suppose, for example, that he carries a device which will smash a glass
if the flashes arrive simultaneously (or if the rear flash arrives first). If the observer at
rest on the platform finds that the glass is not smashed, then it is not, irrespective of
which reference frame we adopt for the purpose of calculating time and space intervals.
It follows that, in the reference frame of the train (i.e. that in which the train is at
rest), the front flash arrives at the passenger before the rear flash does. Also, the two
scorch marks at the front and back of the train are equidistant from the passenger in
the middle of the train, and the light pulses have the same speed (by postulate 2). It
follows that the firecracker explosion at the front of the train must have happened first,
before the one at the back, and not simultaneous with it, in the reference frame of the
train.
We infer that simultaneity is a relative concept: it depends on reference frame. It also
follows that in the reference frame of the train, the train and the platform are not of
the same length: the train must be longer than the platform.
influence nor be influenced by A. It consists of all events with a spacelike separation from A.
The single most basic insight into spacetime that Einstein’s theory introduces is the relativity
of simultaneity: two events that are simultaneous in one reference frame are not necessarily
simultaneous in another. In particular, if two events happen simultaneously at different spatial
locations in reference frame F, then they will not be simultaneous in any reference frame moving
relative to F with a non-zero component of velocity along the line between the events. An
example is furnished by “Einstein’s train,” see box.
By careful argument from the postulates one can connect timing and spatial measurements in
one inertial reference frame to those in any other inertial reference frame in a precise, quan-
titative way. In the next chapter we shall introduce the Lorentz transformation to do this in
general. Arguments for some simple cases were presented in The Wonderful World; these are
summarised in Appendix 1.
2.3 Matrix methods
By writing down the Galilean transformation using a matrix, we already assumed that the reader
has some idea what a matrix is and how it is used. However, in case matrices are unfamiliar,
we will here summarize the matrix mathematics we shall need. This will not substitute for a
more lengthy course of mathematical training, but it may be a useful reminder.
A matrix is a table of numbers. We will only need to deal with real matrices (until chapter 17)
so the numbers are real numbers. In an “n × m” matrix the table has n rows and m columns.
Here is a 2 × 3 matrix, for example:
µ ¶
1.2 −3.6 8
. (2.9)
2 4.5 2
If either n or m is 1 then we have a vector; if both are 1 than we have a scalar.
A vector of 1 row is called a row vector; a vector of 1 column is called a column vector:
 
1
row vector: (1, −3, 2), column vector:  −3  .
2
The sum of two matrices, written as A + B, is only defined (so it is only a legal operation) when
A and B have the same shape, that is, the two matrices have the same number of rows n, and
they also have the same number of columns m (but n does not have to equal m). A + B is then
defined to mean the matrix formed from the sums of the corresponding components of A and
B. To be precise, if Mij refers to the element of matrix M in the i’th row and j’th column,
then the matrix sum is defined by
M =A+B ⇔ Mij = Aij + Bij .
This rule applies to vectors and scalars too, since they are special cases of matrices, and it
agrees with the familiar rule for summing vectors: add the components.
The product of two matrices, written as AB, is only defined (so it is only a legal operation)
when A and B have appropriate shapes: the number of columns in the first matrix has to equal
the number of rows in the second matrix. For example, a 2 × 3 matrix can multiply a 3 × 5
matrix, but it cannot multiply a 2 × 3 matrix. The product is defined by the mathematical rule
X
M = AB ⇔ Mij = Aik Bkj . (2.10)
k
It is important to note that this rule is not commutative: AB is not necessarily the same
as BA. The rule is important in order to have a precise definition, but the use of subscripts
and the sum can leave the operation obscure until one tries a few examples. It amounts to the
following. You have to work your way through the elements of M one by one. To obtain the
element of M on the i’th row and j’th column, take the i’th row of A and the j’th column of
B. Regard these as two vectors and evaluate their scalar product: that is, ‘dive’ the row of A
onto the column of B, multiply corresponding elements, and then sum. The result is the value
of Mij .
The only way to become familiar with matrix multiplication is by practice. By applying the
rule, you will find that if a k × n matrix multiplies a n × m matrix then the result is a k × m
matrix. This is a very useful check to keep track of what you are doing.
The whole point of matrix notation is that much of the time we can avoid actually carrying
out the element-by-element multiplications and additions. Instead we manipulate the matrix
symbols. For example, if A + B = C and A − B = D then we can deduce that C + D = 2A
without needing to carry out any element-by-element analysis. The following mathematical
results apply to matrices (as the reader can show by applying the rules developed above):
A+B = B+A
A + (B + C) = (A + B) + C
(AB)C = A(BC)
A(B + C) = AB + AC.
We shall mostly be concerned with square matrices and with vectors. The square matrices will
be mostly 4 × 4, so they can be added and multiplied to give other 4 × 4 matrices. A square
matrix can multiply a column vector, giving a result that is a column vector (since a 4 × 4
matrix multiplying a 4 × 1 matrix gives a 4 × 1 matrix). For example
    
1.2 −3.6 8 2 1 20
 2 4.5 2 0.5     
   −3  =  −9.5  .
 −1 5 1 −0.5   2   −12 
−2 3.2 3 −5 −4 14.4
A square matrix can be multiplied from the left by a row vector, giving a result that is a row
vector (since a 1 × 4 matrix multiplying a 4 × 4 matrix gives a 1 × 4 matrix).
Matrix inverse
Many, but not all, square matrices have an inverse. This is written M −1 and is defined by
M M −1 = M −1 M = I (2.11)
where I is the identity matrix, consisting of 1’s down the diagonal and zeros everywhere else.
For example, in the 4 × 4 case it is
 
1 0 0 0
0 1 0 0
I=  0 0 1 0 .

0 0 0 1
The identity matrix has no effect when it multiplies another matrix: IM = M I = M for all
M . Inverses of non-square matrices can also be defined, but we shall not need them.
There is no definition of a ‘division’ operation for matrices (in the sense of one matrix ‘divided
by’ another), but often multiplication by the inverse achieves what might be regarded as a form
of division. For example, if AB = C and A has an inverse, then by premultiplying both sides
by A−1 we obtain A−1 AB = A−1 C, and therefore B = A−1 C (by using the fact that A−1 A = I
and IB = B).
The inverse of a 2 × 2 matrix is easy to find:

µ ¶ µ ¶
a b 1 d −b
M= ⇔ M −1 = .
c d ad − bc −c a
Here the inverse exists when ad − bc 6= 0 and you can check that it satisfies (2.11).
There is also a general rule on how to find the inverse of a matrix of any size—you should
consult a mathematics textbook when you need it.
The inverse of a product is the product of the inverses, but you have to reverse the order:
(AB)−1 = B −1 A−1 .
Proof: (AB)(B −1 A−1 ) = A(B −1 B)A−1 = AA−1 = I and you are invited to show by a similar
method that (B −1 A−1 )(AB) = I.
Transpose and scalar product
The transpose of a matrix, written M T , is the matrix obtained by swapping the rows and
columns. To be precise:
A = MT means Aij = Mji .
For example, the transpose of the matrix displayed in (2.9) is

 
1.2 2
 −3.6 4.5  .
8 2
The transpose of a row vector is a column vector, and the transpose of a column vector is a
row vector.
The following results are useful:
(A + B)T = AT + B T (2.12)
(AB)T = B T AT (2.13)
(AT )−1 = (A−1 )T (2.14)
Note the order reversal in (2.13). You can easily prove this result using (2.10). Then (2.14)
follows since if M is the inverse of AT then we must have AT M = I, taking the transpose of
both sides gives M T A = I T = I and hence M T = A−1 and the result follows.
The product of a row vector and a column vector of the same length is often useful because it
is simple: it is a 1 × 1 matrix, in other words a scalar. If we start with a pair of column vectors
u and v of the same size, then we can obtain such a scalar by
uT v. (2.15)
This comes up often, so it is given a name: it is called the scalar product or inner product of
the vectors. (The inner product of a pair of row vectors would be uvT .) You can calculate
it by multiplying corresponding components and summing. For example if u has components
u1 , u2 , u3 and v has components v1 , v2 , v3 then
uT v = u1 v1 + u2 v2 + u3 v3 . (2.16)
Most science or mathematics students will meet the scalar product first in the context of vector
analysis in space, where one is typically dealing with three-component vectors representing
things like displacement, velocity and force. In this context it can be convenient not to be too
concerned whether the vectors are row or column vectors, and so the dot notation is introduced:
the scalar product is written u · v. In relativity we will be dealing with 4-component vectors
in time and space, and for them we will introduce a special meaning for the dot notation and
for the phrase ‘scalar product’. In chapter 9 we shall also introduce the ‘outer product’ which
enables a square matrix to be obtained from a pair of vectors.
Chapter 22
Appendix 1. Some basic

arguments
22.1 Simultaneity and radar coordinates
If two events happen at the same place in some reference frame, then it is easy to define the time
and distance between them. The distance in this case is zero, and the time is determined by the
number of ticks of a clock situated at the location of both events. (In practice we would also
need to agree some sort of standard of time—this is currently done by observing the oscillations
of the nucleus of a caesium atom in vacuum, but the details are not necessary in order to study
Relativity, we just need to agree that some such standard can be defined. The Principle of
Relativity ensures that the definition applies in all reference frames equally.)
For events happening at different places, the time and distance between them has to be defined
carefully. A convenient method is first to use ‘radar coordinates,’ and then derive times and
distances from those. For any given reference frame F, we can pick a position to serve as
the spatial origin O of a coordinate system. The particle located at such an origin will have
a straight worldline. Now consider an arbitrary event R, not at the origin. We imagine a
‘radar echo location’ scenario. That is, at time t1 the particle at the origin of F sends out an
electromagnetic pulse, propagating at the speed of light c (think of it as a radio wave pulse or
a flash of light, for example). We suppose that the pulse is reflected off some object present
at event R (so R is the event of reflection), and then the pulse propagates back to the particle
at the origin, arriving there at time t2 . The times (t1 , t2 ) constitute the radar coordinates of
event R in reference frame F. Together with the direction of travel of the pulses, they suffice to
determine the position and time of R in frame F. Let’s see how.
First, since the speed of light is independent of processes such as reflection, the outgoing and
625
D
R
Figure 22.1: Identifying simultaneous events on a spacetime diagram. AB is a straight timelike

worldline, so it can be the worldline of a particle in uniform motion, at the spatial origin of an
inertial reference frame F. AR and RB are photon worldlines. D is half way between A and
B. In the reference frame whose time axis is AB (i.e. frame F), R must be simultaneous with
D, because the photon travel times of AR and RB must be equal. Applying the argument to
further events, one concludes that all events along the dashed line are simultaneous in frame F.
incoming pulse must have the same speed c in F. It follows that the outgoing pulse takes the
same time to get to R as the incoming one takes to come back, so R must occur at a time half
way between t1 and t2 , i.e.
t2 + t1
tR =
2
in frame F. Also, since the pulse traveled out and back at the speed c, the distance from O to
R must be
c(t2 − t1 )
xR =
2
in frame F. This pair of equations allows us to convert easily between radar coordinates and
ordinary coordinates. It means that we only need one clock, at the origin, together with the
Postulates of Relativity, to define a complete coordinate system for any given reference frame.
The worldline of a particle at the origin of F serves as the ‘time axis’ of F. The set of events
simultaneous with R (according to reference frame F) consists of all those having the same
value of the radar echo time (t2 + t1 )/2. On a spacetime diagram these form a line that can be
constructed as shown in figure 22.1. If we choose the scales such that photon worldlines have
slope 45◦ on the diagram, then a line of simultaneity or ‘distance axis’ for frame F makes the
same angle with a photon worldline as the time axis of F does. In other words, the distance axis
and time axis of any reference frame are oriented such that the angle between them is bisected
by a photon worldline. This is the graphical representation of the fact that the speed of light
is the same in all reference frames.
The spacetime diagram construction allows us to see the relativity of simultaneity very easily:
sets of events that are simultaneous in one reference frame are not simultaneous on another,
because the lines of simultaneity associated with different frames cut through spacetime in
different directions.
As a result of this, the temporal sequence of events with spacelike separation can depend
on reference frame. The temporal sequence of events with timelike separation is, however,
independent of reference frame, because lines of simultaneity never slope more than photon
worldlines. This preserves the logic of cause and effect, so that Special Relativity is a physically
and logically consistent theory.
22.2 Proper time and time dilation
For any pair of events having a timelike interval between them, the proper time is defined to be
the time between such a pair of events, as observed in the reference frame in which the events
happen at the same place. By definition, the proper time as related to the invariant interval by
s
τ= (22.1)
c
(by using (2.7) with τ = t2 − t1 , and x2 = x1 , y2 = y1 , z2 = z1 ).
If a particle or system is in uniform motion at speed v relative to some reference frame F, then
the time interval between a pair of events at the particle is related to the proper time by
τ
T =p = γτ (22.2)
1 − v 2 /c2
where
1
γ(v) ≡ p . (22.3)
1 − v 2 /c2
This is called time dilation, because T ≥ τ . It means that all systems evolve more slowly if they
are moving then if they are at rest. For example, a rabbit carried along in a rocket moving at
v = 0.999c would not experience the slightest difference in the laws of physics which determine
its metabolism, and yet it would live for 150 years (i.e. 22 times longer than slow rabbits),
when the time is measured by clocks relative to which it has this high speed. This effect has
not been observed for rabbits, owing to experimental difficulties, but it has been observed for
muons and other particles, and for atomic clocks.
h h
d0
v
vth vth
Figure 22.2: The photon clock. A pair of mirrors is attached to a rod of rest length d0 , such
that light can bounce between them. In the rest frame, the time for a round trip of a light
pulse between the mirrors is τ = 2d0 /c. The figure shows the situation observed in a reference
frame travelling to the left relative to this ‘clock.’ A such a reference frame, the mirrors have
a speed v to the right. Let the round trip time in this reference frame be t. To complete
a round trip the light pulse must travel a distance 2h. Clearly since h > d0 we must find
t > τ : time dilation. Pythagorus’ theorem, h2 = d20 + (vth )2 where th = t/2. Therefore
t = 2h/c = 2(d20 + v 2 t2 /4)1/2 /c. Solving for t one finds t = γτ . The argument hinges on the fact
that the events ‘pulse leaves’ and ‘pulse returns’ are just that: events, so t is the time between
the same two events as those whose time separation is τ in the rest frame.
To deduce the time dilation factor, one can argue from the invariance of the interval (see
exercises) but the conceptually most simple derivation is arguably the ‘photon clock’ argument,
see figure 22.2.
22.3 Lorentz contraction
If a pair of events has a spacelike separation, then there exists a reference frame in which they are
simultaneous. The distance between the events, as observed in such a reference frame, is called
the proper distance L0 . Owing to the fact that we normally study the evolution of particles
and systems, the concept of proper time is much more useful in practice than the concept of
proper distance. However, distance is also needed in order to get a complete description when
changing from one reference frame to another.
A physical object can be regarded as a set of worldlines (those of the particles of the object).
If these wordlines are straight and parallel then the object has constant velocity and fixed size.
The spatial size of an object is defined as the size of the region of space it occupies at any
instant of time. Owing to the relativity of simultaneity, this concept is a relative one, i.e. it
is well-defined only once a reference frame is specified, and the value obtained for the size can
depend on which reference frame is adopted.
Suppose we choose some direction in space, and take an interest in sizes of physical objects
along this direction. The length of an object in its own rest frame (i.e. the frame in which
its velocity is zero), along the chosen direction, is called its rest length. In any reference frame
moving with respect to the rest frame, the object has some non-zero velocity v. In such a
reference frame the length of the object, along the chosen direction, is given by
L0
L= , (22.4)
γ(vk )
where vk is the component of velocity along the chosen direction. Since L ≤ L0 this is called
Lorentz contraction or space contraction. It means that an object in motion is contracted along
the direction of motion compared to its size when at rest. For example, a rabbit carried along
in a rocket moving at v = 0.999c would have physical dimensions approximately 15 cm × 20
cm × 1 cm, when measured by rods relative to which it moves at this high speed.
To deduce this result one can use a photon clock once again, but now the clock is oriented along
the direction of motion. In other words we are using a radar method again.
Consider a simple rod or stick. Suppose a pulse of light is emitted from one end of the stick
and travels to the other end, where it is reflected and comes back to the first end. In the rest
frame of the stick, the time between emission and final reception must be 2L0 /c where L0 is
the stick’s rest length. The events of emission and final reception happen at the same place in
Figure 22.3: A simple set of events suited to reasoning about the Doppler effect. Two particles
(with relative velocity less than c) pass one another at the origin O. At event A the first particle
sends a light signal to the second particle, where it arrives at event B.
the rest frame of the stick, so their time separation in that frame is the proper time between
them, τ = t2 − t1 . The time between these events in any other frame must therefore be
2L0
T = γτ = γ .
c
The stick singles out a direction in space by its own axis, and we now consider a reference frame
in which it moves in that direction. In such a reference frame, the emitted light pulse moves
at speed c and the far end of the stick moves at speed v. It follows that the time taken for
the light pulse to reach the point of reflection is L/(c − v) and the time taken to come back is
L/(c + v), where L is the length of the stick in the new reference frame. Hence
L L 2cL c2 − v 2 L0
T = + = 2 ⇒L= γL0 = .
c−v c+v c − v2 c2 γ
This is eq. (22.4) in the case where vk = v. The general case is treated in chapter 3.1.
22.4 Doppler effect, addition of velocities
Suppose two particles, moving along a line, pass one another at event O, then at event A the
first particle sends a light signal to the second, where it arrives at event B; see figure 22.3.
We take O as the origin of position and time. If the relative velocity of the particles is v then,
in the reference frame F of the emitter, the reception event takes place a distance d = vtB away,
so the signal travel time is d/c = vtB /c. It follows that tB = tA + (v/c)tB , therefore
tA
tB = . (22.5)
1 − v/c
Events O and B take place at the same place in the reference frame F0 of the receiver, so their
time separation t0B is a proper time in that frame, so tB = γt0B , hence
s s
t0B 1 1 1 − v 2 /c2 1 + v/c
= = = . (22.6)
tA γ 1 − v/c (1 − v/c)2 1 − v/c
We can use this result to deduce the Doppler effect for light waves. All we need to do is suppose
that the emitter emits regularly space signals, once every time tA in his reference frame, then
the above argument applies to all the signals, and the receiver will receive them spaced in time
by t0B as given by (22.6). This set of signals could in fact be one continuous stream of light
waves, with period tA . Then the event A could be, for example, “the electric field of the light
wave at the emitter is at a maximum”, and the event B could be “the electric field of the light
wave at the receiver is at a maximum”. Since we define the period of a wave to be the time
interval between successive maxima, it follows that (22.6) relates the periods observed at the
emitter and receiver. By taking the inverse, we obtain the relationship between the frequencies.
Hence if light waves of frequency ν0 are emitted by a particle, then the frequency observed by
any particle moving directly away from the emitter at speed v is given by
s
1 − v/c
ν= ν0 . (22.7)
1 + v/c
This is called the longitudinal Doppler effect or just Doppler effect. It permits one to deduce,
for example, the speed of a star relative to Earth, from the frequency of the received light, if
one has independent evidence of what the emitted frequency was.
Now we shall use the Doppler effect to deduce a formula concerning relative velocities. Suppose
F0 moves relative to F with speed u, and F00 moves relative to F0 with speed v (i.e. v is the
speed of F00 as observed in frame F0 ), all motions being along the same direction, then we can
calculate the speed w of F00 relative to F by using the Doppler effect formula three times, as
follows:
s s s
0 1 − u/c 00 1 − v/c 0 1 − w/c
ν = ν, and ν = ν = ν.
1 + u/c 1 + v/c 1 + w/c
It follows that
µ ¶µ ¶
1 − u/c 1 − v/c 1 − w/c
=
1 + u/c 1 + v/c 1 + w/c
and after a little algebra we obtain
u+v
w= . (22.8)
1 + uv/c2
This is the formula for “relativistic addition of velocities.” The generalization to velocities in
any direction is presented in chapter 3. The formula predicts that w is never greater than
c as long as u and v are both less than or equal to c; this is in agreement with the Light
Speed Postulate. The result predicted by classical physics is w = u + v; the relativistic formula
reproduces this in the limit uv ¿ c2 .
Note that the velocities in the formula (22.8) are all what we call ‘relative velocities’ and they
concern three different reference frames. There is another type of velocity that can be useful
in calculations, which we shall refer to as ‘closing velocity.’ The concept of ‘closing velocity’
applies in a single reference frame, and it refers to the rate of change of distance between two
objects, all distances and times being measured in a single reference frame. When both objects
are moving relative to the reference frame, a closing velocity is not necessarily the velocity of
any physical object or signal, and it can exceed the speed of light. For example, an observer
standing at the collision point of a modern particle accelerator will observe a bunch of particles
coming towards him from the right at a speed very close to c, and another bunch approaching
from the left at a speed very close to c. The positions of the two particle bunches can be
written z1 = d − vt and z2 = −d + vt in the rest frame of such an observer, where v ' c and d
is a constant (equal to half the distance between the bunches at t = 0). He finds the distance
between the bunches to be z1 − z2 = 2d − 2vt. The rate of change of this distance is 2v. The two
particle bunches may then be said to have a ‘closing velocity’ of 2v, and this can easily exceed
the speed of light. Nevertheless, the relative velocity of the bunches is w = 2v/(1 + v 2 /c2 ),
which is less than c. The relative velocity is the velocity which an observer moving along with
one bunch will find the other to have.
An example of closing velocity can be found in equation (22.5). We can obtain that equation
by looking at the situation at time tA . At this moment, the emitter considers that the receiver
is a distance vtA away, and he finds that the signal sent out at speed c has a closing velocity
with the receiver of c − v, therefore it takes a time vtA /(c − v) to close with the receiver. Hence
tB = tA + vtA /(c − v) which you can verify agrees with (22.5).

Relativity Made Relatively Easy: Andrew Steane

Uploaded by

Copyright:

Available Formats

Relativity Made Relatively Easy: Andrew Steane

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Relativity Made Relatively Easy: Andrew Steane

Uploaded by

Copyright:

Available Formats

Relativity made relatively easy

2.1 Classical physics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

2.2 Special relativity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

2.2.1 The postulates of Special Relativity . . . . . . . . . . . . . . . . . . . . . 20

2.2.2 Central ideas about spacetime . . . . . . . . . . . . . . . . . . . . . . . . 21

2.3 Matrix methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

3 The Lorentz transformation 29

3.1 Introducing the Lorentz transformation . . . . . . . . . . . . . . . . . . . . . . . 29

3.2 Derivation of Lorentz transformation . . . . . . . . . . . . . . . . . . . . . . . . . 32

3.4 Lorentz invariance and four-vectors . . . . . . . . . . . . . . . . . . . . . . . . . . 34

3.4.2 Lorentz invariant quantities . . . . . . . . . . . . . . . . . . . . . . . . . . 38

3.5 Basic 4-vectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

3.5.1 Proper time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

3.5.2 Velocity, acceleration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

3.5.3 Momentum, energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

3.5.4 The direction change of a 4-vector under a boost . . . . . . . . . . . . . . 47

3.5.6 Wave vector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

3.6 The joy of invariants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

3.7 Moving light sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

3.7.1 The Doppler effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

3.7.2 Aberration and the headlight effect . . . . . . . . . . . . . . . . . . . . . . 55

3.7.3 Stellar aberration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

3.7.4 Visual appearances* . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

4.1.1 Transformation of force . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

4.2 Motion under a pure force . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66

4.2.1 Constant force (the ‘relativistic rocket’) . . . . . . . . . . . . . . . . . . . 69

4.2.2 4-vector treatment of hyperbolic motion . . . . . . . . . . . . . . . . . . . 74

4.2.3 Circular motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76

4.2.4 Motion under a central force . . . . . . . . . . . . . . . . . . . . . . . . . 77

4.2.5 (An)harmonic motion* . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82

4.3 The conservation of energy-momentum . . . . . . . . . . . . . . . . . . . . . . . . 82

4.3.1 Elastic collision, following Lewis and Tolman . . . . . . . . . . . . . . . . 83

4.3.2 Energy-momentum conservation using 4-vectors . . . . . . . . . . . . . . . 83

4.3.3 Mass-energy equivalence . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87

4.4.1 Elastic collisions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93

4.5 Composite systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96

4.6 Energy flux, momentum density, and force . . . . . . . . . . . . . . . . . . . . . . 98

5 Further kinematics 101

5.1 The Principle of Most Proper Time . . . . . . . . . . . . . . . . . . . . . . . . . . 101

5.2 4-dimensional gradient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101

5.3 Current density, continuity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

5.4 Wave motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

5.4.1 Wave equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110

5.4.2 Particles and waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110

5.5 Acceleration and rigidity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113

5.5.1 The great train disaster1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115

5.5.2 Lorentz contraction and internal stress . . . . . . . . . . . . . . . . . . . . 116

5.6 General Lorentz boost . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119

5.7 Lorentz boosts and rotations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119

5.7.1 Two boosts at right angles . . . . . . . . . . . . . . . . . . . . . . . . . . 121

5.7.2 The Thomas precession . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123

5.7.3 Analysis of circular motion . . . . . . . . . . . . . . . . . . . . . . . . . . 126

5.8 The Lorentz group* . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129

5.8.1 Further group terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . 134

5.9 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134

6 Relativity and electromagnetism 135

6.1 Maxwell’s equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135

6.1.1 Moving capacitor plates . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137

6.2 The fields due to a moving point charge . . . . . . . . . . . . . . . . . . . . . . . 143