Summary of Special Relativity

MIU PHYS 340 Robert D. Klauber www.quantumfieldtheory.info

Wholeness Chart 1. Overview of the Development of Special Relativity

Details Comment
Observers in inertial frames only. (But, accelerating object can be handled
Frames considered
by special relativity.)
Chronology
Speed of light: not same for all (not invariant)
Pre-Einstein:
Laws of nature:
Galilean
transformation F=ma valid for all observers (invariant)
Maxwell’s equations not valid for all (not invariant)
Problems before Implied light speed not obey
Michelson-Morley experiment
1905 Galilean transformation
1) Speed of light same for all observers (invariant) Invariant in form = vectors
Einstein postulates
2) Laws of nature same for all observers (invariant in form = covariant) in eqs change covariantly
Lorentz transformation (instead of Galilean) Resulted in Lorentz
contraction, time dilation,
ct  
1
 ct  vc x  ct 
1
 ct   vc x simultaneity not the same
1 v / c
2 2
1 v / c
2 2
for all (not invariant),
Result of 1 1 E=mc2, and more.
postulate #1 x   x  vt  x  x  vt  
1  v2 / c2 1  v2 / c2 Reciprocal: each observer
sees other frame with
y  y y  y
Lorentz contraction, time
z  z z  z dilation, etc.
Impact on Maxwell’s equation valid for all (invariant in form = covariant)
postulate #2 F=ma not valid for all (not invariant in form = not covariant)
dx 
So, Einstein du   u   = proper time
New 4D law of mechanics: F  m where u is 4-velocity d
changed mechanics d
on object (see below)
Only mechanics and e/m
Result of ↑ in 1905 All laws of nature same for all observers (invariant in form = covariant)
known then.
True for weak and strong
Result of ↑ up to force laws. A “must have”
modern day: Invariance in form of laws of nature is now a general principle used in all
for any new proposed
physics. Any law must be covariant under Lorentz transformation.
Postulate #2 valid theory (SUSY, GUTs,
strings, etc.)
Minkowski in 1908 Space and time = 4D spacetime continuum
Concepts and Relations
x = (ct, x, y, z) = (x, x , x , x3) contravariant components
1 2
4D position vector
x = (‒ct, x, y, z)= (x, x1, x2, x3) = (‒x, x1, x2, x3) covariant components
(s)2 = ‒ (ct)2+(x1)2+(x2)2+(x3)2 same for all observers between same Not seen before Minkowski
Invariant interval two events. (We need a minus sign for (ct)2 to get a Lorentz invariant because assumed + sign for
“length” for the position vector between two events.) (ct)2. s not then invariant.
2
Time  passing on standard clock at rest with respect to object. Found from invariant
Proper time  on an
t interval between 2 events
   1  v / c t
2
object   t  at x  y   z   constant
 on object world line.

Length measured with meter sticks at rest with respect to object. Found from invariant
Proper length Lo of 1 interval between 2 events at
an object L0   L  L t   0 and t  0 ends of object at same time
1  v / c
2
in each frame.

w = (w, w1, w2, w3) contravariant components

4-vector
w = (w, w1, w2, w3) = (‒w, w1, w2, w3) covariant components
3
Magnitude of a  w 2   w w  w w  w0 w0  w1w1  w2 w2  w3 w3 Einstein convention after
4-vector  0 2nd equal sign.
 w0 w0  w1w1  w2 w2  w3 w3

Covariance of a To qualify as a legitimate four-vector, its magnitude must be Lorentz

4-vector invariant. Components can vary, but not magnitude.

 1 0 0 0  w0 
 1 
    0 1 0 0  w 2  w w   w0 w1 w2 w3   w2 
0 0 1 0
 0 0 0 1   w3 
 w 
0  w 
Minkowski metric 0
 1 0 0
and 4-vector length
  w w w w   0 1
0 1 2 3 0 0  w1     w w Compare to (w)2 above
0 0 1 0  w2 
 0 0 0 1  w3 
 
 s 2   x x (assumes initial s0 = 0, so s= s ‒ s0 = s) Compare to (s)2 above
u 0  c c
 1  1  1

dx u  v  1 v  dx i Always tangent to particle
4-velocity u    2  v 2   v 2  vi  Newtonian velocity 
d u  v
2 dt world line
  1    
u 3  v 3  c v 3 
     
Invariant. Same for any
4-velocity squared (u)2 = uu = ‒ c2 Massive particles.
particle and any observer.
Massive particles p= mu. ↓ Valid for all particles
 p0   E / c
 1  1  E = relativistic energy; pi =
4-momentum  p   p  mc 2 mvi
p   2   2  E   mc 2  12 mv 2  … p 
i
relativistic 3-momentum
p 2 2
p
    v v
 p3   p 3  1   KErel 1  
     c  c
(p)2 = pp = ‒ m2c2 Massive and massless particles. Invariant. Same for any
4-momentum
E 2
E 2 observer, any velocity.
squared p  p   2  pi pi   2  pi pi  p  p  m 2 c 2 Different for different mass.
c c
4D unit basis
e = e0, e1, e2, and e3. Like i, j, k in 3D
vectors
4-vectors A= A0e0 + A1e1 + A2e2 + A3e3 = Ae Same A, diff frames: Ae = Aˊeˊ
Invariance vs Invariance = no change for different coordinate systems (observers) s invariant, not conserved
conservation Conservation = no change over time E conserved, not invariant
3

Spacetime Diagrams See figures below.

Concept/Entity Timelike Interval (AB, Fig. 1) Spacelike Interval (AC) Lightlike Interval (AD)
Region on spacetime diagram Inside light cone Outside light cone On surface of light cone
Space vs time components ct > x ct < x ct = x
(s)2 = ‒( ct)2 + (x)2 negative positive zero (null)
s imaginary real zero
Travel from first event to second? Yes No Only light can.
Find proper time  from Yes, for a particle traveling No. Particle would have
 = 0 for a photon.
(s)2 = (c)2 ? from first to second event to travel faster than light.
Yes. At or below light speed,
Can first event affect (cause) the No. A signal would have Yes, but only an electromagnetic
a signal from 1st event can
second event? to travel faster than light. signal. All others too slow.
reach 2nd event.
No. Inside light cone can
Yes. Can have xˊ axis No. Observer would have to travel
Can the two events be never have the xˊ axis (the
extending from first event at light speed between events to see
simultaneous for some observer? axis where all events occur at
to second. them simultaneous.
the same time)
Is the time order of events the No.
Yes Yes
same for all observers? (See Fig. 2 below.)
s invariant? Yes Yes Yes
Is the primed time axis timelike,
Yes, timelike. No. No.
spacelike, or null?
Is the primed space axis timelike,
No. Yes, spacelike. No.
spacelike, or null?
Is the light path for all observers
No. No. Yes, null.
timelike, spacelike, or null?

path of clock
edge of
fixed in S'
light cone ct ct' ct ct"' ct ct"" x""
edge of ct'
ct
light cone
ct'C > 0 C C x"' ct""
A =0 C
ct"'
C =0
C x'
D
B
x' A A A
x x x
ct'A = 0 ct"'
A =0
x ct""
C <0
A

Figure 11-1 Kinds of Intervals Figure 11-2. Order of Spacelike Separated Events Different for Different Observers
(Not for timelike such as AB. Cannot rotate space axis through both events,
so never simultaneous nor reversed in order.)
4
From MIU PHYS 530 Robert D. Klauber www.quantumfieldtheory.info

Wholeness Chart 2. Electromagnetism: Classical 3D + 1 vs Relativistic 4D Spacetime

Equation numbers are for Griffiths, Introduction to Electromagnetism, 4th ed.
Entity 3D + 1 4D Comment
 0 E1 / c E 2 / c E 3 / c 
 
  E1
/ c 0 B 3
 B 2
 Antisymmetric.
E/m field
N/A F     (12.119) F = F(t,xi) =
tensor 2
E / c  B
3 1
B 
0 F(x)
 3 2 
  E / c B  B1 0 

 0 B1 B2 B3 
 
Antisymmetric,
   B E 3 / c E 2 / c2 
1
E/m dual field 0
N/A G   (12.120) G = G(t,xi) =
tensor 2
 B E / c
3 2
 E1 / c 
0 G(x)
 3 2 2 1 
  B  E / c E / c 0 

Electric field 3-vector E or Ei cF01, cF02, cF03 Components of the

4D e/m field
Magnetic field 3-vector B or Bi F23, F31, F12 tensor
Charge Q Q Invariant

Proper charge Q Q Q Invariant,

  0   (V  0) page 565 0  (V  0) (12.121), (12.122)
density V0 V V0  = (t,xi)= (x)

Q Q Q 0
Charge
 same as above    (12.122) Not 4D invariant,
density V V V 1 v / c
2 2
1  v 2
/ c 2 since V not
0

Current 0 v Griffiths uses u for

J=v page 565 J  v   0 u i (12.122)
density 1 v / c
2 2 our v,  for our u

J=[J0, J1, J2, J3] = [ c,v1,v2,v3] (12.124)

 Not 4D invariant,
4-current N/A
= 0u 
J = [‒ c,v ,v ,v ] = 0u

1 2 3
J=J(t,xi)=J(x)
Continuity  J 
equation  iJ  0 below (12.124) 0 (12.126) x0 = ct
(charge) t x 

 E  1 
0

 B  0
Maxwell’s B F  G 
E  
(7.40) 
 0 J  0 (12.127)
equations
t x x
E
  B   0 J   0 0
t
Lorentz force Griffith’s K
F  q  E   v  B  below (12.129) K = quF (12.128)
law is our F

Scalar & Griffith’s V

vector  and A A= [/c, Ai] T = [A0, A1, A2, A3] (12.132) is our 
potential A = A(x)
5
E/m field A A Antisymmetric.
tensor via N/A F    (12.133)
x x F = F(x)
4-potential

Maxwell’s    A   2  c0   A    A 
For 3D + 1,
equations ct   different units, see
        0 J (12.134)
via 4-     A
2
x  x  x  x  Jackson 1975 pg
potential       A    2 2   2 A  0 J 220 (6.32), (6.33)
 ct  c t

 A
Lorenz gauge  i A   1c pg. 570 0 (12.136)  = V/c
t x

 2
Maxwell’s    2   c 0  For 3D + 1,
c 2 t 2   
equations  A  0 J  (12.137) different units, see
in Lorenz 2 A x x Jackson 1975 pg
gauge   2 A  0 J 220 (6.37), (6.38)
c 2 t 2

NOTE: For no source charges or 3-currents (i.e., J = 0) last equation above ((12.137) in Griffiths) is just the wave
equation
   for   2   2   2   2   2
A 0   A  A  A  A 0
x x 1in x direction x0 x 0 x1 x1 x2 x 2 x3 x 3
= 0 since A2 only depends on x1 ,t
1
take x as x  2 A2  2 A2  2 A2 2  A
2 2
    0   c c  wave speed
c 2t 2 x 2 t 2 x 2