Special Theory of Relativity

An Introduction

DJM 8/01

Background
Most basic concept of relativity is as old as Galilean and Newtonian Mechanics. Crudely speaking, it is: The Laws of Physics look the same in different frames of reference In Newtonian Relativity: This applies to Mechanics In Einsteins Relativity: This applies to all Laws of Physics When applied to all laws, difficult issues arise. Implications seem to violate intuition masses increase, lengths are shortened, time expands Has generated fascination, has brought glamour... Bottom line: STR is corroborated by experiment Until 20th century, Newtons Laws (F = dp/dt, Fab = -Fba, conservation of energy and momentum) were held true (and still are) and absolutely believed in (not any more, as written originally)

Background
Recall: Newtons Laws are valid only in an inertial frame
An inertial frame of reference is one in which Newtons 1st Law holds; it exists only in empty space; it is at rest w.r.to stars)

Basic Newtonian mechanics issue: A state of a system is specified at some time to by giving position coordinates (xo,yo,zo) and velocity (vox,voy,voz) at to. Can calculate position (x,y,z) and velocity (vx,vy,vz) at a later time t by knowing all forces acting on the system and applying Newtons laws. Often desirable to specify such a state in terms of a new set of coordinate axes, which is moving with respect to the first Three important questions arise:
How do we transform the state from the old to the new frame? i.e. How do we convert x1,y1,z1,t1 x2,y2,z2,t2 ? What happens to Newtons Laws under this transformation? What happens to E&M theory (Maxwells Equations)?

Theory of Relativity concerns itself with these questions

Background
Albert Einstein formulated the modern Theory of Relativity. He proposed two such theories: The Special Theory of Relativity
1905 - Deals with the case of an inertial frame of reference moving with constant velocity with respect to another inertial frame
Its consequences are most important as v c

The General Theory of Relativity

1915 - Deals with the case of an inertial frame of reference accelerating with respect to another inertial frame. A relationship results between accelerated motion and gravitational effects It is the current theory describing the gravitational interaction Not based on quantum mechanics Requires knowledge of tensor calculus

The S1 and S2 Inertial Frames

y y
Event y2

y1

vt1=-vt2
x2

v = constant

S1
z z

S2

x1

x2 = x1 - vt1 x1 = x2 + vt2 t1 = t 2

Newtonian Relativity

(Some History)

Galilean Transformation of space-time

Two observers (1,2) are in their own, separate, inertial frame of reference S1 and S2
S2 moves with respect to S1 with v = constant along x-axis

Each observes the same event giving position and time as measured in their own frame of reference (x1,y1,z1,t1; x2,y2,z2,t2)
Each has own meter stick and clock When temporarily at rest, sticks same length, clocks synchronized i.e. at the instant t1 = 0, then t2 = 0 and x in S1 = x in S2

When S2 is moving with respect to S1, the state (x and t) of the event as seen by the two observers is related by:
x2 = x1 - vt1 y2 = y1 z2 = z1 t2 = t1 v -v gives x1 = x2 + vt2 y1 = y2 z1 = z2 t1 = t2 (S1 S2)

Newtonian Relativity

(Some History)

Transformation of velocities and accelerations follow directly by taking derivatives

v2x = v1x - v and a2x = a1x v2y = v1y a2y = a1y v2z = v1z a2z = a1z This is the Newtonian answer to Question #1 (intuitive) Important implicit assumptions: space is absolute, time is absolute x dx Note (definition): v x = = lim etc. dt t 0 t Implies that, for meaningful vx, measurement of x and t must be made with respect to a single (the same) reference frame When using Galilean Transformations this is not important because time is absolute and, therefore, t1 = t2

Newtonian Relativity

(Some History)

All laws of mechanics are the same in all inertial frames

Example: Newtons 2nd Law under Galilean transformation When a mass accelerates, let the force acting on it be measured by S1 and S2. They find, respectively: F1 = ma1 and F2 = ma2 But a1 = a2 Therefore F1 = F2 And, therefore, Newtons 2nd Law retains its mathematical form in both inertial frames (is the same, invariant). All other laws of mechanics are also invariant, as they are derivable from 2nd Law Note assumption: mass m is the same in both S1 and S2 i.e. m1 = m2

This is Newtonian answer to Question #2 Corollary: Impossible to prove by mechanical tests that one frame is at absolute rest rather than in motion with v = const. w.r. to another one. (Mechanics looks the same in both)
Corollary: All inertial frames are equivalent; Length, time and mass are independent of relative motion of the observer

Failure of Newtonian Relativity

(Some More History) Maxwells (4) equations describe electromagnetism successfully Maxwells equations predict the existence of e.m. waves propagating through free space with speed c = 3.00 x 108 m/s
Question #1: With respect to what frame is c to be measured? Question #2: Through what medium do e.m. waves propagate? 19th century answer: Ether hypothesis - existence of a massless but elastic substance permeating all space
Electromagnetic waves propagate through the ether c is to be measured with respect to the ether

An experiment is needed to test the presence of ether!

Michelson-Morley Experiment (1881)

Premise: If one believes in the ether (and that c is 3x108 w.r. to it), and in Maxwells Eqns, and in Galilean transformations for E&M, then if one measures the speed of light in a frame moving w.r. to ether, one should be able to measure v of this moving frame. Plan: Measure earths speed as it moves through ether Result: Null; can not detect any motion of the earth through ether

Failure of Newtonian Relativity

(Some More History) Implications:
Question (remember #3?): Do Maxwells equations obey Newtonian Relativity? (i.e. do they stay invariant under Galilean transformations, under the ether hypothesis?) Short Answer!: No! If they did, equations would produce c+v for the speed of light (and in a moving space ship, electrical or optical phenomena would be different than in a stationary space ship; then, some one on the spaceship could measure the speed of the space ship)
Contradicted by Michelson-Morley experiment.

Another problem: Maxwells equations seem to predict the same value for c for all inertial frames. Therefore:
If the source of the e.m. waves is moving, the e.m. wave is predicted to move through space with the same c. But: This is contradicted by Galilean transformations. Experiments support c = 3.00 x 108 m/s in all frames

What to Do?
Are Maxwells equations not the same in all inertial frames?
Maxwells equations could be wrong (only 20 years old!) If so, Galilean transformations would be retained and Maxwells equations should be fixed to obey Galilean transformation and A preferred inertial frame must exist which should produce different values for c as one moves through it. Keep looking!

Are Galilean transformations wrong?

Then, keep Maxwells equations as they are; experiments OK Either hypothesis would have to be abandoned Absolute space and absolute time would have to be abandoned

What about c?
All moving frames seem to give same value: 3.00 x 108 m/s Poincare: A total conspiracy of nature is itself a law of nature

Einstein to the rescue:

Fruitless to continue to show experimentally that a special inertial frame exists. Proposes Two Postulates.

Special Theory of Relativity Einsteins Two Postulates (1905)

Postulate I (Principle of Relativity)
All laws of physics must be the same (invariant) in all inertial reference frames (Can not detect absolute uniform motion)

Postulate II (Constancy of c)
The speed of light in vacuum is constant (same value, c = 3.00 x 108 m/s) in all inertial reference frames, regardless of motion of source or observer

Some Observations
Problems solved by STR:
Keep Maxwells equations Accept measurements of c, all producing 3.00 x 108 m/s No need for either
build a theory on experimental results, not the other way around!

New issues arise:

Galilean transformations are abandoned
need new transformation equations

Newtons laws dont obey Postulate I any longer

have to be changed Absolute space and absolute time must be abandoned Constancy of mass (and conservation of mass) must be abandoned Laws are OK the way they are as long as v << c

New and non-intuitive predictions appear Experimental verification needed Acceptance by scientific community is slow

Consequences - Simultaneity
Usual Meaning: Two events occurring at the same time Simultaneity is not an absolute concept anymore
It depends on motion of the observer Taken for granted in Newtonian physics
But it doesnt matter, since v << c in Newtonian world

Precise new definition needed:

Two events in an inertial frame of reference are simultaneous if light signals from the locations of these events reach an observer positioned halfway between them at the same time

Consequences:
Two events simultaneous in one inertial frame are not simultaneous in another frame in motion w.r.to the first t(x); t1 t2; t1 and t2 are intimately intertwined with x1, x2

Consequences - Time Dilation

Clocks slow down when seen by observer moving w.r. to them!
Consider the duration (time interval T) of an event Let event be: a light ray goes from the origin of S2 to a mirror at y2 = d, is reflected there, and returns to the origin. If S1 and S2 are momentarily at rest, both agree, using their own clocks and meter sticks, the duration of the round trip (proper time, i.e. measured in the rest frame of the event, at a fixed location) is: y y 2d = = T T Next: let S2 move along x-axis of 01 02 Mirror c
d v=0

x zz

S S1 2

S1 with v = constant. S1 will measure the time interval again. Call it T1

Consequences - Time Dilation

As seen by S1, light sent by S2 follows path shown, taking time t to get from the origin of S2 to the mirror and back New time interval T1, measured by S1, is T1 = 2t, where

(ct )
y

= d 2 + (vt ) solve for t : t =

2

d 2d , T = 1 c c2 v2

1 1 v c2
2

or

T1 = T02 where =
ct vt vt ct v = constant

1 v2 1- 2 c

S1
z z

x x

Since 1 always, T1 T02

S2

(Relativistic T > Proper T)

Time Dilation Example: Muon () Lifetime

When at rest in the laboratory (i.e. in its rest frame) the has (average) lifetime o = 2.2 x 10-6 s, before it decays.
At the speed of light it would travel ~600 m before decaying

We know are produced high in the atmosphere, ~10 km up, from collisions of cosmic rays with the atmosphere How are we able to observe them on the earth?
As measured by an earth based observer, the life time of the dilated (non-proper time interval): Since a high energy (relativistic) then

= o

is

>> 1 and >> o

has speed v c

Therefore it has enough time to reach the earth and be observed

Consequences - Length Contraction

A length gets shortened when measured by observer in motion with respect to it!
Consider the length between the earth and a star. A space ship (S2) travels from the earth to the star moving with constant speed v along x-axis of an observer at rest on the earth (S1) Earth observer measures the distance to the star L01 (proper length) and computes the time it takes for the space ship (the moving frame) to get there as T1 = L01/v; (T1 is non-proper T). Space ship observer, at rest in S2, sees the star moving with -v; measures time T02 (proper time, at fixed place) it takes to reach the star and computes distance (non-proper L) to be: L2 = vT02 The two time measurements are related by T1 = T02 Therefore, substituting T02 in L2: L2 = vT1/, or:

L2 =

L 01

Since

> 1 always, L2 < L01 (Relativistic L < Proper L)

Length Contraction Example: Human Space Travel

A 25 year-old astronaut wishes to travel from earth to Alpha Pegasus (distance L0 = 100 lt yr away), using a fast spaceship
Will he/she get there in his/her lifetime?

Use v = 0.95c for spaceship speed w.r. to earth ( = 3.2) Astronaut measures T02 = L2/v = (L0/)v = 100/(3.2*0.95) = 32.8 yr (measures proper time interval, computes contracted length) Therefore, astronaut will be and feel 25 + 32.8 = 57.8 years old upon arrival, will get there alive and well. Earth observer computes: T1 = L01/v = 100yr/0.95yr/ltyr = 105 yr for astronaut to get there, (measures proper length, computes dilated time interval), T1 = T02 = 3.2*32.8 = 105 yr
Biological, (e.g. heart beat) and other processes on spaceship will slow down as seen by earth observer

Lorentz Transformations
Recall, Galilean transformations are not valid when v c Need new equations to transform (x1,y1,z1,t1) (x2,y2,z2,t2) Assume relativistic transformation for x1 x2 is the same as the classical one except for an unknown constant , therefore, x2 = (x1 - vt1) (1) and x1 = (x2 + vt2) (2)
y2 = y1 z2 = z1 t2 = ? y1 = y2 z1 = z2 t1 = ?

must be the same in both equations (Postulate I). Substitute x2 from (1) in (2) and solve for t2

x1 2 (x1 vt1 ) x x 2 x1 = = c (Postulate II) t2 = (3). Form 2 ; Set t 2 t 2 t1 v

= 1 v2 1 2 c

Solve for .

Substitute in (1), (2), (3) and get final result:

Lorentz Transformations
x2 = x 1 vt 1 v2 1 2 c y 2 = y1 x1 = x 2 + vt 2 v2 1 2 c y1 = y 2 t2 + vx 2 c2 v2 1 2 c

z 2 = z1 vx 1 2 c t2 = v2 1 2 c t1

or

z1 = z 2 t1 =

Note: Lorentz Transformations are valid for any v.

As v 0, they reduce to the Galilean Transformations, Therefore, G. T. are good enough for Newtonian Mechanics

v > c produces imaginary, non-physical numbers. Therefore c becomes a limiting speed for physical phenomena

Lorentz Transformation of Velocities

By differentiation,
v v v 2x = 1x v 1 2 v1x c v1y v 2y = v 1 2 v1x c v1z v 2z = v 1 2 v1x c

or

v 2x + v v 1 + 2 v 2x c v 2y v1y = v 1 + 2 v 2x c v 2z v1z = v 1 + 2 v 2x c v1x =

If v2x = c, then v1x = c

i.e. if v = c in one frame, then v = c in any frame

These equations refer to measurements done in different frames. Vector addition of velocities in the same frame remain the same

Lorentz Transformation of Velocities Example

Two spaceships A and B approach the Earth from opposite directions Each ship moves with v = 0.80c with respect to the Earth How fast is each ship moving with respect to the other?
i.e. what is the speed of one spaceship measured in the rest frame of the other?

Must establish S1 and S2: Let S1 be the earth, let S2 be spaceship A, moving w.r.to S1 with speed v = +0.80c Event is spaceship B, moving with speed vB1 = -0.80c in S1 Speed vB2 of spaceship B (event) relative to A (S2) is:

v B2 =

v B1 v - 0.80c 0.80c = = 0.98c (- 0.80c )(0.80c ) v v 1 B1 1 c2 c2

Note: vB2 (Galilean) vB1 v = -0.80c 0.80c = -1.6c

Relativistic Dynamics
Newtons Laws are invariant under Galilean transformations but not under Lorentz transformations. They must be changed Experiment shows that Newtons Laws:
Force laws (with p = mv) Conservation of energy Conservation of momentum

may retain their (mathematical) form IF the classical concept of mass is modified: m constant, independent of relative motion between object and observer. Conservation of mass is abandoned!

In relativistic physics, let:

m = m(v), that is:

A mass m moving with speed v with respect to an observer is different than the mass m0 of the same object when at rest with respect to the observer (proper mass or rest mass)

Can show: m =

m0 or, m = m Since > 1 always, m > m 0 v2 1 2 (Relativistic m > Proper m) c

Relativistic Dynamics
Newtons Laws (now relativistically correct):

r r K = W = 0 F d s Definition of Work, K.E. r r [pi ] initial = [pi ] final Conservation of Momentum

s

r r p = mv m = m r r dp F= dt

New Definition of Momentum, where Relativistic Mass, then : Second Law of Motion

Note: As v c and m >> m0, even a small v requires a large F to produce it. Example:
Going from v = 0 to v = 0.01c (v = 0.01c), p = 0.010m0c. Going from v = 0.98c to v = 0.99c (same v = 0.01c), p =2.10m0c (or 210 times more). Therefore F is 210 times more in the second case

What about Energy/Conservation of Energy?

Relativistic Energy
Kinetic Energy (in one dimension):

m0 v v m0 vdv v v v v dp m0 c 2 2 K = 0 Fds = 0 ds = 0 vdp = 0 vd m c = = 0 0 2 2 2 dt v v v 3 1 1 1 2 c c2 c2

Or: K = mc2 - m0c2 = (m - m0)c2 = m0c2 - m0c2 (1) Note: mc2 = m(v0) ; m0c2 = m(v=0) Interpret m(v=0) as Rest Energy E0, E0 = m0c2
No classical equivalent, but not in conflict with Newtonian physics
Can add rest energy on both sides of any Newtonian energy balance without destroying it

Interpret K + E0 as Total Energy E,

E = mc2

from (1).

i.e. E = K + E0 = mc2 = m0c2 = E0

Examples
Find the velocity of a 6 GeV electron
K = 6 GeV 6x103 MeV E K + m 0c 2 K 4 1 1 1.174268x1 0 = = + = + = Find : = 2 2

m 0c m 0c 0.511MeV 1 1 Find v: v = c 1 2 = c 1 = 0.999999996c 4 2 (1.174268x10 )

E0

Find the rate at which the sun loses mass if it radiates energy at the rate of 3.85 x 1026 W
P = E/s = 3.85 x 1026 W M/s = (E/c2)/s = = [(3.85 x 1026)/(3.00 x 108)2]/s= 4.28 x 109 kg/s

Conservation of Energy
The rest energy m0c2 is of more than casual interest
Processes exist where the total rest mass is not conserved Experiment shows that a change E0 is compensated by a change K so that the total energy change E = 0, therefore:

The Total Energy E is the conserved energy quantity now.

Law of Conservation of Energy applies to total energy E In relativistic physics there are no separate laws of conservation of mass and energy

If there is Potential Energy involved, then E = K + V + E0 K is equal to the increase in the mass of the object when it moves with speed v As v 0, K 1/2 m0v2, as it should, (binomial expansion) Note: K can not be obtained from K = 1/2 mv2 by using m = m0

Energy - Momentum Relationship

In Newtonian physics: K = p2/2m relates K to p To find corresponding expression in relativistic physics, use:

m0c2 and p = m0v with =(1 - v2/c2)-1/2 to derive: E2 = p2c2 + (m0c2)2 (very important/useful) E2 - p2c2 = (m0c2)2 is invariant (same value in all frames)
E= Interesting limiting cases:
2

p2 1 p2 p2 2 2 If v << c: 1+ E = m 0c 1 + 2 2 =m 0c 2 2 + ... = m 0 c + 2m 0 2 m0 c m0 c 2 Get K = p /2m

If v c: E >> E0, therefore E pc and K E

If m0 = 0, E = pc, E = K and must have v = c otherwise E = mc2 = m0c2 = 0 If v = c (and E 0), must have m0 = 0 (and E = pc) otherwise m = m0 (0/0 is OK!)

Units: Mass, Energy, Momentum

E2 = p2c2 + (m0c2)2
In nuclear/particle physics preferred unit of energy is MeV 1 MeV = 1.6 x 10-13 J Since mc2 or m0c2 represent energy, their units are MeV
Therefore, mass m can be given in units of MeV/c2 Example: Rest mass of a proton is mp0 = 938MeV/c2; This means: rest energy of a proton is mp0c2 = 938 MeV

Similarly: pc has units of MeV

Therefore, p can be given in units of MeV/c Example: A proton with (total) E = 2 GeV has momentum p:

E2 (m0c2 ) = p= c2
2

(2000MeV)
c2

938MeV = 2 c

MeV (2000 938 ) = 1766 MeV/c c

2 2

Example
Compute the momentum of a 500 MeV proton
From E2 = p2c2 + (m0c2)2 = p2c2 + E02 And from E = K + m0c2 Compute:

p= =

1 1 2 E2 E0 = (K + m 0c 2 ) 2 (m0c 2 ) = c c

Redo, using system of units where c = 1

1 (500 MeV + 938 MeV)2 (938 MeV)2 = 1090 MeV/c c

Convenient since, now, p, E, m will all have units of energy Here E2= p2 + m02 and E = K + m Therefore:

p=

(K + m ) m
2

2 0

= (500MeV + 938MeV) 2 9382 MeV 2 = 1090 MeV

in the regular system of units, would write p = 1090 MeV/c

Relativistic Two-Body Kinematics Elastic Scattering K

m1
1

P1

Given: m1, m2, K1, 1 Find: K1, K2, 2

m1

K1 p1

m2

1 2
m2 K2 P2

Conservation of Energy (using c =1): E1 + m2 = E1 + E2 (1) Conservation of Momentum: p1 = p1cos1 + p2cos2 (2) 0 = p1sin1 - p2sin2 (3) E 1 = K 1 + m1 and E12 = p12 +m12 Recall:
E1 = K1 + m1 E2 = K2 + m2 and and E12 = p12 +m12 E22 = p22 +m22

Solve (1), (2), (3) for K1, K2, 2

Relativistic Two-Body Kinematics Elastic Scattering Results (1)

' E1

E '2

cot '2 =

+ m ) p cos A 4m [(E + m ) p cos ] = 2[(E + m ) p cos ] A (E + m ) p cos A 4m [(E + m ) p cos ] = 2[(E + m ) p cos ] (1 + )cot ( + ) (1 ) + cot A (E
1 1 2 1 ' 1 2 1 2 1 2 1 2 2 1 2 ' 1 2 1 2 2 1 2 ' 1 2 1 2 1 ' 2 2 2 2 2 2 1 2 2 1 2 ' 2 2 1 2 2 1 2 ' 2 1 2 ' 1 2 1 2 2 1 2 ' 1

1 12
1, 2, are given by:

Where A1, A2,

Relativistic Two-Body Kinematics Elastic Scattering Results (2)

A1 = 2(m12 + m1m 2 + m 2 K1 ) 1 = 2 = = A1p1 A 2 = 2(m 2 + m1m 2 + m 2 K1 ) 2

(E

2 2 2 ) ( ) m A 4m E m p + + 1 2 1 1 1 2 1 2

] ]

(E

(E1 + m 2 ) p12 m2 ) A2 4m 2 1 + 2 2
2

A 2 p1

1
2 v cm 1 2 c

E1 + m 2 = m12 + m 2 (2E1 + m 2 )

Relativistic Two-Body Kinematics Elastic Scattering - Observations

Note: Not any 1, 2 are possible in two-body kinematics
Square root in E1, E2 must be > 0

A12 4m12 (m1 + m 2 ) + 2m 2 K1 2 ' sin 1 4m12 p12

2

] ]

(m1 + m 2 ) + 2m 2 K1 4m 2 A2 2 ' 2 2 sin 2 2 4m 2 p 2 1

2

Note: If m1 = m2 then 1 + 2 < /2 (unlike classical mechanics)