Module II: Relativity and Electrodynamics

Lecture 1: From electrodynamics to Special Relativity

Amol Dighe TIFR, Mumbai

Outline

Faradays law and Lorentz force

Motivations for Special Relativity

Lorentz transformations

Coming up...

Faradays law and Lorentz force

Motivations for Special Relativity

Lorentz transformations

Faradays law of induced emf

Maxwells equations without external sources
E=0 B=0 E = B/ t B = 0 0 E/ t (1)

Faradays law
E = B/ t E d = / t (2)

The total emf induced in a closed loop is equal to the rate of change of magnetic ux through the loop. Is this valid in all situations ?

Wire loop cutting through magnetic eld lines

In the rst case, the magnetic ux through the loop clearly changes, so emf is induced in the wire loop, and current ows. In the second case, the wire cuts through the magnetic lines of force, but the total magnetic ux through the loop never changes. But current still ows ! (momentarily, when the loop starts moving, till the charge buildup cancels the EMF) In the third case, both the source of magnetic ux and the wire loop move together. The ux lines are not cut by the wire. But current still ows ! (and now, continues to ow) Maxwells equations must be incomplete then...

Lorentz force
Force on charge q in the presence of E and B
F = q ( E + v B) This is an experimental result. It may be interpreted as an effective electric eld: E =E+vB. This relation is not contained in the Maxwells equations (at this stage) The combination of two relations E = B/ t , and E =E+vB (5) (4) (3)

is sufcient to take care of all situations. (See Feynman lectures for a detailed discussion).

Faraday Disc: a problem

Bar magnet along the axis of the conducting disc Conducting loop as shown in the gure

Does the current ow through the loop when:

The magnet is stationary and the disc is spinning ? The disc is stationary and the magnet is spinning about its axis ? Both the disc and the magnet are spinning with the same angular speed ?

Coming up...

Faradays law and Lorentz force

Motivations for Special Relativity

Lorentz transformations

Motivations from electrodynamics

The Lorentz force law seems to be valid whether the conductor is moving in a magnetic eld, or a magnet is moving near a conductor, only the relative speeds count. However with Galilean transformations x = x vt and t = t , the wave equation for a scalar potential (x , t ), 1 2 2 2 2 2 dx c dt (x , t ) = 0

does not stay invariant. So an EM wave in one frame is not an EM wave in another frame ? Indeed, the wave equation in vacuum for E is 2 E + 0 0 ( 2 E/ t 2 ) = 0 , leads to a wave travelling with speed c = 1/ 0 0 , which does not depend on the speed of the medium.

Motivations from measurements of the speed of light

Michaelson-Morley experiment (1887):

Speed of light is independent of the speed of the medium through which light is travelling. (Simulation applet at http://galileoandeinstein.physics.virginia.edu/more_stuff/ ashlets/mmexpt6.htm ) Maxwells equations also have been telling us the same thing !

An historical perspective
After the Michaelson-Morley experiment, FitzGerald wrote a terse paper in [The Ether and the earths atmosphere, Science 13: 390 (1889)], postulating that lengths may be contracted along the direction of movement through ether. In 1892, Lorentz wrote a more quantitative paper [In Dutch: De relatieve beweging van de aarde en den aether, Amsterdam, Zittingsverlag Akad. v. Wet., 1, p. 74 (1892)], in which he calculated that contraction by the factor 1 v 2 /(2c 2 ), would explain the MM experiment. This is the Lorentz FitzGerald contraction. Einsteins celebrated paper ["Zur Elektrodynamik bewegter Krper", Annalen der Physik 322 (10), 1905], used the consistency of Maxwells equations as his main motivation to come up with the revolutionary concept that space and time are unied. [Sahas translation: The Principle of Relativity: Original Papers by A. Einstein and H. Minkowski, University of Calcutta, 1920, pp. 1-34]

Coming up...

Faradays law and Lorentz force

Motivations for Special Relativity

Lorentz transformations

Linear transformations of space and time

A transformation of coordinates (x , y , z , ct ) (x , y , z , ct ) should be linear, since it should not depend on where the origin of the coordinate system is. From another point of view, the innitesimal transformation (dx , dy , dz , c dt ) (dx , dy , dz , c dt ) has to be linear since the quadratic terms will vanish in the innitesimal limit. For relative velocity v along x direction, one does not expect y and z to change, thus y = y and z = z . For the other coordinates, one takes the most general transformation to be x = ax + bct , ct = px + qct . (6)

Let a beam of light starting at (x , y , z ) = (0, 0, 0) at t = 0 reach (x , y , z ) at t . Then (ct )2 x 2 y 2 z 2 = 0. In the primed frame, this light has started at the origin at t = 0, and has reached (x , y , z ) at t . Then (ct )2 x 2 y 2 z 2 = 0. Combining these two gives (p2 a2 + 1)x 2 + (q 2 b2 1)c 2 t 2 2(pq ab)c t x = 0 . (7)

Deriving Lorentz transformations

Since the above expression has to be valid for alll x and t , all the three terms in Eq. 7 have to vanish individually. This allows us to substitute p = sinh 1 , q = cosh 2 , a = cosh 1 , b = sinh 2 ,

to take care of the rst two terms. The third term then yields sinh(1 2 ) = 0, i.e. 1 = 2 = . Thus, p = b = sinh , q = a = cosh .

The coordinates then transform as ct cosh sinh x sinh cosh = y 0 0 z 0 0

0 0 ct x 0 0 . 1 0 y 0 1 z

(8)

Lorentz transformations in terms of and

Let S be a frame moving with a speed v along x -axis, relative to an inertial frame S . The motion of the origin of S, as seen in S, is x = vt . The transformations obtained earlier yield the coordinates of the origin of S, as seen in S, to be x = sinh ct , and ct = cosh ct . Thus, tanh = v /c = . This gives cosh = 1 1 2 =, sinh = 1 2 = . (9)

The net Lorentz transformations then are ct 0 0 ct x x 0 0 = . y 0 0 1 0 y z 0 0 0 1 z

(10)

Some remarks on Lorentz transformations

The inverse Lorentz transformations are ct 0 0 ct x 0 0 x = y 0 0 1 0 y z z 0 0 0 1 . (11)

Clearly, these are obtained simply by changing the sign of v or . We have only talked about Lorentz transformations with boosts along x direction. However using the isotropy of space within each reference frame, we can always rotate the frames such that the relative motion between them is along the x axis. Formally speaking, 1 0 0 0 0 0 1 0 0 0 0 0 0 0 (12) = 0 0 R1 0 1 0 0 R2 0 0 0 0 1 0 where R1 and R2 are 3 3 rotation matrices.

Take-home message from this lecture

Even without the actual measurement of the speed of light, the invariance of Maxwells equations indicates that the speed of light should be independent of the motion of the medium. Given the speed of light is the same in all frames, there is a unique set of linear transformations that achieves this, which are Lorentz transformations.