1. Jul 27, 2006

yogi

A long spaceship with a clock C1 at the front and a clock C2 at the rear is moving at a uniform velocity wrt to earth after being launched from the earth. Its motion wrt to earth is tangent to a radial vector perpendicular to the earths surface. We now synchronize C! and C2 by finding the midpoint P of the spaceship and from P we emit a light flash. When the flash reaches C1 we set it to zero and when it reaches C2 we set it to zero - since the distance from P to C1 is the same as the distance from P to C2, the two clocks are now synchronized in the moving frame of the spaceship.

We next rotate the spaceship 180 degrees so that C1 is at the rear and C2 is at the front - we can do this ether by rotating the entire vehicle, or by exchanging them along a linear line, or by initally having placed them on a turntable - we again send a flash from the midpoint P and discover that it arrives at C1 before C2.

By measuring the amount of de-synchronization, we calculate the velocity of the spaceship wrt to space!:

Somethings obviously wrong - but what

2. Jul 27, 2006

Garth

Why?

Garth

3. Jul 27, 2006

HallsofIvy

Am I missing something here? Since the two clocks are equidistant from P in the spaceships frame of reference and the light flash is sent from P, again in the spaceships frame of reference, why should the flash of light arrive at one before the other?

4. Jul 27, 2006

Jorrie

I think what is wrong is your assumption that you will find the clocks de-synchronized after the rotation. You stated it as if light moves in an aether with speed dependant upon direction relative to the ship.

5. Jul 28, 2006

yogi

Well actually light does move at a constant speed wrt to space - and it is measured as isotropic relative to any inertial frame - but the fallacy, as noted - is nothing changes when the clocks are reversed.

6. Aug 1, 2006

baryon

Considering Michelson-Morley are we?

7. Aug 1, 2006

clj4

Actually this is a basis of a class of experiments that attempt to measure light speed anisotropy. But the theoretical underpinning is NOT SR but rather the Mansouri-Sexl test theory. In that theory, light speed is assumed to be anisotropic and iit predicts a change in synchronization.

Correct, you have rederived one such experiment. What happens in reality is that no desynchronization gets detected (as expected) , therefore there is no way of detecting the speed of the ship wrt space (actually the MS theory uses a preferential frame wrt which light speed IS isotropic). Generally CMBR is chosen as such frame. Either way, SR predicts no desynchro, the experiment (if run) will detect no desynchro, so everything is cool.

8. Oct 16, 2011

Petyab

This experiment depends on the magnetism present at the two locations. In one location an alternating electric field, example at the bridge, could cause the hands on the clock to move slightly slower or faster than the hands on the clock of the other and vice versa. This is related to the net torque of the electric field. Say the electric field is acting with a net force of something like 1x10^-30J. This force will roughly cause an acceleration of an electron to be 9.7181729*10^-12m/s^2. A negligible force. For nuclear clocks however this is enough to cause a difference.