# Light Clock Experiment Flawed?

1. Dec 30, 2006

### swerdna

Hi - I’m new here and I'm probably a bit of an impostor in that I’m not a physicist, scientist or even academic. I’ve been trying to increase my understanding of Relativity and often see a light clock experiment that is used to demonstrate time dilation. Given the extent that this particular experiment is used, I expect it has been thoroughly analysed and validated. Unfortunately, I can see an obvious flaw in the experiment. Am I wrong, or am I right?

The experiment:

First scenario -There are two identical, synchronised light clocks, A and B. In each clock a blip of light bounces vertically between two horizontal mirrors. The clocks record a second every time their blip hits a mirror. When both clocks are stationary, both blips hit the mirrors at exactly the same time and the clocks show the same time to a stationary observer (I know that stationary means uniform motion).

Second scenario - Clock A and the observer remain stationary but clock B moves in a horizontal direction. The light blip in clock B now has to travel both horitzontaly as well as vertically and has to cover a greater distance to hit each mirror. Clock B therefore now runs slower than clock A.

The flaw?

The blip is light, and light wouldn’t inherit the velocity of the moving mirrors of clock B in the second scenario. It would want to continue bouncing vertically in a stationary position. For example, if the horizontal mirrors were spinning discs in a stationary clock, the blip would still bounce up and down vertically. In the Second scenario the blip would only move in the horizontal direction of the moving mirrors if it was fired from whatever created it at an angle from the vertical. This means the clocks are no longer identical. Effectively, the mirrors are now further apart in clock B than they are in clock A. To look at it in another way, if the blip is fired at an angle from the vertical, the mirrors don’t even need to be moving. A series of stationary mirrors could be positioned to bounce the zig-zagging blip. In fact, two long stationary mirrors could be used for both clocks. All this experiment is proving is that a direct route is shorter, and therefore quicker, than an indirect route. Don’t see where Relativity comes in to it.

I hope I have explained things clearly enough that you can understand what I mean.

2. Dec 30, 2006

### bernhard.rothenstein

light clock

3. Dec 30, 2006

### Galileo

Swerdna, you worry too much about the technical aspects of the thought experiment. say the clock is in a train. The light in the clock is bouncing up and down as seen from someone inside the train. It was started that way, ok? Then if it goes up and down in the train it HAS to go zigzag wise as seen from someone on the ground.

I hope I understood your argument correctly. You do not set the light clock in motion, for then I can imagine the light continuing up and down. You are looking at the same clock from two different frames.

4. Dec 30, 2006

### JesseM

If you have a device which emits light in a single direction, like a flashlight or a laser, then when that device is moving relative to us, it will continue to emit the beam in the same direction in the rest frame of the device, which means the beam will appear to be emitted at a different angle in our frame. For example, if you have a flashlight that is pointed straight up, and it's moving on a train at close to light speed relative to us, then we'll see the beam being almost diagonal rather than straight up, although it will still point straight up in the frame of an observer in the train. This can be understood as a consequence of the fact that the laws of physics work the same way in every inertial reference frame--the observers sitting on the train shouldn't be able to tell their speed relative to the earth by looking at which direction the light beam from a flashlight pointed up comes out, the beam should always come out at the same angle as long as the train is moving inertially.

Last edited: Dec 30, 2006
5. Dec 30, 2006

### Janus

Staff Emeritus
The problem is that you are assuming that there is some absolute reference system by which the light can judge its motion. This is in violation of the postulates that the experiment is based on:

1. The laws of Physics are the same for all inertial frames.
2. The speed of light is invarient for all inertial frames.

It is the first one that is givng you trouble. In essence, it says that there is no experiment that you can perform that will tell you if you are moving at a constant velocity.
This is true of the light clock itself. If you are an observer alongside clock B, if what you assumed is true, then the light from clock B would travel at an angle from your perspective. (even though you aim the blip to hit a mirror exactly across from the light source, it would miss.) IOW, you could tell how fast you were moving by measuring the difference between where you aimed the light and where it actually hit. This violates the first postulate above. For the postulate to hold, the light must travel directly across from your perspective, just as the light from clock A does fro an observer next to it.

Put another way, you are assuming that light behaves in one way, while the experiment postulates that it behaves in another.

How does it really behave? To find out, we do experiments. The light clock example is just a thought experiment, but it leads to predictions that can be tested by real experiments. To date, every experiment performed validates the light clock experiment postulates.

6. Dec 30, 2006

### swerdna

Thanks for your answers. I see what you mean that clock B was not accelerated as a running clock from the rest frame of clock A, but it was set running when it had the rest frame of the horizontal motion (on the train). Given there’s no fixed point in the universe that’s a reference for actual “stationary”, the clocks must be able to operate correctly in any rest frame. So light travels at c regardless of the motion of the object that emitted it, but it somehow has the uniform motion of that object. That’s an interesting thing to get my head around, but I can’t dispute it so it must be so.

Given this however, I still don’t understand why it is, in the given light clock experiments, that the blip of clock B, that’s travelling diagonally between the vertical and horizontal, is shown to have the same speed on the as that of clock A on the vertical. In other words, it’s effectively shown that it was fired on the diagonal (which it wouldn‘t have been). EUREKA! I’ve just got it! I’ve just suggested that the speed of the light blip on the diagonal should be going faster than the speed of light (DOH!). If we were talking about some magic bouncing ball I would be correct, and the speed of the ball on the diagonal would have a combination of the diagonal and horizontal speeds. The thing I missed was that light is already at the legal speed limit. Thanks for helping me sort this out. These eureka moments are always a mixture of disappointment and elation. The elation is always the greater by far!

I still need to do some more thinking on the fact that light independently travels at c but adopts the uniform motion of whatever emitted it. Maybe this will help me to understand how the motion of the emitting object can cause light to become redshifted.

ETA - I'm very happy now - HAPPY NEW YEAR!

Last edited: Dec 30, 2006
7. Dec 31, 2006

### robphy

Does this animation http://www.phy.syr.edu/courses/modules/LIGHTCONE/LightClock/VisualizingProperTime-y-pair-A-with-photons.avi [Broken] address your question?

Last edited by a moderator: May 2, 2017
8. Dec 31, 2006

### MajorComplex

So I'm guessing the light clock experiment is to test time dialation with accelerating masses correct? Wouldn't this suggest that space can manipulate our perception or recording on time and that it doesn't directly manipulate the speed of time it's self?

Last edited: Dec 31, 2006
9. Dec 31, 2006

### swerdna

Thanks, I’ve viewed heaps of this type of animation. Even made quite a few of my own.

Last edited by a moderator: May 2, 2017
10. Dec 31, 2006

### swerdna

Seems to pose more questions than it answers doesn’t it. If the theory of curved space is correct, it would seem to follow that light is never actually moving in a straight line, so is it always going slower than its actual potential top speed? Perhaps if we could correct the curviture of space somehow, light would travel faster than c.

I also can't help wonder if this whole thing is more to do with perception anomalies (optical illusions) rather than that time is actually slowing down. Early days as yet for me.

11. Dec 31, 2006

### robphy

Have you seen or produced something like this [from the lower half of the webpage quoted]
time dilation (with length contraction and relativity of simultaneity):
http://physics.syr.edu/courses/modules/LIGHTCONE/LightClock/VisualizingProperTime-M-multi-no-mtrail-v=8-A.avi [Broken]
twin paradox (clock effect)
http://physics.syr.edu/courses/modules/LIGHTCONE/LightClock/VisualizingProperTime-TwinParadox-v=8-A.avi [Broken] ?

Last edited by a moderator: May 2, 2017
12. Dec 31, 2006

### swerdna

Thanks robphy - Yes had seen them, but they're always worth another look.

Last edited by a moderator: May 2, 2017
13. Dec 31, 2006

### swerdna

I’ve been giving some thought to the light clocks and don’t see how they can actually work when the light blip is moving both vertically and horizontally. Let’s slow things down a bit and use a bouncing ball in place of the light blip. The ball has the properties of light in that it always travels at a constant speed, and lets say that speed is 100 (can’t go faster or slower). When the clock has no horizontal motion the ball bounces vertically at 100. If the clock has a horizontal motion of 25, the ball can then only bounce vertically at 75 (total 100). For the ball to travel at the correct angle and speed to intersect with the second mirror, I believe it needs to be able to travel at a combination of the vertical and horizontal speeds (125). In other words, if the ball can’t travel faster then 100, it could miss hitting the second mirror. I have made and attached an animation to hopefully more clearly demonstrate this. Once again, I suspect there is something I’m viewing incorrectly.

#### Attached Files:

• ###### light-clock.gif
File size:
15.5 KB
Views:
174
14. Jan 1, 2007

### robphy

For one thing, velocities are vectors. The magnitude-of-the-sum of two velocities [i.e. the resultant speed] is not generally the sum-of-the-magnitudes of the velocities. You must use the Pythagorean theorem for perpendicular velocities. A horizontal velocity of 25 plus a vertical velocity of 75 yields a resultant of sqrt(25^2+75^2). Practically every derivation of the light clock discusses this.

15. Jan 1, 2007

### Janus

Staff Emeritus
To continue on what robphy has already stated.

Therefore, what you would do is start with the fact that you know that the ball's total velocity is 100 and it's horizontal velocity is 25 and then find the vertical velocity from that with

$$V_v = \sqrt{100^2-25^2} = 96.82$$

16. Jan 1, 2007

### swerdna

Thanks Robphy and Janus. Thought the answer might lie somewhere in the math. Perhaps I should expand on the first sentence of my OP. Not only am I not a physicist, scientist or even academic but I have also had a very poor education. Hated school and left aged 14 with no qualifications. Although I hated school I have always enjoyed learning. I’m mature chronologically but immature educationally. I love math at the level I understand it, but that level is well below where it needs to be to understand all this stuff. Unfortunately, my busy lifestyle doesn’t allow much spare time improve my math abilities, but this is obviously what I need to do.

Share this great discussion with others via Reddit, Google+, Twitter, or Facebook