I have a friend who launches the occasional cube sat; I am personally a little on the sceptical side of time dilation so it seemed like an interesting opportunity to try an experiment; please note that I'm not a physicist and so it's very likely that I'm missing something!
To the best of my knowledge, no "all optical" clock (one which uses no oscillating atoms or electronics in the signal creation process) has been created and put through sustained acceleration in order to directly test the theory of time dilation - they all appear to involve control electronics or atoms which may be distorting the results.
(Rough!) Optical clock design - a feedback loop using optical interference
This clock uses interference in order to make an optical pulse train (a clock signal) at point A. It can be followed through like this:
- The laser light passes through the splitter
- It bounces around the mirrors
- It interferes with itself at the splitter and cancels out
- The dark beam goes 'around' the mirrors
- It's no longer interfering with itself; the output brightens again
- Repeat. The signal oscillates between bright and dark at a frequency determined by the distance around the mirrors
Next, place this on a small satellite such that signal A can be read from Earth with minimal additional timing electronics on board. Compare the tick rate with a clock on Earth.
Hypothesis: Maybe atomic clocks are flawed. Atoms are expected to increase in mass at high speeds so this could be changing their oscillation period or affects decay length.
From the point of view of the satellite, the ticking occurs normally. From the point of view of Earth, the ticks would gradually slow down as satellite speed increases.
A counter theory
The Universe entirely ignores atom velocity when a photon is emitted from it; c is still constant. From either Earth or the satellite, the ticks would continue at the same rate but would reach a point at which the clock breaks down when the mirrors fall out of alignment. At extreme speeds, the first mirror has essentially moved out of the way before the light could bounce off it. As a side note, colour shifting does still appear to hold.
The important distinction is that at LEO speeds there is a measurable difference.
So, the question itself: Would such an experiment be worthwhile? What am I missing?