There was quite interesting discussion which was partially about time dilation in rotating non inertial system: http://www.sciforums.com/threads/time-dilation-gravity-force-vs-gravitational-potential.156181/ I was believing, that due to Equivalence principle the linear acceleration and also non inertial fictitious force do cause time dilation, but well educated members of this forum teached me, that there is practically no time dilation due to GR in rotating non inertial system, only dilation caused by speed which can be calculated with SR. Actually Im no professional physicist, so I do need not take mainstream interpretation of GR like god given script to avoid professional suicide and I like to see what experiments say on that matter. So I checked experiments and the outcome was quite interesting: There was experiment with muons, which seems to confirm that there is no time dilation due to centrifugal force, but only due to speed of particles: http://www.nature.com/nature/journal/v268/n5618/abs/268301a0.html Unfortunately, I couldnt find any free full text version of the paper I really do mistrust short intepretations of papers in popular texts. Also there is second experiment confirming it, so called Mossbauer rotor experiment. What is interesting, recently was this experiment repeated and they found a significant discrepancy compared to original result. And they interpreted this additional effect (behold) as proof of Equivalence principle and based some new theory on it. So this is not mainstream and this is why Im placing this thread into Alternative theories. Here are related papers: https://www.researchgate.net/public...onfirmation_of_the_corrected_result_by_Kundig https://www.researchgate.net/public...ossible_explanation_of_the_extra_energy_shift There was some kind of rebuttal to the published anomalous results: http://arxiv.org/abs/1602.04212 But to my surprise, the rebuttal did not dismissed the anomalous results, but rather tried to explain it by (mainstream) GR. But if I understand it right, it means, that there is time dilation caused by rotating system. I know, it is quite complex topic, but I wonder what are your views on it.

The muon lifetime measurements confirm Special Relativity. The Mossbauer rotor experiment is a tricky one. In the 60's it was assumed close to 1/2 v²/c² which would be the special relativistic Doppler effect, but that discounts that the clock on the rim is not in the same inertial frame. And the measurements were also assumed "close enough". If you look at figures 3 and 4 of https://www.researchgate.net/public...onfirmation_of_the_corrected_result_by_Kundig you will see that the experimental values came nowhere close to any of the lines, which demonstrates a systematic problem with the experiment or analysis. But the analysis in http://arxiv.org/abs/1502.04911 and http://arxiv.org/abs/1602.04212 seems both more persuasive and has been published in journals of good reputation that for this particular setup of a rigid rotator and clocks not at the same radius that 2/3 v²/c² is the value supported by General Relativity's general coordinate transform. See also section 2 of http://relativity.livingreviews.org/Articles/lrr-2003-1/ I have not vetted these calculations personally. Nor have I attempted to recover the GR analysis in a strictly SR basis.

As noted in the other thread, Clock Hypothesis - time dilation dependent on relative velocity not acceleration - has been confirmed to very high accuracy for both ultra-relativistic circular and linear accelerated 'clocks': https://en.wikipedia.org/wiki/Time_dilation_of_moving_particles#Clock_hypothesis https://en.wikipedia.org/wiki/Clock_hypothesis Hence the first article by C. Corda cited in #2: http://arxiv.org/abs/1502.04911 is imo misleading in at least two respects. The wording "...Using Einstein Equivalence Principle on the equivalence between the gravitational "force" and the pseudo-force experienced by an observer in a rotating frame of reference..." implies 'gravitational force' is at the heart of the 'new result' that k = 2/3 not 0.5 re low speed transverse Doppler expression ΔE/E = -kv²/c² as used there. And second that somehow GR has to be used to get and understand such result. Wrong on both counts. I gave link recently to two articles pointing out only SR need be used for analysing arbitrary accelerated motion. Again: http://www.springer.com/cda/content...369097-c1.pdf?SGWID=0-0-45-1113969-p173959770 http://math.ucr.edu/home/baez/physics/Relativity/SR/acceleration.html Regarding the (non)relevance of 'gravitational force' - a complete red herring. Except in the loose and indirect and experimental setup-specific sense that in the rotating frame, a 'straight' light beam i.e. gamma ray will appear to undergo a trasnsverse arcing motion in the plane transverse to rotation axis. Hence in order for synchronization between source and detector, allowance has to be made for such arcing under 'artificial g' experienced in the rotating frame. In the case that receiver is located anywhere on the rotation axis of circularly orbiting emitter, above factors are present and an emitted photon must be aimed slightly against being exactly transverse to the direction of instantaneous relative motion in order to 'hit' the detector. But that is just what transverse Doppler involves and in that situation of detector lying on the rotation axis, the usual low velocity transverse Doppler result necessarily holds and k = 0.5 regardless of centripetal acceleration, as it must. Without trying to work out the exact experimental configuration that Corda bases his k = 2/3 result on, such result must be owing to the fact that source and detector are not linked in the manner above described i.e. detector is not on the rotation axis. And this must be introducing a bias that adds in a small additional component of first-order i.e. longitudinal Doppler shift. The combined purely transverse Doppler + component of longitudinal Doppler is obviously then giving the net 2/3 'newly discovered GR effect'. Or you could opt for the overthrow of both SR and GR as 'usually understood' via a low velocity experiment - while somehow explaining how well the Clock Hypothesis has held up to experiments at ultra-relativistic energies and enormous proper accelerations. Good luck with that.

My last post was harking back to scenario in earlier thread with a simple source-target situation. A closer look at the experimental setup of fig.1 in http://arxiv.org/abs/1502.04911 (same as for newer version at http://arxiv.org/abs/1602.04212) and it's quite different to the case of just an orbiting emitter and stationary detector. Instead source is on the rotation axis and absorber is orbiting. Stationary detector then counts the change in detection rate of transmitted gamma-ray photons, as a function of changed Mossbauer absorption owing to orbital motion of absorber. This changes the particular details somewhat but net effect must still be entirely predictable by SR kinematics alone - and assuming no substantial effects from mechanical stresses. In lab frame absorber itself is time dilated according to k = 0.5 not 2/3. Hence by factor √(1-v²/c²) = 1/γ. But circular motion destroys symmetry of Lorentz transformation and consistency requires reciprocity not equality of relative time dilation as measured between both frames. Hence in absorber frame the source is net blueshifted by the inverse i.e. by factor γ. Which is inclusive of that since gamma rays from source are emitted radially in lab frame, in absorber frame the interception angle is slightly raked 'head on' by ~ tanθ ≈ θ ≈ v/c with v the peripheral speed. Hence a slight component of longitudinal blueshift thereby. One would expect Mossbauer absorption peek was maximum for zero relative motion between source and absorber, and close enough to a symmetric drop off for either a slight redshift (orbiting emitter & stationary absorber), or the reciprocal slight blueshift (orbiting absorber & stationary emitter). If that case held consistency must have a final lab frame determined time dilation according to k = 0.5 not 2/3. That it evidently is not in the given experiment(s) suggests a preliminary asymmetry in absorption curve when relative v = 0, hence an asymmetry in absorption change between the two possible regimes. Also, the earlier quoted experimental result was significantly lower at ~ k = .58, with estimated error bars way too small to account for the difference. All in all, things don't consistently add up to a supposed 'GR effect' or 'YARK effect' and poor experimental design has to be suspected. I'll back Clock Postulate as confirmed to far greater levels of confidence than such discordant results claiming logically impossible causes.