Monday, December 21, 2009

Mickelson-Morley experiment revisited

The famous Michelson-Morley experiment carried out for about century ago demonstrated that the velocity of light does not depend on the velocity of the source with respect to the receiver and killed the ether hypothesis. This could have led to the discovery of Special Relativity. Reality is not so logical however: actually Einstein ended up with his Special Relativity from the symmetries of Maxwell's equations. Amusingly, for hundred years later Samuli Pentikäinen told me about a Youtube video reporting a modern version of Michelson-Morley experiment by Martin Grusenick in which highly non-trivial results are obtained. If I were a "real" scientists enjoying monthly salary I would not of course pay a slightest attention to this kind of stuff. But I am not a "real" scientists as many of my colleagues are happy to testify (without "":s of course) and have therefore nothing to loose. This gives me the luxury of thinking and I can even try to understand what is involved assuming that the discovery is real.

To my best knowledge there is no written document about the experiment of Martin Grusenick in web but the Youtube video is excellent. The only detail, which might give a reason to suspect that fraud might be in question is when Grusenick states that the mirror used to magnify and reflect the interference pattern to a plywood screen is planar: from the geometry of the arrangement it must be concave and I have the strong impression that this is just a linguistic lapsus.

Addition: Andre K. told that in German video Grusenick uses the expression "einfach" (simple) instead of the counterpart of "planar". Also pronunciation might have led to wrong pattern recognition as I know from my own experience as a person talking very badly broken english;-).

Addition: The change of the distance between beam splitter and mirror in the vertical position might explain the observations in terms of existing physics. A simple estimate however shows that this effect is by a factor of order 10-3 too small. I am grateful for Samppa for suggesting the estimate.

The reader willing to learn in more detail how Michelson-Morley interferometer works can look very short video sketching how the interference pattern is created. This longer video describes in more detail the principles involved.

I do not bother to transform latex to html since a lot of formulas are involved and automatic translators do not work properly. Instead, I give a link to a pdf file representing the results of Grusenick and their analysis and interpretation in detail.

The results are following.

  1. The findings of Grusenick can be understood if the radial component grr of the metric of Earth at the Earth's surface deviates from Schwartschild metric by a factor 1+Δ, where Δ is of order Δ ≈10-4.

  2. If one requires that Gtt vanishes for the modification of Schwartshild metric, Δ(r) behaves as Δ(R) R/r outside Earth's surface in good approximation. If the gravitational fields of stars, say Sun, have similar radial component grr, the predicted effects on planetary orbits are significant only for elliptic orbits sufficiently near to the surface of the star.

  3. In General Relativity the presence of non-vanishing "pressure" terms Grr, Gθθ, Gφφ in Einstein tensor with a vanishing energy density are difficult to understand. In TGD framework these terms could be due to the sub-manifold constraint forcing the allowed space-time surfaces to be extremals of Kähler action with Einstein equations with energy momentum tensor of matter (not containing the contribution of Kähler action) being satisfied.

  4. The effect of the gravitational field of Sun on the interference pattern measured at Earth surface can be visible (fraction of order 10-2 about the effect of Earth itself) and the experiments indeed demonstrate a diurnal variation of the interference pattern.

  5. The extended Michelson-Morley interferometer could provide a new high precision tool to measure the behavior of grr as a function of the distance from the Earth and to test the proposed model.

Addition: I am grateful for Frank Pearce for informing me almost year later (20.11. 2010) that he has carried out the Grusenick experiment again. There is small movement of the interference pattern during vertical rotation but nothing comparable to that detected by Grusenick so that the effect is very probably an artifact. An excellent video Vertical Michelson Morley Interferometer Experiment 11 12 2010 about Pearce's version of the experiment can be found here.

For details and background see the chapter TGD and GRT of "Physics in Many-Sheeted Space-time".


Anonymous said...

I saw the Michelson-Morley youtube video by Sampo Pentikainen. Amazing video! Thanks Matti for the informations on your blog apart from the awesome video. I am a college sophomore with a dual major in Physics and Mathematics @ University of California, Santa Barbara. By the way, i came across these excellent physics flash cards. Its also a great initiative by the FunnelBrain team. Amazing!!!

Matti Pitkänen said...

I agree. This video is really amazing! As a matter fact, the experiment was done by Martin Grusenick, a young fellow from Germany. I learned about the video from Samppa.

Matti Pitkänen said...

A further comment. It is better to be as skeptic as possible. The changes of the distance are rather small: of order 20 microns. The distance between mirror and splitter in vertical position might shorten by the contraction due to the weight of the system. One should estimate this shortening to see whether it is same order of magnitude.

Anonymous said...

If it is the Coriolis effect, then it also would appear on the horizontal plane when the apparatus is rotated.

If the diffraction pattern changed, then light velocity has changed? What can act on light? Is the photon mass or massless? The solar wind is affected, but not light? Why the diurnal difference?

I can imagine this is similar to magnetic flux. The change in flux can result in the change in velocity dB=dv . Since rotating parallel to earth surface plane result in no change, but rotating in a plane perpendicular to earth surface resulting in cutting the flux lines and velocity change resulted. How is light treated differently then?

If this is the case, 'aether field' (as dark matter) = magnetic field = gravity field?

There are much in common with gravity, magnetism, electricity, sub atomic forces; they're all forces that act on matter. This is the idea of unification; the same field with different density level (hbar?) and field geometry (topology, spacetime sheats?).

Why is the aether concept still so actual?

The effects of gravity was hardly seen by NASA, so how can it be seen here in this small scale? It must be the magnetic field or sheats, interfering with light. Do light also create a magnetic force around the vectors? (Faradays cage?) Low power lasers are used for acupuncture use, for instance. How can that light interfere then? How are those cyclotrone frequencies created?


The experiment is repeted I see. With pictures.

Anonymous said...
The first attempt to detect the absolute medium were based on interferometry
(the famous experiment of Michelson and Morley [56,57]) but it produced a nil
result. Many were quick to interpret the nil result of the Michelson and Morley
experiment as evidence that the absolute continuum did not exist. Actually,
the only thing which the nil result proves is that there is a contraction of the
lengths in the direction of motion as suggested by Lorentz and FitzGerald.

Matti Pitkänen said...

This is reply to both comments of Anonymous.

Ether explanation is excluded. Nothing happens in plane parallel to Earth's surface. Since Earth rotates this situation should be encountered only twice during 24 hours unless the experimenter is at North or South pole(;-).

General Relativity predicts this effect for Scwartschild metric but it is by a factor 10^-4 too small. I propose in the pdf file an explanation based on the assumption that the radial component of the metric outside Earth surface differs from the empty space value.

If energy density outside Earth is assumed to be zero the predicted deviation decreases like 1/r. It does not affect appreciably planetary orbits if present also for other astrophysical objects. For small objects used to measure G this effect could explain the surprisingly large 1 per cent variation in determinations of G.

Quantum classical correspondences suggests a physical interpretation for the deviation from the empty space metric. The non-vanishing Einstein tensor in regions containing matter in usual sense would describe the presence of virtual particle exchanges including graviton.

Matti Pitkänen said...

In TGD framework the effect can be also interpreted in terms of change of light velocity. This has nothing to do with ether hypothesis.

Gravitation makes space-time surface curved and makes the distance from A to B longer so that the time taken by photon travel from A to B is longer than along the geodesic of 8-D space. c decreases.

This allows to predict the observed slow decrease of light velocity in solar system as apparent effect and the rate of decrease is predicted correctly (see this). In the recent case the time to travel in radial direction would become longer.

One must however remain skeptic: it might be that the length variation is due to weight of the system contracting it in vertical direction.

André K. said...

About the problem with the planar/concave lens: I watched the video in German. There he says "einfache Linse" which just means "simple lens". Probably some confusion with planar-plane-simple here...


Matti Pitkänen said...

Thank you. This must be the explanation.

Anonymous said...

It took me a few days to see that Grusenick's shift was in fact due only to differential stress on the two mirrors, which reaches a peak at the two angles where he sees the shift change in direction (zero point). It is at those points that the major load shifts from one mirror to the other. At the points he asks why no zero point, look and note that both mirrors are perfectly symmetrical to the line perpendicular to the earth, thus there is no difference in stress between the two mirrors. The shift continues in the same direction because gravity is pulling both mirrors in the same direction, but the greater force swaps between the mirrors at the so called zero point and the shift in the pattern then changes direction. Very interesting experiment, though. I am no longer at all interested in building a device that really works, my interests shift faster than Grusenick's interferometer.

Matti Pitkänen said...

You might be right. I am not quite sure whether you speak about the same effect that I thought as a possible trivial explanation of the Grusenick's findings that is compression due to the own weight of the system. My rough estimate was considerably smaller than the effect.