Declan Traill BSc
Interferometer Experiment trying to detect Anisotropy in the speed of light through the Earth's reference frame.
1310nm DFB Laser source
Optics box with NIR camera and USB cable
Optics Box - opened up, showing optics elements
NIR camera, NPBS cube, orthogonal PM fibers and X/Z Translation tables for alignment
Theoretical Calculations for the expected fringe shift with rotation angle of the interferometer.
Orthogonal interferometer arms - wooden structure
Interferometer arms, each wound with 15 turns (10m) of PM optical fiber.
Interferometer with inline 1x2 coupler and polarizer connected to PM fiber arms.
The full interferometer setup mounted on a rotating table, with laser source and battery connected and fibers from orthogonal arms connected to the closed optics box.
The first detection results imaged by the NIR camera. The fringes move in one direction, then pause and move in the other direction, as the orthogonal arms are rotated slowly in one direction, then pause and rotate in the other direction. The fact that the interference fringes move at all during rotation indicates that there is an anisotropy in the speed of light in the Earth's reference frame (that is moving through the Aether at ~486km/sec).
Correction!! : These initial result turned out to be in error as it was discovered that there was mechanical strain on two optic fiber connectors whilst rotating the interferometer. This caused a much greater effect (more fringe shifts) than was expected or modeled (as you can see in the above videos). The following video (below) was made after securing the two fiber connectors to prevent the mechanical strain. The number of fringe shifts during rotation decreased significantly after this change.
Expected fringe shift (Y-axis) over a full rotation (angle on X-axis)
The second detection results imaged by the NIR camera. The fringes move in one direction, then pause and move in the other direction, as the orthogonal arms are rotated slowly over about 100 degrees in one direction. The modeling shows there there is an expected fringe shift of around 2 fringes in each direction over this rotation interval. The video appears to show this. The fact that the interference fringes move at all during rotation indicates that there is an anisotropy in the speed of light in the Earth's reference frame (that has been found by other experiments to be moving through the Aether at ~486km/sec).
Expected fringe shift (Y-axis) over the approximate range of angles (225° to 325°) that the interferometer was turned during the above video clip. The modeling has been done for an Aether wind speed of 420km/sec which seems to match the observations here, better than the 486km/sec number (which would have a peak number of fringes of around 3 rather than 2).
Caution!! : These secondary results may be in error as there are still fringe shifts occurring due to movement of the rotating table and USB cable connected to the optical box, so it is difficult to prevent unwanted fringe shifts due to these factors and see just the effect due to rotational motion.
I took some further measures to reduce unwanted movement of the interferometer and cables (which can cause interference fringe shifts) and performed another run of the experiment, this time rotating the interferometer from an angle of 270° through to 360°.
The following graph shows the expected fringe shift, from the model, over this range of angles, if the optic-fibre-mode interferometer is subject to differing light travel times in the two orthogonal arms of the interferometer (as a gas-mode interferometer is). Also shown is the captured video clip of the interference fringe during this rotation. It is clear that NO fringe shift occurred due to the rotation.
Small Title
The modelling for this latest run of the rotating optic fiber interferometer experiment.
It is quite possible that there is no fringe shift due to rotation, despite the modeling indicating that there should be. This may be due to an effect that only occurs in single-mode optical fibers and would normally result in a fringe shift in a gas-mode interferometer. Please read evidence of this effect being seen in a similar experiment, here. See section 4.1 on page 16:
A New Light-Speed Anisotropy Experiment: Absolute Motion
and Gravitational Waves Detected
Reginald T. Cahill
Additional Note:
In light of my new findings about light propagation in various types of interferometers and cable types, I have fully modeled Cahill's "Flinders University Gravity Wave Detector" (see pages 15-24 in his paper. The link to the paper is above this box). I find that the maximum time difference that his detector would be capable of recording, for an Aether wind speed of 486km/sec would be ~0.028 ps (Picoseconds), which is a lot less that the reported change of 55 ps. What is more, this maximum time difference is the difference between the SUM of the two cable times, between a North/South and East/West orientation of the detector, not simply the difference in times between the two cable arms. The simple difference between cable arms is always exactly zero! Cahill seems to assume that because there is no difference in the travel times in the orthogonal arms of a single-mode optic fiber interferometer, that the travel times in opposite directions in such a fiber would be the same (i.e. the light would travel at c/n isotropically) - but this is invalid, as light signals are known to exhibit Fresnel Dragging and Sagnac type effects in single-mode optic fibers, so the one-way-speed-of-light times in such fiber would be different in different directions. I think Cahill's mathematical model is incorrect. I have posted a link to my modeling of his detector, incorporating everything that I have learned, on the right side of this box.
My accurate modelling of Cahill's Flinders University Gravity Wave Detector when orientated in a North/South configuration (aligned with the Aether wind)..
My accurate modelling of Cahill's Flinders University Gravity Wave Detector when in its circular calibration configuration.
Michelson-Morley and Miller (Mt Wilson) accurate modelling.
This is a diagram the depicts how the light signal travels in the interferometer arm that is perpendicular to the direction of motion through the Aether. This is applicable for gas-mode interferometers such as that used by Michelson-Morley and Miller (Mt Wilson) - (see the link to here to my paper that fully explains their results by my model) - but seems not to be the case for single-mode fiber interferometers such as used in this experiment and that conducted by Cahill his paper titled 'A New Light-Speed Anisotropy Experiment: Absolute Motion and Gravitational Waves Detected" (see link to the paper above).
Single-mode optic fiber interferometer modelling with new equation for L1 distance.
Note: The reason why light behaves differently in single-mode optical fiber than it does in a gas-mode interferometer (or in a coaxial cable where light speed anisotropy can also be observed) is that in single-mode fibers the light is confined to a single mode of transmission and cannot move laterally within the fiber. The transmission of light progresses along the fiber by the normal absorption/re-emission process by the glass molecules and travels at speed c in the space (vacuum) between them. The process of absorption/re-emission slightly slows the speed of the light's propagation through the fiber. This results in an overall speed through the fiber of c/n when the fiber is stationary. It is known to be the case that this is the method of the light's propagation, as single-mode optical fibers do exhibit Fresnel Dragging and longitudinal measures of the one-way-speed of light through them have revealed a transmission time that is dependent on the orientation of the fiber. However, the light is constrained from moving sideways whilst traveling through the space between molecules of the glass in the optical fiber - so it cannot follow the exact same 'staircase' pattern of propagation as I showed in the diagram above (with the L1 and L path lengths shown). It is a design requirement that light cannot deviate sideways in single-mode fiber so that there is as little dispersion as possible of the optical signal as it travels down the fiber, so it should be no surprise that it continues to work as intended when the fiber is moving rapidly (in a perpendicular orientation to the direction of motion) through space. The shape of the 'staircase' path is altered by the interface between the core and cladding in the single-mode optical fiber. This interface constantly redirects the signal back along the fiber. So, the light is forced to continue to travel at c/n through the fiber such that no travel time difference between orthogonal interferometer arms is observed and the anisotropy in the speed of light in the Earth's reference frame cannot be measured by this means.
Analysis:
Why are light travel times down orthogonal arms of a single-mode optic fiber interferometer the same, despite the Aether wind passing through the interferometer's frame - what is the mechanism that explains this?
Light travels at c with respect to the Aether’s reference frame (when in the vacuum between molecules) rather than at c relative to the moving interferometer (and single-mode fibers).
Whatever the cause, the apparent effect is that over the additional distance that light travels through the vacuum (when the interferometer is moving through the Aether, rather than stationary) it seems to travel at c/n rather than c.
What are the possibilities? :
(1) Light encounters more glass molecules when traveling down the perpendicular arm of the interferometer when it is moving at speed v through the Aether, than it does when the interferometer is stationary in the Aether frame.
- How can there be additional molecules along the light’s path when the fiber is moving? The light passes down the same length of fiber when it is moving or stationary, so the number of glass molecules encountered during its journal – thus, the total delay by those molecules – would be the same.
So, this possibility is DISCOUNTED.
(2) Light is slowed for longer by each glass molecule in the perpendicular arm of the interferometer when it is moving at speed v through the Aether, than it does when the interferometer is stationary in the Aether frame.
- Maybe the process that delays the light (whilst momentarily absorbed by the glass molecules) in single-mode optic fibers is different than in normal bulk glass or water, for example, and this process changes when the single-mode fiber is in motion? If this was so, then why wouldn’t it be slower in both arms of the interferometer, rather than just in the perpendicular one? There is, of course, a slight slowing of the propagation process due to motion – the Relativistic time dilation given by the Lorentz factor – but this amount is already accounted for in the model, effects both arms equally and has the wrong magnitude to account for the effect anyhow. There is no obvious reason why the glass molecules would oscillate for longer before re-emitting the light waves during the light’s propagation down the fiber, but only when the fiber is oriented perpendicular to the direction of motion of the interferometer through the Aether.
So, this possibility is POSSIBLE but UNLIKELY and would need explaining how.
(3) Light travels at a slower speed than c through the space (vacuum) between molecules in the perpendicular arm of the interferometer when it is moving at speed v through the Aether, than it does when the interferometer is stationary in the Aether frame. This could only occur if the Permittivity/Permeability of the space between molecules was different when the fibre is moving through the Aether, but only when the fiber is oriented perpendicular to its direction of motion through the Aether. There is no reason why this should be the case.
- There is no reason why light would not be traveling at c in the space between molecules.
So, this possibility is DISCOUNTED.
(4) Light travels a further distance through the vacuum between molecules in the perpendicular arm of the interferometer when it is moving at speed v through the Aether, than it does when the interferometer is stationary in the Aether frame.
- The Huygens wavelets (the relative timing of the spatially distributed wavelets) that cause the angled propagation direction of the wavefront of the light when it is traveling down the fiber that is perpendicular to the Aether wind direction, would have to be altered by something in the single-mode fiber, causing the propagation angle of the light into the vacuum between glass molecules to be slightly different. This is a possibility, as the wavelets that form the beam that propagates into the space (vacuum) between glass molecules originate from two different refractive index glass materials (the core and the cladding of the fiber) and so the light would propagate at slightly different speeds through each of the two optical materials – thus, forming the light wavefront at a slightly different angle than normally would be the case for a single refractive index material. However, if the light propagates through the vacuum between glass molecules at an angle that is different than for a vacuum-mode interferometer, then the light's direction would not match the motion of the fiber through the same space. So, the light signal would be angled into the cladding of the fiber, rather than directly along the fiber. This may not matter, though, as the core/cladding interface will bend the signal back into the correct direction (i.e. along the fiber) – that is how the two layered fiber (each layer with different refractive indices) works. If that angle is too great, however, the signal would get extinguished and would stop propagating down the fiber.
So, this possibility is POSSIBLE but would need explaining how.
It seem that possibility (4) is the most likely, but needs further work to develop the idea.
Proof of Concept:
I have written a computer model to simulate the Huygens wavelets that propagate through the glass material of the core and cladding in a single-mode optic fiber. The light is modeled propagating along the fiber from bottom of the screen upwards. The Aether wind is modeled as moving from left to right. These three animations demonstrate the difference between:
(1) No difference in refractive index between the Core and the Cladding glass material.
(2) A difference of 20%.
(3) A difference of 40%.
In all cases the Aether wind is modeled as being at 30% of the speed of light. In reality, its speed is about 0.16%.
Note: these simulations are deliberately extreme in order to clearly show that there is an effect on the direction of the resulting wave-front due to the difference in refractive indices. This demonstrates that the suggested mechanism (discussed above) is a possible explanation for the results obtained from a single-mode fiber interferometer.
On the right of each of the animations from the model is a still image of the final result from the modeling, with the direction of the resulting wave-front highlighted with yellow lines.
A .zip file containing the PC .exe file of this simulation is available here (link below) to download, if you would like to try the model for yourself...
Simplifying the model by modelling just a single pulse (set of wavelets) starting at the bottom of the screen, but adding in the row of glass molecules at the interface between the core & cladding in the optical fiber (where the refractive index changes) gives the following results. When the light from the initial Huygens wavelets reach the molecules at the interface between the core/cladding of the fiber, new Huygens wavelets are initiated and the pattern of all the Huygens wavelets emerges as time evolves. The change in the angle of the resulting wave-front is clearly seen as the refractive index difference between core and cladding material increases.
The Aether wind is again being modeled here as 30% of the sped of light, and the refractive index ratio between the core and cladding ranges from 1:1 in the first results and 1.5:1 and 2:1 in the next two.
Here is the latest, improved version of the Huygens model with a number of controls that can be changed by the user.
