A modification of the above existing idea.

I just had an idea that kind of merges EM Cavities, Dielectrics, and may possibly be related to the Woodward idea but not in an obvious way. At the moment it seems like a closed system, so far.

We start with a cavity that lets light in one way (cinnamon roll shape) and was suggested by some one in the early threads. I can not remember where, but am grateful for their suggestion of the cavity shape. I think I remember rfmwguy commenting on their cavity shape. The light should travel around one way in a circle.

The next idea is to fill half this cavity with a dielectric such that when the light enters the dielectric it slows down. I assumed momentum is conserved so to slow down the light I changed the effective mass of the light by adding "dm".

I then considered the circular path the light takes and assumed a force F = m*a = m*v^2/r which isn't relativistic but the photon doesn't approach infinite mass at light speed anyways. Maybe I am making the wrong assumption here. The force around the circular path appears to be different for the photon in the dielectric than for free space.

I found another paper called, "An Effective Photon Momentum in a Dielectric Medium:

A Relativistic Approach" which appears relatively new 2015. Department of Physics, Weber State University, Ogden, UT. I can't seem to find it on googlescholar but the link is here:

http://physics.weber.edu/galli/EffectivePhotonV2-1.PDF Their paper suggest the effective momentum P_eff = P/n where n is index of refraction so that unless I am mistaken they don't take momentum to be conserved. Maybe upon entry and exit of the dielectric some momentum is transferred. I am not certain about equation 10 where they say it reduces to Snell's law when u->0. With u->0 I think it reduces to sin(theta2)/cos(theta2)=sin(theta1)/[n*sqrt(1-sin(theta2)^2/n^2)] where (1-sin(theta)^2)=

**cos**(theta)^2, which is close but I don't think it is the same, unless I am mistaken. Regardless I think their effect may only enhance the Force difference in the idea I was contemplating (

**or come out the same possibly if momentum is transfered to the dielectric upon entry and back to the photon upon exit. - momentum is conserved.)**. Image attached below.

It is interesting it requires a cavity and it appears the Q of the cavity may enhance the effect. At the moment it looks like a closed system and possibly related to the Woodward idea based on the light changing in mass.

**If the photon changes in relativistic mass upon entering a dielectric then the parallel to the Woodward effect (called by another name by Woodward) then the force of changing the photons path after it changes mass is the parallel. **Edit: When I think about it, there must be some force from speeding up and slowing down of the photon when entering/exiting the dielectric that would work against the estimated thrust. There is the possibility it could cancel out any thrust altogether.