I also wanted to draw your attention to an interesting problem related to the controversial EM drive, and NASA experimental results. Actually, as you know, there are experiments performed all over the world in this direction, including in UK, and in China (the Chinese recently reported positive results in space, the existence of thrust ).
In Scientific American, August 1, 2009, there is a nice article written by Eduardo Gueron ,
"Surprises from General Relativity : 'Swimming' in Spacetime ".
https://www.scientificamerican.com/article/surprises-from-general-relativity/#Basically, in curved spacetime, the procedure used to determine the center of mass of a system of particles (or a system characterized by a certain mass distribution ) can be ambiguous, leading to different results for different procedures (and this does not happen in flat spacetime ). That means that a physical system with a variable mass distribution of its components, can actually travel through a curved spacetime. There is no violation of the conservation of momentum. Even if there is a displacement of the center of mass of the system, due to a sequence of maneuvers, there is no continuing momentum associated to the system. At the end of the sequence, the system is stationary, as before the sequence started. It is essential that the inhomogeneities of the gravitational field and the size of the system should have the same order of magnitude.
What does this have to do with the EM drive?
We consider the equivalence between mass and energy. Gravity is not only sensitive to variable mass distributions, but also to variable energy density regions. This is also related to the problem of the indefiniteness in the location of the electromagnetic field energy (this is also a problem emphasized by Feynman). Basically, we make an analogy here. The variable distribution of the regions of high energy density of the electromagnetic field in a resonant cavity is equivalent to a variable mass distribution of the system. In the tronconic EM resonator, the variable distribution of electromagnetic energy density regions inside the resonant cavity is essential. For each sequence of distributions, the displacement of the center of mass of the system will be very small, but a rapid succession of such sequences will lead to a measurable effect.
For a cylindrical resonator, a nice numerical simulation can be found at the following link:
http://demonstrations.wolfram.com/CylindricalCavityResonator/The calculations have to be done for the EM drive tronconic resonator though . So, instead of a variable mass distribution of its components, the system will have a variable distribution of the electromagnetic energy density regions inside the tronconic resonant cavity. In curved spacetime, if properly designed, such a system will be able to travel through the curved spacetime. A starting reference would be:
http://dspace.mit.edu/bitstream/handle/1721.1/6706/AIM-2002-017.pdf?sequence=2I emphasize that nothing in this model assumes any new physics, but the original references must be studied for a solid mathematical model, not the popularization references mentioned above. Once the intuition is clear, there is an open road for improving the design, using this model.
We reach the following conclusions. Shawyer (the original inventor of the system) has suggested that thrust is caused by radiation pressure imbalance between the two faces of the cavity. This explanations was assessed as incorrect by a number of well established physicists. Other explanations require new physics, and imply phenomena that have not been observed experimentally. The system will work only in curved spacetime. Gravity Probe B showed that the curvature of spacetime in Earth's vicinity is small but measurable. I suspect that there will be a measurable effect in orbit, close to Earth. This system cannot be used for interstellar travel. There is no constant force acting on the system (and no acceleration), just a measurable displacement of the center of mass. Also, the curvature of spacetime in interstellar space within our galaxy is probably too small, in order to make the system feasible.
This would be an interesting experiment though, related to the location of the energy of the electromagnetic field, very close the Feynman's suggestion.