...wait...
> The current design is calculated to provide a thrust of about .012 Newtons (equal to the weight of a paperclip, or a raisin) for one kilowatt of power, similar to the drives used to maneuver satellites.
It requires power input. Doesn't that mean this isn't a closed system?
Conservation of momentum only applies to mechanical processes, such as force and collisions. Converting energy breaks out of those bounds, and so it no longer applies - and the emdrive breaks no physics.
At least, that's how I've always understood it.
Not true for the EmDrive.
These relationships are derivable via Noether's theorem, one of the most foundational results in physics.
What am I missing here ?
Rockets, differently to balloons, do not depend on an interaction with a gravitic field for their operation. They work by ejecting reaction mass opposite to the flight direction and relying on Newton's third law of motion. Most of the time, the energy to eject the reaction mass is generated by combustion, but it doesn't really matter. Rockets can also be driven by the reaction mass's pressure inside its tank.
In general, any system can be turned into a closed system by including all sources of force in the description. Of course, the more forces are included in our model, the less useful it gets. Therefore, physicists and engineers must decide which forces can be ignored. Else, even experiments would become useless because without a model, the experience gained from experiments cannot be generalized.
And for the second question, mass is energy and energy is mass (as per relativity). Adding or removing energy from a system makes it respectively heavier or lighter; if you heat up an object so that afterwards it has exactly the same atoms but just more internal energy, then it will be slightly heavier than when it was cold - the difference is very small (m=E/c^2, and c^2 is large) but nonzero.
If you're looking from outside at the behavior of an isolated black box, it's impossible to detect if inside it some mass is "lost" in a nuclear reaction i.e. converted to energy, as both the inertial and gravitational mass of a system is equivalent to the sum of both rest mass and energy, and that total is conserved. If that box pushes out energy/mass to outside or gains energy/mass from outside, then of course it is not a closed system. But converting rest mass to energy and vice versa in a chemical or nuclear reaction would not make it a non-closed system as long as all the inputs and outputs stay in the system, as the total gravitational effect of the system before and after the reaction would be exactly the same.