> Unless they believe the machine breaks some of the fundamental physics laws, the device must emit something.

I presume you're referring to the third law of motion, but if so, you've mischaracterized it. Nothing in the laws of motion state that engines must "emit" something. My car drives along quite efficiently without relying on any emission of mass. So does a sailboat.

Of course, these are still third-law compliant. My car exerts force against the earth (slowing it slightly), the sailboat takes energy from the wind.

To be third-law compliant, this alleged "reactionless" drive would need to apply force against something. But ejecting matter or energy to gain moment in the opposite direction is only one way to do that.

The biggest red flag about this is that there's no physical explanation. It's true that an explanation would imply some new physics. But it wouldn't be breaking the existing physical laws of motion, it would just be discovering some new thing to apply force against (virtual particles, or dark matter/energy, or whatever).

Actually, the conservation of momentum is local, so the car must emit some particles. The bottom of the tires emits (virtual) photons that are absorbed by the road. When you see it globally, the tires just push the road. (You can't go to the local mechanical shop and talk about virtual photons.)

In the middle of the space, where there is only vacuum around, you can't use the "virtual particle that get absorbed almost immediately" trick. So you need to emit some real particles (like a photon).

To explain this experiment, you can't push against the virtual particles, because they collapse and create some real particle (like a photon) to carry the momentum away and then you get the momentum/energy upper bound. (It's more complicated, particles appear and disappear, but the net effect is the same.)

Most of the "explanations" use virtual particles, but as an alternative explanation it's possible that this is pushing "dark matter". Perhaps this is a neutrino fan. (I'm not sure if neutrinos are technically dark matter, but they are "dark" enough and very abundant). It catches some neutrinos from the sun, change their direction a little, and then the neutrinos exit the chamber unnoticed. The change in the neutrinos direction gives the momentum to the thruster. I think that this doesn't break any theoretical law in an obvious way. The problem is that the dark mater is "dark" because it's very difficult to measure and to interact in any way. It's very difficult to change the direction of neutrinos, but at least it's an engineers' problem, not a physicists' problem. (In particular, neutrinos can travel through the earth.) The easier explanation is a measurement error.

Like I said, there would have to be some breakthrough physics here, like what "vacuum" actually means, or how dark dark matter is, etc.

I'm just pointing out that there are a number of logically possible solutions that don't involve violating the conservation of momentum, which may or may not be physically possible.

I believe that the dp=dE/c requirement for thrust isn't valid with static fields: simply consider an electromagnet a large distance D away from a permanent magnet. For a very small energy cost (P<<F * c), you can generate a magnetic field to propel yourself forward. The opposing magnet won't move for D/c seconds, so you're transporting momentum at a much lower cost.* , which I believe is what lukev refers as "other logical possibilities" -- the car is another example of this.

Of course, in free space magnetic field decays as 1/D^3, so eventually you will be better off spending F * c power to generate a force F. I do not disagree with "The device must emit something" though, since obviously momentum has to be carried by some particle, although I'm not a physicist to attest the specifics precisely.

* - I believe in this case the energy cost is so low because you're "storing" energy in the field, energy you're spending acts more like a trigger. As a thought experiment consider a stream of photons carrying a large amount of power P through a cross section. Now suppose your "device" is a mirror aligned parallel to the stream. Now for an arbitrarily small energy cost, you can align the mirror to the stream and cause the photons to produce a force P/c, "triggering" the system to relocate energy.

> For a very small energy cost (P<<F c), you can generate a magnetic field to propel yourself forward. The opposing magnet won't move for D/c seconds, so you're transporting momentum at a much lower cost.*

The mechanical moment, i.e. the moment of the object you can see (for example magnets, electrons, ...), is not conserved. The total moment is conserved, i.e. the moment of the objects and the moment of the fields, in this case the electromagnetic field. The electromagnetic field is quantized, and the particles are the photons. So in this case the photons travel from the electromagnet to the static magnet, and they carry away some moment and energy, so the moment is conserved locally.

It's important to consider the moment of the fields, because some of the photons don't go to the other magnet, but go to other directions and are never absorbed, the thecnical name is "radiation". If you don't consider it, then the system will lost some moment and some energy.

> As a thought experiment consider a stream of photons carrying a large amount of power P through a cross section. Now suppose your "device" is a mirror aligned parallel to the stream.

This is a solar sail. The solar sail is a real device. But I think that this experiment is not trying to build a complicated version of a solar sail. To test a solar sail, you should put the device in a strong light. I suppose that the test chamber is in the dark.

While it's true that most reactionless drives break some of the most basic physical laws, that's not the case for all of them. An Alcubierre drive is a good example of a propulsion method that works within the current laws, yet does not emit anything.

The idea that the designs they tested (and EmDrive) would be reactionless is exactly why everybody is so baffled and excited by this. If they are, they must have some way to maintain those basic laws, through a process we don't know about yet. Aside from the practical uses, it could lead to a whole new physics to study.

Alcubierre drive is not a thruster, you can't attach it to a spaceship to move it. It's something like: "If you put a spaceship inside this, and it stabilizes the gravity field using that, then it will travel faster than light in an external reference frame." It travels at a constant speed, so it doesn't need to emit particles to change the momentum. (On the other side, it uses exotic matter that no one has ever seen, so it may be impossible to construct.)

When you say "reactionless" you mean that the device doesn't emit anything and it breaks the momentum conservation law? The previous experiments tried to give some theoretical explanations, but usually they were equivalent to emitting photons or gravitons. The conservation of momentum is a very fundamental law. It's possible to change the physics laws, but you need to have very good experimental evidence. (Hey! I believe in the Higgs boson!) This is like the faster than light neutrinos. It's very exiting if it were true, but the easier explanation is an experimental error.

I asked a particle physicist to compare the validity of this to the faster than light neutrino claim that came up and his response was:

"Although several people are drawing that comparison, they're really not in the same class. OPERA did an extremely thorough analysis of the faster than light neutrino result before they published. In the end it turned out to be a very subtle experimental error that cause the problems.

In contrast, this is just junk science from start to finish. The initial claim was likely an outright lie to lure investors."

But there's three independent labs working on this concept, plus NASA. _Who_ is lying, the EmDrive folks? And if they are, then how are NASA, the Chinese and the other lab all replicating their results?

By not testing it in a vacuum, which would eliminate most of the immediate problems with thrust being produced by outgassing from heated wires (10kW is more power then your stove top) or coronal discharge.

I don't know what's going on with whoever is doing this at NASA, but the joke of testing IN A VACUUM CHAMBER while not actually pumping it down is ridiculous. These people are scientists who have access to the device - how did they not turn it on? For lack of a suitable capacitor? That's a nothing part for such a massive potential breakthrough.

The OP (Wired article) says that the vacuum chamber was indeed pumped down:

While the original abstract says that tests were run "within a stainless steel vacuum chamber with the door closed but at ambient atmospheric pressure", the full report describes tests in which turbo vacuum pumps were used to evacuate the test chamber to a pressure of five millionths of a Torr, or about a hundred-millionth of normal atmospheric pressure.

Yeah just saw that part. The abstract doesn't mention it, which is super-weird, and neither did the original Wired article.

But if they found the effect persisted when pumped down then that puts this way further into "interesting" territory.

EDIT: Ok I've just been through the paper, but it's buried at the end - "Vacuum compatible RF amplifiers with power ranges of up to 125 watts will allow testing at vacuum conditions which was not possible using our current RF amplifiers due to the presence of electrolytic capacitors"

There is a big description of the test rig, and how it can be pumped down to vacuum during tests, but they do not explicitly say that they pumped down to vacuum before running tests on the thruster. And the vacuum is never mentioned again until the very end, where we get this one ambiguous sentence. They might be referring to a new optimized thruster they couldn't test, or they might be saying they couldn't test in vacuum at all - it's never made clear and it's pretty damn important.

This was not a huge experiment for NASA, most likely the budget wasn't there for vacuum-capable capacitors and the time to convert the equipment. It was either go with what they had, or don't go at all. I do hope that, with a tentatively positive result, they can now get a bit more budget to perform the test again with an improved setup.

Just for the record: I agree.

I was just comparing that there was also a lot of excitement.

In an Alcubierre drive the total system does accelerate forward. The spaceship itself won't, but the total system ('bubble', contents and ship together) will increase in momentum to an outside observer. I agree that it may be impossible to construct, but it's theoretically possible.

According to the experiments, the drive seems reactionless so far. It's possible that it does emit something that we can not measure, it's also possible that it transfers momentum to something else, or that it works on something else entirely. We simply don't know yet. There is potential here, if there does turn out to be experimental evidence for it being truly reactionless, for that to require an addition to physics. Even if the evidence turns up something else, it'll at least be interesting.

I agree that there are a lot of ifs, ors and buts. That's why everybody is calling for independent confirmation. Just like the neutrino case, that's how science is supposed to work. But any potential for expanding our understanding should be given proper attention and enthusiasm.

I think it's fair to place Alcubierre drive and something that's been physically built and tested in two very different camps.

I understand this thread was about the theory, but from an engineering perspective that's a substantial distinction.

NASA suspects the thrust comes from "quantum vacuum virtual particles". These literally come from nowhere and would be the balance in the conservation of momentum. See Casimir Effect or Hawking radiation for examples where quantum vacuum virtual particles have an effect.

Both of those conserve momentum. Black holes receive the momentum from their Hawking radiation (but in normal space are generally spherical - it's been proposed a 1000kg black hole might be able to be balanced and fed in a reaction chamber to give thrust), and the Casimir effect conserves momentum because it acts equally on the conductive plates but in opposite directions.

The problem here is where are they getting the thrust from and what mechanism is causing virtual particles to become real? And where in the energy for that coming from? Because you're paying not just KE but also the rest energy cost to make them real with Hawking radiation.

Quantum Vacuum Virtual Plasma: http://xkcd.com/1404/

Just to be clear, because I think it's easy to forget this:

> If you are using a nuclear reactor, a small par of the mass of the atoms is transformed into energy and you put that energy in the photons

This applies equally to nuclear reactors and to AA batteries (like you mention earlier). All forms of energy have mass.

Relativity says the speed of light is the same in all reference frames. Newton's Third Law says any action has an equal and opposite reaction. These two ideas are incompatible. If you shine a light beam from a moving object, it doesn't make a faster moving light beam. The reaction is not equal and opposite. Since light carries momentum, Newton's ideas about conservation of momentum might be wrong too.

Relativity is self-consistent and incorporates conservation of momentum.

(Photons can carry differing amounts of energy, and therefore momentum. This is exhibited as differing wavelengths. Specifically, light being emitted from a moving object is blue/redshifted depending on the relative motion of the observer.)