[mikeash]

Yes, which is why violating a known law isn't automatically disqualifying. But it is highly suggestive, and requires extremely good evidence to surmount. So far, such evidence has not been forthcoming, so this remains a good reason to doubt.

Every so often, something comes along that overturns an established law, like the photoelectric effect or the Michelson–Morley experiment. But the vast majority of the time, the laws remain intact and it's the claim which is somehow flawed. (See the Pioneer anomaly for a recent example where new physics was long considered as a potential explanation, but it ended up being a completely mundane effect.)

[danielmorozoff]

This is not the first experiment undertaken by Nasa/Pentagon/Boeing or the DoD. As well as similar measurements confirmed by other research teams (including foreign teams).

I totally agree it requires drastic evidence, but it could be explained by another physical action not necessarily the violation of Newton's laws or effects that supersede those laws like quantum effects

https://drive.google.com/file/d/0B7kgKijo-p0idV9tcmVIVzZrdTQ...

http://www.ibtimes.co.uk/emdrive-exclusive-roger-shawyer-con...

http://www.slideshare.net/KurtZeller/investigation-of-a-part...

[mikeash]

If it's another physical action within known physical laws, then nothing interesting is happening and it's a failure of the experimental setup. Yes, many experiments have been done, and none have conclusively disprove it, but at the same time none of them have conclusively shown that something is going on either. The detected effect remains well within the realm of potential error, so until that uncertainty is resolved, "error" remains a much better bet than "new physics."

[achileas]

>If it's another physical action within known physical laws, then nothing interesting is happening and it's a failure of the experimental setup

I'm not sure what makes you think this. There's no failure in science when an effect is observed and it's explainable within the frame of current theory. If that was the case every experiment I've done would have been a 'failure.'

[mikeash]

With the experiments that showed a result, that result is not explainable within the frame of current theory. The whole purpose of the experimental setup is to isolate the test article from any possible interactions that could exert force on it, other than the one being tested. If the results seen so far are due to mundane interactions, then that isolation didn't work. Is that not a failure of the experimental setup?

[achileas]

IIRC, the recent tests are to reduce experimental error, not 'mundane interactions.' Just because there may be a way to explain it within the frame of current theory that the original researchers did not pursue or understand, still doesn't mean the experiment failed.

[shuntress]

So, thermal radiation slowed the Pioneer spacecrafts. 'Thermal' is a type of Electromagnetic (EM) radiation, right?

Doesn't that make it an EM Drive?

Why the cone apparatus? What is preventing the intentional design of a spacecraft that uses thermal radiation (or Thermal Recoil Force) similar to the Pioneer spacecraft?

e: The replies indicate the answer is that this could be built but the effect is to weak to be practical. Which I find to be a completely acceptable answer, thanks.

[mikeash]

That makes it a photon rocket, which is nothing new. It's a logical consequence of the fact that light carries momentum, which is part of Maxwell's electromagnetic theory from 150 years ago. It's not very practical unless you have an extremely dense power source (nuclear fission won't cut it, you need fusion or antimatter) so it's more of a theoretical toy at this point.

What's interesting about the EmDrive's claims is that it supposedly produces far more thrust for the power that goes into it than would be possible from a photon rocket.

(The Pioneer anomaly wasn't because nobody realized that thermal radiation would act as a photon rocket, but rather that the effect wasn't correctly calculated. The theory is pretty simple, but it gets complex when you apply it to a real spacecraft with complex shapes, non-blackbody materials, and temperature gradients.)

[Dylan16807]

Light has effective mass. It's well known that you can shoot light in one direction for a tiny amount of thrust the other way.

The device under consideration is special because if it works it is a reactionless drive. Nothing is coming out of it.

[triplepoint217]

Light actually has momentum but not mass. Calling it effective mass isn't far off, but sometimes my pedantic physicist kicks in.

[Dylan16807]

The proper term I'm used to is "relativistic mass", but I didn't want to make the discussion complicated. Would you not phrase it that way?