Phil Plait, Bad Astronomer, has a blog post about this today. In short, don’t get too excited about this. He thinks the “thrust” will likely eventually be traced to experimental error along the lines of the faster-than-light neutrinos from a while back.
I like this quote from the John Baez Google Workspace Updates: New community features for Google Chat and an update on Currents site:
That is the problem (as addressed upthread) in taking the test data very seriously; the scale of possible error is significantly greater than the measured force itself, and so measuring and filtering the sources of error is the major challenge of the exoeriment. And to do this reliably requires a good prediction of the theoretical performance. At least from the website, that information isn’t available. So, the purportively positive result is nice, but by no means conclusive, and argues for improving experimental methods and apparatus (and scaling the level of thrust) in order to validate or falsify the result.
Stranger
It’s unfortunate that Habeed did that since it’s a distraction from his absolutely valid point.
Here, I’ll build a free energy device. The original reporting on the EmDrive reported 0.72 N of thrust on 1 kW power. I haven’t been able to find NASA’s equivalent data, but the actual numbers don’t matter here; I’m only using them to make the point clearer.
Suppose we have a long tube with a capsule in it and magnetic decelerators on each end. The capsule is externally powered (via laser, etc.) and we’ll suppose weighs 72 kg. The tube is a billion meters long (hey, I didn’t say this was a practical device).
So, as inputs:
F = 0.72 N
m = 72 kg
d = 1000000000 m
P = 1000 W
We derive:
a = F/m = 0.01 m/s^2
t = sqrt(2d/a) = 447,000 s
v = at = 4,470 m/s
From here, we can see that the input energy is:
E = Pt = 447,000,000 J
While the kinetic energy of the capsule (which can be completely captured) is:
KE = 0.5mv^2 = 719,000,000 J
We’ve gained energy there. The system is self-contained and just cycles back and forth, so we can gain energy forever.
The device must operate the same in any inertial reference frame if it’s not to violate relativity.
Irrelevant. When a conventional rocket (whether electric drive or otherwise) produces thrust, then in a frame where the rocket is moving at high speed, we see two things happen: the KE of the payload increases, while the KE of the ejected propellant decreases (since it has slowed down relative to us). This is absolutely critical if the books are to balance. The rocket is spending a small amount of energy to convert the KE of the propellant (put there by prior energy expenditure) into KE for the rest of the payload.
The only propellantless drive that actually works is a photon drive. They’re dreadfully inefficient and you need relativity to understand why they don’t violate conservation of momentum/energy.
Photon drives aren’t propellantless. The photons are the propellant.
They’re propellantless in the sense that photons are massless and can be generated in infinite quantity. I don’t care whether this counts as really propellantless or not; that’s just semantic quibbling.
In any case, they do require energy to create, and therefore the mass of the ship must go down with every photon ejected. Furthermore, the ejected photons will be redshifted according to an observer at “rest” (i.e., on Earth or whatever). As such, you can’t use a photon drive to build a similar free energy device to the ones Habeed and I described.
I’d have to run the math a bit more to totally convince myself, but I’m fairly sure that any drive claiming to be more efficient than a photon drive and with no/massless propellant will violate some law of physics.
Strangelove, the various tethers and methods to interact with the magnetic field of the earth do work and make a “propellant-less” drive that only works, weakly, in low earth orbit. The earth itself is the other half of the problem and where the books are balanced.
The proponents of all these massless propellant drives claim that, somehow, the rest of the universe is being interacted with. While being able to build an engine that can somehow push or pull off the rest of the universe or interact with expanding spacetime is improbable, the books would balance. The “free” energy comes from the rest of the universe, analogous to how a tidal generator extracts energy from the kinetic energy of the moon.
Or, as Stranger’s link notes, charged particles in the vicinity. Effectively propellantless but not in the same sense as this drive.
I don’t really see any way to make sense of it. There has to be something to “grip”, whether a magnetic field, or quantum foam, or whatever. One property of basic physics is that it takes more energy to push yourself from a moving surface than a stationary one. This is why, for instance, cars accelerate more slowly as they speed up (assuming constant power). This must be the case, else KE=0.5mv^2 couldn’t be true.
Ok, fine. We accept that like a car, our drive shows less thrust the faster it goes. But that assumes we started with the quantum grippiness at rest. Why should that be the case? Is it stuck to Earth somehow? Our Solar System? Our galaxy?
It all sounds very much like Ether.
In fact, I think I could demonstrate that at the efficiencies claimed, it would have to Earth–maybe even the surface of the Earth. That seems rather difficult to explain.
Or, it could just be gripping the chamber walls through normal EM interactions…
Well Mach’s Principle claims that inertia itself comes from a mass instantly interacting with all the mass in the rest of the universe, and that general relativity can’t account for rotational relativity without it.
Or something like this. Again, Strangelove, you are almost certainly correct. With that said, if you could build a drive that somehow “grabbed on” to the cumulative effect of the countless masses in the universe, it would work. Umm, interestingly, it would also develop less thrust as you went faster because you now have a different relative velocity to the thing you are pushing off of - the rest of the universe.
It’s worse than that, though. Suppose we pick the CMB rest frame for the quantum foam stuff. The Earth moves at about 370 km/s relative to this frame.
At this speed, a 1 kW thruster should only produce 0.0027 N, not 0.72 N (the number reported for the EmDrive).
Maybe the NASA test shows less efficiency. Their whitepaper mentions 30-50 micronewtons of thrust but not the power level. If the thrust exceeds P/370000, then any “interacts with the universe” line of reasoning is bogus.
Of course, there’s no reason to believe the CMB rest frame is privileged, either. It shows how old the universe is but not how big.
I just popped in to say that it is very interesting to read an animated discussion where I only tenuously comprehend the statements made by both sides and where the matter in question could be anything from experimental error through to ground-breaking stuff that will both rewrite the physics books and open up huge possibilities in satellite technology, but in either case is a very small observation in absolute terms.
At the moment the questions are bigger than the answers and that is fascinating!
I should add that my result was only for traveling opposite that of the movement relative to the CMB; in other directions you could have greater thrust. In fact, you could have thrust without any energy expenditure at all if there were a way to metaphorically stick a paddle in the underlying flow.
All of this implies that there should have been some anisotropy in the results (just as with the Michelson-Morley experiments to detect the ether). None was reported, as best I can tell.
Watch out if you build a device that can really lock into the CMB quantum foam. It might suddenly fly out of the room at 370 km/s.
XKCD has picked up on this today.
[QUOTE=XKCD]
I don’t understand the things you do, and you therefore may represent an interaction with the quantum vacuum virtual plasma.
[/QUOTE]
One thing that occurs to me is that most methods of turning electric power into something else are at least in theory reversible. Some electric motors can be used as generators. A LED can act as a photocell (though admittedly not a very good one). A speaker can also act as a microphone (again, not very well). A Peltier junction can generate a temperature difference from electricity, or generate electricity from a temperature difference. The physics are reversible. The one exception is using electricity to generate heat, because you can’t just reverse entropy like that.
This suggests to me that if it’s possible to build a device that accelerates when connected to an electric power source, then it might also be possible to make a device that generates electricity when accelerated by an external force. Of course, if that’s possible, then you’ve just generated a free energy device. In Relativity it’s impossible to distinguish between gravitational acceleration and inertial acceleration, so just set your device up in a gravity field and you’ll get free power out.
Not necessarily: It might need to be accelerated relative to something else, and if that something else is also stationary relative to the surface of the planet, then there’s no relative acceleration.