A British inventor invented an engine that can provide thrust in outer space without a propellant. This engine has been tested by a Chinese team, and more recently by NASA:

The engine itself is The EmDrive: and is FAR too complicated for me to understand. If this engine actually is legitimate, what would be the consequences for Science and Space exploration?

  • $\begingroup$ On a parallel tack - space.stackexchange.com/questions/3591/… $\endgroup$
    – Everyone
    Aug 3, 2014 at 5:25
  • 2
    $\begingroup$ That's not a parallel track. Emitting photons is a lousy way to produce thrust, but it does produce thrust. The EmDrive doesn't even emit photons. It's a closed device. That's what makes it impossible. $\endgroup$ Aug 3, 2014 at 23:55
  • $\begingroup$ @DavidHammen I'd say the "drive" does emit photons. The micorwaves induce currents in the plates, which generate heat. I'd certainly expect infrared radiation being emitted. If that is the reason for the claimed propulsion I can't determine, since I don't have the necessary hardware. But everyone remembering the anomalous acceleration of spaceprobes ("Pioneer anomaly") will be cautios--turns out the effect was caused by the photons of anisotropic thermal radiation... $\endgroup$
    – Jens
    Dec 9, 2016 at 10:56
  • $\begingroup$ @Jens -- It does not emit photons (non-spherically, that is), and even if it does, the thrust is way off. The 720 mN of thrust claimed to have been observed by a team from China (since retracted) would have required over 200 gigawatts if that thrust was photonic. Even the 50 μN level claimed by Eagleworks would have required 15 kilowatts if that thrust exists and if it's photonic. $\endgroup$ Dec 9, 2016 at 12:31

4 Answers 4


Looking through the linked article, several forces are mentioned, including 720 mN and 30-50 uN. These are extremely low levels, but still can have some use. At first glance, these seem similar to the numbers provided by Ion Drives, and Ion Drives also use a large amount of power, so let's compare the two of them, and see how they compare. From Wikipedia, I find that the thrusters typically operate around 2-3 kW of power, with a thrust around 70 mN for that given power level. The numbers according to the website provided for the Chinese test are 720 mN with a power level of 2.5kW. So according to that article, they are potentially more energy efficient than ion drives, at least according to the Chinese. I say potentially because the NASA levels reported are significantly different.

I would assume this would replace the Ion Drive as the long duration maintenance thruster for maintenance, when it has been fully vetted, and be used for some deep space missions. This will increase the time and capabilities of these missions, but not provide a radical difference than what we have now. It might make more possible long durations missions, such as Dawn, or even interplanetary missions, but the power is far too little to do something like lift off of a planet, for instance. I imagine there will be other uses found as well, ones that we can't easily think of at the moment.

It will save some on the fuel. Dawn, for instance, had 425 kg of Xenon for use as fuel. That is a significant portion of the spacecraft, which can be used for other things, or just make the mission cheaper. Xenon has a cost of about \$1200 per kg, and thus just the fuel alone costs around \$600,000.

  • $\begingroup$ Another point to consider is that this new engine is fuel-less, meaning one can eliminate the fuel (that would otherwise be used for something like an ion engine) from a spacecraft's payload if using this engine, which can lead to significant weight and financial savings. $\endgroup$
    – ajp15243
    Aug 1, 2014 at 14:24
  • $\begingroup$ 425 kg on Dawn, so yes,that is significant. $\endgroup$
    – PearsonArtPhoto
    Aug 1, 2014 at 14:27
  • $\begingroup$ Slightly more efficient? The numbers you gave are 10,000 times better efficiency (10 times the thrust, 1/1000 the power input required) Actually I believe your stated 2.5W power is incorrect, the link mentions 1.0kW and 2.5kW tests. $\endgroup$
    – Ben Voigt
    Aug 1, 2014 at 14:29
  • 2
    $\begingroup$ That was a typo. So it's only 10x more efficient, by the Chinese test. I gave it slight because the NASA test showed a much smaller thrust... $\endgroup$
    – PearsonArtPhoto
    Aug 1, 2014 at 14:33
  • $\begingroup$ I think that explanation should go into the answer. To me, even if the absolute numbers are small, a 10x improvement is not exactly "slight". $\endgroup$
    – user
    Aug 1, 2014 at 17:54

Call me unimpressed.

NASA did not "validate" this "impossible" space drive.

First off, this was a conference paper, not a peer reviewed scientific journal paper. Even if the results had been published in a peer reviewed scientific journal, I would still not call it "validation." The peer reviewed literature is where science starts, not ends. To be "validated", the results would have to be widely replicated elsewhere.

But let's dig deeper. From the conference paper,

Thrust was observed on both test articles, even though one of the test articles was designed with the expectation that it would not produce thrust. Specifically, one test article contained internal physical modifications that were designed to produce thrust, while the other did not (with the latter being referred to as the “null” test article).

This supposed EmDrive Cannae Drivea is a highly extraordinary claim; it violates conservation of momentum. Extraordinary claims require extraordinary evidence. The authors can't even reject the null hypothesis. That's not extraordinary evidence. It's more along the lines of "Move along, move along, there's nothing to see".

a The article being tested apparently is not Shawyer's EmDrive. It is a different device, Guido Fetta's Cannae Drive, which uses the "quantum vacuum virtual plasma" (whatever that is) as the means for obtaining thrust.

It's the media that made the connection to the EmDrive. That connection is nowhere to be found in the conference papers by White et al. or by Fetta. In particular, Fetta's patent claims no prior art, which must mean it is somehow distinctly different from Shawyer's EmDrive, at least in the eyes of the patent office.

  • $\begingroup$ This drive also explains the CERN super-luminal neutrinos... $\endgroup$
    – DJohnM
    Aug 2, 2014 at 20:42
  • 1
    $\begingroup$ Just repeating the words of the article. $\endgroup$
    – Nzall
    Aug 4, 2014 at 13:05
  • 2
    $\begingroup$ In general, claims of "no prior art" aren't normally investigated all that much by patent offices. After all, lots of "doesn't do what it claims to do" items are patented. $\endgroup$
    – aramis
    Oct 21, 2015 at 0:25

The Drive's Impacts

If the drive works as described, it could reduce the costs to launch to those in the range needed to commercialize manned flight. They claim to be able to design a 2G shuttle using it for primary thrust.

Note that a consistent 0.01G (0.1 m/s² vehicle would be capable of making it to mars in a matter of days rather than months. It would be about 15 days, instead of 6 months, at closest approach. A 1G drive reduces closest approach to about 35 hours.

The drive as described uses energy conversion from frequency to kinetic force. Assuming the papers are correct, the measurements weren't fudged, and the apparatus actually operates, it's a potential revolution.

The drive isn't without consumables, however. The drive requires cryogenic cooling. This is as surely a limit as the fuel for an ion thruster. In theory, a sufficiently potent heat pump and radiator system could be used to provide both closed cooling and (very very) miniscule additional thrust. In practice, cryogenic coolants generally are open cycle; using them involves venting them.

The Impact of NASA Confirmation

The impact of NASA replicating the chinese and inventor's results indicates that the drive may have actual validity, and that the physics may be incomplete.

Primarily, it is a credible agency giving a credible result of "apparently works" to a drive that's revolutionary in concept - both concept of design and concept of mode of operation.

Secondarily, it's a clear "Conservation of Momentum needs refinement" result. Until the specific method is evaluated better, it's just a "Useful anomaly" rather than a deal breaker.

Cascade Effects

If NASA finds it cheap and credible, then it will get tested further (tho' not, of need, by NASA). If that testing bears out the initial findings, then it will lead to commercial use of the drive.

If the drive bears out in orbital flight the tested ratings, then the cascade through physics will include many new theory tweaks trying to explain it, and a lot of tests, until eventually a consensus explanation arises.

Ongoing Testing Notes

Nov 2015

Just a note that NASA is apparently still testing the device. source: IFLScience, 3 Nov 2015

24 Apr 2016

on 13 Apr 2016, M.E. McCulloch released a paper, Testing quantised inertia on the emdrive, explaining how the drive works without violating conservation of energy, based upon. Source: arXiv.org > physics > arXiv:1604.03449 Paper is up for peer review. The basis is Unruh Radiation Effect, a not yet widely accepted bit of theoretical physics, which can be used to predict thrust of an EM Drive based upon the chamber, the power, and the wavelengths, and changes in the effective inertial mass of the photon due to frequency shifts as it bounces around. McCulloch does not, however, actually define the maximum theoretically ideal Q value...

Here's the Conclusion of McCulloch:

More than eight tests in four independent labs have shown that when microwaves resonate within an asymmetric cavity an anomalous thrust is generated pushing the cavity towards its narrow end.

This force can be predicted fairly well by using a new model for inertia (MiHsC) which assumes that the inertial mass of the photons is caused by Unruh radiation whose wavelengths have to fit exactly inside the cavity so that the photons’ inertial mass is greater at the wide end. To conserve momentum a new force appears to to push the cavity towards its narrow end, and the predicted force is similar to the thrust observed.

MiHsC suggests that the thrust can be increased by increasing the input power, the Q factor, or using a dielectric. As a direct test MiHsC predicts that the thrust can be reversed by making the length L equal to the width of the narrow end.


If it works as advertised, it can have quite an impact on satellites.

They need to make small corrections to their orbits every now and then. Over years, it adds up to quite a bit of fuel, which had to be ferried to orbit at launch.

This new system's only input seems to be energy, which is plentiful at space. If the power comes from an increase in the size of solar panels, you'd need to do the math to see if it's worth it. But it could come from off-peak power in some satellites which are not used full-on all the time. These adjustments have a fairly flexible schedule.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.