Aug 01 2014

NASA Tests EM Drive

I’m skeptical. I know, you’re shocked. When you recover, take a look at this article about NASA “validating” an allegedly impossible drive.

The bottom line is that I just don’t believe it. I could be wrong. I hope I’m wrong. I don’t necessarily think the results of NASA’s test are untrue, just that I don’t think they have “validated” that the propellantless drive is what proponents say it is.

My reaction is identical to the claim made in 2011 that a team of researchers found that neutrinos travel faster than light. I didn’t believe those results either. The researchers were very careful, they rigorously reviewed every aspect of their experiment, and only announced the results when they were confident they ruled out all error. The physics community didn’t believe it, but they did their due diligence. After further analysis, it was found that the results were an error – an artifact introduced in the experimental setup. Initial skepticism was vindicated.

The claims made for a machine that can provide thrust without propellant is as unlikely and at variance with the laws of physics as neutrinos traveling faster than light or free energy machines. Sure, it’s always possible that our understanding of the universe is incomplete in a way that allows for one of these phenomena to be true, but our current understanding calls for extreme initial skepticism. Such a stance has a very good history to support it.

At the same time, I wouldn’t invest a dime of my own money in a company claiming to have invented a free energy machine, and I don’t think our taxes should fund such research either. It’s worse than playing the lottery.

Producing thrust without propellant is similar. It seems to violate the conservation of momentum. Conservation laws in physics are among the most reliable and solid of the laws of the universe that we have discovered. Scientists should not idly speculate about violating them.

The idea here is to create a drive that converts energy into thrust without the need for accelerate propellant in the opposite direction. This would transform space travel, and make things like hovercars a reality. Perhaps the biggest limitation of space travel is that you have to carry around your fuel and propellant with you. This means you have to accelerate the fuel and propellant also, and the fuel for accelerating that fuel, etc. Most of the energy spent in getting to orbit is lifting the fuel you need to get into orbit.

For most rockets used today, the fuel is the propellant. There are drives, however, that generate energy to accelerate a separate propellant, such as an ion drive. There are also designs that do not need propellant, such as solar sails or using a laser to push a spaceship. Ram drives are designed to scoop up their propellant as they go (from the thin hydrogen gas in space). Many of these are theoretical technologies or provide very little thrust.

Imagine, however, if we had a device that could turn energy directly into thrust without the need for propellant, by exploiting some exotic feature of physics. That is what some people claim they have developed.

NASA has recently tested one design and published the results in a paper entitled: Anomalous Thrust Production from an RF Test Device Measured on a Low-Thrust Torsion Pendulum. In the paper they do not speculate about the physics involved, just test the drive. They found: “Approximately 30-50 micro-Newtons of thrust were recorded from an electric propulsion test article consisting primarily of a radio frequency (RF) resonant cavity excited at approximately 935 megahertz.”

They are correct to term this “anomalous thrust,” which just means they don’t know where it is coming from. They are not concluding it is propellantless thrust or that that their test confirms the speculation of proponents. I also find it highly significant that that anomalous thrust is teeny tiny – 30-50 micro-Newtons. One Newton is the amount of force required to give a 1-kg mass an acceleration of 1 m/s/s. So this thrust is 30-50 millionths of a Newton.

It is hard to draw any firm conclusions from such tiny anomalies. The fact that the force is so small means that even very subtle errors in the experimental setup, or unknown factors affecting the measurement, could be sufficient to explain the results. You can’t simply extrapolate from such small effects and assume they will scale up. That is the perpetual mistake of perpetual motion machine claimants – they find tiny anomalies and naively believe they will scale up.

The other possibility is that the anomalous thrust is genuine and is produced by some subtle physical effect, so subtle that this is all you are going to get and it won’t scale up. There may not be more thrust to be had.

Proponents argue that even a tiny thrust is useful in space and over years can provide a significant cumulative acceleration. This, of course, would also provide the ultimate test. If engineers can build a ship using this form of thrust and actually use it to accelerate a probe, that would be impressive evidence. This is similar to my challenge to free energy gurus – call me when you are running your home, or even a large motor, off your device (actually doing work, without any “supplemental” energy source).

For now my attitude toward the NASA test of this alleged propellantless drive is the same as it was toward the faster-than-light neutrinos. I think it is far more likely that this will turn out to be some experimental artifact than a truly new phenomenon. This will need careful independent replication, enough to get the physics community excited, before I will get excited too.

I do sincerely hope that I am wrong. I want my flying car just as much as the next guy.

29 responses so far

29 thoughts on “NASA Tests EM Drive”

  1. I am pretty skeptical of the result as well, but there is much in the write-up here that’s, I think, misleading. Dr. Novella is pooh-poohing the result based on the minuteness of the force, and that’s not really a good way to think about it. The relevant figure of merit is how much force 30 microNetwons is relative to the precision of a torsion pendulum, and, it ought to be a lot; a pretty similar setup has been used to make incredibly precise measures of the gravitational constant, for example. And they tell you that — the instrument’s precision is well below what they saw, and far far more precise torsion pendula have been built. In addition the claim of the inventor isn’t that you’re violating conservation of momentum — if you’re spewing microwave photons out the back of the thing, you aren’t, and his little EM generator is supposed to be like a little etalon, “leaky” on one side. I’m really, really unhappy about how it’s generating its power and the resonant energy inside the thing, and it all seems like hoo-hah to me. But it’s not that you are violating conservation of momentum by not having a propellant — things powered by solar sails don’t have propellant either, after all, light has momentum — so much as I’m not seeing any reasonable mechanism to generate power in the first place; there’s no gain medium inside his little resonant box nor any obvious way that it’s getting excited. That’s just not how I want to think things work.

  2. Pugg Fuggly says:

    I hope nobody is actually listening to this drivel. Everybody knows Steve and his cronies are all paid shills for Big Propellant.

  3. Kid – perhaps I wasn’t clear. I said I don’t necessarily doubt the result. It’s not about the precision of the instruments and the measurement. It’s about the source of the anomalous thrust. Such a tiny result means that a subtle effect, not accounted for, could be causing it.

    I agree that if radiation is being ejected by one side, that causes thrust and does not violate conservation of momemtum. I don’t think that is what is being claimed. That would be unremarkable – hey, we got a tiny thrust out of asymetrical radation. They are claiming a new effect that does indeed violate conservation of momentum.

  4. worlebird says:

    I was trying to wrap my head around the size of this supposed effect. 30 – 50 micronewtons is not the kind of measurement I do everyday, so it’s a little hard to conceptualize how small this really is.

    A little Googling and some back-of-the-enveleope calculations suggests that it takes about 50 millinewtons to pick up a standard US nickel (5 cent coin). A millinewton is 1000 micronewtons, which means that this drive was supposedly generating enough thrust to lift 1/1000th of a US nickel off the ground.

    Even that is hard to conceptualize, but it gives an idea of just how tiny this effect is, if it even exists. I can easily see how any of a hundred different things could end up causing such an anomaly without actually being a true, useful, new method of thrust. I mean, an errant breeze could have caused that kind of force. I’m sure they accounted for that, but still, it’s obvious this is going to require some major reproduction of results to ever be taken seriously.

  5. ferrousbueller says:


    The device isn’t meant to spew microwave photons out the back. It allegedly generates propulsion through the reflection of photons trapped between parallel plates in a tapered waveguide. I’m going to do a bit more reading on the subject tonight and see if I can make more sense of the claims.

  6. stixx23 says:

    I think the key passage from the NASA study is this:

    “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).”

    So it sounds like there must be something in the design of the experiment itself that is showing results. Unless they are so good at creating thrust from nothing that they can even do it on an engine that isn’t supposed to work.

  7. garyacrowellsr says:

    This seems important:

    “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 would tell me that whatever they theorize as the source of the thrust, it’s probably wrong.

    The first thing that comes to my mind is some sort of magnetic field interaction – earth’s field, residual magnetism of the chamber, something induced by the device? Run it again with the device enclosed in a significant mu-metal shield.

  8. DS1000 says:

    The article also mentions but glosses over the very important fact that the force measured by the NASA tests was 1/1000th that measured by the Chinese and the original inventor for the same power input. Sounds a lot like the common point Steve makes about alt med and other supernatural phenomena; the better the test the smaller the effect size. Time to use our Science Based Physics thinking in lieu of the Evidence Based Physics presented in the article!

  9. mrgamehelper says:

    With respect to the technical report, one thing that I find unusual is, they stated “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.” So they were measuring a non-zero thrust for something that should have been their control. I would have expected that the authors would have attempted to explain or remove this effect first, well before they began attributing any thrust to the emDrive system. Until this can be accounted for, this experiment doesn’t show anything other than this cavity can produce a non-zero thrust (whatever the cause may be). But making any claims about what is causing it appears to be a pretty big leap of faith.

    I’m also not so sure why they expect to see a net force from a random process. To me, this sounds a lot like standing on ice skates and swinging your arms back and forth; your feet move back and forth on the ice a lot, but you never actually go anywhere.

  10. Sylak says:

    Yeah, anyway those positive resulta will be cover up by the green alien lizarda king probably. lol

    I remember you talking about this On SGU. I don’t get the whole concept on how it is suppose to work. all this sounds it is fishy. I reminds me Of that Rossi guy with his scam machine.

    I understand the need to test this. but Playing the lottery is a good analogy for this.
    NASA should work on making the Ion drive more efficient and powerful I want my TIE interceptor.

    But do the photons of the micro wave could push on the pendulum to create the thrust ? like a laser would on a solar sail?

  11. Gborr says:

    I just registered because this thing is driving me nuts, mostly the boundless lack skepticism I see everywhere (or worse yet, the amount of flak skeptic get for “being negative assholes”).

    The amount of disinformation on the web is also getting out of hand. By now most headlines refer to the machine as “fuel-less engine” instead of propellant-less, casually “deduce” that the effect is because of quantum physics, the three questionable tests by the same Chinese researchers are always cited as “three independent research groups” and any kind of skepticism is generally dismissed by saying “But these are NASA researchers you are criticizing! Are you telling me you are smarter than people working for NASA?”

    Seriously, I have no problem with the test results themselves. If they are correct, hooray, if not then the scientific method will figure out where they made a mistake. The way the news media and the Average Joe gobble it all up without even a hint of critical thought on the other hand just gets under my skin…

  12. Stormbringer says:

    While I have the same reservations about the success of this drive very basic error checking is the next thing that has to be done. I would suggest this test; set up the fixture and get your best average reading, rotate as little of the thruster as possible to have the direction of the thrust 180° from the original direction and redo the same series of tests. After getting these readings while not changing the thruster reverse the environment as much as possible; instrumentation, recording equipment, power supplies, etc… So it looks as much as possible like a mirror of the first test. For the last test rotate the thruster back to the original position. If the thruster is creating thrust you should have the same force each time but the direction should be reversed for the second and third test.
    While not a perfect why to eliminate errors it tries to minimize external influences.

    If they do find that they are creating a thrust maybe they are not pushing against the 4% of the universe that we normally do.

  13. Pete A says:

    Thrust is produced whenever there is an asymmetric nonlinearity in the system transfer function. E.g. a loudspeaker produces a small net thrust due to the nonlinear nature of air compression/rarefaction. In electronic systems, asymmetric nonlinearity manifests as a DC offset; in mechanical systems it manifests as a net trust (force).

    It makes me wonder if the tests on the EM Drive have properly taken this into account.

  14. hackettk says:

    You seem unaware that photons carry momentum. The formula is p = h / lambda, where p is the momentum, h is Planck’s constant, and lambda is the wavelength of the photons (roughly 0.32 meters). Also, photons carry energy. The formula is e = h f, where f is the frequency of the photons (935 MHz), and e is the energy of the photon. Using these elementary formula, one can compute the amount of microwave power required to, for example, produce 50 microNewtons of thrust. That amount of power is: 14,989.6 Watts – lots of photons – about 2.42 x 10^28 per second at this frequency. Not too much power, though, – about 15 microwave ovens worth. Of course, that number assumes all the microwave photons are used to produce thrust.

    How much microwave power did they or the Chinese use in their experiments? It sounds like the microwaves are not being radiated, but are trapped in a cavity. If trapped in a cavity, the material of the cavity will heat up, which heats the surrounding air, producing air currents, and also the hot cavity radiates thermal energy (those photons also carry momentum). Asymmetries in heating the air, or the thermal radiation could also account for a net thrust. This would mean, however, that much more microwave power would be required, as the observed thrust would be the vector sum of thrusts pushing in different directions. If the thrust balance only measured one component of thrust (thrust only in a single direction), that could also be an issue. The key to this effect is that the cavity will have some time constant for heating, which would be observed as a time-changing thrust.

    If they are using less microwave power than 15 kilowatts (assuming the 50 microNewton number), then suspicion falls on other effects such as instrumentation, forces from power feed currents interacting with magnetic fields, and the like. The possibilities for error are quite numerous.

    Otherwise, they would be claiming an effect that violates conservation of energy and momentum. The likelihood is then high that there is some error in the experiment or measurements.

  15. hackettk says:

    Further to my last. It would be good to see the apparatus. It is difficult in these types of experiments to provide power to an object in a torsion balance in a force-free manner. And, it is important to note, if measuring a change in force, that very minor temperature changes (well below what a thermal imager can sense) can create large forces due to thermal expansion.

    For example, a 1 m long copper rod will expand by about 16.5 microns with a 1 C change in temperature. The amount of force required to stretch a 1 cm^2, 1 m long copper rod by this tiny amount is about 214 Newtons.

    To generate a 50 microNewton force from thermal expansion of a similar copper rod attached to something would only require 2.3 10^-7 degrees C temperature change, assuming I’ve done this right. The amount of force generated scales like the cross sectional area of the rod.

    The point is that very minor changes in temperature, not measurable with a thermal imager, can create huge mechanical forces from thermal expansion.

  16. joshuagenes says:

    If you accelerate ions in a halfway though a circle and de-accelerate the ions back through the other half of the circle the acceleration and deceleration with respect to the aparatus would cancel each other out. But what about the relative mass? High speed accelerating ions with a high relative mass due to speed at a close fraction of the speed of light and a lower relative mass at a lower speed during deceleration might have some effect. Tell me what you think.

  17. Paulz says:

    This is just more proof that any result that comes out of China needs to be treated with extreme suspicion until proper replication.

    Also, NASA is seriously dropping the ball of late, including with their Warp Drive.

  18. SARA says:

    I’m not clear on why the Chinese test was questionable?

  19. Bronze Dog says:

    It’s something of a sci-fi staple these days. I’ve played my share of sci-fi video games where your ship, jetpack, or mecha has some sort of boost function that runs directly from a generator and refills when not in use. It’d be nice if we could generate thrust without the need to expel mass, but as always, just because we want something doesn’t mean the universe is obligated to provide it.

  20. Maximilian says:

    Testing my ability to comment. I’m new.

    Also @Steven Novella, are you alerted when someone comments on one of your posts? Because I have been considering posting a comment on one of your posts from a few years ago, but I wasn’t sure if you would ever see it.


  21. Pete A says:

    @hackettk, good points. Wouldn’t it be better to use, say, an ultraviolet laser of the same power to produce thrust? I ask this because microwave photons have insufficient energy to cause ionisation.

  22. Plittle says:

    Important to note that the machine tested by NASA is not the same as the one tested by the Chinese. Different “engine.” Different inventor. They claim to be using the same basic principles, but they are two different pieces of equipment.

  23. prebys says:

    There are a few things that should set off pretty loud alarm bells in even mildly reasonable people:

    The initial claims were made by a private company, who didn’t allow anyone to independently test their device. These claims can be completely discounted until someone else confirms them, so let’s look at the “confirmations…

    Before doing a test, the Chinese did an ANSYS calculation, which they say predicted a net thrust. Since ANSYS incorporates Maxwell’s Equations as we currently understand them, any result which predicted a net thrust in a closed system is – by *definition* – evidence you made a mistake. This shows the Chinese group lacks a basic knowledge of E&M, so I’d take anything else they say with a grain of salt.

    The NASA group saw the same effect on their test article as the control device. This is – again, by definition – evidence you don’t understand your experimental setup. Also, the effect claimed by NASA is *more than 10,000 times smaller* than the effect claimed by the Chinese. Labeling this a “confirmation” of anything is ludicrous.

    Bottom line: hang on to your wallets.

  24. tcmJOE says:

    It doesn’t matter what frequency you use if you’re building a photon rocket (getting momentum from light power), the formula ends up being momentum = energy/speed of light. If you take this as “per second” you find out that to get 1 N of thrust, you need ~300 MW of power.

    The Cannae drive seems to give 100 times that, so it’s not just the EM waves leaking and providing the thrust.

    But honestly, this thing was not tested in vacuum but at ambient pressure*. A small differential in heating could easily lead to air currents in the testing chamber, making the thing act like a Crookes radiometer. That would seem to be the most obvious explanation to me. They also do very minimal error accounting in the report, which really causes me to doubt any of their conclusions.

    *They state in the NASA abstract that it was at ambient pressure. In the paper that’s been floating around, there’s a second talking about their capability of taking the pressure down to a millionth of a Torr but then later state that their RF equipment was not capable of going to such lower pressures.

  25. Eric Tergerson says:

    @steve -The more and more people look at the paper, the more porous it seems to become.

    The best summary I came across was briefly summarized by John Baez, who then goes onto explain in much more detail. But his tl;dr:

    “1. They tested a device that was designed to work and one that was designed not to work. They both worked.

    2. They tested the devices in a “vacuum chamber”, but they didn’t take the air out.

    3. They didn’t carefully study all possible causes of experimental error… like their devices heating the air.”

    I saw elements of his criticism originally on slashdot:

    Full Disclosure: I’m biased- I bet on a “negative verdict” on the validity of this whole thing- the loser buys a pint at our next Skeptics in the Pub-

  26. Pete A says:

    Please read Ethan Siegel’s take on this “impossible space engine?”:

  27. kim says:

    I fully believe NASA test result true.
    1. Many people think EMdrive does not violate the energy conservation law, but quantum and Newton’s law.
    2. I think EMdrive does not violate any law.
    3. I published one small hand book which can explain and back support physically EmDrive at August 26 1998. South Korea.
    Title is “Einstein’s photon does not exist.”
    4. By this book, modern physics problem is from Einstein’s photon of general relativity theory.
    5. Einstein’s mistake was the thinking that EM, photon, acts as just the particle with relativity mass. Impacting force when EM moving is F from real particle momentum and reaction force is F’.
    Modern physics tragedy is from this.
    6. According the theory on my book, EM acts just like a particle with relativity mass, but not the real particle. Impacting force when EM moving is zero that is from no changing velocity, but Kq1q2/r2 from Coulomb force F between each orbit electrons and fundamental particle, and reaction force is also F’.
    Here, By EM moving some orbit electrons on front position charge reverse periodically by space surrounding of EM periodic polarity and act to near fundamental particles by reverse force.
    So, EM moving is not real particle force F, but Coulomb force F of particles on the way distance gap. It is the result of EM polarity surrounded orbit electron or not yet on the forwarding way.
    This Kq1q2/r2 is from charging by EM 3 dimension space occupying.
    I named it “Hogun Phenomenon” according my name.
    7. I think EMdrive theory is not resonant cavity thrust, but an electron on the position where surrounded by polarity of reached EM periodic is generating Coulomb Force with near fundamental particles by periodically. I should say that EMdrive power be needed to design with cross section on direction and frequency and case material.
    8. I think this “Hogun Phenomenon” can explain faster speed than light, EmDrive, mirror, transparency, not be affected by gravitation, Photoelectric effect, Compton effect etc.
    I am sure it can solve difficult modern physics problem and prove that nothing violates Newton’s law.
    Feel free to get in touch with me for more information.

  28. stevehigh says:

    The skeptic plays an invaluable role in keeping the progress of science from degenerating into the antics of the denizens of an insane asylum. For my part, I am merely hopeful that the span of years I have left will be sufficient to allow me to be witness to a paradigm shift similar to the one brought forward by Einstein. So naturally my ears perked up when I heard of the propellantless drive. Could this be my moment?
    It is my custom whenever I encounter something like this to quickly proceed to what the skeptics are saying, because my time is limited. Hence I found myself here. Your cautionary words are well written and will probably be proven valid.
    I do ask myself if there is a point at which skepticism takes on a hardened inflexible edge that becomes an impediment to scientific progress. If you will allow me, I did notice an element of “cognitive dissonance” in your post that might induce a thoughtful person to wonder whether the progress of science is truly being served.
    You admit to being a skeptic, but you also confess that you sincerely hope to be proven wrong. You point out that it is foolish to invest money in schemes like this and you strenuously object to spending taxpayers’ dollars in what you characterize as a lottery.
    My point on cognitive dissonance would be this. How could you be hoping that the propellantless drive will be proven valid while at the same time be so insistent that it is wrong to expend resources in evaluating it?
    I think this is a valid question given that in most areas of science skeptics appear to have the upper hand when it comes to deciding whether a potentially paradigm-shifting idea should be investigated.

  29. John Reeves says:

    Nice, I got an opportunity to use the Crackpot Index on kim!

    I was generous but it’s at least 40 points (mentions Einstein a lot, mentions Einstein was wrong), and names a phenomenon after himself. There are also possibly a lot of claims that are widely accepted as false, which would add a lot, but it’s hard to parse the english.

Leave a Reply