≡ Menu

EmDrive Back in the News

Martin Tajmar’s presentation at the American Institute of Aeronautics and Astronautics’ Propulsion and Energy Forum and Exposition in Orlando yesterday has been getting plenty of press. Tajmar is looking at the device now commonly called an EmDrive, studied by Sonny White’s team at Eagleworks (Johnson Space Center) and advocated by Roger Shawyer, Guido Fetta and Chinese experimenters as a way of producing thrust in a way that seemingly violates conservation of momentum.


Tajmar (Dresden University of Technology) offers a paper entitled “Direct Thrust Measurements of an EmDrive and Evaluation of Possible Side-Effects” in his presentation on apparent thrust produced by the test device. As he told WIRED (which announced that The ‘impossible’ EmDrive could reach Pluto in 18 months), the current work will not close the story. From the paper itself:

The nature of the thrusts observed is still unclear… Our test campaign can not confirm or refute the claims of the EmDrive but intends to independently assess possible side-effects in the measurements methods used so far. Nevertheless, we do observe thrusts close to the magnitude of the actual predictions after eliminating many possible error sources that should warrant further investigation into the phenomena. Next steps include better magnetic shielding, further vacuum tests and improved EmDrive models with higher Q factors and electronics that allow tuning for optimal operation. As a worst case we may find how to effectively shield thrust balances from magnetic fields.

Image: Physicist Martin Tajmar. Credit: Dresden University of Technology.

An example of something needing attention is that the thrust measurements linger even after the power is turned off. Such behavior is indicative of thermal effects, but it is premature to reach that conclusion.

A thruster that operates through methods we do not understand naturally seizes the attention, because we seem to do away with the need for a propellant, which would make all manner of missions possible that would otherwise be achieved only through more costly chemical rocket methods. And if we are uncovering something that gets at ‘new physics,’ so much the better, as productive things happen when we find anomalies that lead to deeper investigation and, if we are lucky, a formulation of new principles.

Will that happen here? What needs to be emphasized is that this is work in progress, as Tajmar himself points out, so we cannot draw premature conclusions. We’re at the beginning of a process that includes peer review analysis and publication of papers widely disseminated in the physics community, as well as replication of experimental results examined in those papers. Finding out that momentum is not necessarily conserved would be a result so startling that it would demand the highest level of scrutiny, especially in terms of possible systematic errors — i.e., are there effects being registered which we can account for through the experimental apparatus itself? Tajmar knows this and says as much in his paper.

A bit of background: If you’ll check our book Frontiers of Propulsion Science, you’ll see that Martin Tajmar did an independent series of replication experiments on work performed by James Woodward (the ‘Woodward effect’), while working at the Austrian Research Center’s department of electric propulsion physics. While that work produced a null result, Tajmar went on to pursue experiments with rotating superconductors and, for a time, believed his apparatus was producing anomalous gravitomagnetic forces. Replication experiments that researchers at EarthTech in Austin planned to perform were abandoned because of what they believed to be flaws in the experimental apparatus Tajmar was using, including issues with the laser ring gyro Tajmar used that produced systematic noise that was being misinterpreted as a positive anomalous force signal. Tajmar continued the work for a time but eventually ended the experiment.

Does a similar fate await the EmDrive? We learn as we go, and if we can find ways to reduce or eliminate the problem of onboard propellant, we will utterly change the game of deep space. So, as experiments continue, let’s look for analysis in the journals as the work is subjected to peer review, and let’s insist on the same degree of caution we would use with any result that seems to contradict known physical law. If the effect Tajmar is studying is genuine, science will ferret it out, a process that is usually time-consuming and often subject to misinterpretation.

Addendum: George Dvorsky’s piece No, German Scientists Have Not Confirmed the Impossible EMDrive cites Eric Davis, Tau Zero founder Marc Millis and physicist Sean Carroll (Caltech), and is well worth your time.

An article that brings a determinedly neutral perspective to the matter is Suggestion: The EM Drive Is Getting the Appropriate Level of Attention from the Science Community. Thanks to Sonny White (NASA JSC) for the link to this one.


Comments on this entry are closed.

  • Charlie August 2, 2015, 1:33

    I appreciate with everybody is said by way of explanation. I would just like to add to what Studio Project commented on a let me undo that led me to question whether or not the fact that the speed of light is constant really has any effect (or plays in a fact) on the operation of this device. My understanding of ‘radiation pressure’ that it is a property of the energy of the photons involved in the actions-reaction due to ideas of Newton’s third law. A rocket engine operates in a standard manner by the fact that you have gases that act on the nozzle and the nozzle ask on the gases and this is the action reaction that is always spoken about. I assume that when you speak of radiation pressure in this context, the same basic idea is in play. Is that what others here see about this particular device?

  • Charlie August 3, 2015, 0:58

    EMdrive paper and what original inventor Roger Shawyer believes is happening


  • Robert August 3, 2015, 12:57


    “If there is an effect, this class of experiments can’t demonstrate it.” “unreliable experiments can’t demonstrate anything reliably”

    Well, that about says it all! Let’s be clear, you are not just saying this particular experiment has unreliable results and not just that all the previous experiments done to date have unreliable results but that each of these experiments from All the teams, including NASA, are inherently unreliable and are thus worthless. In fact, you are asserting that no one has yet conceived of an experiment that would be reliable enough to test this presumed effect. That is a very convenient armchair scientific judgment if one does not like the data! Why should I hold your judgement higher that the judgment of all the dozens of researchers working in this area?

    “You asked if I think people should continue working on this topic. Should is a loaded word. If it were me holding the pursestrings then no. ”

    I have seen scientists actually interfere in others work they deemed unworthy. They try to get the research stopped or paper and patents rejected.

    “However as a private research effort, why not? People are pretty much free to do whatever self-funded research they want “.

    Much of the most exciting research is being done privately, quietly and behind the scenes these days. But it brings up the point that many scientists, scraping for funding, do work which tends to be safe, conservative and frankly uninteresting. Basically, it amount to public welfare for research.

  • Dave August 4, 2015, 4:02

    Robert, I don’t think I asserted anything about whether anyone has conceived an experiment that properly tests this effect. All they would have to do is engineer an experiment that defeats all of the significant error contributors.

    When measuring tiny forces on kg-class test articles it is not effective to use static force techniques like these where the test article is held in a test stand and the reaction force is measured via mechanical contact with a transducer. You can read back on this and other threads to see what some of the reasons are, and there are many more that haven’t been listed. There are simply too many confounding effects in this configuration. Static force methods are pretty much only used when the force to be measured is enough bigger than these error sources: the general practice is to use precision load cells with care taken regarding all of the mechanical boundary conditions, calibration, configuration drifts, etc. These guys did not. You can find numerous examples of effective technique in the scientific and engineering literature.

    For tiny forces there are better methods that are far more sensitive. These methods have been used to explore subtle effects, many orders of magnitude smaller than the forces reported here. Again, you can find plenty of papers online.

    For example, again, Cavendish’s 200+ or so year old gravity experiment beats this one by a factor of about a thousand in sensitivity (quality). I would think you’d be hungry for at least this standard to met.

    The state of the art for precision force measurements is incredibly sensitive, limited by the noise term from fluctuations due to the Brownian motion of the top layer of atoms in a test article. Relative to that standard, this experiment is something like using a yardstick to measure the size of an atom. Or stated in terms of this site, like using a pony when you want to get to alpha Centauri within a few years. I think the reason you sense strong reaction to this and similar papers is because commenting engineers & physicists are well aware of the exquisite level of the state of the art – for some this is their bread and butter.

    I don’t know that you should or shouldn’t hold my judgement any higher than anyone else’s. That’s got to do with how you crowd-source your own judgement-making. But for me, my policy is generally to assess others’ scientific judgements on the technical merits. By carefully attending to this I often learn something new.

  • Larry Kennedy August 4, 2015, 8:49

    Of course the theory at least is apparently wrong since in the NASA test both the “drive” and control article “worked”. It’s my understanding that both were built by proponents of the theory. Am I wrong about that?

  • Robert August 4, 2015, 13:52


    Thanks for your thoughtful reply. I do not disagree with you in principle but I do disagree in practice over this case. Tajmar is a talented researcher and his paper was balanced with a strongly caveat end conclusion. Likewise, the NASA team is very talented and competent. I am sure they are well aware of all these kinds of issues and sources of error. Critiques are important in research when they amount to published results refuting another paper and even then there is no guarantee the critique is always correct. But informal dismissive critiques in the press have the effect of hindering progress. Look at what the Wiki page says about the RF thruster;

    “As of 2015, there has been some hype about such engines in popular media, but few scientists take the claims about these designs seriously. Neither of the inventors of named drives have been able to reliably demonstrate thrust from one of their own theoretical designs. None of the experimental research showing positive thrust have been published in peer reviewed journals.[further explanation needed] There is concern that all results seen so far are simply misinterpretations of spurious effects mixed with experimental errors. And as negative results are almost never published, the existence of a few positive experiments may be due to publication bias. The research teams that have seen tentative results are continuing their work to remove potential sources of error, and see if they can explain the observed thrust using traditional physical models.”

    The second sentence is flat out untrue and is the merely the editors opinion. The sentence about peer review is false yet even if true is irrelevant. Not yet being peer reviewed does not make a thing untrue. Then the writer states “there is concern…spurious effects…..experimental errors.” Again, opinions and expressed in such a way as to bias the reader against it. Then the writer charges the teams with selection bias. Where is the evidence? I think he just made that up. Finally, he admits that teams are still working to resolve the issues but asserts they need to explain any real thrust using “traditional” models. So, it is only legitimate if it can be explained by “traditional” models and it’s rejected if it leads to new physics?

    The editors of this page are biased, hostile and doing a disservice to science and progress. That is true whether or not the effect is real.

    Also, rather large magnitudes of this effect have been seen by the team from Northwestern Polytechnical University (NPU) in China. These researchers measured 700,000 uN, basically, almost one Newton.

  • Robert August 4, 2015, 13:55

    Larry, there has been a serious misconception propagated by critics that the NASA test is meaningless because the “null” experiment also produced thrust.
    One can get that opinion if one relies on second hand accounts. The fact is that the “null” test in question was a test of whether or not the grooves Fetta claimed were essential to his design were actually necessary. They were not. NASA tested two version of that thruster. The actual null device, a resistor in place of the thrusters, produced no thrust.

  • Robert August 4, 2015, 14:11

    “For example, again, Cavendish’s 200+ or so year old gravity experiment beats this one by a factor of about a thousand in sensitivity (quality). I would think you’d be hungry for at least this standard to met.”


    I forgot to mention that I actually measured G in undergrad physics lab with two other students. We measured it three ways and got reasonable good results notwithstanding the vibrations from students walking down the hall. So if I can measure G in a basic undergraduate physics lab, and that was decades ago, not with the benefit of modern electronic sensors and computers, I think these professionals can measure a few micro Newtons.

  • Larry Kennedy August 4, 2015, 14:14

    @ william f collins
    The NASA tests were inconclusive as far as the effect. My point was that the test article that was supposed to demonstrate the drive and the control article which wasn’t had similar results. This seems to say something about the alleged underlying theory.

  • Robert August 4, 2015, 15:17

    “The NASA tests were inconclusive as far as the effect. My point was that the test article that was supposed to demonstrate the drive and the control article which wasn’t had similar results. This seems to say something about the alleged underlying theory.”


    That is not the case. The null test in question was about Fetta’s design with grooves vs. a design without grooves. NASA proved the grooves make no difference. Sometimes one finds that the reason one believes why something works is not the correct reason it works. It still works but for a different reason.
    That was the issue behind the working “null” device. I know there are folks out there who would like to say that if Fetta’s original theory was wrong, it’s all bunk anyway but they would be unreasonable.

  • Larry Kennedy August 4, 2015, 15:58

    @ Robert
    Indeed you are right, so it only indicates that Guido Fetta doesn’t understand what’s happening.

  • Larry Kennedy August 4, 2015, 17:40

    I agree. I’ve always held the position that that there could be something happening here. We KNOW we have things to learn. My objection has always been that some people have made all sorts of claims about something still unproven and unknown.

  • Dave August 4, 2015, 23:54


    I congratulate you on your successful gravity experiment. You beat the Tajmar experiment noise floor by a wide margin. Why are you still impressed by their results when you’ve done so much better in undergraduate lab? You even saw real environmental noise!

  • Robert August 5, 2015, 15:08


    Thanks. Why? Probably because I really respect scientists working their hearts out doing experiments a lot more than critics who merely talk about those experiments claiming the experiments and the experimenters are wrong. The arguments between a published or presenting scientist and his critic are not of equal standing unless the critic publishes or presents a scientifically valid critique in an equivalent journal or forum.

  • Dave August 6, 2015, 3:42

    Robert, you’ve given me a lot to respond to in that one short paragraph!

    I sense you are making a social, not a scientific, argument regarding personal vetting norms instead of a technical argument on the merits of the experiment. That aside, technical critiques don’t need to be published to be correct. How would that even work? There is an objective reality out there, and that reality doesn’t suddenly shift when something goes from unpublished to published.

    I don’t agree with you that the relative balance of respect implies who is right and who is wrong. That is a completely unscientific point of view. But I sense this is an important part of your view of this experiment. To me the balance of respect seems like a rapid & risky shortcut for forming an initial judgement, one that certainly ought to be quickly checked by exercising due curiosity before making any kind of commitment.

    “critics who merely talk” – I don’t think you know if any particular critic merely talks or has something to back it up; in any case critics are not obliged to do anything more than state their criticism. If you believe otherwise I’d be curious to know how you think they entered into a larger obligation and with whom.

    Critics with first-hand technical objections (and I tend to give zero weight to any others) by and large have sufficient experience designing, performing, and analyzing experiments to know what they are talking about. Do you claim they are raising irrelevant objections, and making things up? Any made up or irrelevant objections will quickly be found out, that is the process. The process is aimed at what is correct, not who is correct. You seem dismissive of the hard work some people have put in over their careers to achieve high levels of expertise, but respectful of the experiment team’s hard work – which was not even at the same quality level as your own few hour experiment in undergraduate lab. I think it is good and generous to give the benefit of the doubt, but there has to be doubt. Maybe you are allowing perceptions of politeness into the mix, which while understandable, is irrelevant.

    There is virtually no difference in standing between Tajmar publishing his AIAA conference presentation online and commenters publishing experimental critiques in this online forum. The two media venues have roughly equivalent channel-imposed entrance gates & quality control: both are moderated but neither is peer reviewed. In fact the conference has less standing: presenters did not have to submit their paper, but merely an abstract that could be revised, while here our submitted comments are final and moderated.

    I will agree though that statistically, conference papers are likely to have a higher scientific quality than comment sections. But that doesn’t mean that all conference papers are higher quality than all online comments, right?

    You are now the one asking for a high bar by calling upon commenters to publish papers. First of all, that work would be unfunded; spending a few minutes of personal time to comment online is one thing – writing and publishing a paper is another.

    Second, none of this effort would matter since it’s old, rehashed ground. The criticisms are for all intents and purposes already published in one form or another, some for hundreds of years, people have been doing this (designing experiments & precision mechanics) for a long time. The criticisms have been available for the team to consider since day one. In fact, because they are already in the literature it would be hard to publish them again – papers need to be covering new ground.

    Third, the experimenters need to mount a credible effort before others will spend the time. The thought here is that if the experimenters can’t be bothered to come up to speed, then why believe they would consider any serious objections? Besides, only a credible effort has a chance at being interesting and why spend time on things that aren’t interesting?

    Fourth, and most importantly, it’s up to the experimenters to convince us to update our understanding of physics & how to explore new phenomena, not the other way around, just as it is up to us to convince others when we present our own work. Why should they get a pass on this essential part of the scientific process?

  • Mark August 6, 2015, 13:00

    “no one has yet conceived of an experiment that would be reliable enough to test this presumed effect.”
    Here’s one:
    Put a solar powered drive in orbit and let it eat.
    See where it is a year later.
    If it’s in a higher orbit, then it worked.
    If it’s on the ground, then it probably didn’t work.
    Then we could stop spending mental effort in pointless debate.

  • Larry Kennedy August 6, 2015, 13:33

    I’ll go along with that idea , as long as You’re paying.

  • Alex Tolley August 6, 2015, 14:16

    That isn’t a tight enough experiment. c.f. Pioneer anomaly.

  • Robert August 7, 2015, 12:26

    The whole debate will be moot if newer results show macro level thrust in the Newton range with superconducting cavities.

    BTW, do those who are very skeptical also criticize Young Bae’s experiment with photon recycling that pushed a macro sized object with a laser beam?
    Do you say that was just a heated air effect or really photon thrust?

  • Robert August 7, 2015, 13:20


    I wrote a long, point by point response to you last post then decided that I could save us both a lot of time and effort by just saying that on virtually all points I either disagree or don’t accept your premise for this case. Yes, it’s up to the experimenters to convince the rest of us but let’s agree to let them try. We shall just have to wait and see where this goes. Thanks for the discussion.

  • Sergio August 7, 2015, 13:52

    There is another electromagnetic engine, called PNN.
    It hasn’t reached NASA’s attention (yet), however its thrust is visible at naked eye. It works with a different principle of Emdrive but like its counterpart is propellantless and reactionless. For the moment it could produce a thrust of 1/32g. Its inventor, Emidio Laureti, is going to held a public demonstration once the prototype will reach 1g of thrust (take off)
    Here’s the link (with video): https://neolegesmotus.wordpress.com/2015/08/04/aspss-electromagnetic-engine-tds-vf2/
    This engine is the result of 23 years of researches.

  • Robert August 10, 2015, 13:08


    Plate #3 reminds me a lot of a so-called Lifter, a lightweight device able to lift itself based on ion wind effects induced by very high voltage in air. Decades ago there was a researcher T.T. Brown who believed a highly charged asymmetric capacitor was a link between electromagnetism and gravity and thought he invented a new space drive. It turned out to just be ion wind and only works in an atmosphere. But a lot of hobbyists enjoy making Lifters and there is quite an
    undercurrent of conspiracy theory’s associated with the technology being suppressed and such.


  • Eniac August 13, 2015, 1:32


    Again, propellantless, not reactionless.

    Consider me having said propellantless. What matters is that no propellant is expelled in the opposite direction so that the net momentum change can be zero.

    … , the device would accelerate forever but as the velocity increases, the acceleration the power could support diminishes to but never quite gets to zero and in the limit approaching a terminal velocity.

    Now this is interesting. If the thrust/power ratio of the propellantless drive were to decline with increasing velocity, the perpetual motion argument could be avoided. I had just never heard that being said by anyone, and certainly there are mission plans out there that assume a constant thrust/power ratio.

    This version seems to violate the relativity principle, though, as the efficiency of the drive would be at a maximum when the drive “stands still”. “Still” relative to what? To my knowledge, the concept of an absolute reference frame, of “standing still”, aka the “ether”, has long been eliminated from physics. Are you going to reintroduce it?

    Lastly, if you do postulate an absolute reference frame and a velocity-dependent thrust/power ratio that avoids the perpetual motion problem, you will find that in the end you are still stuck with the rocket equation, unless you also have a magical energy source that requires no fuel. This is because propellantless is not the same as fuelless, and fuel has the same impact as propellent in terms of having to carry it with you.

  • Mark Zambelli August 24, 2015, 2:45

    Hmm… it’s not often we get presented with such a polarizing concept as this EM Drive, especially when it seems to have survived duplication by several teams. I have donned my sceptical-hat but do find the waves this drive is creating rather fascinating as the story continues.

    @Alex Tolley from August 6, 2015 at 14:16,

    Before the other ‘Mark’ stated his thought experiment, to which you mentioned the Pioneer anomaly, I had been thinking along those lines also. However, if the experiment were to be performed along the lines of the Gravity
    Probe B experiment (pop the ‘thruster’ in a box and measure any movement relative to the box) or even a space-based LISA type setup then can we eliminate the effects Yarkovsky or YORP in principle? Took forty years to get the gravity probe test aloft but I don’t want to wait that long ;)

  • Sergio September 2, 2015, 4:23


    Hi Robert,
    I believe ASPS’s engine is not a lifter.
    All their prototypes are sealed in a casing and they’ve always claimed that the device doesn’t move through movimentation of external or internal masses. Furthermore, after 23 years they would have surely realized that the propulsion was due to lifter effect.

  • Charlie November 2, 2015, 14:54

    A passing thought: (What follows is a re-post of the story linked at the bottom of this message)

    Discovering Propellantless Propulsion (2)
    Benjamin T. Solomon Become a fan
    Member, Nuclear and Future Flight Propulsion Technical Committee, American Institute of Aeronautics & Astronautics; founder, Xodus One Foundation

    Of course, discovering propellantless propulsion is based on the assumption that the physics of propellantless propulsion does exist. The now defunct Alcubierre Warp Drive and string theories, point to the need for care and empirical proof that the physics of propellantless propulsion does exist.

    Let’s deconstruct.

    In part 1 of this article I had suggested that a gravitational field can be structured into three parts, the source, the field & the field effect as determined by the massless g=tau.c^2. In addition that the mass source could be replaced by the Omega function (a yet to be determined function that behaves like the mass source). Thereby, having an engine generate the Omega function would clearly be the first step to propellantless propulsion.

    To prove that the physics of propellantless propulsion exists, we need to show three conditions are met. First, that gravitational acceleration is independent of its gravitating mass source, i.e. that the field structure is all that is required to accelerate an object. Second, that the omega function can be non-mass function. And third, to modify a gravitational field in a manner that departs from it natural direction.

    The first condition, acceleration is independent of the gravitating mass was proved correct with g=tau.c^2. Mass does not appear in this formula. Sure there is a linkage between mass and acceleration but g=tau.c^2 breaks this linkage into its component parts. Tau is the change in the time dilation transformation divided by the distance across this change in the local spacetime of the object. Note, Einstein used the local spacetime of a falling object to prove the Principle of Equivalence.

    To effect an acceleration on an object we only need to modify the spatial gradient of the time dilation transformations in the object’s local spacetime. This spatial gradient field is named the Non Inertia (Ni) Field. In a gravitational field these transformations are governed by the Newtonian Gravitational Transformations (NGT). From an engineering perspective, at lower velocities, NGT and General Relativity produce the same effects. First condition met.

    The second condition, that a non-mass omega function exists, is easiest to prove if we can prove that gravity is not caused by mass. This I did. In my book Super Physics for Super Technologies, I proposed that mass is a proxy for the amount of matter causing the gravitational fields. That part of matter that causes the gravitational fields is the quark motion in the atomic nuclei/nucleons. That is, it is subatomic motion, not mass, that causes spacetime to deform into gravitational fields. Additionally, in Bad, Bad, Bad Physics, I showed how the motion of the Laithwaite Big Wheel experiment alters the gravitational field. That both nano-scale and macro-scale motion create/modify gravitational fields. This is a significant step forward compared to contemporary physics which requires mass without explaining how mass causes this spacetime deformation. Second condition met.

    The third condition, ability to modify the gravitational field, is vindicated by both the Laithwaite experiments, and the Podkletnov experiments. Podkletnov showed that gravity shielding was observable but others could not reproduce his experiments. Not because he was wrong but because these physicists could not reach the experimental conditions required of them. Third condition met.

    To strengthen this case, electron motion in a magnetic field can be modeled exactly with Ni Fields. Nature provides an electromagnetic method to producing accelerations.

    Therefore, the physics of propellantless propulsion does exist. QED.

    The three parts of the gravitational field (source, field and effect) inform us on how to design propellantless propulsion. In A Theory for Momentum Exchange, I had explained that Nature exhibits shape-motion duality or velocity by particle shape deformation per NGT and LFT (Lorentz-FitzGerald Transformation). These are instantaneously concomitant effects. Note, unlike General Relativity which requires the altering the structural shape of spacetime with Ni Fields we are altering this shape-motion duality of matter in the spacecraft to effect motion.

    Thus, using fields to alter particle shape deformation will produce the concomitant velocity and acceleration effects. This implies that a wrap field that wraps the spacecraft would only need to modify subatomic particle shapes just as gravity does. This wrap field propulsion is safe as all matter in the wrap field is simultaneously moving in the same direction and no relative motion exists between objects within the spacecraft.

    The wrap field engine design should allow for two field properties, field modulation, and field vectoring. Field modulation is the ability to modify the field strength, and therefore the magnitude of the acceleration. Field vectoring is the ability to modify the field direction and therefore, the direction of this acceleration. Imagine if we could affect directional changes akin to the Phased Array Radar, the direction change would be practically instantaneous.

    More in a future post. I will continue to post as much of the non-proprietary information I have researched. This is so that by placing much of my work into the public domain, the US Patent Office (USPTO) cannot issue patents to other private interests for technology designs based on my work.