@Antzi's answer to EM drive requirements links to the recent Ars Technica article NASA’s EM-drive is a magnetic WTF-thruster; Test reveals that the magic space unicorns pushing the EM-drive are magnetic fields.. According to that article, results of a study of the proposed, "observed", and "measured" Em Drive phenomenon were recently presented at Space Propulsion 2018 by a group from the Technische Universität Dresden's Institute of Aerospace Engineering. The results indicate that the previous observations were wrong, and go on to address why.

If I understand the highly stylized article, the magnetic shielding of the earlier experiments was inadequate, and allowed some field to penetrate. This then interacted with DC current running in power leads. The RF amplifier uses ampere-scale (probably many) DC current and was a likely suspect, and discussed in footnote 3 of the original paper. I'd proposed a similar problem in The "Em Drive" paper is out - need some help understanding it but I'd assumed they'd competently shielded the Earth's field and so asked about the magnetic field produced by the current itself.

But at this point I'm just working off of the Ars Technica article.

Are there any technical write-ups of the Dresden work available? Slides of the presentation perhaps, or a video of it?

  • 1
    $\begingroup$ Strictly speaking, the original paper also discusses compensating for the Earth's magnetic field as well, and the setup used seems like it ought to work, so I don't know why it doesn't. I'll need to read the new paper to see if that gives any insights. $\endgroup$ May 22, 2018 at 19:57
  • 1
    $\begingroup$ @NathanTuggy It will probably take a while longer for this to really be concluded. I suppose an x, y, z triplet of Helmholtz coils could have been added to minimize the field near the apparatus, making any further shielding even more effective. $\endgroup$
    – uhoh
    May 23, 2018 at 2:14
  • 1
    $\begingroup$ The latest JBIS was a special on "Advanced Propulsion Concepts" and has a number of papers on Mach and microwave cavities. $\endgroup$
    – user20636
    May 23, 2018 at 13:29

2 Answers 2


Yes,you can find the PDF here

The abstract:

Propellantless propulsion is believed to be the best option for interstellar travel. However, photon rockets or solar sails have thrusts so low that maybe only nano-scaled spacecraft may reach the next star within our lifetime using very high-power laser beams. Following into the footsteps of earlier breakthrough propulsion programs, we are investigating different concepts based on non-classical/revolutionary propulsion ideas that claim to be at least an order of magnitude more efficient in producing thrust compared to photon rockets. Our intention is to develop an excellent research infrastructure to test new ideas and measure thrusts and/or artefacts with high confidence to determine if a concept works and if it does how to scale it up. At present, we are focusing on two possible revolutionary concepts: The EMDrive and the Mach-Effect Thruster. The first concept uses microwaves in a truncated cone-shaped cavity that is claimed to produce thrust. Although it is not clear on which theoretical basis this can work, several experimental tests have been reported in the literature, which warrants a closer examination. The second concept is believed to generate mass fluctuations in a piezo-crystal stack that creates non-zero time-averaged thrusts. Here we are reporting first results of our improved thrust balance as well as EMDrive and Mach-Effect thruster models. Special attention is given to the investigation and identification of error sources that cause false thrust signals. Our results show that the magnetic interaction from not sufficiently shielded cables or thrusters are a major factor that needs to be taken into account for proper µN thrust measurements for these type of devices.

  • 1
    $\begingroup$ I think it's worth noting that this is a conference paper, and as such, does not appear to have been subjected to a full peer review (this is common in this field). $\endgroup$
    – Chris
    May 22, 2018 at 2:50
  • 5
    $\begingroup$ @Chris I guess they will make a full peer reviewed paper once they finish the study. It's still very much WIP, which allows vice and other bottom grade news papers to use misleading titles such as "OMG SOMEONE IS ACTUALLY MAKING EM DRIVES AND THEY WORK" $\endgroup$
    – Antzi
    May 22, 2018 at 3:02
  • $\begingroup$ @Antzi : this is common in other areas as well. Once the Nature published an article stating that vitamin C can slightly alleviate one of the many harmful side-effects of radiotherapy used in cancer treatment. How did the newspapers report it? "scientists finally unveil what doctors don't want you to know: vitamin C instantly cures all forms of cancer!", linking to the original Nature article, knowing that most readers don't read the article itself. $\endgroup$
    – vsz
    May 22, 2018 at 10:13
  • $\begingroup$ The abstract seems to show that they are setting up a research infrastructure for testing it properly as opposed to claiming that the drive actually works. $\endgroup$
    – Baldrickk
    May 22, 2018 at 12:54

The paper is pretty decent (though the results of the experiment are somewhat disappointingly inconclusive), but the Ars Technica article is insultingly bad, missing quite a few important technical aspects and implications, then loudly proclaiming the author's deep and abiding loyalty to SCIENCE!

Why do I say the experimental work done so far was inconclusive? Because when you're trying to show that a hypothetical effect does not exist and is just noise from various known sources, you absolutely have to show that once all the known sources of noise go away, there's nothing left worth looking at. Otherwise, you can certainly demonstrate the presence of noise, but that doesn't prove that there is no signal buried in it. But the paper is quite clear that there's a fairly strong remaining amount of apparent signal (if one that is rather bafflingly inconsistent), and also quite clear that the apparatus isn't completely shielded and therefore at least one known source of noise is almost certainly sneaking through.

In short, the paper describes yet another apparatus that currently gives noisy data on the EmDrive, and uncovers yet more details that need to be handled to eliminate the noise properly. This is good and necessary work, if rather lackluster, but Ars Technica's tone of "Booyah, science just blew the EmDrive up!" is unsupported at best, and self-important tribal signaling at worst. Nothing about the EmDrive has been disproved. This is clear from the Conclusions section:

We continue to improve our measurement setup and thruster developments in order to finally assess if any of these concepts is viable and if it can be scaled up.

  • $\begingroup$ Thanks for your observations and thoughtful answer! This is why I asked for a better source; while Ars Technica often publishes well written pieces on science and technology, this isn't quite one of them. The author has done better but still seems to be trying too hard in other articles, although this could be coming at the direction of the management. $\endgroup$
    – uhoh
    May 23, 2018 at 11:33

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.