Historical research guide · 0529 · 2015
Direct Thrust Measurements of an EMDrive and Evaluation of Possible Side-Effects
Martin Tajmar · Georg Fiedler
In one minute
In 2015 Martin Tajmar and Georg Fiedler built their own EMDrive and put it on two different scales. Roger Shawyer’s thruster is a sealed, tapered copper cavity in which microwaves resonate, and the claim is that the taper turns radiation pressure into a one-way push with nothing leaving the craft. Working in close cooperation with the inventor, the Dresden pair first modelled the cavity numerically for high efficiency, then built a tunable copper breadboard and drove it with a commercial 700 watt microwave magnetron mounted on the cavity’s side. They measured it first on a knife-edge balance in room air and then on a torsion balance inside a vacuum chamber, with thermal and electromagnetic shielding, because the point of the campaign was to hunt the side effects that could counterfeit a thrust. Forces the size of the earlier claims did appear, about twenty micronewtons — but also in directions that should give nothing. Their verdict is careful, and it names a suspect: the magnetic pull on the power feed lines.
Connection
This is the campaign that turned the EMDrive from an argument into a measurement problem, and chapter 8 needs exactly that: the question is no longer whether a cavity moves a scale but whether the movement reverses when the cavity does. Chapter 1’s evidence ladder is built out of papers like this one, which report a resolution alongside a result and then name the artefact that has to be killed next.
From the source
The hardware was built to the inventor’s own design intent, not against it: the authors developed a numerical model to design their cavity for high efficiency in close cooperation with the EMDrive’s inventor, then built a copper breadboard whose resonance frequency could be tuned to match the magnetron attached on the side of the cavity, and measured the assembly’s Q factor before powering it.Abstract, sentences three and four
The same thruster was weighed twice on two unlike instruments — a knife-edge balance at ambient pressure and a torsion balance inside a vacuum chamber, both with thermal and electromagnetic shielding — precisely so that an artefact peculiar to one transducer could not survive as a result.Abstract, sentence two; and Tajmar’s review chapter, section 5, figure 8, which shows the three setups
Driven by a commercial 700 watt magnetron at 2.45 gigahertz into a cavity whose Q factor was only about 50, the experiment produced forces of the same size as the earlier claims — about twenty micronewtons in either sign — but it produced them in orientations that should give no thrust at all, which the authors report as a null measurement within a resolution that sits at the level of the claimed thrusts.Abstract, sentence six
In the lead author’s later account of this campaign the two transducers separate cleanly: on the beam balance in room air the thruster did read differently pointing up, down and horizontally, but on thermal drifts of hundreds of micronewtons, while the same thruster on the torsion balance in high vacuum showed no thrust above the twenty micronewton error bar — a thrust-to-power ratio below 0.03 millinewtons per kilowatt against the roughly one millinewton per kilowatt claimed elsewhere.Tajmar, Neunzig and Weikert 2021, section 2 and the comparison table row for Tajmar 2015; sheet at /library/stm-0d21110734
The authors state the scope of their own result rather than overstating it: the campaign can neither confirm nor refute the EMDrive, and its purpose was to assess independently the possible side effects in the measurement methods used so far — a purpose it met by identifying the experimental areas that need attention before any firm conclusion is possible.Abstract, sentences seven and eight
The named suspect is the wiring, not the cavity: the magnetic interaction of the power feeding lines running to and from the liquid metal contacts is identified as the most important side effect not yet fully characterised, and the one that has to be evaluated to improve the resolution — which is what the Dresden group then did, moving the amplifier outside the chamber and feeding radio-frequency power over a dedicated liquid-metal contact in the campaigns that followed.Abstract, final sentence; followed up at /library/stm-92d585638d section 5.1
The way in
https://doi.org/10.2514/6.2015-4083How to cite it
Martin Tajmar, Georg Fiedler (2015) Direct Thrust Measurements of an EMDrive and Evaluation of Possible Side-Effects doi:10.2514/6.2015-4083