This experiment, which is being presented as an initial exploration of the feasibility of using microwave-generated plasmas in a sort of electic jet engine, has been attracting some attention in the general media, for example Ars Technica and Bloomberg News. To put it mildly, I am highly doubtful of the method used to estimate thrust, or even that it is demonstrating any thrust at all, but am I justified in that skepticism?
The experiment involves a vertical quartz-glass tube with an inside diameter of 24 mm. A compressor pumps air into the bottom, and immediately above that, the tube passes through the apparatus to ionize and heat the air, forming a plasma with a temperature up to 1,000 °C. The tube extends upwards beyond that for about another 23 cm, and the upper end is open to the air. The thrust measuring device is a hollow steel sphere, diameter 75 mm, with a hole in it allowing small steel balls to be dropped into it to change its weight. In use, this sphere is balanced on the end of the quartz tube, and the weight is adjusted until the airflow causes it to wobble (there is a video in the figures section showing this.) The actual measurement of the "jet propulsion force" is the minimum weight of the sphere needed to prevent this wobbling.
On the basis of these measurements, the authors claim to have demonstrated, in this device, a "total propulsion pressure" of up to 24 kN/m², which they say is "comparable to those of a conventional jet engine of an airplane". I am not sure what "total propulsion pressure" is, but my back-of-envelope calculation suggests that they might be thinking of a jet engine's static thrust divided by the nozzle area. Regardless of whether that is so, their figure is simply the weight of the ball divided by the area of the tube, with the corresponding figure for the no-microwave situation subtracted.
Where does one begin to analyze this device? For one thing, at the point where the ball is not yet wobbling, there is presumably no flow (or just a little leakage.) Surely this is a static situation, with little or no relevance to what thrust might be generated? Furthermore, does the weight at which the ball will begin to wobble depend essentially entirely on how much pressure the compressor can reach with its outlet blocked? And pressure given here is only 1/4 atm. -- how much thrust can you hope for, from such a small expansion ratio, even with the best possible nozzle design and a 1,000 °C plasma?