If we assume that the claims for the EmDrive, Woodward effect and/or quantum vacuum plasma thruster are born out (notice I did NOT mention the Dean Drive), what would be the upper limit of potential applications based on predicted if not measured force per unit of input power and present-day power generation technologies (including nuclear)?

Some of the stated results seem to suggest that for practical power input, only modest forces yielding tiny fractions of a g could be applied to a spacecraft, making reactionless drive (if it actually works) suitable for nothing more than orbital stationkeeping or long duration robotic spaceflight.

Other numbers seem to suggest far greater capability: enough force per unit power to yield more than 1g vehicle acceleration for current power generation technologies. One could imagine flying cars, easy flight into space, and interplanetary vehicles completing trips in hours.

What is the source of the disparity in claims? Are there errors in the physics/math? Are the differences due to different mechanisms of operation (e.g. Woodward effect vs quantum vacuum plasma)?

Of course this is all based on claims which have yet to be rigorously proven out.

What are the claimed yields in force per unit power and what is the potential scalability for each of the proposed reactionless drive concepts/principles? Are we talking space probe microthrusters, flying cars, or flying battleships? How close is each principle to a peer reviewed science based answer as to whether it is a reality with a reference design or merely some experimental error, anomaly, or hoax? Are we talking months, years, or decades?

By the way... where does the input power go in these drives? If a test article is simply pushing against a stationary test instrument, is all input the power converted to heat in the device, or does some of it disappear into the quantum vacuum foam or whatever (as with the putative momentum transfer)?

  • $\begingroup$ I'm not familiar enough with Emmy Noether's first theorem to derive an upper bound, but the violation of the conservation of motion must mean there's a non-zero spatial asymmetry. But if there's such an asymmetry, it's smaller than we can currently measure. $\endgroup$
    – MSalters
    Aug 6, 2014 at 21:24
  • 5
    $\begingroup$ -1, for two reasons. Reason #1: This question is counting chickens before they are hatched. The most likely explanation of this supposed impossible drive is experimental error. Reason #2: This question is essentially asking "What do the laws of physics say will happen if the laws of physics are wrong?" $\endgroup$ Aug 10, 2014 at 23:56
  • 3
    $\begingroup$ Related question: how many angels can dance on the head of a pin? $\endgroup$
    – Kirkaiya
    Oct 21, 2014 at 22:46
  • 2
    $\begingroup$ @RussellBorogove, what you describe is a photonic drive. It's the most efficient drive possible in terms of mass expended per delta-V, but the least efficient possible in terms of energy expended per delta-V. Great if you've got plenty of cheap energy (eg. nuclear fusion, antimatter), but horrible otherwise. $\endgroup$
    – Mark
    Sep 8, 2015 at 7:59
  • 1
    $\begingroup$ And photonic drives take no less than 300MW/N according to project rho, so that doesn't fit the observations. $\endgroup$ Sep 22, 2015 at 1:38

2 Answers 2


For the EM-Drive, the disparity comes from the test results of the different test apparati.

at least 5 test run versions have been made:
— The 2 test rigs by the developer
— The Chinese replication
— The NASA test (tested in the same program as two cannae thrusters - see Brady, et al)

Each version produces a different thrust. This may be due to non-optimal configuration, sub-optimal materials, or improper frequency setting for the chamber.

Note that each replication was a ground-up build from specifications, rather than borrow the developer's version and test it, as far as I can tell.

Also note: the claimed peak efficiencies require a superconducting chamber; which said chamber is in development, and was not tested by NASA nor the Chinese, both of whom appear to be using room temperature devices.

I can say that the SPR model and another SPR Model don't look the same as the NASA one (photo at TheRegister.co.uk). They are similar in material, and show a generally conical profile, but there are visible differences in construction. The chinese replication looks different still (photo at next big future blog)

While I have seen and heard references to a European lab confirmation, I cannot find who that is at the moment.

For the Cannae drive... NASA replications in Brady, et al, show a difference between the slotted and unslotted drive - but again, the NASA test articles were manufactured in-house, not the same articles tested by others.

Anomalous Thrust Production from an RF Test Device Measured on a Low-Thrust Torsion Pendulum David A. Brady, Harold G. White, Paul March, James T. Lawrence, and Frank J. Davies retrieved 14 aug 2014 from libertariannews.org; abstract at NASA.gov. Note: the paper is clear that the cannae and the trucated conic chamber were tested.


from the report of yang Juan and NASA I computed that EmDrive was reported as 200-1000mN/kW and Cannae Drive as 1.4-5mN/kW

since the null reactor of Fetta worked despite Fetta's theory, and since it is about 0.5% of EmDrive measured performance, until new evidence one can consider the possibility that Cannae Drive is a parasitic EmDrive (it works because of a tiny asymmetry)...

The impact of EmDrive, and even Cannae Drive is huge for satellites and proves, provided you consider it is easy to install solar panel providing few kW, or nuclear reactor of few kW/MW. continuous acceleration of few Newton, and even milli-Newton are much higher than what ion-drive propose.

for satellite the main cause of satellite expiration is when they have no fuel... if they can correct their trajectory with an EmDrive, their life will be the one of their solar panel.

for heavier application, flying cars, rocket launching, one have to base prediction on Shawyer (see his IAC2013 paper). He estimate from his theory (probably broken, no problem if test succeed) that using superconducting cavities one could sustentate tons per kW... This is only projection of his theory, so unlike his reproduced test, this is to take with more caution.

note that EmDrive theory explain that momentum is conserved, that if the thrust is used to accelerate the reactor then the energy of the microwave is consumed (through Doppler effect). Conservation of energy and momentum is enforced according to EmDrive theory... This is only theory, so...

Be careful it can be an artifact, but a smart artifact linked to microwave resonance (see Emdrive FAQ), not a dumb one.

hope this helps.


Yang Juan EmDrive test http://www.emdrive.com/yang-juan-paper-2012.pdf NASA Cannae Drive test: http://arc.aiaa.org/doi/abs/10.2514/6.2014-4029

much data is missing to judge of the quality of Cannae Drive test. For EmDrive test, the FAQ on EmDrive site gives details.


Not the answer you're looking for? Browse other questions tagged or ask your own question.