I hit another Wikipedia tunnel of information, starting with the first link and progressing to the last:

So, after reading everything (until hitting my knowledge limit when starting the papers)...

The Knudsen pump also referred to as "thermal transpiration pump" or "Knudsen compressor" is a gas pump that utilizes no moving parts. Instead it uses thermal transpiration, the phenomenon that gas molecules drift from the cold end to the hot end of a narrow channel.

Wikipedia has a link to a paper about using this to control attitude, but has this ever been used?

enter image description here

The paper seems to conclude it's possible at a 150km high orbit, but not beyond 200km. So has this been done? Any more information (in more laymans terms) on how this attitude control system would work? That paper was admittedly dense for my brain.

My overall impression is that this is a relatively cheap non-mechanical method for transferring liquids/gasses, using a low flow rate and temperature gradients. If you end up scaling it by making an array of them, higher and higher throughput an be achieved alongside other benefits (and, of course, complications).

My main question is:

  • Has this principle of a non-mechanical pump been used in space exploration?
  • If it has not:
    • Why not?
    • What may be applications of this in terms of space exploration.
  • If it has:
    • What was the application (this does not have to be related to the papers)?
    • Has the attitude application been seriously considered beyond theoretical?

Related: Is atmospheric skimming for propellant feasible?


1 Answer 1


Knudsen pumps are not used in space exploration because they do not function at the altitudes required by nearly all satellites, and do not function at all when completely outside an atmosphere.

As for why not, it's hard to parse the intended conclusion from the summary beyond "it may be possible at certain altitudes", but consider also some remarks about Knudsen Pumps from the intro to that same paper:

"...demerits like low pressure and the severe condition for the pumping system to be workable. [...] Researchers are are very doubtful of the use of the Knudsen compressor in space, since a gas in space is highly rareified."

In other words, it won't work without at least some atmosphere - the conclusion of the authors is that "150km altitude is possible, while the use at 200km is difficult." Presumably after 200km it's no longer workable at all. So it's completely useless for any probe or satellite in interstellar space or around a body with thin atmosphere.

Now compare that narrow operational band (150 to 200 km) to the (approximate) range of Low Earth Orbit (LEO) - 160 to 2000 km. (For reference, the altitude of geosynchronous orbit is 35,786km).

In other words, the Knudsen effect could theoretically barely work at the extreme low end of LEO, and absolutely won't work for higher orbits. And that extreme low orbit isn't desirable for almost any application, because it has significantly more atmospheric drag than higher orbits. The Wikipedia article on orbital decay notes that:

"Due to atmospheric drag, the lowest altitude above the Earth at which an object in a circular orbit can complete at least one full revolution without propulsion is approximately 150 km (90 mi)."

So Knudsen pumps only work at altitudes just barely beyond the range where powered flight is still required (below which they are totally unnecessary), but not at all at higher ranges where atmospheric drag is no longer a major concern (above which they are totally useless).

Given the cost of developing and proving any space exploration technology, it's not surprising that one with very, very few applications has been largely ignored in favor of solutions that work under a much broader range of conditions.

  • $\begingroup$ I didn't even consider how the prerequisite temperature gradient would be created or maintained in space to make the pump operate. But any heating/cooling method would require additional mass and energy consumption, and that source could be used for other control methods as well, including ones that might well be more efficient for their relative masses. $\endgroup$
    – brichins
    Oct 25, 2019 at 18:32
  • $\begingroup$ Isn't there a natural temperature gradient from the sun? The sun-facing side is hot, the opposite, is the cold side. $\endgroup$ Oct 29, 2019 at 18:43
  • 1
    $\begingroup$ @MagicOctopusUrn Yes, but it's not controllable or constant - your craft orientation might be fixed to the sun (unlikely) and produce a constant power source, but more likely it would be facing the planet and therefore rotating constantly. It might also pass through the planet's shadow, making the pump regularly unusable (though that could be worked around). Any collector for using that heat would likely need to be separate from the pump itself, and probably articulated for control, adding complexity and weight - which offsets the appeal of the "no moving parts" pump. $\endgroup$
    – brichins
    Oct 30, 2019 at 19:18
  • $\begingroup$ Great points! Could it possibly be used at L1 for station keeping seeing as the gradient there would be far more constant (and controllable)? You could face the craft in any direction to get any desired gradient to push the craft in any conceivable direction, right? Just by using the sun's rays. Then you'd obviously need your own source of atmosphere (instead of your own heat source). Guess that kind of defeats the purpose, it'd be easier just to bring mono-propellant at that point... $\endgroup$ Oct 30, 2019 at 20:15
  • $\begingroup$ @MagicOctopusUrn You just described an ion thruster :D $\endgroup$
    – brichins
    Oct 31, 2019 at 21:04

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.