I know air circulation was a tricky problem on the ISS . I think that in a rotating station the air would be moving with the station, but the air against the “floor” would be moving faster or have greater force exerted on it than the air on the ceiling in the same way the force on your feet would be slightly more than on your head.

Wouldn’t this rotation naturally result in constant air circulation or at least movement, even if the station had walls separating it into multiple sections (each section would experience this same effect)?

Would it be enough of an effect to be able to meet health and comfort needs with an air circulation system no more complex than in an office building?

Note: I’m not referring to how the air conditioning would cool the air, replace oxygen, etc. Just how the circulation of air could happen.

  • 1
    $\begingroup$ Have a look at @AtmosphericPrisonEscape's answer to Radial variation of atmospheric pressure in rotating O'Neill cylinder-like ship? (Rendezvous with Rama) It's not a duplicate but it will be interesting reading! My feeling here is that the rotation will interact with temperature gradients (heat sources and sinks) to produce some unpredictable amount of localized circulation, but that wouldn't produce enough mixing to prevent build up of CO2 and other contaminants in some areas. $\endgroup$
    – uhoh
    Commented Oct 12, 2019 at 9:37
  • 1
    $\begingroup$ If the air in a rotating space station would circulate on its own, gas centrifuges for isotope separation would not work. $\endgroup$
    – Uwe
    Commented Oct 12, 2019 at 9:41
  • 2
    $\begingroup$ I was going to mention centrifuges. The rotation would create a pressure gradient towards the edge. But after everything settles out it would just rotate together as a chunk. Except for heat sources, like the sun-exposed side, which could create currents. And Coriolis forces would act on those currents. But with the likely size of the station, and partitioning into smaller sections, I don't think anything interesting or useful would come from that. $\endgroup$
    – Greg
    Commented Oct 12, 2019 at 13:08
  • $\begingroup$ @Uwe no, centrifuges (gas, test-tube, or any other type) don’t work by circulation. They just produce a strong radial acceleration and the denser or heavier bits ever so slowly diffuse outwards. Gas centrifuges do use a gentle flow but that’s done as a way of cascading them. $\endgroup$
    – uhoh
    Commented Oct 12, 2019 at 14:31

1 Answer 1


Would air circulate in a rotating station on its own?

Short answer, no.

Rigid body rotation of an air-filled station would result in a static and very very slight gradient in pressure (being highest at the point farthest from the center of rotation) but rotation alone wouldn't produce any forces that would sustain circulation. Friction would slow any initial motion to a stop and all you'd have is diffusion.

If we didn't have temperature and humidity gradients driving weather on Earth, the atmosphere would just rotate like a rigid body along with the rest of the planet, which is mostly what it does near the surface already. At the equator for example, air near the Earth's surface rotates lock-step with the planet at 465 m/s, appearing to be nearly at rest with everything else, unless there's a typhoon...

However, a realistic station will likely have temperature and humidity gradients, and those combined with the artificial gravity may result in random, unpredictable, and very slow circulation motion in some places, but it's not something you can rely on for comfort.

  • 3
    $\begingroup$ Note that the strength of the convection currents will locally be as strong as they would be on the surface of a planet with equivalent gravity. Here on Earth, convection currents aren't always enough (that's why central air exists), but quite often they are. Many electronic devices are passively air cooled...they will stay cooler in centrifugal gravity than they will in freefall. $\endgroup$ Commented Oct 13, 2019 at 19:48

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.