Assuming a spinning space station shaped like a wheel. With the inhabitant living on the inside of the Wheel to get spin gravity. Something like the station discussed in this post.

The needed speed and radius to achieve a certain gravity can be calculated using this page.

How large would the radius have to be for the people living in the station to feel that the ground is flat. I'm thinking that there are two aspects that to the feeling of flat ness that can be discussed separately.

  1. The ground looks reasonable flat
  2. It feels like flat ground when walking on it
  • $\begingroup$ I'm pretty certain I've seen a question very similar to this recently, maybe here or maybe on Space Exploration. $\endgroup$
    – user
    Commented Jan 5, 2017 at 12:38
  • $\begingroup$ I only searched worldbuilding, not space. I'll head over and see what I can find. Thank you for the link. $\endgroup$
    – Marius
    Commented Jan 5, 2017 at 12:57
  • $\begingroup$ Couldn't you just build flat surfaces to walk on inside the wheel? Like, for a simple example, a hexagon in a circle. Yes you would have these "corners", but if you had a large space station then the "corners" could be as innocuous as a speed bump, or a little dip in the road. If this is an acceptable approach to answering the question, then math can be used to find the optimal shape and size ratio/combination. $\endgroup$
    – Inbar Rose
    Commented Jan 5, 2017 at 13:23
  • $\begingroup$ I don't think hexagon would be a good approximation. In a hex only the center of each flatt area would have spin-gravity directly down. As you moved away from the center spin-gravity would gradually move away from perpendicular to the ground, and I assume you would perceive it as going up or down a slope. $\endgroup$
    – Marius
    Commented Jan 5, 2017 at 14:25
  • 2
    $\begingroup$ I agree that the hexagon would make things worse - as you got towards the corners, it'd act like a convex slope down. The corners of a hexagon won't be innocuous, as the interior angles of a hexagon are always 60 degrees (in other words, two 30 degree downward slopes meeting in the middle). You could of course introduce more corners (and sides) until they are fairly unobtrusive - but you're just moving back towards actually having a circle. $\endgroup$ Commented Jan 5, 2017 at 15:20

4 Answers 4


Looking flat

This will be a problem if there are long lines of sight - you will notice the ground sloping upwards even if it does so 10 km away. We can see ground sloping down and Earth's radius is over 6 Mm. The shorter the line of sight, smaller the radius can be without people noticing. Also, humans are used to ground sloping up due to hills, so I guess that as long as you restrict the line of sight to about 0.2r (so you see ground slope by at most 10 degrees), people will be (mostly) fine as far as sight goes. With somewhat claustrophobic design, this allows for a radius on the order of 100m. Also, in a station with a radius of approximately 200m and low ceiling, the floor would disappear behind the "horizon" of the ceiling close to the 40m line. Making the station 10 times larger would allow for view range of several hundred meters, which (judging by human depth perception limits) should be more than comfortable.

Feeling flat

I guess humans lack the ability to feel a difference of less than 1 degree (or at least can ignore it). While standing still, the length of your feet would require a radius of about 15m. Walking would increase this to about 50m and spreading your arms would require about 100m radius. Running involves losing contact with the ground completely, so I expect the problem shifts to you needing to gain angular velocity as you accelerate. At full sprint humans reach up to 10m/s, meaning that if we wish to keep angular speeds below 1 degree/s we'd need a radius of about 600m.


If you wish to simulate Earth-like gravity, physics is your enemy. You'd want little variance between the gravity at floor and head level - the difference is proportional to change in radius, so a 1% variation would require about 200m radius. Also, precession forces are significant above 2 RPM, requiring over 200m radius to maintain 1g.

NASA research

According to the wiki, NASA research led to the conclusion, that radii over 500m (implying about 1 RPM) are comfortable for people.

Good news

Humans adapt to almost anything non-lethal. Reducing gravity and forcing the brain to get used to some weirdness can easily push the required radius down. Visitors would hate it, but the locals would consider it natural. Other than that, a radius of 500m or more should be fine for most people.


Wikipedia pages on artificial gravity, space habitats and a few related concepts.

(Yes, I know, wiki is a bad source but this is not academia so give me a break.)

  • $\begingroup$ Could you actually link the articles you are using as your sources? $\endgroup$
    – Mołot
    Commented Jan 6, 2017 at 1:05
  • $\begingroup$ I forgot to keep a list. Will try to go back and add one later, it's 2AM here and I need to sleep. $\endgroup$ Commented Jan 6, 2017 at 1:06
  • $\begingroup$ A good answer should include the minimum size so that coriolis forces on the inner ear are below detection threshold. $\endgroup$
    – Innovine
    Commented Jun 13, 2020 at 19:53

They are two very different things. Virtually any radius would feel flat when walking on, as long as the circle wasn't so small as to actually be noticeably circular underfoot.

Looking flat is a whole other issue altogether. On a clear day you can see maybe 20km - so you're going to need a circle big enough that a curve is not really visible at that sort of distance. If it's allowed to look a bit hilly (say 6 degrees slope up) then your circle will need a circumference 60 times that - 1200km, or a nearly 400km diameter circle.


The problem of "flatness" is that a line, horizontal at eye level in point A, will inevitably hit the deck. So you will see the "ground" in front of you, and the brain will tell you that you have a steep hill ahead, climbing towards the "sky".

enter image description here

To avoid this, you must either:

  • have a radius so great that even with good visibility, the "dip" towards the ground of the horizontal light ray is not perceivable. For this, not even the radius of the Earth is enough.

  • reduce visibility. Assume that a dip of less than X% is not perceivable, then you can calculate the dip from the radius, and from there determine how much to restrict visibility; or, assuming a minimum required visibility, determine which radius will allow the dip to be below X% at that distance.

  • cheat! It should be possible to dope the atmosphere with a combination of a denser, innocuous gas like sulphur hexafluoride and water vapour. The purpose of this is to change the atmospheric refractivity index with altitude (i.e. distance from the ground) in such a way that a light ray traveling from "higher" to "lower" altitudes is refracted upwards. On Earth you can get this effect by heating the air nearer the ground, which causes light rays to bend upwards so that when looking at the distant ground, you actually see the sky, which the brain interprets as the road being "wet". The problem, of course, lies in having and maintaining a stable gradient; otherwise, you'll have a disturbing "heated air" shimmering effect in the distance.


This doesn't answer all the points if the question.

I once tried to design a ring space station that would take one day to make a full rotation (1,1574 * 10^-5 Hz) and where "gravity" (normal acceleration) would be constantly 9,81 m*s^-2.

I gave up on keep going with the world building because the radius gave 1.854.992.492,543 m = 0,01239 AU.

I'm pretty sure that a straight line wouldn't be noticed. Half a degree would be 16.187.863,297 m long. That's more than Earth's diameter.

-Edit: I've recently watched a video about artificial gravity and, as the angular speed would be 0,00069 rpm, there would be no canal sickness. Also, the Coriolis effect would be minimum.

  • $\begingroup$ "1 Hz" is wrong on many, many levels. 1 Hz = 1 cycle/second. Leaving aside that revolutions ≠ cycles, that would be 60 RPM, whereas just before you say that you want 1/1440 RPM. $\endgroup$
    – Matthew
    Commented Jun 14, 2020 at 15:27
  • $\begingroup$ Oops, sorry, I'm correcting it right away $\endgroup$ Commented Jun 14, 2020 at 16:10

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .