# Why do space settlement designers use tori instead of cylinders for space settlements?

Almost all space settlement designs featuring artificial gravity by centrifugal force use the inner side of the outer part of the torus:

Why is it that I see hardly any cylindrical-shaped artificial gravity habitation areas like this one:

Isn't the surface area under 1g of centrifugal acceleration for cylinders more than that of tori? Won't tori have variable fluctuations of gravity?

If my point is valid, then why do people suggest tori for space settlement designs?

• There are certainly proposed designs for cylinders. One example is en.wikipedia.org/wiki/O%27Neill_cylinder In fact, here is a list of cylinders; en.wikipedia.org/wiki/… – uhoh Dec 25 '17 at 12:16
• A torus is the optimal shape for a rotating pressure vessel. The torus may contain a cylindrical floor for walking. If a cylinder like your drawing is used, it should be closed at the upper and lower end by a torus cut in half. But this shape is less optimal than a torus to hold the pressure of the atmosphere inside. – Uwe Dec 25 '17 at 13:30
• Kalpana One is a proposed cylindrical space colony. nss.org/settlement/space/kalpana.htm – Dean Calahan Dec 25 '17 at 15:46
• Most space settlement designers are artists asked to come up with a concept with very little thought given to engineering. They may also be influenced by earlier work (ISTR some early designs for space colonies that were tori). – Hobbes Dec 25 '17 at 20:51
• Constant centrifugal acceleration from head to toe of a standing astronaut in such a space station is not possible anway. – Uwe Dec 26 '17 at 15:07

Despite the obvious reason - a torus looks much more elegant than a barrel - I can come up with several points that favor a toroid:

• Rounded shapes can withstand the pressure difference between inside and outside much more easily. Large areas with sharp corners like at the end caps of the cylinder are difficult to make really tight. The parts of the ISS are actually cylindrical, but their diameter is tiny compared to that of a rotating station.

• Usually a fixed, easy to access, non-rotating part of the station is needed for thrusters, docks, experiments in micro gravity. This comes naturally with a torus, but is harder to build with a cylindrical shape.

• To get the same inhabitable space, the radius of a torus is larger than the one of a cylinder. This makes the toroid rotate more slowly which helps with joints and traversing between fixed and rotating parts. Also, it reduces the curvature of the floor.

• You mention the floor area at precisely 1g is larger in a cylinder. That's true, but there is no actual reason that the same acceleration is needed everywhere. Even if it varies by 10% throughout the station, I doubt that this can be felt by any of the astronauts. In your example double-toroid the inner ring has a reduced gravity, but this might even be a good thing, e.g. to have storage places where heavy objects can be moved more easily.

• Point #1--yup--but this doesn't preclude a design where you take your bagel-shaped station and slice it like you would a bagel, inserting a cylindrical station in it. – Loren Pechtel Dec 29 '17 at 22:32
• Point #2--you need a non-rotating center in either case. The only advantage of the bagel-station is more docking area compared to the volume. – Loren Pechtel Dec 29 '17 at 22:33

There has been a space colonization discussion started by Gerard O'Neill. The idea was to not colonize a planet or moon, but rather orbital space, particularly L5 or the trojan asteroids. They envisioned massive spinning colonies, miles across. They came in all designs, including torus and sphere, but most of them were cylinders. (with dome end caps) The advantage of the cylinder or sphere was efficiency of materials use. This worked well for huge structures because, with the radius of 1 to 3 miles, the rotation period was low enough to reduce the coriolis effect to negligible levels. The coriolis effect can cause a kind of sea sickness. In smaller structures, a torus, or better yet dumbbells can increase the radius for the same purpose. So that may be why you see a lot of toruses.

A spinning object is not stable if it is spinning around it's long axis. So if you imagine a colony the proportions of a long can, spinning around it's middle, after a while it would wobble and wind up spinning end over end. So that favors a torus, dumbbells, a sphere, or squat cylinder.

However O'Neill favored pairs of long cylinders attached together with beams at the poles. That way they would balance out and be able to keep spinning on the long axis.