Thats quite cool.

You could have an entrance at the bottom and just walk up to higher gravities.This could almost be thrown in the "Manned Missions to Ceres" thread. Ceres has 3% gravity.One issue, with a structure as large as I imagine this you might only be able to put it exactly at the planets pole to be aligned with the planets rotation or else have to deal with continual twisting forces in the bearings. Ceres rotates every 10 hours for example...or would it be possible to engineer the axis to precess like a wobbly spinning top exactly in line with the planet's rotation?

or else have to deal with continual twisting forces in the bearings.

MORE artificial gravity! WooHoo!! Apologies if this has been done before on this site or elsewhere. Please send me the link if so. I did some math to come up with the "floor shape" for a spinning AG habitat sitting on a planetary surface. By floor shape, I mean the curve that enables someone to stand orthogonal to the surface and not feel any (significant) tugs to the side.(It's not a hard problem--would make a good homework question for a calculus-based freshman physics class--but it was difficult for me, anyway. It was also a lot of fun exploding the heads of myself and a few friends as we whiteboarded it out.)The curve is a parabola, y = (w^{2}/2A_{g})r^{2} where:w = angular velocity of the habitatA_{g} = the planet's gravitational accelerationr = distance from the hab's center of rotation.You might ask, why the heck would you even do that? Well, it might turn out that lunar gravity, for instance, isn't enough to support normal bone and muscle health. Or, adults might be fine, but children--either gestating or growing up--might need more gee to develop correctly. So there might be a reason to do it, or there might not. Either way, the math and physics are interesting.I’m doing a single-variable integration, setting w and A_{g} as constants. Someone more edumacated might come up with a more general solution in which these are variables.Here goes:At any point on the floor, there are two vectors: A_{g} and centripetal acceleration A_{c}.A_{c} = rw^{2}.The normal (to the floor) acceleration A_{n} is the sum of A_{c} and A_{g}. This is the “target” acceleration, most likely 9.81m/s^{2}. Its slope is A_{g}/A_{c}.The slope of the “floor curve” is perpendicular to this vector, so it is A_{c}/A_{g} = rw^{2}/A_{g}.Since you now have a slope, you can integrate to get the curve. The integral of y’ = rw^{2}/A_{g} is y = (w^{2}/2A_{g})r^{2}.That's it! Now you can write an equation for any combination of angular velocity and local gravity to design “floors” for rotating habitats on the Moon, Mars, or any other planetary body.For example, say you’d like a 1rpm habitat on the Moon, with a max “gravity” equal to Earth’s, 9.81m/s^{2}.A_{c} = sqt(9.81m/s^{2} - 1.63m/s^{2}) = 9.67m/s^{2}r = A_{c}/w^{2} = 9.67m/s^{2}/(2pi/60s)^{2} = 882my = (w^{2}/2A_{g})r^{2} = ((pi/30)^{2}/1.63)r^{2}y = 0.003365r^{2}Subbing in, you get y = 2618m.Now say your rotating pressure vessel is a torus, and the cross sectional diameter is such that the “bottom” point on the floor (min r) is 20m below the “top” (max r).r_{min} = sqt(2598/.003365) = 879mA_{cmin} = (879m)((pi/30)^{2}) = 9.64m/s^{2}A_{nmin} = sqt((9.64m/s^{2})^{2} + (1.63m/s^{2})^{2}) = 9.78m/s^{2}Again: if you see a horrible math or physics mistake (or a sup/sub problem or typo amounting to the same thing), please let me know. Don’t be shy. A diagram is attached to help visualize it.

First time this was proposed, to my knowledge, was by Peter Diamandis in: http://ssi.org/ssi-conference-abstracts/space-manufacturing-6/ in 1988.

Quote from: KelvinZero on 03/09/2015 07:46 PMor else have to deal with continual twisting forces in the bearings.Meh, the twisting-acceleration on the axis due to precession is 0.05% of 1g at Ceres' equator¹. (Can't work out the actual torque-force without knowing the angular momentum of the structure. But in linear terms, it's around a pound of force per ton-equivalent of mass. Your shaft will need to handle more than that for normal vibration and mass changes from people walking around.)[At the moon's equator¹, it's just 4 billionths of 1g. Rounding off error's rounding off error.]However, IMO, you wouldn't hang your structure from a central boom-arm, you'd support it on a ring-track. That opens up your maximum radius beyond the structural strength of the support arm. As a side effect, that track could deal with a huge amount of torque force, because it's then compressive on the rim, not twisting on the axis.¹ Equator is the worst case. Force drops as you increase latitude.

The fact that the "floor shape" is a parabola is incidentally something that has lead to some lunar space telescope proposals. I recall hearing about a lunar telescope proposal that would be essentially a rotating platter of mercury.

[...]1. A tower/tether that holds it in place for the spin and2. A track that it rests on to stop and only imparts force normallyThe most foolproof way to keep the track's force always normal would be to have it act perpendicular to the tether. Like this:

For this case, it would seem wasteful to use bearings to move the station's axis of rotation along with the rotation of the asteroid. Nonetheless, you still might want to use it for shielding and have something which slightly resembles a "surface" station for that reason.

As we look at larger masses, at some point this becomes unreasonable. 1 rotation per minute means that the body's gravity would flip relative to your ground that often. 1% change in gravity, applied over a full vertical circle, is enough to knock some things off your table.

you would just run it for 1-2 hours and then stop it to change passengers...

{url=http://www.crystalinks.com/precession.gif}http://www.crystalinks.com/precession.gif{/url}.

Come to think of it, one could build a ring around Ceres, attach it just enough to maintain position relative to the asteroid. use the asteroid to land or moor craft to and use a surface, maglev train to spiral up to the ring, flipping over bofore contact with the ring. Once on the ring, the "Train" would shunt to oneside to avoid impacting the spiral ramp.

Quote from: AlanSE on 03/10/2015 04:40 PMFor this case, it would seem wasteful to use bearings to move the station's axis of rotation along with the rotation of the asteroid. Nonetheless, you still might want to use it for shielding and have something which slightly resembles a "surface" station for that reason.Again, use a simple torus tensile structure. The force of precession is so low that a guide-track would be a fairly light structure. Gravity is so low that essentially it's a free-flying toroidal space-station, with an extra guide-track to keep it centred in its ring-tunnel.Anything else is overkill.