Is there a way to build a rotating multi G living platform?

[Ant]

I am trying to build a rotating station that has a thin (5m radius) long (600m) cylindrical shaft with two flat habitation module cylinders (100m radius x 20 m long) at each end and a 1/3 G gravity lab cylinder (100m from the middle of the shaft).

My problem is that the bar bell habitation modules only seem to allow spherical habitation modules at the end and does not have a way to add modules in the middle of the bar.  

On the other hand building my rotating station from Cylinder parts seems to work until I try to spin them.  When I try to spin the station it seems to want to spin each component cylinder on its own instead of calculating the CG and the moment of inertia for the whole structure.

BTW, if we need to write a mod to allow this to happen, it would be nice to think ahead and make it possible to create moving modules (for example allowing the 1/3 G lab above to move along the long cylinder) as this would allow modelling of several important functions like:
* An adjustable gravity platform (that depends on how far it is from the CG.
* An elevator for transportation between modules.
* A ballast that allows the operators to actively control the CG location.

Other useful features that can be modeled for long rotating cylinders (and cables) are the electrical resistivity of the structure and controllable electrical conductors that can interact with local magnetic fields to exchange energy and momentum between the station and its orbit.

[JoeStrout]

You are correct. The bells in a barbell configuration are always spherical; you can't construct a different shape as you're trying to do.

You can, however, combine multiple barbells to have both 1/3 G and 1 G bells.

[Ant]

Spherical bells make sense for storage tanks but they are not very useful for habitat areas. The inner half sphere where the centrifugal G points inward instead of toward the surface is useless as a habitat area and a planar surface would be much more efficient.
Even for the other half sphere, much of the area is wasted compared to a cylinder design.

[JoeStrout]

True, but spheres hold pressure much better than cylinders. You could still "flatten" the interior of your spherical bell by building it up in the middle (putting utility space or whatever below the ground). My intuition is that this would still be cheaper/easier than using cylinders at this scale.

[Ant]

A thin 100m radius sphere would probably require some local truss supports in any case to prevent local buckling.
A thin and flat 100m radius x 20 m high cylinder with slightly larger truss supports would have no trouble holding one atmosphere of pressure.

A double shell design with two layers that are 2 to 3 meters apart (forming a floor and ceiling) would require even less structural material for the truss and provide double the living area for a cylinder Most of the area of a two layers sphere on the other hand would create vertical or diagonal walls (with respect to local "gravity") which will not be as useful as a living area.

Talking about dual purpose thin structures, how hard would it be add a large parabolic antenna (or mirror) that are connected with a truss to the cylinder?

What about the structural stability of a 100mx100m solar array, does the weight/cost calculations include the supporting truss?
Do they scale up with the size of the array and the rotational acceleration?