How can I define a rigid ring in a cylinder (shell)?

Problem:
I have got a cylinder that is clamped at the bottom edge. the top edge is supposed to be able to move in the global coordinate system - but the top nodes shall be rigid body (a stiff rigid ring). It is possible to model this by using *rigid body, but the problem is that a reference point has to be defined (leading to a movement of all nodes bound to this RP) and the model has to work for not symmetric loads on the shell (and buckling as well) as well.
Can anybody help?
thanks,
anton

 

[Drej]

I can see your problem using the *RIGID BODY keycard. The reference node in the middle of the ring - or wherever - in order to describe movement is your problem. The best thing I can suggest is to link the top nodes together using either:

(i) a set of beam elements with very high second moment of areas and elastic stiffnesses or
(ii) a set of link/spar/spring elements, with very high stiffness

both should work equally well, but you will not be burdened with the reference node.

Hope this helps,
- - D R E J - -

 

[fatiguetech]

Try MPCs to link the deformations of all the nodes in the rigid area to one node in the rigid area. The correct *MPC should be TIE and will allow you to tie DOF at the nodes to the displacment of a single node. It should not matter what node you use, as long as it allso is in the rigid area.

 

[corus]

The simplest way of making part of the body rigid is to tie the freedoms of a group of nodes together using *equation. otherwise simply alter the material properties of that region so that it is effectively infinitely stiff in comparison to other materials.