A hole in the middle of a 2D axisymmetric Mooney Rivlin rubber model is enlarged by a block that is displaced radially outward. Surface to surface contact is defined between the block and the hole. The solution fails to converge due to an element formulation error. Can you suggest anything that would help?
|
The following changes led to a successfully converged solution for this particular model: 1) CONTACT: a) Switch from Augmented Lagrange to Lagrange + Penalty b) Fixed penetration tolerance = 0.001 c) Define symmetric contact 2) PLANE182 element formulation: a) Switch from full integration to enhanced strain 3) Nonlinear options: a) NSUB,1000,1000,5 (start w/very small displacement increment) A steady state analysis was used (after trying unsuccessfully to run transient). Q2: A second issue is that it converges, but the contact behavior is not correct. There should be separation of the contact surfaces at the top of the contact surfaces and as the surface is pushed out radially, only contact at the bottom corner should remain. I changed the Contact Detection Method to "On Nodes - Normal from Contact." It was "On Nodes - Normal to target." This gave the correct contact surface separation but the solution did not converge at about time = 1.5. I've tried a few other changes but nothing has helped so far. Let me know if you have any other ideas.. A2: Redefine the order of the hyperlastic Mooney Rivlin model (was 9, changed to 5). The model now converges and the contact behavior is correct as well. |
||
|