Analysis of an sphaerolager (metal/rubber component) in Ansys

[scarules2006]

Hello everybody!
I'm trying to analyze an sphaerolager in Ansys( the sphaerolager is subjected to an radial displacement,as last step in which it doesn't converge). Like in the case of an bush, which I've simulated earlier (also metal/rubber component), it gives me a hard time to converge to a solution. In the case of the bush, the solution came after I've used Mooney Rivlin 3 parameters model, otherwise the analysis would have worked only to 70%(max, by using Neo Hookean model, using other models worked less than 70%). Can someone please guide me to what can I follow or where to look for tips so that I can obtain an convergence to this problem? I've tried changing the laws for rubber, my elements have an aspect ratio less than 5, I've change the definition of contacts but nothing seem to work and I really don't know why or where to look. I get the same error, that my elements have become highly distorted.
Thank You very much!
Any idea would be greatly appreciated!

 

[btrueblood]

Are you using the nonlinear geometry solver as well as nonlinear material model?

Do you know if the element distortion is beyond the limit of the nonlinear geometric solver to handle; is there an alternate element you could use? Does Ansys support re-meshing of the distorted part?

 

[scarules2006]

Hi trueblood!
Thank you for your answer.
I should have specified a few things from the beginning, I'm new to FEA(kind of, I made some linear analysis when I was student in Femap,in those analysis the problem was result interpretation not convergence) and also new to Ansys and I'm using Ansys Workbench 15. Regarding your first question yes, I'm using nonlinear material model but to the other questions I don't know how to answer.Can you be more specific?I can remesh everything and try to solve it again, I've tried that by using all sorts of methods and all kind of elements.
Kind regards,
Marius

 

[scarules2006]

Another thing, the analysis is static structural.

 

[btrueblood]

Marius,

Sorry I'm not an expert in Ansys either. I do know that the more advanced versions of Ansys have solvers that can handle "geometric nonlinearities". The most common example of that term is large displacements, where the small angle rule (sin x is approx. equal to x) is violated. There are other problems that can occur when the element shape becomes badly distorted, generating singularities or truncation errors due to the subtraction of two near-singular values. Not having used the nonlinear solvers, I can't say how the method in Ansys works; I would suspect that it involves special element types as well as methods for handling large displacements. A technique I have heard of being used (but don't know if Ansys handles it) is to halt execution well before the nonlinearity begins (usually by solving at some fraction of the load/displacement), then re-mesh the deformed body, map the stress field onto the new elements, and continue the simulation with the next increment of load/displacement.

 

[scarules2006]

Thank you btrueblood! I appreciate that you are trying to help me!
I've managed to solve the problem.It was from my mesh, more precise from the mesh of a body(rigid) used in a previous step, that I've put in contact with my sphaerolager, in this same analysis.
Best regards!