UMAT - dstran

[Erwinsche]

I try to implement en elasoplastic material equation in ABAQUS by using an UMAT.

After the warning "NEGATIVE EIGENVALUE" ABAQUS gives an incredible large strain increment (dstran) to my UMAT. Due to the large strain increment my UMAT diverges.

I think the UMAT works correctly, so there must be a problem in cooperation UMAT-solver.

Has somebody any experiences with this problem?

Merry Christmas

 

[xerf]

Some remarks:

1. Test UMAT on a simple model with 1 or only a few elements.

2. Use fixed time increments.

3. If the material undergoes fully plastic deformation (i.e. the yielding criterion is fulfilled at all integration points)and if the model is subjected to force/stress loading (as opposed to prescribed displacements) you can get non-convergence. This depends on the hardening module and typically occurs for (elastic-) perfectly plastic materials.

 

[Erwinsche]

Thank you XREF,

indeed I work with an elastplastic model. So can you explain, why non-convergence occurs? Are there any solutions?

 

[Erwinsche]

Thank you XERF,

indeed I work with an elastplastic model. So can you explain, why non-convergence occurs? Are there any solutions?