Hi all,

Ive got 2 little questions (that may seem stupid!). Im trying to implement in the Abaqus FE code a hyperelastic potential written in principal stretches. Ive already written an UMAT subroutine but there are problems of convergence (and I really dont know from where it comes from). So, I was wondering if writting an UHYPER subroutine wouldnt be simpler than an UMAT (since the expressions needed by Abaqus are easy to compute compared to the expressions required in an UMAT because no tangent matrix to define). The problem is that one has to define the partial derivative of the strain energy density with respect to the first and the second invariant of BBAR and my model is written in terms of principal stretches. How can I proceed to compute these expression easily and properly?

In fact, the implementation of a hyperelastic potential is not my final goal. I would like to implement a visco-hyperelastic model and my second question is : can I still use an UHYPER subroutine to do that or must I go back to an UMAT subroutine? I mean, is a UHYPER subroutine only appropriate for a classical hyperelastic model and can it be used for more complicated model (but still based on hyperelastic models)?

Thank you very much!

Have a nice day

Vincent