Firstly thank you for adding me to the group. Here's my problem:
I'm trying to write a UMAT for damage in 3-d beam elements. The problem is that the damage initiation and propagation are defined based on the axial strain values. When the umat is being called at every integration point (3 IP and 17 section points for each element), due to the discrepancy in strain values, it results in errors. I'd like to evaluate damage by computing an average strain (3*17) in the element and assign damage to each integration point based on that value. I was wondering if I could use a similar technique like yours to achieve this purpose. Thank you for your time for going through this email.
At the moment, as the umat gets called at each material point, I evaluate strain and assign damage based on that value. This is not numerically stable as I see. Thank you for your inputs.