Gradient crystal plasticity FEM implementation

Hi everyone,

I am recently implementing gradient crystal plasticity using FEM. As I understand, in order to obtain gradient of some plastic parameters (e.g. gradient of slip), additional dof (e.g. slip) should be added to each node and be interpolated with shape funcitons. And unlike conventional plasticity, which slip increment is computed locally at every integration point with stress update algorithm, both diaplacement increment and slip increment is computed globally.

My question lies in that: I have serious convergence issue with such kind of method. I cannot get physical result of slip increment. I am wondering if anyone has experience with gradient plasticity and could provide some hint on dealing with numerical issues of it. Thank you so much.

Best,

Hanglin