Should the MTS and power law viscoplastic models use the time step?
[imported from redmine]
The viscoplastic model functions that compute the strain rate are using the time step as one of their parameters. It's used for some sort of cut-off: there's a hardwired strain of 1.0e-20 that is divided by dt to get a rate. This doesn't sound like the correct thing to be doing to me, as the result now depends on time step, which it shouldn't. I think this aspect of the viscoplastic models needs to be improved.