search

jrleeman / rsfmodel

Rate and state frictional solver
MIT License
31 stars 12 forks source link

State law #11

Closed jrleeman closed 9 years ago

jrleeman commented 9 years ago

Better encapsulate the state evolution relations and make things a bit more independent. Addresses #7 and decoupling.

jrleeman commented 9 years ago

Ok, I used the velocity_component method again, but had to do a couple of small modifications. I think overall this is a nicer solution though? It does leave an array of the state evolution in each instance of StateRelation at the end, which is rather nice.

dopplershift commented 9 years ago

Yeah, looks good to me. Some good things I notice:

jrleeman commented 9 years ago

Maybe before we merge we should consider naming RateSate to something like Model. Maybe it would even make sense to have self.system = ExternalSystem() when we create that object? Just trying to think how to make it easiest to interact with, but maybe that's messy.

On numbered variables. Should we have any since the state relations are now getting set as a list?

dopplershift commented 9 years ago

Looks like another good step forward. And no, there shouldn't be any numbered vars now. To be clear, I thought those were positives (sorry that was unclear).

jrleeman commented 9 years ago

Maybe we're ready to merge this one and hit a few other issues then?