Implement constitutive equations for viscosity contributions

[adigitoleo]

The effective viscosity should be a harmonic mean of the viscosities caused by whichever rheologies are of interest (e.g. dislocation creep, diffusion creep, Peierls creep, brittle (Drucker-Prager) deformation, ...) Viscosities for these rheologies usually depend on stress and often also on strain rate, temperature, etc.

• add pydrex.viscosity with methods to compute viscosity contributions based on constitutive equations Add pydrex.viscosity module #201
• add Fluidity steady-state example that generates a VTU with all requisite fields
• hook up the deformation regime selector (the existing DeformationRegime enum) Add diffusion and yielding #195

[adigitoleo]

If I don't get the Fluidity example included in time (before the beta release due date) then I'll drop it from the beta release. All of the other components are here, I think, and we already have some other Fluidity examples that can be a useful starting point for variable viscosity models.

[Patol75]

What is the main issue with the FLuidity example?

[adigitoleo]

What is the main issue with the FLuidity example?

I will need to run again and check, I think there was a problem with the temperature IC using some halfspace methods from utils, and an annoying thing was that it didn't like the mesh coming from my 2D mesh generator generated using our mesh module. Actually let me push out the beta release milestone until next week (01/10), this is worth including.

[adigitoleo]

For the mesh thing, the workaround is to just generate it from a .geo file instead, so I guess that works for now.