Improvements to TOV evolutions, pt 2

Follow up to !107 (merged).

Atm it seems that we can stably, and also fairly good, evolve a stable TOV in Cowling approximation with resolution that is comparable (in one dimension) with that of Deppe's paper, that is, n = 5, k = 16-24. What is interesting is that it seems like we don't need an AV for the stable star.

The key change seems to have been the 'freezing' of the atmosphere, combined with the velocity limiter from the Dumbser paper. Freezing atmosphere means that we restrict the points at which we are allowed to impose the atmosphere within ERK substeps. Which points are allowed to evolve during substeps is determined from the initial data of each step. The idea is to enforce a finite propagation speed for the atmosphere treatment.

Whether that works also in dynamical scenarios needs to be checked. With regards to the finite propagation speed. I think one can cook up a nice argument in relation to a CFL like condition to show that freezing the atmosphere makes sense (=^= light cones).

added 1d GRHD parameters labels

added 3 commits

• 2a923aaf - rescale smoothed viscosity by the extrema of D
• de8cc83a - fix time step computation to use maximum of smoothed_mu
• 966fc3c5 - introduce c2p_freeze_atm flag which allows to define when an atmosphere can propagate

Compare with previous version

added 3 commits

• d7723642 - fixup logic of freezing atm
• ede888d3 - init c2p_freeze_atm
• 8e535ab8 - update grhd ref tests to use c2p_freeze_atm_reset = false to restore previous behavior

Compare with previous version

added 1 commit

• e20025c3 - wip: stash some tov examples

Compare with previous version

Just realized that the current implement of the GRHD equations is wrong: I seem to have forgotten for the volume element

det(\gamma_{ij})

in the :spherical1d formulatoon.

Will merge this as is, because the changes here only affect c2p and not the evolution equations. I will follow up with a PR to fix the equations.

merged

mentioned in commit 4531207e