Skip to content

SRHD: Fix FV slope limiter method to use up-to-date primitives in evolution

Florian Atteneder requested to merge fa/srhd-fix-fv into main

Due to the cons2prim methods being tied to D, S, tau registered as dynamic_variables, we need some juggling to get the primitives of the current step.

This inconvenience is due to

  1. not (yet) separating the time stepping from the flux computation for the FV methods,
  2. and the 'hard coding' of the primitives to the cons2prim methods.

Side note: We face the same issue with the GRHD version.

This PR also adds some more reftests to cover the blastwave1 exercise. To this end we also updated the FV solver to work with non-periodic bdry conditions with bdry values taken from initial data.

Merge request reports