robust reconstruction methods for MHD #1431
Description
Describe the proposal
Consistent with what I found, Athena++ developers have found that the extrema-preserving versions of PPM, while performing beautifully for test problems, just don't really work for real simulations.
We should test whether our reconstructions work with either the DiskGalaxy problem or a supersonic driven turbulent box.
Some new WENO variants may help. There is WENO-MZ and WENO-AO-AH.
Describe alternatives you've considered
Always use PLM (with minmod, as we do in Quokka) for MHD sims.
Additional context
PrincetonUniversity/athena#690