This is done and in PR #273. Last week I discovered that the issues I was having with it were just due to some errors in how I was computing the mean magnetic pressure during each time step (I was including ghost cells) and now that’s fixed and the test passes. It did highlight some issues with MHD I/O though.
Previously MHD was outputting one extra cell in each direction for each of the magnetic fields when it should have just been outputting one extra in the direction of the field. In PR #270 I fixed that, added a restart test for MHD and hydro, and replaces all custom
PI macros with built in
There have been some compiler warnings with MHD builds for awhile, mostly related to non-MHD code. In PR #272 I fixed them with a mix of converting
char arrays to
std::strings and the use of the
[[maybe_unused]] attribute specifier.
I added initial conditions and a regression test for the Orszag-Tang Vortex detailed in Gardiner & Stone 2008. This test passes. The PR will be up once previous PRs are merged as it depends on some of the utility functions implemented in them.