I implemented MHD time steps for the 3D case and refactored the hydro time steps to move the time step calculations into their own “crossing time” function to bring the hydro builds to parity with the MHD builds, reduce code duplication, and improve readability. I also wrote tests for the two crossing time functions and the function that they call to compute the cell centered magnetic field.
I started on implementing the VL+CT (Van Leer + Constrained Transport) integrator by modifying the current 3D Van Leer integrator. So far I understand how the integrator works and I’ve gotten all the required arrays allocated and deallocated. My next step is to work on modifying the interface reconstruction kernels to work with MHD. I need to do some re-reading on exactly what kind of reconstruction the VL+CT integrator expects next week before I can move forward on modify the reconstruction kernels.
- Read Wibking & Krumholz 2021
- Fixed a bug in MHD initial conditions where only one cell was being set instead of an entire face