|
MPI-AMRVAC 3.2
The MPI - Adaptive Mesh Refinement - Versatile Advection Code (development version)
|
This template uses the standalone mf physics module, not the legacy magnetofriction loop embedded in MHD. Its six evolved variables are three artificial magnetofriction velocities and B1:B3; there is no density.
It restarts from either a PotentialField or legacy MagnetofrictionalRelaxation snapshot. usr_transform_w imports only the three magnetic components and initializes the artificial velocity to zero. It then drives the lower boundary with the B-only sequence B_0001.dat, B_0002.dat, ... . The frame time is stored in each file; driving_time_scale in usr_list controls observational seconds advanced per simulated second and defaults to 12. No physical velocity boundary files are read. The observed B is imposed in the innermost lower ghost layer, while farther B ghost layers and the artificial mf velocity are extrapolated. Only the two magnetic frames bracketing the current observation time are kept in memory; the small array of frame timestamps is retained for lookup.
The Python DataDriven.ipynb workflow copies this template and writes time_dependent_mf.par with portable project-relative paths.