Here are some simulations that use the RadFLAH (radiation-fluxlimiter-aware hydro) variant of unsplit Hydro. They differ from generic 3T simulations in that there is no phyiscal distinction between the temperatures of electrons and ions; that is, electrons and ions are understood as forming a “matter” component. The “matter” component can have a different temperature from radiation.
./setup radflaHD/BondiAccretion -1d -auto +spherical -nxb=16 +mgd mgd_meshgroups=1 \ species=h1 ManualSpeciesDirectives=True +parallelio +uhd3tr \ -without-unit=physics/Hydro/HydroMain/unsplitNote that this setup differs significantly from the one described in Krumholz et al (2007), by using 1D instead of 3D geometry, and by excluding the central region from the domain instead of representing the accreting mass at the origin as a sink (or “star”) particle.
./setup radflaHD/RadBlastWave -1d -auto +spherical -nxb=16 +mgd mgd_meshgroups=1 \ species=h1 ManualSpeciesDirectives=True +parallelio +uhd3tr
For Case II of Zhang et al (2011), the following variant, using the experimental ExpRelax MGD solver implementation, may give better results:
./setup radflaHD/RadBlastWave -1d -auto +spherical -nxb=16 +mgd mgd_meshgroups=1 \ species=h1 ManualSpeciesDirectives=True +parallelio +uhd3tr RadTransImpl=ExpRelax
For both cases, the user has to create an appropriate parameter file; several sample *.par files can be found in the Simulation/SimulationMain/radflaHD/RadBlastWave directory.
Example configuration:
./setup radflaHD/SupernovaRad1D -1d -auto +spherical +mgd mgd_meshgroups=1 -nxb=16 \ -objdir mesa_sn_2T --maxblocks=1000 +pm4dev +uhd3tr +newMpole \ species=neut,prot,h1,he3,he4,c12,n14,o16,ne20,mg24,si28,s32,ar36,ca40,ti44,cr48,fe52,fe54,ni56 \ ManualSpeciesDirectives=True \ +parallelio -without-unit=physics/Eos/EosMain/multiTemp/Multigamma RadTransImpl=MGD \ -parfile=flash_working.par