#1071: ensure atmosphere is properly set after ID ------------------------------------+--------------------------------------- Reporter: knarf | Owner: Type: defect | Status: review Priority: major | Milestone: ET_2012_11 Component: EinsteinToolkit thorn | Version: development version Resolution: | Keywords: ------------------------------------+---------------------------------------
Comment (by rhaas):
I like the idea of actually marking atmosphere points by a mask and not not only by their density since it will actually freeze their evolution.
I would like to point out though that this will change current behaviour even for ID thorns that know about atmosphere. Right now the atmosphere mask is really mostly a "do-not-evolve" mask (or set-to-atmosphere) since it is set in GRHydroUpdateAtmosphereMask based on densrhs and taurhs. Also currently the atmosphere mask is cleared after each regrid so if we want to adopt the new logic in the patch of marking all points where rho<rho_min in the atmosphere mask (which would be what the name implies but not what it currently does) then we should also schedule GRHydro_InitAtmosMask in PostRegrid (and all other places where currently SetupAtmosphere is scheduled). Also we schedule AtmosphereReset in MoL_PostStep which clears the atmosphere mask (since wk_atmosphere .eq. 0 is the default) in normal runs.
With the patch one should be a careful I believe as to what happens in the cells that border the atmosphere.