You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
the sasvw problem is the only known candidate at this time
identify where material species is defined in the input and output
surfaceModel section of gitrInput.cfg and surface.nc in output folder.
identify where these options are parsed and used in the codebase
in gitr.cpp, within conditional blocks if( f.surface_model > 0 ), three sections can be found.
First section obtains the necessary size
Second section allocates multiple sim::Array classes of proper size to hold probability distributions
Third section populates the memory from netcdf input, and formats the probability distributions using make2dCDF and regrid2dCDF.
these are passed to the reflection operator as parameters
in surfaceModel.h and surface_model.cpp
all relevant logic is contained in the reflection class operator()
identify the outcome of the surface model and work backwards to determine needed changes - the "endpoints"
two endpoints seem apparent:
First, the new vx, vy, and vz of the particle that is reflected or sputtered.
Second, the grossErosion, grossDeposition, and other arrays are updated with particle weights.
study how the reflection operator works to determine what needs to be changed to add multi-species
the places where interp3d occurs appear to be where the probability distributions are actually sampled
the r7-r10 variables appear to be where the Monte Carlo randomness comes from
open questions
some of the arrays in surface.nc output seem to be modified in geometry_check as well as surface_model. How do these functors work together to update the surface model information?
do any of the outputs in surface.nc need to be material-species aware? Do new bins need to be added to distinguish between the particle species that would currently end up in the same bin?
what are the next steps to add multi-species?
The text was updated successfully, but these errors were encountered:
approach:
define a multi-species test problem
the sasvw problem is the only known candidate at this time
identify where material species is defined in the input and output
surfaceModel section of gitrInput.cfg and surface.nc in output folder.
identify where these options are parsed and used in the codebase
in gitr.cpp, within conditional blocks if( f.surface_model > 0 ), three sections can be found.
First section obtains the necessary size
Second section allocates multiple sim::Array classes of proper size to hold probability distributions
Third section populates the memory from netcdf input, and formats the probability distributions using make2dCDF and regrid2dCDF.
these are passed to the reflection operator as parameters
in surfaceModel.h and surface_model.cpp
all relevant logic is contained in the reflection class operator()
identify the outcome of the surface model and work backwards to determine needed changes - the "endpoints"
two endpoints seem apparent:
First, the new vx, vy, and vz of the particle that is reflected or sputtered.
Second, the grossErosion, grossDeposition, and other arrays are updated with particle weights.
study how the reflection operator works to determine what needs to be changed to add multi-species
the places where interp3d occurs appear to be where the probability distributions are actually sampled
the r7-r10 variables appear to be where the Monte Carlo randomness comes from
open questions
some of the arrays in surface.nc output seem to be modified in geometry_check as well as surface_model. How do these functors work together to update the surface model information?
do any of the outputs in surface.nc need to be material-species aware? Do new bins need to be added to distinguish between the particle species that would currently end up in the same bin?
what are the next steps to add multi-species?
The text was updated successfully, but these errors were encountered: