Fix egs_kerma FD bugs and expand its capabilities #610
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
😲 Nice! |
|
@rtownson this pull request requires updating the egs++ documentation once it is merged! |
|
@ftessier, @blakewalters and @rtownson I know this looks like a huge undertaking, but I have validated the results over and over again and the kerma and fluence results agree very well with other MC codes! 😉 |
rtownson
approved these changes
Jul 14, 2020
blakewalters
approved these changes
Jul 27, 2020
ftessier
force-pushed
the
fix-update-egs-kerma
branch
from
723f152 to
569aa9c
Compare
|
Rebased on develop. |
ftessier
force-pushed
the
fix-update-egs-kerma
branch
from
569aa9c to
884e35f
Compare
|
Working on the commits, not changing any code. Work in progress... |
- Fix bug in kerma calculation with the forced detection variance
reduction technique: Use proper equation based on the integral over
the path across the scoring region.
- Forced detection (FD) scoring is requested by defining an FD geometry
rather than via a flag. The rationale behind this is that in most
cases the FD geometry will direct the ray-tracing towards the scoring
regions (using a howfar call) which might only cover a small solid
angle. In cases of sources embedded inside scoring regions the actual
calculation geometry can be defined as the FD geometry. The logical
checks to activate FD are:
0 - Is there an FD geometry? AND
1 - Is particle inside a scoring region? OR
2 - Is particle outside FD geometry and directed towards it? OR
3 - Is particle inside FD geometry?
If the FD geometry is not judiciously chosen, time is wasted
ray-tracing photons which might no be crossing scoring regions.
- Forced detection can now be requested for individual calculation
geometries.
- A default FD geometry can also be provided. In that case, calculation
geometries without their FD geometry will use this one. If no FD is to
be used, this option must be omitted.
- Scoring kerma in more than one region is possible if their individual
masses are provided.
- Masses for each scoring region are required, unless one is only
interested in kerma in the whole volume, in which case only the mass
of the whole scoring volume is required (original approach).
- Dropped terms such as cavity region, cavity mass, cavity geometry in
favor of scoring region, scoring region masses, FD geometry, etc. Old
inputs will fail with a message instructing the user what input key
should be used.
- Fluence is still calculated for the whole sensitive volume. Option for
estimating fluence in each scoring region might be added in the future.
The E*muen file name can now start with an environment variable name.
Account for the possibility of multiple entries into the scoring volume when ray-tracing along the photon's direction when using forced detection (exponential track-lenght estimation). This is relevant for hollow geometries such as spherical or cylindrical shells. The scoreInCV method calls itself recursively if the photon re-enters the FD geometry or is still inside it.
Use a while loop to continue navigating through geometry when a re-entry to the scoring volume is possible. The scoreInCV method was called recursively when there were multiple entries to the scoring volume, which caused memory allocation errors.
- Use new input keys for scoring regions - Use $EGS_HOME for E*muen file location - Change photoelectric cross sections to xcom (unrenormalized) - Set number of histories to 4 million - Refer to expected efficiency gains with FD over TL estimation.
Add two new examples:
- depth dose in water for a 5 keV photon point source
- photon fluence outside a thick lead shield for a 20 MeV electron
point source
Rename emuen_icru90_1.5MeV.data to emuen_rho_air_1keV-1.5MeV.data, and
add two new E*muen/rho files calculated with the g application and the
mcdf-xcom option:
- water, from 1 keV to 1.5 MeV
- air, from 1 keV to 20 MeV
Fix the table of contents (TOC) of application documentation. Links to common inputs was defaulting to the title of first section of the "Common Inputs" page, which is "Introduction". It then appeared twice in the TOC of application documentation. Add optional text "Common inputs" to the TOC.
ftessier
force-pushed
the
fix-update-egs-kerma
branch
from
884e35f to
5138aaf
Compare
|
Cleaned up commits, reordered, squashed some small ones, etc. No changes to the files (except EOL white space removal and astyle formatting). |
ftessier
approved these changes
Aug 19, 2020
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Introduces new features in
egs_kermaand fixes several bugs in the forced-detection algorithm. Thelinear track-lengthscoring has not been modified as it was working as expected.Bug fixes:
based on the integral over the path across the scoring region.
Common input syntaxpage from documentedegs++apps manuals.New features:
without their FD geometry will use this one. If no FD is to be used, this option
must be omitted.
howfarcall) which might only cover a small solid angle. In cases of sources embedded inside scoring regions the actual calculation geometry can be defined as the FD geometry. The logical checks to activate FD are:If the FD geometry is not judiciously chosen, time is wasted ray-tracing photons which
might no be crossing scoring regions.
are provided.
in kerma in the whole volume, in which case only the mass of the whole scoring
volume required (original approach).
E*muenfile name can start with an environment variable now.cavity region,cavity mass,cavity geometryin favor ofscoring region,scoring region masses,FD geometry, etc. Old inputs will failwith a message instructing the user what input key should be used.
fluence in each scoring region might be added in the near future.
- depth dose in water for 5 keV photon point source
- photon fluence outside a thick lead shield for 20 MeV e- point source
E*muen/rhofiles calculated withg-appandmcdf-xcomoption:- water 1 keV to 1.5 MeV
- air 1 keV to 20 MeV
emuen_icru90_1.5MeV.data->emuen_rho_air_1keV-1.5MeV.data