Problem 1823

Created attachment 381 [details]
Detailed documentation of the observations and of the simulation settings to reproduce them

No backscattered electrons are observed with G4eSingleCoulombScatteringModel in Geant4 10.2 (public release, 4 December 2015) and geant4-10-02-ref-01 (internal reference tag), while they are observed with G4eCoulombScatteringModel. 

The electron backscattering fraction generated with G4eCoulombScatteringModel is consistent with experimental measurements (see slides presented at CERN Simulation weekly meeting on 8 December 2015, https://indico.cern.ch/event/398322/session/1/contribution/3/attachments/1201521/1748756/ebssft7.pdf).

The author of G4eSingleCoulombScatteringModel Mauro Tacconi, whom we communicated with in December 2015 and January 2016, observes backscattered electrons with test58; we do not, using a clone of test58 PhysicsList (nor using clones of TestEm5 and G4EmStandardPhysicsSS physics configurations). 

bssim test code, made available to Geant4 maintainers at the time of a previous problem report, has been updated to include tests exhibiting this problem; the DOI of the updated code is http://dx.doi.org/10.5281/zenodo.45511.

Extensive documentation of our tests and their outcome is included in bssim git repository, in the file documentation_mott.docx. For convenience, the same documentation is attached in PDF format.

As a cross-check, we verified that backscattered electrons are produced in the bssim environment not only with single Coulomb scattering through G4EmStandardPhysicsSS, but also with G4EmStandardPhysics. The validation results in  https://indico.cern.ch/event/398322/session/1/contribution/3/attachments/1201521/1748756/ebssft7.pdf confirm that backscattering is observed with all the other configurations included in our tests.

Presumably the apparent mismatch between our observations and the authors' observation of backscattered electrons is due to a communication problem regarding the instructions how to use G4eSingleCoulombScattering. Hopefully this report will contribute to clarify the situation and to identify the proper way of using G4eSingleCoulombScatteringModel in the bssim environment to observe backscattered electrons.

We thank Geant4 electromagnetic maintainers in advance for investigating this issue.

Maria Grazia Pia and collaborators