Created attachment 747 [details] Brief presentation describing the issue in the detector simulation. Reporting a data - MonteCarlo discrepancy observed when simulating X-ray interactions in a photoelectric based gas detectors in the 2-8 keV energy range. More specifically, the shape of simulated output spectra from monochromatic x-ray input beams show a prominent difference with experimental data in its left tail. The discrepancy is specific of the Livermore Pysics lists and has been observed with versions 10.7 and 11.0 of Geant4, while it was not present with version 10.3, which has a very good agreement with our data. We tracked down this discrepancy to a variation in the fluorescence emission from copper around ∼1 keV, which seems to be much larger in version 10.7 and 11.0 compared to version 10.3
Created attachment 784 [details] Plot of fluo lines excess also in version beta 11.1.b01 Copper fluorescence lines, from different G4 versions. As previously stated, version 10.3 matches IXPE data. Plot obtained by using G4EmLivermorePhysics in all Geant4 versions mentioned.
Using version beta 11.1.b01 does not solve the problem, as shown in the new plot.
Dear Koichi, Sebastien, please assign this problem to Luciano Pandola. Vladimir
Reassigned
Created attachment 809 [details] Plot of gamma-lines from 2 keV gamma-rays on G4_Cu (10.3 vs. 10.6) Zhuxin Li and I are working to reproduce the problem, but with no success. We run a job of 1e7 2-keV photons impinging on G4_Cu and the attachment shows the fluorescence spectrum. The difference in the intensities between 10.3 and 10.6 are below 3% and similar numbers are also found for the current development version of Geant4. Furthemore, we are unable to reproduce a few of the gamma-lines shown in your distribution https://bugzilla-geant4.kek.jp/attachment.cgi?id=784 and specifically the isolated line at 0.5 keV (marked as "v10.3") and the doublet at about 1.1 keV ("v10.7" and "v11.1.b01"). We are going to double-check with Auger electrons, but we would like to be sure that the plot of https://bugzilla-geant4.kek.jp/attachment.cgi?id=784 refers to fluorescence gamma only and to pure G4_Cu material.
(In reply to Luciano Pandola from comment #5) > Created attachment 809 [details] > Plot of gamma-lines from 2 keV gamma-rays on G4_Cu (10.3 vs. 10.6) > > Zhuxin Li and I are working to reproduce the problem, but with no success. > We run a job of 1e7 2-keV photons impinging on G4_Cu and the attachment > shows the fluorescence spectrum. > The difference in the intensities between 10.3 and 10.6 are below 3% and > similar numbers are also found for the current development version of Geant4. > Furthemore, we are unable to reproduce a few of the gamma-lines shown in > your distribution https://bugzilla-geant4.kek.jp/attachment.cgi?id=784 and > specifically the isolated line at 0.5 keV (marked as "v10.3") and the > doublet at about 1.1 keV ("v10.7" and "v11.1.b01"). > We are going to double-check with Auger electrons, but we would like to be > sure that the plot of https://bugzilla-geant4.kek.jp/attachment.cgi?id=784 > refers to fluorescence gamma only and to pure G4_Cu material. Yes, the plot refers to fluorescence gammas from G4_Cu material: there is a geometry constraint that points to a pure G4_Cu volume, and a particle definition constraint that considers only gamma particles, for the production of such plot. Just as a reminder, the comparison is between versions 10.3 and 10.7, Livermore polarized physics for photoelectric effect; options utilized are: SetFluo(true) SetAuger(true) SetDeexcitationIgnoreCut(true)
