The NDL database used by neutron_hp models seems to have some problems. The energy spectrum of gamma emmited after interaction of 14 MeV neutrons with light nucleis (carbon, aluminium, oxygen...) is very different to those obtaines experimentally. After some investigation, I found that the database, derived from ENDF libraries contain the LR flag of File 3 ENDF format, but don't use it. I've written a patch in G4NeutronHPInelasticCompFS.cc to take it into account, and this is sufficent to reproduce the gamma ray for 12C simulation. Problems are however always present for other nuclei. I've written a PERL code which translate ENDF database to NDL format. Using JEFF3.1 library gives much better results for 16O, but G4 code can't manage the case of LTT=3 and LI=0 of file 4 ENDF data, which is often present in recent libraries. Further investigation is on the way but any help would be very usefull...
I've also encountered that problem! So any news is appreciated... (In reply to comment #0) > The NDL database used by neutron_hp models seems to have some problems. The > energy spectrum of gamma emmited after interaction of 14 MeV neutrons with > light nucleis (carbon, aluminium, oxygen...) is very different to those obtaines > experimentally. > After some investigation, I found that the database, derived from ENDF libraries > contain the LR flag of File 3 ENDF format, but don't use it. > I've written a patch in G4NeutronHPInelasticCompFS.cc to take it into account, > and this is sufficent to reproduce the gamma ray for 12C simulation. > Problems are however always present for other nuclei. > I've written a PERL code which translate ENDF database to NDL format. Using > JEFF3.1 library gives much better results for 16O, but G4 code can't manage the > case of LTT=3 and LI=0 of file 4 ENDF data, which is often present in recent > libraries. > Further investigation is on the way but any help would be very usefull...
I am also facing this problem, using G4NeutronHPInelastic with G4NeutronHPInelasticData. For the moment, I tried only C12 and O16. With C12, we know that with 14-MeV neutrons, one should only observe the 4.44-MeV line. The simulation with Geant4, taking a graphite target, leads to 4.44-MeV, 7.65-MeV, 9.64-MeV and 12.71-MeV lines. (I got a good qualitative agreement Geant4/experiment when I removed by hands the lines in G4NDL3.10/Inelastic/Gamma/z6.a12 which correspond to non measured gamma ray lines...) On the other hand, gamma-ray lines are missing for O16: from simulation, one gets only the lines corresponding to {n+O16->n'+O16*;O16*->O16+gamma}. Are missing the lines coming from {n+O16->alpha+C13*;C13*->C13+gamma} and {n+O16->n'+alpha+C12*;C12*->C12+gamma}: the exited states of C13 and C12 seem not to be taken into account after O16 breakup.
Thank you for reporting problems and sorry for delayed response. LR Flag in MF3 Currently G4NDL does not have this entry. If we use this ENDF flag, not only Final State generators (NeutronHPs) but also many G4NDL files will need to be change and validated as well. Because of the large effort involved I have no plan for making this change. However we are investigating new HP models which will solve this problem. parameter LTT=3 (Legendre distribution for low energy and Tabulated distribution for High energy) Currently no G4NDL file except for JENDL_HE files has this value (3). Thus only Elastic FS can handle the parameter. (This is introduced at V8.2) It is not difficult to introduce it into Inelastic FSs, however there is no G4NDL file which uses this parameter. In general your messages is not a bug report but instead a result of design limitations of the original G4NDL file format. If these improvement is important to you, please send your request to Geant4 technical forum and then will attempt to allocate man power to this project. Tatsumi
