I found a small bug in the file processes/hadronic/models/neutron_hp/src/G4NeutronHPCapture.cc In "ApplyYourself( const G4Track &, G4Nucleus &)" The problem: The variable "index" is declared twice. This has the following effect: for composite materials, ONLY the index of the first element is ever used to determine which capture process applies. The method has code to loop over cross-sections to determine which element captured the neutron-- but with the multiple definition of index, this code doesn't have any effect. Only the first element captures. The fix: CHANGE { xSec = new G4double[n]; G4double sum=0; G4int i, index; TO { xSec = new G4double[n]; G4double sum=0; G4int i; SMALL REQUEST: The second argument to this method, G4Nucleus &, was presumably included to allow the user to specify which nucleus is capturing a neutron. Currently the argument is not used, but if this functionality could be added, that would nice:) Best Regards, Bryan Tipton Caltech
Thank you for reporting this. Will be fixed in the next releae. As to the G4nucleus argument: This makes part of the geant4 model interface that is common to all hadronic final state generators. For the neutron transort, it is 'dummy'. It normally carries the information of the element scattered off. For the neutrons, the cross-section depends on isotope-wise information, though, hence this is not used. If you want to use the class stand-alone, you can 'feed' it with single element and/or isotope materials, and you get the effect, you would like to have. Many greetings, Hans-Peter.
