I am simulation the interactions of ~5GeV/n C12 onto a G4_Soft_Tissue target using the FTFP_BERT_HP + binary_ion physics list. Frequently the simulation crashed with the exception messages attached below. Essentially, it looks like it is trying to simulate the C12 + H1 interaction with the FTFP model but failed to complete. Is this a limitation or a bug? ------------------------------ In /afs/cern.ch/sw/lcg/external/geant4/9.6.p02/share/source/processes/hadronic/models/parton_string/management/src/G4VPartonStringModel.cc, line 81: ===> G4VPartonStringModel::Scatter(): fails to generate strings -------- EEEE ------- G4Exception-START -------- EEEE ------- *** G4Exception : had006 issued by : G4HadronicProcess::PostStepDoIt In /afs/cern.ch/sw/lcg/external/geant4/9.6.p02/share/source/processes/hadronic/models/parton_string/management/src/G4VPartonStringModel.cc, line 81: ===> G4VPartonStringModel::Scatter(): fails to generate strings Call for FTFP Target element H Z= 1 A= 1 Unrecoverable error in the method ApplyYourself of ionInelastic TrackID= 1 ParentID= 0 C12[0.0] Ekin(GeV)= 51.9031; direction= (0.062103,0.414857,0.907765) Position(mm)= (0.178202,1.19666,1.26898); material G4_TISSUE_SOFT_ICRP PhysicalVolume <Layer-2> ApplyYourself failed *** Fatal Exception *** core dump *** -------- EEEE -------- G4Exception-END --------- EEEE ------- *** G4Exception: Aborting execution *** Abort
I also have met this problem for days. Not only FTFP_BERT_HP but also QBBC, there's also this problem. And I have tried several version of geant4,such as 9.5.p01,9.5.p02,9.6.p01,9.6.p02 But nothing works. Anybody could tell me what is the primary cause of this bug??? Thanks.
I cannot reproduce the bug. I tryed C-12 + G4_H at C-12 momentum 12*4.8 GeV/c, 12*5. GeV/c, 12*6. GeV/c. All is working properly. More information is needed. Question: What is "material G4_TISSUE_SOFT_ICRP"? Comment: Simulation of Nucl-Nucl interactions must be in Geant4 9.6 released 17 May 2013 (patch-02), if all was doing well. It was not implemented in a full form in the previous releases: 9.5.p01,9.5.p02,9.6.p01 I do not know, what to do? Vladimir.
Hi Vladimir, G4_Soft_Tissue is a Geant4 NIST predefined material you can use straight away like G4_H. I was using v9.6-p02. Fan
Hi Fan! Up to now, I have no ideas how to find a bug for Nucleus+H interactions. There is only complicated way: In G4FTFModel.cc (geant4/source/processes/hadronic/models/parton_string/diffraction/src) at the begining there is a lines //#define debugFTFmodel //#define debugReggeonCascade //#define debugPutOnMassShell //#define debugAdjust //#define debugBuildString Uncomment them. Close 7 lines with G4int Uzhi;G4cin>>Uzhi; Save and compile (make) FTFmodel in geant4/source/processes/hadronic/models/parton_string/diffraction/ At run time there will be a lop of printings. I am interesting in the last pages. If you can do these with your code it will be great! Vladimir.
Email from Alberto Ribon with explanation of the situation. --------------------------- Subject RE: problem report 1516 From Alberto Ribon To Vladimir Uzhinsky Cc dwright@slac.stanford.edu Date 15.10.2013 09:39 Hi Vladimir, unfortunately the guy who reported the crash, Fan Lei, is not at CERN, so I cannot help you! However, your latest development of nucleus-nucleus interactions in FTF has been included for the first time in G4 10.beta (June 2013 release), NOT in G4 9.6 (nor in following patches). In fact, such development was too late to be fully tested and included in G4 9.6 (released on 30 November 2012). Moreover, we never introduce new developments in patches (because patches should contain only bug fixes, not new features!). So, the bug reported refers to the old, naive treatment of nucleus-nucleus interactions by FTF, in which the projectile fragmentation was very crude; while you were trying to reproduce the problem with the new, better FTF code for treating nucleus-nucleus interactions. In conclusion, in order to reproduce the problem, please take exactly G4 9.6.p02 (you can download the source code from the Geant4 web page), and then build it. Cheers, Alberto --------------------------------------------- So, the problem is resolved in G4.10 Its solution in G4.9.6 has no meaning. Now we can recomment to use G4.10 for simulation of nucleus-nucleus interactions.
