[Clas12_software] RE : GEMC and particles.

[PROCUREUR Sébastien]

Dear Maurik,
Concerning the pid number, here is the (Geant4?) convention Maurizio uses:
  // Nuclear codes are given as 10-digit numbers +-100ZZZAAAI.
   //For a nucleus consisting of np protons and nn neutrons
   // A = np + nn and Z = np.
   // I gives the isomer level, with I = 0 corresponding
   // to the ground state and I >0 to excitations
According to this, your 1000020040 appears to be an alpha.
 
Cheers,
Sébastien.

________________________________

De: clas12_software-bounces at jlab.org de la part de Maurik Holtrop
Date: jeu. 28/10/2010 22:50
À: Maurizio Ungaro
Cc: Clas12 Software
Objet : [Clas12_software] GEMC and particles.



Hello Maurizio,

I have been spending some time looking into the internals of GEANT4 and the different particle models that it uses, particularly in the context of electro-production processes. I have the following observations that are probably relevant to the wider gemc community:

Method: send of order 3E9 electrons through a thin (0.25% X0) tungsten target and look at the particles produced by GEANT4.
Note: It is difficult and time consuming to check the actual cross sections of the particles produced. I am interested in very small angle scattering, where there is little data, even for e+ and gamma production.


1) The standard choice of physics model "QGSC_BERT" does fairly well scattering electrons and producing gammas and positrons, but does not produce anywhere near the correct amount of pions, and no kaons at all. I do see protons and neutrons.

2) The build-in list called "gemc", does the same level of bremsstrahlung, but does not produce any pions or kaons, and too few protons and neutrons. I am surprised about this, since the class HadronPhysics does invoke the G4ElectronNeuclearReacion class. This appears to be a bug?

3) The "LHEP_BERT_HP" appears to give more realistic pion and kaon production. I need to check the actual cross sections though, which is on my "to do" list. The "LHEP_BERT" list also suffers from insufficient pion production. The HP part (high precision neutron tracking) seems to do more than just do a better job with the neutrons.

I am starting up some QGSP_BERT_HP jobs to see if these improve pion production similarly as for the LHEP list.

For ALL the physics options, I get a not insignificant number of particles with pids larger than 1E10, most of the time this pid=1000020040, sometimes other values around it. This is quite outside the range of the permitted pids. I have no clue yet what is going on there and need to spend some debugging time there.


Any comments on this, especially some additional insight, would be greatly appreciated.

Cheers,
        Maurik



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/clas12_software/attachments/20101029/6d9e7979/attachment-0002.html>