TRAK310 Altering the order of tracking secondary particles

When secondary particles are produced during particle transport, if the user wants to transport them in turn, they must be stored by a call to GSKING or GSKPHO onto the temporary stack JSTAK. As explained in [TRAK310], this is a LIFO structure, where the last particle introduced is the first to be extracted to be transported. It may be sometimes interesting to alter the order of tracking, for instance to decide whether the current event is worth completing or whether it should be skipped to go to next one.

It is possible in GEANT to alter the behaviour of the JSTAK structure using one of the words of the stack, normally reserved for the user. Stack ordering is activated via the data record SORD 1. This will set to 1 the variable ISTORD in common /GCSTAK/, which is by default initialised to 0.

When stack ordering is activated, the routine GLTRAC, which fetches the next particle to be tracked from JSTAK, instead of the last one, will select the one with the maximum value of the user word. This user word must be set by the user in the variable UPWGHT in the common /GCTRAK/ prior to the call to GSKING to store each secondary.

If, for instance, all the protons must be tracked first, in order of production, and all the other particles after, again according to their time of production, the following code should appear in GUSTEP:

      SUBROUTINE GUSTEP
+SEQ,GCTRAK,GCKING
      .
      .
      .
      DO 10 IG=1,NGKINE
*---         Younger particles tracked first
*---         Add to the current time of flight the time delay of
*---         the secondary
         UPWGHT = -TOFG-TOFD(IG)
         IPART = GKIN(5,IG)
         IF(IPART.EQ.14) THEN
*---         This is a proton, add a 100 microseconds to have it
*---         tracked first
            UPWGHT = UPWGHT+1E-4
         ENDIF
*---         Now store the particle in the GEANT stack
         CALL GSKING(IG)
  10  CONTINUE

Users should be aware of the fact that ordering the stack will change the results, because transport depends on random numbers which will be different for the same particle in the two cases.

TRAK399