Re: Kolb's quenching parametrization

From: Anna Ferrari <>
Date: Mon, 12 Apr 2010 01:11:56 +0200 (CEST)

Dear Vassili,

 the treatment of the quenching is implemented in FLUKA, up to
 now, only following a second order Birks parametrization.
 In this case it can be set directly in the input card.

 You want to use the Wright parametrization, that foresees only one
 parameter (eq.5 in the Kolb article): even if this is the case, I think
 that it's useful to briefly remind, first, how to proceed in the case of
 a correction 'a la Birk'. This could help to better address a solution in
 your case.

 In general the quenching correction must be applied at two different

 a) to the binned energy, that is scored via USRBIN or EVENTBIN:
    in this case you use the TCQUENCH card (see the manual);
 b) in the track reconstruction, to the deposited energy along the track:
    in this case you have to use the USERDUMP card, with sdum=UDQUENCH;
    what(2) and what(3) will be the first and the second Birks parameters,
    in g/(MeV cm2) (see the manual also in this case).

 As I understood, the two corrections are absolutely independent: if you
 are interested only to score the track variables, you can also not to use
 the first card.
 Let'see, now, how to retrieve the quenched values in the case b).
 In mgdraw.f you have to save the corrected values both along the track
 steps and for a "spot" deposition, that means:
 1) in MGDRAW, by doing the sum of the 'quenched' depositions along the
   track. Instead of summing the original energy depositions dtrack(k)
   (with k=1,...mtrack) you have to sum the corresponding quenched
   values, that are simply given by the dtquen(k,1) output variable of the
   QUENMG routine (the second number refers to the fact that you can load
   more than one set of parameters, here I refer to the set 1).
   The routine above is called because the setting as in b) forced the
   LQEMGD variable to be .TRUE.

 2) in ENDRAW, by considering, in the same way, the value of dtquen(1,1)
   (here the first 1 means that we have only one "spot" deposit)
   instead of the value of the variable 'rull'.

Let's come, now, to your correction. First of all: QUENMG is not a user
routine and you cannot modify it.
You can think to act on the energy deposits: you don't activate the
quenching via USERDUMP and you apply your correction BOTH to the dtrack(k)
energy depositions in MGDRAW and to the 'rull' spot deposition in ENDRAW.
BUT: while the first correction is (maybe) easy to do, I have doubts about
the meaning of the second one (maybe the authors can hlp us). What I know
is that the treatment of the quenching in case of spot depositions
is not trivial to compute, because we have always to do assumptions,
depending of the kind of spot. For example a spot can be due:
 - to particles under the transport thereshold (dE/dx is in this case set to
   the remaining energy);
 - to residual nuclei recoils (dE/dx is inthis case rescaled
   from the corresponding value, of a proton of the same energy);
 - to charged particles generated from low energy interactions, that are not
   in general transported (they are transported only for interactions
   on hydrogen and few others): in this case dE/dx corresponds to their
   production energy; .....

For this kind of complexity, I don't know if a meaningful correction
can be done simply in ENDRAW...

 Hope it helps,

> Dear FLUKA experts,
> I wanted to implement the quenching parametrization proposed by N.R.Kolb
> et al (NIM-A 368, 1996, pp.745-749), but didn't find where I can get
> dE/dx (or, what is equivalent, the "energy deposition length") for each
> energy deposition. What common is used by QUENMG?
> Regards,
> Vassili
> P.S.: Actually the parametrization I want to use is proposed by
> G.T.Wright,
> Phys. Rev. 91 (1953) 1282.
Received on Mon Apr 12 2010 - 13:58:05 CEST

This archive was generated by hypermail 2.2.0 : Mon Apr 12 2010 - 13:58:05 CEST