Re: [fluka-discuss]: Heavy Ion collision using DPMJET in FLUKA framework

From: Francesco Cerutti <Francesco.Cerutti_at_cern.ch>
Date: Thu, 13 Nov 2014 22:43:14 +0100

Dear Sourav,

i. your previous crash is due to the fact that your colliding nuclei are
... not colliding, since in SPECSOUR you input the same lab momentum for
both, whereas they should have opposite direction (i.e. negative
z-component for one of the two).

ii. Then please note that SPECSOUR requires the total momentum, while
HEAVYION in the BEAM card requires momentum or kinetic energy per nucleon
(in fact per nuclear mass unit). Moreover, in the presence of SPECSOUR,
the only relevant parameters in BEAM are the particle species and the
energy/momentum, with the latter just used for transport initialization
purposes (the upper limit of stopping power tabulations is defined based
on it).

iii. You do not need EVENTYPE (obsolete), IONTRANS (ion transport and
interaction already on by default for HEAVYION beams), nor the DPMJET card
(whereas you obviously have to use - as I believe you are doing - an
executable where dpmjet is linked, like flukadpm3 generated by the ldpmqmd
script). You need instead a PHYSICS card with SDUM=LIMITS, specifying an
upper threshold for nucleon CMS momentum (110 GeV/c is fine for your
case). Standard USERDUMP is not an useful option here.

iv. Note also that the simulated Au-Au collisions are nuclear non-elastic
reactions, disregarding electromagnetic dissociation (which has a much
higher cross section, but in the present release is not yet available
for source collisions).

v. Your XYP plane is useless since you set it exactly upon the RPP
downstream face, this way your regAu4 is empty (i.e. has got zero volume).
The two VACUUM ASSIGNMA are odd, since the first in fact applies only to
regAu4 (FROM regAu4 TO regAu3, but the respective region numbers are
such as 2[regAu3] < 3[regAu4]) and the second anyway redefines the regAu4
material (keeping in mind that this region is meaningless as just said).

vi. Concerning CPU time, a 100GeV/n Au + 100GeV/n Au event takes on my
machine a bit more than half a second, so 1E5 collisions require less than
20 hours of CPU, i.e. less than 2h on 10 cores, which looks to me as a
reasonable time.

Kind regards

Francesco

**************************************************
Francesco Cerutti
CERN-EN/STI
CH-1211 Geneva 23
Switzerland
tel. ++41 22 7678962
fax ++41 22 7668854

On Mon, 10 Nov 2014, Sourav Tarafdar wrote:

> Dear Fluka experts,
> In continuation of the problem I came up with couple of days ago I want to add few more
> things.  After removing “SPECSOUR” card from my input file while keeping DPMJET physics
> process and executing FLUKA after linking DPMJET to it , “core dump” problem got removed.
>  So basically the input file is with just unidirectional Au beam with momentum of
> 100GeV/nucleon. For p+p collision by keeping “SPECSOUR” card it takes forever to finish
> even 1 FLUKA cycle.  Is it some limitation regarding SPECSOUR card for invoking heavy ion
> collision or something related to it is missing in my input file ? I am hereby attaching
> my input file.  The collision vertex is defined in vacuum within +/- 0.001 cm along X-Y-Z
> coordinate and it has been divided into upstream and downstream along Z axis by XYP cut
> plane. 
>
> Thanks
> -Sourav
>
>
> On Nov 9, 2014, at 12:44 PM, Sourav Tarafdar <Sourav.Tarafdar_at_weizmann.ac.il> wrote:
>
> Dear Francesco,
>   Thanks for the suggestions. I implemented your suggestions as far as defining the
> collision vertex is concerned , i.e., inclusion of BEAM card and HI-PROPE card on
> top of SPECSOUR card. Also I removed Au material region relevant for my collision
> vertex.  However I got the error on my terminal which looks like
>
> ======================= Running FLUKA for cycle # 1 =======================
> $FLUPRO/flutil/rfluka: line 359: 13280 Floating point exception(core dumped)
> "${EXE}" < "$INPN" 2> "$LOGF" > "$LOGF"
>
> Just wondering whether I am missing something in my input file ? My error file and
> output file are completely empty.
>
> Thanks
> -Sourav
>
>
> On Nov 9, 2014, at 12:28 AM, Francesco Cerutti <Francesco.Cerutti_at_cern.ch> wrote:
>
>
> Dear Sourav,
>
> 1) Shall I have to define Target region of Au  material at
> collision vertex
>  ? 
>
>
> not at all (unless your collisions are supposed to take place in gold
> instead of vacuum...). You defined the second beam species in the
> SPECSOUR card, where you have also said that your first beam is made by
> HEAVYIONs. So you need a HI-PROPE card to specify the HEAVYION nature.
> And you still need a BEAM card (where you shall put HEAVYION - making
> redudant your SPECSOUR WHAT(11) setting - and an energy per nucleon
> exceeding 100 GeV/n for transport initialization purposes).
>
> 2) Which scoring card should be used in order to retrieve
> kinematic
> information of produced particles and nuclear remnants
> after H.I. collision
> ? So far I have used ???USRYIELD??? card. 
>
>
> USRYIELD does not work yet for the SPECSOUR collision event. You can
> inspect the properties of the products by customizing the usrein.f
> routine - which is automatically called before source particles start
> to be transported - in order to look at the FLKSTK content (to this
> purpose remember to add the
>      INCLUDE '(FLKSTK)'
> statement).
>
> Cheers
>
> Francesco
>
> **************************************************
> Francesco Cerutti
> CERN-EN/STI
> CH-1211 Geneva 23
> Switzerland
> tel. ++41 22 7678962
> fax  ++41 22 7668854
>
> On Sat, 8 Nov 2014, Sourav Tarafdar wrote:
>
> Dear Fluka users,
>   I have been trying to simulate Au+Au collision at C.M.
> energy of 200
> GeV/nucleon using DPMJET3 within FLUKA framework. For
> defining the colliding
> beam I used SPECSOUR card with PPSOURCE type.  I am mostly
> interested in
> getting the kinematic variables (px, py, pz, energy , pid
> etc) of the
> particles after the collision and also the remnants of the
> nuclei after
> collision. My questions are
> 1) Shall I have to define Target region of Au  material at
> collision vertex
>  ? 
> 2) Which scoring card should be used in order to retrieve
> kinematic
> information of produced particles and nuclear remnants
> after H.I. collision
> ? So far I have used ???USRYIELD??? card. 
> Any suggestions will be of great help.
> Please find the attached input file for FLUKA and the
> output file after
> executing FLUKA using the command $FLUPRO/flutil/rfluka -e
> flukadpm3 test
> Any suggestions will be of great help.
> Thanks
> -Sourav
>
>
> <fluka_dpm_wotarg.inp>
>
>
>
>
Received on Fri Nov 14 2014 - 00:12:23 CET

This archive was generated by hypermail 2.3.0 : Fri Nov 14 2014 - 00:12:24 CET