Re: Follow-up on the topic: run stopped (UBCHCK) and voxel geometry

From: Paola Sala <paola.sala_at_mi.infn.it>
Date: Mon, 11 Oct 2010 12:32:49 +0200 (CEST)

Hello
we need a bit ore information
If the voxel file is too big, could you send, plese, the .inp and .out files?
Also , could you run the debugger and analyze the core file?
There should be a core.xxxx file in the fluka_### subdirectory. Type:

gdb $FLUPRO/flukahp fluka_###/core.xxxx
gdb> bt

you'll get the traceback of the code at the moment of the crash. The
various levels are flagged with numbers. Identify the level
corresponding to ubchk and type (N=level number)
gdb> frame N
gdb> print *iubn
gdb> print *ixyz
Try also to identify the routine and line where ubchk was called, looking
at the traceback.

If you manage, please send us the info
Paola

> Dear Alfredo,
>
> Thank you for your answer concerning the ubchk error.
>
> I tried now to run the input with the voxel geometry with Cartesian
> binning
> (USRBIN) and made the bins very large. Nevertheless, the run stopped.
>
> Based on this, new questions arised concerning this problem:
>
> In the attachment I added the two *.err files that are produced out of
> these
> runs (one run is done with protons "protons001.err" and one with neutrons
> "neutrons001.err"). The *.err files contain many warnings or messages;
> maybe here you can see something suspicious?
>
> Is it generally possible that I can find out the coordinates of the
> particle
> that causes the run stop? Then I could find out if it is always the same
> location / region where something is happen that the simulation doesn't
> run further.
>
> The UBCHK error might be related also to the following facts:
> The input used for the simulations contains a voxel geometry with 900
> different organs and 7D+07 voxels. When I open the input in Flair v0.8.0,
> the last 4 organs (VOXEL897 - VOXEL900) are highlighted
> in red with a question mark.
> Further, in the flair output terminal it is written that these VOXELs
> do not exist and that the body "VOXEL" is invalid (see attachment
> flair_input.jpg).
> However, in the output of a FLUKA run, the voxel geometry is identified
> correctly and all organs (including VOXEL897 - VOXEL900) are listed
> (see attachment fluka_output). The whole geometry can be plotted in the
> flukaGUI with out any problem.
> Simulations of energy deposition (with Cartesian binning, USRBIN) within
> this voxel geometry run through for photons. Simulations of the same voxel
> geometry (with Cartesian binning) stop for neutrons, protons and Helium
> particles.
> Simulations without Cartesian binning run through in all cases (ph, p, n,
> He)
> and energy deposition can be scored in the VOXELs 897-900.
> Why FLAIR gives these warnings? Is it because something is going wrong
> with Flair or is there a serious problem with the voxel geometry?
>
> Thank you in advance for your help.
>
> Best regards,
> Andrea
>
>
> PS: Basically I can send you the input, but the corresponding "golem"
> (*.vxl file) has around 140MB. How could I accord you this large file?
>
>
>
> --------------------------------------------------------------
> Re: run stopped (UBCHCK)
> . This message: [ Message body ] [ More options ]
> . Related messages: [ Next message ] [ Previous message ] [ Maybe in reply
> to ]
> From: Alfredo Ferrari <alfredo.ferrari_at_cern.ch>
> Date: Fri, 17 Sep 2010 16:01:30 +0200
> Dear Andrea
> the message means something went really weird during usrbin scoring.
> That routine (ubchck) checks for memory out of bound when scoring
> usrbin's. It is nothing depending on the user input unless you
> used some "strange" usrbin settings (eg extremely thin bins or similar).
> Could you please send the input file and random number resulting in the
> problem?
>
> +----------------------------------------------------------------------+
> | Alfredo Ferrari || Tel.: +41.22.76.76119
> | CERN-EN/STI || Fax.: +41.22.76.69474 |
> | 1211 Geneva 23 || e-mail: Alfredo.Ferrari_at_cern.ch |
> | Switzerland || |
> +----------------------------------------------------------------------+
> On Mon, 30 Aug 2010, Zechner.Andrea_at_mi.infn.it wrote:
>> Dear FLUKA experts and users,
>>
>> I am trying to run an input containing a voxel geometry surrounded by
> concrete
>> walls (Geometry modelling was done using cutting surfaces as it is
>> recommended). In case it plays a role, the beam particles are low energy
>> neutrons. The input is running but after a certain number of particles
> the run
>> stops.
>>
>> I receive the following error message written at the end of the *.out
>> and
>> .*err file:
>> "Abort called from UBCHCK reason IXYZ-1 Run stopped! STOP IXYZ-1"
>>
>> The geometry is well checked against errors. The same geometry input ran
>> successfully with different voxel geometry some time ago.
>> Is the problem associated with the USRBIN cards? Cartesian and region
> binnings
>> are included in the input, but the usage of these cards is ok.
>>
>> Could somebody please explain me the reason / meaning of this error
> message?
>>
>> Thanks in advance!
>>
>> Andrea Zechner
>>
> Received on Sat Sep 18 2010 - 11:34:58 CEST
> . This message: [ Message body ]
> . Next message: L.Al-Sulaiti_at_surrey.ac.uk: "RE: fluence spectra"
> . Previous message: Giuseppe Battistoni: "Re: Proton scattering"
> . Maybe in reply to: Zechner.Andrea_at_mi.infn.it: "run stopped (UBCHCK)"
> . Contemporary messages sorted: [ by date ] [ by thread ] [ by subject ] [
> by author ] [ by messages with attachments ]
> This archive was generated by hypermail 2.2.0 : Sat Sep 18 2010 - 11:35:05
> CEST
>
>
>
>

Paola Sala
INFN Milano
tel. Milano +39-0250317374
tel. CERN +41-227679148
Received on Wed Oct 13 2010 - 00:17:38 CEST

This archive was generated by hypermail 2.2.0 : Wed Oct 13 2010 - 00:17:38 CEST