Re: [fluka-discuss]: Floating point exception scoring ACTOMASS with two step method

From: Marcus Matthias Morgenstern <marcus.matthias.morgenstern_at_cern.ch>
Date: Mon, 14 Dec 2015 10:29:30 +0100

Dear Vittorio,

I checked with the latest 64-bit respins, 2011.2c.2 and 2011.2c.3. I'm
using the gfortran version and tested with gcc-4.8.3 and gcc-5.2 version.
 From a discussion with my colleagues I guess the problem is the
conversion of Bq/cm^3 to Bq/g via the density of the material in the
scoring region. Apparrently, for vacuum it would then divide by zero
which would yield a floating point exception if this is not captured in
the code. Could this maybe the reason?

Thanks,
Marcus

On 12/13/2015 09:21 PM, Vittorio Boccone wrote:
> Dear Marcus,
> can you report which FLUKA version and flavour (version, release
> number and architecture) are you using and which Linux distribution
> (gfortran/gcc version)?
> Best,
> Vittorio
>
> On Fri, Dec 11, 2015 at 12:00 AM, Marcus Matthias Morgenstern
> <marcus.matthias.morgenstern_at_cern.ch
> <mailto:marcus.matthias.morgenstern_at_cern.ch>> wrote:
>
> Dear fluke experts,
>
> I came across an issue when scoring ACTOMASS with two step method,
> i.e. setting material for decay to vacuum. This seems to only
> occur for ACTOMASS, while if I score ACTIVITY is runs fine. I
> compiled a simple example to reproduce the error (it doesn’t have
> any physics meaning). I attached the input file and backtrace. Is
> this some kind of bug?
>
> Thanks,
> Marcus
>
>



__________________________________________________________________________
You can manage unsubscription from this mailing list at https://www.fluka.org/fluka.php?id=acc_info
Received on Mon Dec 14 2015 - 12:08:53 CET

This archive was generated by hypermail 2.3.0 : Mon Dec 14 2015 - 12:08:58 CET