Please provide the used software versions.
FLUKA: 4-5.1
Flair:
Operating system:
Compiler:
Please I am facing a problem.
When increasing the number of particles from 10.0e6 to 50.0e6 per cycle, the run does not finish and shows error. For 10.0e6 worked fine. Below are screenshots from results.
What could be the problem?
Thanks
Please upload all relevant files. (FLUKA input file, Flair project file, user routines, and data files)
Dear Luis,
thanks for your question. Could you please upload your input file, Flair file and user routines, so I can understand the possible issues better?
Kind regards,
Barbara
Thanks @bhumann
please find enclosed.
I am staying tuned.
electron_photon_lu_2.flair (4.2 KB)
electron_photon_lu_2.inp (4.2 KB)
.abaq enclosed. it has been saved as .txt in order to be possible to upload here but it is in fact .abaq
electron_lu-177_2.txt (642.1 KB)
Dear Luis,
I have launched a simulation using your flair file with 50e6 primaries per cycle and it finished without problems. Since the same input runs fine on my side, most likely the crash is specific to your run environment rather than to the input itself.
First thing to check: did your run hit a wall-time limit, or did you completely fill up your disk quota? Either of these will silently stop a run.
Next, look at the actual termination message rather than the warning in your screenshot: what do the last lines of the *.err file say, and is there a STOP/Abort message at the end of the *.log file? That message would indicate why the run stopped.
The “errors” you see in your error-file screenshot are only warnings that the code has trouble transporting a particle from one region to the next. They are almost certainly not what stopped your run. When you run cycles with fewer primaries you simply see them less often, because the likelihood of such a transition at this exact position is much lower.
On the warnings themselves: the trouble is in transporting the particle from region 15 to region 14, which are named reg_12 (region 15) and reg_11 (region 14) in the *.out file. Both are part of your .abaq geometry, which suggests the mesh may have overlaps or gaps at their interface. Re-running the FLUKA_mesh_fix.py script (to be found here) should help resolve this.
More generally, I would advise you to reduce the number of primaries per cycle and instead increase the number of cycles and spawns. This gives you better cycle-to-cycle statistics (and therefore more reliable uncertainties), and it keeps the simulation short enough that it is far less likely to run into a wall-time limit.
I hope this helps.
Cheers,
Barbara
Thanks @bhumann , I’ll check those files.
Greetings @bhumann
Please see below what those two files say. First .err and second .log. and third .out.
I kept to 50e6 particles per cycle, but I increased cycles to 10, and I increased spawns to 4.
But still the run stops and does not finish.
Thanks.
Dear Luis,
Probably I have been unclear in my previous reply regarding the number of primaries, cycles and spawns. I would suggest reducing the number of primaries while increasing the number of cycles to achieve similar statistics.
So instead of launching 1 cycle with 50e6 primaries, rather launch 5 cycles with 10e6 primaries. Generally, running simulations with more cycles than one is encouraged, as explained in this lecture.
Unfortunately, I cannot tell from the screenshots what the reason for your crashing simulations is. Could you please upload the according files if you would like to investigate the issues further?
Cheers,
Barbara
