[Laurens]

Hi all,

I'm running a single cylinder engine case in Converge with spark ignition, but I am running into some problems. I ran a simulation with a certain spark timing, and I'm now trying to run a simulation with a later spark timing. This time the simulation stopped 53 CADaTDC (~60CADaSpark) in the third cycle (2213 CAD). The simulation is supposed to run until 2548 CAD. Converge does not write anything into the .error file and the slurm-#.out does not display any odd information either. In converge.log (verbose) I can not find anything either. When I check some values like cell count, pressures, temperatures and dt(sec), nothing stands out other than the differences I expect from different spark timing. Does anyone have any tips where to look for the cause of the stopped/crashed simulation?
General info:
-I run the simulation on a high performance cluster.
-I make the case in v3.1 and export to v3.0 to run it on v3.0 on the cluster
-The combustion model used is SAGE
-Turbulence model is RANS RNG k-eps
Thanks in advance!

[bb91936762]

Hey Lauren,


This is Basu Parmar from CONVERGE support. The errors in v3.0, should appear in "err.JobID" file. I cannot exactly comment on what happened. Do you have a restart file close to or where it crashed? If so try making a logfile of the simulation run. It will be something like:

mpirun converge-mpich &> logfile.out (use the mpi statement based on the version you are using)


This may show you the error. Please let me know if you still cannot diagnose it.

Thanks,
Basu

[Laurens]

Hi Basu,

Thank you for responding.
Maybe I shouldve been clearer in my first statement. The error file is made, but does not contain information (0 bytes), in my opinion this meant that the simulation did not crash according to converge, but it did stop somehow! The converge.log is full of information, but not any info that points out to me that something is going wrong.

I restarted the simulation and it continued till completion with no weird phenomena in any graphs or values that I post-processed.

Friendly regards,
Laurens

[bb91936762]

Hey Laurens,


Since you mentioned you restarted the simulation with the exact same setup, same no of cores and mpi and it continued till completion the errors may not be with CONVERGE. There can be many other reasons for this, including problems with the cluster, out of run time etc. Let us know if this happens again.

Thanks,