Sirocco fan tutorial
Hello,
I am attempting to run the Sirocco fan tutorial on Baram 26.1.1 for Windows 11, and am having a problem. I read the Fluent mesh in, and set up the problem as the tutorial indicates, but running the steady state first part of the problem quits with an error:
Reading field p_rgh
Creating reaction model
Combustion model not active: combustionProperties not found
Selecting combustion model none
Creating MRF zone list from MRFProperties
creating MRF zone: MRFCellZone_rotating
Radiation model not active: radiationProperties not found
Selecting radiationModel none
No finite volume options present
job aborted:
[ranks] message
[0-39] terminated
[40] process exited without calling finalize
[41-63] terminated
Needless to say, no window appears when changing Time to Transient.
It further states that " buoyantSimpleNFoam ended prematurely and may have crashed. exit code 3".
I'm confused by this result - the tutorial states the problem will use buoyantPimpleNFoam, not Simple.
Am I doing something wrong?
Thanks,
Tom
Update to this issue: I have a 64 core AMD Threadripper Pro CPU, with 128 logical cores. If I set the parallel environment to 64 cores, the steady state problem fails. Any setting from 32 or less cores, and the problem solves. 64 or higher, and it doesn't solve. Is this an MSMPI issue? I have the latest interation installed. I'd like to use as many cores as I can to keep processing time down. This has not happened on any other problems in Baram 26.1.1, and I've used 64 cores on other problems.
Hello,
Thank you for using BARAM and for the detailed bug report.
The Sirocco fan case is a relatively small mesh, so it's important to use an appropriate number of CPU cores to ensure the domain is decomposed correctly and the simulation runs without issues.
In particular, OpenFOAM is known to crash when any decomposed partition contains zero cells.
We reproduced your setup with 64 cores and confirmed that processor 40 was assigned zero cells during decomposition (You can verify this by checking the decomposePar log in the console) — which is exactly why the error log shows process 40 as the first to terminate.
That said, if neither OpenFOAM nor BARAM has logic to catch and warn about this kind of edge case, that is a deficiency on our end. We will look into whether a safeguard or warning can be added.
Regarding the transient setting window — it does not appear immediately upon switching the time setting to Transient.
It becomes visible when the project is saved or when you navigate to another page.
We agree the current tutorial wording is misleading on this point and will revise it accordingly.
Thank you again for taking the time to report this so thoroughly.
Please don't hesitate to reach out if you run into any other issues!
Best regards,
Hello,
Thank you for the response. I have found this issue on another problem as well - sloshing tank. In fact, sometimes the problem will solve with a specific number of cores, for instance, 16, then won't solve again until you change it. With sloshing tank, I had to go down to 4 cores to get it to solve. Is that an OPENFOAM issue or MSMPI? It would be nice if it were at least repeatable.
Thanks for your help!
Tom