Problem to run OpenFOAM in parallel

Hi !

I am just trying to run the “flow-over-heated-plate-two-meshes” tutorial in parallel but I got the following error with OpenFOAM :

Starting time loop
---[preciceAdapter] Loaded the OpenFOAM-preCICE adapter - v1.2.3.
---[preciceAdapter] Reading preciceDict...
---[precice]  This is preCICE version 2.5.0
---[precice]  Revision info: no-info [git failed to run]
---[precice]  Build type: Release (without debug log)
---[precice]  Configuring preCICE with configuration "../precice-config.xml"
---[precice]  I am participant "Fluid"
---[precice]  Connecting Primary rank to 1 Secondary ranks
---[precice]  Setting up primary communication to coupling partner/s
---[precice]  Primary ranks are connected
---[precice]  Setting up preliminary secondary communication to coupling partner/s
---[precice]  Prepare partition for mesh Fluid-Mesh
---[precice]  Gather mesh Fluid-Mesh
---[precice]  Send global mesh Fluid-Mesh
---[precice]  Receive global mesh Solid-Mesh
---[precice]  Receive global mesh Solid-Mesh-nodes
---[precice]  Broadcast mesh Solid-Mesh
---[precice]  Filter mesh Solid-Mesh by bounding box on secondary ranks
---[precice]  Mapping distance min:2.05197e-06 max:0.00124439 avg: 0.000585094 var: 1.22318e-07 cnt: 66
---[precice] WARNING:  239 of 400 vertices of mesh Solid-Mesh have been filtered out since they have no influence on the mapping.
---[precice]  Filter mesh Solid-Mesh by mappings
---[precice]  Feedback distribution for mesh Solid-Mesh
---[precice] ERROR:  The received mesh Solid-Mesh-nodes needs a mapping, either from it, to it, or both. Maybe you don't want to receive this mesh at all?

I didn’t modify anything in the tutorial…

Configuration :

  • WSL
  • Ubuntu 20.04
  • Precice 2.5.0
  • OpenFOAM 2306

Thank you in advance for your help.

Best regards,

I can reproduce the problem, I fixed it for our next release, thank you!

You can just remove the following line from config:

  <participant name="Fluid">
     <provide-mesh name="Fluid-Mesh" />
     <receive-mesh name="Solid-Mesh" from="Solid" />
-    <receive-mesh name="Solid-Mesh-nodes" from="Solid" />
     <read-data name="Heat-Flux" mesh="Fluid-Mesh" />
     <write-data name="Temperature" mesh="Fluid-Mesh" />

Yes, it works.

Thank you !

I have now a new running parallel issue with a case inspired from this tutorial.

Starting time loop

---[preciceAdapter] Loaded the OpenFOAM-preCICE adapter - v1.2.3.
---[preciceAdapter] Reading preciceDict...
---[precice]  This is preCICE version 2.5.0
---[precice]  Revision info: no-info [git failed to run]
---[precice]  Build type: Release (without debug log)
---[precice]  Configuring preCICE with configuration "../precice-config.xml"
---[precice]  I am participant "Fluid"
---[precice]  Connecting Primary rank to 1 Secondary ranks
---[precice]  Setting up primary communication to coupling partner/s
[LFRXCDW10X0032:08481] 1 more process has sent help message help-btl-vader.txt / cma-permission-denied
[LFRXCDW10X0032:08481] Set MCA parameter "orte_base_help_aggregate" to 0 to see all help / error messages
---[precice]  Primary ranks are connected
---[precice]  Setting up preliminary secondary communication to coupling partner/s
---[precice]  Prepare partition for mesh Fluid-Mesh
---[precice]  Gather mesh Fluid-Mesh
---[precice]  Send global mesh Fluid-Mesh
---[precice]  Receive global mesh Solid-Mesh
---[precice]  Broadcast mesh Solid-Mesh
---[precice]  Filter mesh Solid-Mesh by bounding box on secondary ranks
---[precice] ERROR:  The re-partitioning completely filtered out the mesh "Solid-Mesh" received on rank 0 at the coupling interface, although the provided mesh partition on this rank is non-empty. Most probably, the coupling interfaces of your coupled participants do not match geometry-wise. Please check your geometry setup again. Small overlaps or gaps are no problem. If your geometry setup is correct and if you have very different mesh resolutions on both sides, you may want to increase the safety-factor: "<use-mesh mesh="Solid-Mesh " ... safety-factor="N"/> (default value is 0.5) of the decomposition strategy or disable the filtering completely: "<use-mesh mesh="Solid-Mesh" ... geometric-filter="no-filter" />

The message is not the same as before. I am wondering if it could be useful to for the faces of the interface to be on the same processor ?

BR

What the error message is telling you is essentially to check the orientation of the meshes. Load both in ParaView and visualize them: Do both of them have the same orientation?

Indeed, you are right !
An error of 1000 in scale of the solid mesh…

Now the parallelisation run but the processor still used at 100% by openfoam during the calculix computation :

image

Is there a way to correct that ?

Thank you very much for your support.

Is it? I am not sure I understand the percentages here. How many cores do you have? Is each OpenFOAM thread using each logical core to its 100%? Then why does CalculiX show as 37%?

While waiting, OpenFOAM would still appear as consuming some resources (but it should not be 100%).

This topic was automatically closed 3 days after the last reply. New replies are no longer allowed.