[boffin5]

Finally I am running cases in v2412, after switching from OF8. My case ran okay as a series case, but when I convert it for parallel, it generates data, but in paraView the U, p and T fields don't appear.

This is a chtMultiRegionSimpleFoam case composed of an airplane fuselage that has a duct with a porousZone representing a radiator (image attached).
It starts with a pre-generated polyMesh file in a fluid folder, and another one in the porous folder. Here is my parallel run script:

#!/bin/sh
. ${WM_PROJECT_DIR:?}/bin/tools/RunFunctions        # Tutorial run functions

rm -rf 0
cp -r 0.orig 0

# Place fluid mesh into constant/fluid for decomposing
#-----------------------------------------------
mkdir constant/fluid
cp -r polyMesh.fluid constant/fluid

mv constant/fluid/polyMesh.fluid constant/fluid/polyMesh

decomposePar -region fluid -copyZero 2>&1 | tee log.decomposePar

# Place porous mesh into constant/porous for decomposing
#-----------------------------------------------------
mkdir constant/porous
cp -r polyMesh.porous constant/porous

mv constant/porous/polyMesh.porous constant/porous/polyMesh

decomposePar -region porous -copyZero 2>&1 | tee log.decomposePar

# Check both meshes
#-----------------------------------------------------
mpirun -np 8 checkMesh -region fluid -constant -parallel 2>&1 | tee log.checkMesh.fluid
mpirun -np 8 checkMesh -region porous -constant -parallel 2>&1 | tee log.checkMesh.porous

# Run the simulation
#-----------------------------------------------------
mpirun -np 8 chtMultiRegionSimpleFoam -parallel | tee run.log

runApplication reconstructParMesh -mergeTol 1e-5 -constant -latestTime -allRegions

It runs all the way through, but paraView fails. There is a telling bit in the run log:

writing force and moment coefficient files.

--> FOAM Warning : 
    From virtual bool Foam::functionObjects::fieldSelection::checkSelection()
    in file functionObjects/fieldSelections/fieldSelection/fieldSelection.C at line 179
    Field p not found
--> FOAM Warning : 
    From virtual bool Foam::functionObjects::solverFieldSelection::updateSelection()
    in file functionObjects/fieldSelections/solverFieldSelection/solverFieldSelection.C at line 82
    Valid solver fields are: 5(U epsilon h k p_rgh)Time = 2

It says it can't find p, but I have this BC in both the fluid and porous 0 folders. What is going on here? This is mystifying, since the simulation works find as a series case.

[dlahaye]

Not sure.

You are now using reconstructParMesh.

Should you add reconstructPar (to reconstruct the computed field from the decomposed patches) to visualize the fields?

[boffin5]

Adding reconstructPar solved the problem!!

Thanks dlahaye, and have a Happy New Year!

boffin5

[dlahaye]

I am so happy that reconstructPar started your of your year correctly!

Best wishes, Domenico.