[ginop]

Hi there,


I've got this funny issue:

--> FOAM Warning : 
surfaceFieldValue my_patch: faceZone(samplingPatch):
    Region has no faces

    From function void Foam::functionObjects::fieldValues::surfaceFieldValue::initialise(const Foam::dictionary&)
    in file fieldValues/surfaceFieldValue/surfaceFieldValue.C at line 545.
    From function bool Foam::functionObjectList::read()
    in file db/functionObjects/functionObjectList/functionObjectList.C at line 817
    
--> while loading function object 'my_patch'

when I run the functionObject as:

    {
        type surfaceFieldValue;
        libs ("libfieldFunctionObjects.so");
        writeControl    writeTime;
        writeInterval   1;
        fields          (alpha.water U "p.*");
        writeFields     yes;
        regionType      faceZone;
        name            samplingPatch;
        operation       none;
    }

although if I do either checkMesh:

Checking faceZone topology for multiply connected surfaces...
    FaceZone            Faces    Points   Surface topology                  
    samplingPatch       567      616      ok (non-closed singly connected)

or foamToVTK, I see that it is there?



I am running the interFoam solver from OpenFOAM+ v1812 on Ubuntu 14.04.

Any help would be much appreciated!


Thanks in advance,
Gino

[ginop]

maybe the issue is with the faces being part of an "empty" type boundary? since the method does work on a similar 3D domain case. Someone more experienced than me can confirm this.


although the boundary type is empty, those faces are part of cells that still have scalar and field variables, so I would expect the faces' values to be the same as the cells? Perhaps I totally have the wrong ideas



anyway the above method does not (or should not) work for faces belonging to an "empty" type boundary


Cheers


EDIT;
volFieldValue with a cellZone could be used for 2D cases (see user guide)

[zommuter]

I encountered the same error, and it turns out to be due to parallelization which leaves the interesting patch empty on some nodes empty. I don't know if there is any way to modify the functionObject to skip / consider those empty patches to non-contributing, so for now I had to comment out that part in my controlDict and post-process the reconstructPar-ed case...