[ssze]

Hi all,

I'm exploring switching from openfoam.org's version of openFOAM (of10) to openfoam.com's counterpart (of2306).

I've encountered an issue where, following snappyhexmesh, zero-area faces are present on the edge on one of my baffles, leading to warnings during createBaffles and errors in checkMesh (not satisfying max non-orthogonality or min face area requirements) and a fatal error in simpleFoam. The fatal error states that, for each processor, it is unable to find the 'p' file in the '0' folder. I am running both SHM and simpleFoam in parallel.

Having ran this exact case succcessfully in using of10, I compared my log files and noticed that at the end of SHM, there is an additional step in of10 where it states it is "removing zero-sized patches" and lists the various locations (the blockmesh patches and on the walls I've marked as baffles). In my of2306 simulation, this step seems to be skipped, which I suspect is causing my issues. Is there a way to force of2306's SHM to perform this action within the snappyhexmesh dict? I'm aware of the collapseEdges command and plan to try that as well. To be sure the issue was baffle-related, I ran a version of the case where the problem baffle was omitted and that ran smoothly, with none of the aforementioned errors. I’m running both SHM and simpleFoam in parallel. I’ve copied some of the error messages below:

createBaffles:

 --> FOAM Warning : 
    From static Foam::label Foam::coupledPolyPatch::getRotation(const pointField&, const Foam::face&, const point&, Foam::scalar)
    in file meshes/polyMesh/polyPatches/basic/coupled/coupledPolyPatch.C at line 275
    Cannot determine unique anchor point on face 4((15.2579 -39.4495 0.559797) (15.2579 -39.4495 0.62219) (15.2579 -39.4495 0.62219) (15.2579 -39.4495 0.559797))
Both at index 0 and 3 the vertices have the same distance 2.22045e-16 to the anchor (15.2579 -39.4495 0.559797). Continuing but results might be wrong.

--> FOAM Warning : 
    From static Foam::label Foam::coupledPolyPatch::getRotation(const pointField&, const Foam::face&, const point&, Foam::scalar)
    in file meshes/polyMesh/polyPatches/basic/coupled/coupledPolyPatch.C at line 275
    Cannot determine unique anchor point on face 4((15.2579 -39.4495 0.559797) (15.2579 -39.4495 0.559797) (15.2579 -39.4495 0.497431) (15.2579 -39.4495 0.497431))
Both at index 0 and 1 the vertices have the same distance 2.22045e-16 to the anchor (15.2579 -39.4495 0.559797). Continuing but results might be wrong.

This repeats for a series of locations of the same x and y coordinates and varying z (so along the edge of the baffle) and is always in regards to 'face 4'. There is one face 4 that has a vertex of slightly different x and y coordinates (15.2524, -39.4403), but the anchor coordinates are the same as the rest.

checkMesh:

 ***Zero or negative face area detected.  Minimum area: 0
  <<Writing 76 zero area faces to set zeroAreaFaces
    Min volume = 5.39427e-07. Max volume = 0.0213143.  Total volume = 60580.8.  Cell volumes OK.
    Mesh non-orthogonality Max: 90 average: 8.40599
 ***Number of non-orthogonality errors: 76.
  <<Writing 76 non-orthogonal faces to set nonOrthoFaces

simpleFoam (repeats for each processor):

 Reading field p
[14] 
[14] 
[14] --> FOAM FATAL ERROR: (openfoam-2306)
[14] cannot find file "/home/ubuntu/datafiles/of2306/of2306Test001/processor14/0/p"
[14] 
[14]     From virtual Foam::autoPtr<Foam::ISstream> Foam::fileOperations::uncollatedFileOperation::readStream(Foam::regIOobject&, const Foam::fileName&, const Foam::word&, bool) const
[14]     in file global/fileOperations/uncollatedFileOperation/uncollatedFileOperation.C at line 627.
[14] 
FOAM parallel run exiting

Any help would be appreciated!

[AtoHM]

There is a difference between zero-sized patches and zero or negative area faces. The first one happens e.g. when all faces are moved into another BC. AFAIK you can remove zero-sized patches by running createPatch with an empty createPatchDict.
Zero or negative area faces are a sign of bad mesh quality. Your simulation might run, but with a nonOrtho of > 70 I doubt the results will be meaningful. I strictly stick to the 65 deg limit.


Now for your last error:

cannot find file "/home/ubuntu/datafiles/of2306/of2306Test001/processor14/0/p"

create a log of the decompose utility run and check there, this is where you will probably find the correct spot whats going wrong.