CFD Online Discussion Forums - [snappyHexMesh] Chaotic sHM layering

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)

- OpenFOAM Meshing & Mesh Conversion (https://www.cfd-online.com/Forums/openfoam-meshing/)

- - [snappyHexMesh] Chaotic sHM layering (https://www.cfd-online.com/Forums/openfoam-meshing/184779-chaotic-shm-layering.html)

Chaotic sHM layering

Hi people!

I am struggling to get my layers properly meshed in a complex geometry.

- picture a shows some weird bridges in the middle of...nothing.
- picture b shows no creation of layers at corners all even with featureAngle set to 180.

Here are my sHM settings

Code:

castellatedMeshControls

{

    maxLocalCells 10000000;

    maxGlobalCells 2000000;

    minRefinementCells 10;

    maxLoadUnbalance 0.10;

    nCellsBetweenLevels 5;



    features

    (

        {

            file "verso.eMesh";

            level 8;

        }

    );



    refinementSurfaces

    {

        verso_wall

        {

            level (6 7);

            patchInfo

            {

                type wall;

            }

        }

    }

    resolveFeatureAngle 60;





    refinementRegions

    {

        verso_wall

        {

            mode distance;

            levels ((0.1 6)(0.25 5));

        }



...

        

    }

Code:

addLayersControls

{

    relativeSizes false;

    layers

    {

        verso_wall

        {

            nSurfaceLayers 10;

        }

    }

    expansionRatio                  1.3;

    firstLayerThickness             0.0002;

    //finalLayerThickness             0.5;

    //thickness                       0.01;

    

    minThickness                    0.0002;

    nGrow                             0;



    featureAngle                     180;

    slipFeatureAngle                 80;

    nRelaxIter                        5;

    nSmoothSurfaceNormals             1;

    nSmoothNormals                    3;

    nSmoothThickness                 10;

    maxFaceThicknessRatio           0.5;

    maxThicknessToMedialRatio       0.3;

    minMedianAxisAngle               90;

    nBufferCellsNoExtrude             0;

    nLayerIter                       50;



}

Thanks for any suggestion!

Are you sure, that what you see are not artifacts of paraview? In the settings before actually reading the file, uncheck the split polyhedral setting.

I'm not really sure about the bridge, that looks odd, never seen anything like that.

However I would highly recommend these two things:
1. Use relative sizes for the prism mesh. They are much more robust.
2. Avoid changing surface sizes on a single surface - if you can force the surface to be all one size or over 2 sizes snappy has an easier time making prisms.

@jherb

I don't have the screenshot here but I double checked the view settings in paraView. Didn't change anything.

@me3840
I like to have absolute size for the layers to keep a very precise control as I am running a Low-Re simulation (the layers covering the boundary layer and the first layer according to the targeted y+)

I agree for the consistent size over a surface. cfMesh is working this way and the layers are now perfect.

KateEisenhower suggested gmsh to be even more accurate. But I have very dense stl files and the GUI of gmsh can hardly support it on my laptop.

Hi all,

Just for the record, I didn't say that gmsh is more accurate than cfMesh. Only for the models I am simulating, I think it's easier to create a structured mesh with gmsh than deal with the layers in cfMesh.

Sorry if there was a misunderstanding in our PMs.

Best regards;

Kate

No problem, my apologies.