[NohBalls]

Hello! I am working on running CFD on a road bicycle frame with wheels, to see the effects of certain changes in the frame on its aerodynamics. I have chosen to use OpenFOAM with FreeCad 1.0.2 and CfdOF, and I am using snappyHexMesh.

The model is quite complex, and I have made some effort to remove bodies that add complexity with little relevance to the analysis. Using Fusion 360 I export as STEP and import into FreeCAD. I made a compound of the bicycle model and the rectangular space where the bike resides.

I have tried meshing with other meshers available in FreeCad CfdOF, but snappyHexMesh is the only one to successfully mesh. Upon checking the mesh, I am getting a maximum skewness from 5-10 in most cases, which fails the mesh check. When I try to refine my mesh further to reduce the skew, snappyHexMesh will often freeze on "Introducing zones for interfaces", and not continue.

To summarize, I am running a base element size of 50mm, often refining down to around 0.1 relative element size on the surface refinements of the bicycle components. I have stayed with 3 cells between levels, explicit edge detection, and no edge refinement. When I try and go below 0.1 relative, snappy freezes and does not continue.



Here is a snappyHexMeshDict for a recent mesh of mine, which gave a max skew of about 5.46:

/*--------------------------------*- C++ -*----------------------------------*\
|                                                                             |
|                Generated by the CfdOF workbench for FreeCAD                 |
|                      https://github.com/jaheyns/CfdOF                       |
|                                                                             |
\*---------------------------------------------------------------------------*/
FoamFile
{
    version     4.0;
    format      ascii;
    class       dictionary;
    location    "system";
    object      snappyHexMeshDict;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //

castellatedMesh true;
snap            true;
addLayers       false;

geometry
{
    "Compound_Geometry.stl"
    {
        type triSurfaceMesh;
        name Compound_Geometry;
        file "Compound_Geometry.stl";
        regions
        {
            patch_0_1
            {
                name patch_0_1;
            }

            patch_0_2
            {
                name patch_0_2;
            }

            patch_0_3
            {
                name patch_0_3;
            }

            patch_1_0
            {
                name patch_1_0;
            }

            patch_2_0
            {
                name patch_2_0;
            }

            patch_3_0
            {
                name patch_3_0;
            }

        }
    }

    "MeshRefinement.stl"
    {
        type triSurfaceMesh;
        name MeshRefinement;
        file "MeshRefinement.stl";
    }

    "MeshRefinement001.stl"
    {
        type triSurfaceMesh;
        name MeshRefinement001;
        file "MeshRefinement001.stl";
    }

    "MeshRefinement002.stl"
    {
        type triSurfaceMesh;
        name MeshRefinement002;
        file "MeshRefinement002.stl";
    }

};

castellatedMeshControls
{
    maxLocalCells 100000000;
    maxGlobalCells 2000000000;
    minRefinementCells 0;
    nCellsBetweenLevels 3;
    maxLoadUnbalance 0.20;

    features
    (
        {
            file "Compound_Geometry.eMesh";
            level 0;
        }
        {
            file "MeshRefinement.eMesh";
            level 0;
        }

        {
            file "MeshRefinement001.eMesh";
            level 0;
        }

        {
            file "MeshRefinement002.eMesh";
            level 0;
        }

    );

    refinementSurfaces
    {
        Compound_Geometry
        {
            level (0 0);
        }

        MeshRefinement
        {
            level (4 4);
            faceZone MeshRefinement;
        }

        MeshRefinement001
        {
            level (4 4);
            faceZone MeshRefinement001;
        }

        MeshRefinement002
        {
            level (4 4);
            faceZone MeshRefinement002;
        }

    }

    resolveFeatureAngle 30;

    refinementRegions
    {
    }

    locationInMesh (0.0 0.0 0.0);
    allowFreeStandingZoneFaces true;
}

snapControls
{
    nSmoothPatch 3;
    tolerance 1.0;
    nSolveIter 100;
    nRelaxIter 5;
    nFeatureSnapIter 10;

    implicitFeatureSnap false;
    explicitFeatureSnap true;
}

addLayersControls
{
    relativeSizes true;
    layers
    {
    }

    expansionRatio 1.2;
    finalLayerThickness 0.3;
    minThickness 0.1;
    nGrow 0;
    featureAngle 120;
    nRelaxIter 3;
    nSmoothSurfaceNormals 1;
    nSmoothNormals 3;
    nSmoothThickness 10;
    maxFaceThicknessRatio 0.5;
    maxThicknessToMedialRatio 0.3;
    minMedialAxisAngle 90;
    minMedianAxisAngle 90;
    nBufferCellsNoExtrude 0;
    nLayerIter 50;
}

meshQualityControls
{
    maxNonOrtho 65;

    maxBoundarySkewness 20;
    maxInternalSkewness 4;

    maxConcave 80;
    minVol 1e-13;
    minTetQuality -1;
    minArea -1;
    minTwist 0.01;
    minDeterminant 0.001;
    minFaceWeight 0.05;
    minVolRatio 0.01;
    minTriangleTwist -1;
    nSmoothScale 4;
    errorReduction 0.75;
    relaxed
    {
        maxNonOrtho 75;
    }
}

mergeTolerance 1e-6;

// ************************************************************************* //

Here are the lines produced for a successful mesh with a skew of about 4.83:

18:02:48  End

18:02:48  Executing: checkMesh -meshQuality in C:\Users\nohda\AppData\Local\Temp\meshCase
18:03:20  /*---------------------------------------------------------------------------*\
  =========                 |
  \\      /  F ield         | OpenFOAM: The Open Source CFD Toolbox
   \\    /   O peration     | Website:  https://openfoam.org
    \\  /    A nd           | Version:  12
     \\/     M anipulation  |
\*---------------------------------------------------------------------------*/
/*   Windows 32 and 64 bit porting by blueCAPE: http://www.bluecape.com.pt   *\
|  Based on Windows porting (2.0.x v4) by Symscape: http://www.symscape.com   |
\*---------------------------------------------------------------------------*/
Build  : 12-8e8552546d20
Exec   : checkMesh -meshQuality
Date   : Dec 07 2025
Time   : 18:03:20
Host   : "NOHRIGG"
PID    : 43976
I/O    : uncollated
Case   : C:/Users/nohda/AppData/Local/Temp/meshCase
nProcs : 1
SigFpe : Enabling floating point exception trapping (FOAM_SIGFPE).
fileModificationChecking : Monitoring run-time modified files using timeStampMaster (fileModificationSkew 10)
allowSystemOperations : Allowing user-supplied system call operations

// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //
Create time

--> FOAM Warning : 
    From function static Foam::instantList Foam::timeSelector::select0(Foam::Time&, const Foam::argList&)
    in file db/Time/timeSelector.C at line 269
    No time specified or available, selecting 'constant'
Create polyMesh for time = constant

18:03:20  Enabling user-defined geometry checks.

18:03:31  Time = constant

18:03:31  Mesh stats
    points:           2263607
    faces:            6641921
    internal faces:   6524876
    cells:            2189733
    faces per cell:   6.01296916108037
    boundary patches: 6
    point zones:      0
    face zones:       3
    cell zones:       0

Overall number of cells of each type:
    hexahedra:     2176117
    prisms:        425
    wedges:        0
    pyramids:      0
    tet wedges:    2
    tetrahedra:    0
    polyhedra:     13189
    Breakdown of polyhedra by number of faces:
        faces   number of cells
            4   886
            5   1536
            6   3673
            9   4553
           12   1525
           15   955
           18   61

Checking topology...
    Boundary definition OK.
    Cell to face addressing OK.
18:03:31      Point usage OK.
18:03:32      Upper triangular ordering OK.
18:03:32      Face vertices OK.
18:03:34      Number of regions: 1 (OK).

Checking patch topology for multiply connected surfaces...
                   Patch    Faces   Points                  Surface topology
               patch_0_1     6374     7217  multiply connected (shared edge)
               patch_0_2     7492     8482  ok (non-closed singly connected)
               patch_0_3     1637     1811  multiply connected (shared edge)
               patch_1_0    12051    12272  ok (non-closed singly connected)
               patch_2_0    12051    12272  ok (non-closed singly connected)
               patch_3_0    77440    77880  ok (non-closed singly connected)

Checking geometry...
    Overall domain bounding box (-2.96645712852478 -1.16999995708466 -0.805481016635895) (0.541160702705383 1.16999995708466 1.24515795707703)
    Mesh has 3 geometric (non-empty/wedge) directions (1 1 1)
    Mesh has 3 solution (non-empty) directions (1 1 1)
18:04:12      Max cell openness = 3.4554989943061e-16 OK.
    Max aspect ratio = 7.84703747468256 OK.
    Minimum face area = 2.78960374638515e-08. Maximum face area = 0.000478492562917788.  Face area magnitudes OK.
    Min volume = 2.47891579817589e-08. Max volume = 9.6185816079543e-06.  Total volume = 16.8191569129124.  Cell volumes OK.
    Mesh non-orthogonality Max: 64.7241437838534 average: 2.35993054673786
    Non-orthogonality check OK.
18:04:13      Face pyramids OK.
 ***Max skewness = 4.82949782233829, 1 highly skew faces detected which may impair the quality of the results
    Coupled point location match (average 0) OK.
Checking faces in error :
18:04:15      non-orthogonality >  90 degrees                        : 0
18:04:15      faces with face pyramid volume <     0                 : 0
18:04:16      faces with concavity >  90 degrees                     : 0
18:04:16      faces with skewness >  50 (internal) or 100 (boundary) : 0
    faces with interpolation weights (0..1)  < 0.0001       : 0
    faces with volume ratio of neighbour cells < 0.0001     : 0
18:04:17      faces on cells with determinant <     0                : 0
18:04:17  
Failed 1 mesh checks.

End

18:04:17

And finally, the output when running an unsuccessful mesh, where it froze in an attempt to refine further.

........

Introducing baffles to delete problem cells.

18:19:53  Edge intersection testing:
    Number of edges             : 24981346
18:20:03      Number of edges to retest   : 7176955
18:20:03      Number of intersected edges : 3545085
18:20:06  Created baffles in = 13.715 s
18:20:06  

18:20:06  After introducing baffles : cells:6758398  faces:24981346  points:7820483
18:20:07  Cells per refinement level:
18:20:07      0	2328737
    1	92109
    2	299683
    3	530281
    4	3507588

18:20:07  Remove unreachable sections of mesh
-----------------------------------

18:20:07  Keeping all cells in regions containing any point in 1((0 0 0))
18:20:27  Selected for keeping : 5324409 cells.
18:20:27  Edge intersection testing:
    Number of edges             : 16932619
18:20:33      Number of edges to retest   : 0
18:20:33      Number of intersected edges : 904445
18:20:33  Split mesh in = 26.377 s


18:20:33  After subsetting : cells:5324409  faces:16932619  points:6257126
18:20:33  Cells per refinement level:
18:20:33      0	2102050
    1	92090
    2	272538
    3	515300
    4	2342431

Introducing zones for interfaces
--------------------------------

Surface : MeshRefinement
    faceZone : MeshRefinement
    cellZone : 
Surface : MeshRefinement001
    faceZone : MeshRefinement001
    cellZone : 
Surface : MeshRefinement002
    faceZone : MeshRefinement002
    cellZone : 
18:20:33  Walking from locations-in-mesh 1((0 0 0)) to assign cellZones - crossing a faceZone face changes cellZone
18:20:34  
18:20:34  Found point (0 0 0) in global region 0 out of 1 regions.
18:20:50  Detected 802903 free-standing zone faces
18:20:51

Any and all help, suggestions, comments, and concerns are welcome. Thanks!

[AtoHM]

I have tried several OF solvers with meshes of skew up to 12. None of it fails or produces bad output as long as you are under 12 and its a non-significant ratio of the total cells. Usually its some spot or a few spots with only a few cells.
Go for the simulation, check p is converging and you should be good. 1 cell > 4 is a dream

[Alczem]

Hey,


The brake disc looks all choppy and poorly meshed, are you sure it is propely resolved? I would try to either remove it or thicken it a bit if you cannot afford to refine the mesh at this location.

[NohBalls]

Originally Posted by AtoHM

I have tried several OF solvers with meshes of skew up to 12. None of it fails or produces bad output as long as you are under 12 and its a non-significant ratio of the total cells. Usually its some spot or a few spots with only a few cells.
Go for the simulation, check p is converging and you should be good. 1 cell > 4 is a dream

That is reassuring. I will say I've since refined my mesh further, and despite still having a higher skew, I am able to get my p to converge. Thanks for your help and advice!

Originally Posted by Alczem

Hey,

The brake disc looks all choppy and poorly meshed, are you sure it is propely resolved? I would try to either remove it or thicken it a bit if you cannot afford to refine the mesh at this location.

I have since refined my mesh and the brake disc looks much better. I do agree that this may be worth removing, but I will proceed with it for now.
I did not think to thicken the disc a bit! I will give that a shot.

In hindsight I made this post in a bit of a panic and have since made good progress on my case.
I will create another post once I require some more feedback and advice.
Thanks to you both and best wishes to your new year!