[chandra shekhar pant]

Dear All,


I am trying to mesh a propeller kind of geometry, in which I am having an AMI. To do so I created the 4 STL files, propellerTip, innerCylinderSmall, innerCylinder and outerCylinder, through createdPatch and TopoSet, I had converted them as the AMI, inlet outside cylinder etc (similar to the tutorial). So to add the files, I used the SHM in two steps:
1. In 1 step I put the layers off and castellatedMesh and snap true; and with the meshQuality which looks like: 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;
}
}


2. In second step I put off the castellatedMesh and snap while put on the layers, with different meshquality:
meshQualityControls
{
maxNonOrtho 82;
maxBoundarySkewness 20;
maxInternalSkewness 4;
maxConcave 75;
minVol 1e-13;
minTetQuality -1e+30;
minArea -1;
minTwist 0.02;
minDeterminant 0.001;
minFaceWeight 0.02;
minVolRatio 0.01;
minTriangleTwist -1;
nSmoothScale 4;
errorReduction 0.75;
relaxed
{
maxNonOrtho 85;
}
}


But at the end I am able to get only the 0.0254 layers, with not very high quality mesh (Max skewness = 9.39849, 8 highly skew faces detected which may impair the quality of the results). The two SHM's are attached, could any one help in this regard.



Thanks a lot for your time.

[HPE]

- adding layers via SHM is observed to be the most problematic part of SHM. is it possible for you to proceed your computations without adding layers?
- can you run SHM with all the three stages `true` instead of halving the procedure?

[chandra shekhar pant]

Hello HPE,



Thanks for the comments, my answers:



1. If I am not adding the layers then if I calculate my y+, its very odd, may be in the order of 100000. That's why trying the layers concept, but again this is very hard for me I guess.


2. Yes, I first tried with all the options true (because that is more easy), and what I got is the no layers at all, with very very bad mesh. I think SHM tried to add layers but could not and in the process of doing it, it screwed up the complete mesh, with very high orthogonality and skewness.



Could you please suggest any other way of doing it, some other software which can handle huge geometries?



Thanks a lot!

[john myce]

Hi chandra,

I don't know if you have solved your problem. But it seems that you have a very low feature angle for the addlayer phase (~30). Try with a much higher value like around 150 degrees. (the higher the better).

Also you have nGrow 1;

I don't know how this parameter works but from my experience it is better to keep it at 0 because for the value at 1 SHM fails completely to add layers for some cases.

Cheers.

[chandra shekhar pant]

Hello John,


Many thanks for these valuable inputs. I haven't solved this problem yet, I will incorporate your suggestion and will update in this thread.



Thanks again for your kind help!

[Ship Designer]

Hello everyone, I'm having a very similar problem too, where the mesh is ok just until snappyHexMesh adds layers. After that, checkMesh reports non-orthogonal and highly skewed faces. I don't know how to correct this yet, however first attempts to use cfMesh for meshing a ship hull resistance case, have shown that it is much more reliable at adding boundary layers. I'm not sure if this might help in your case too and I find cfMesh more difficult to use. But I understand that boundary layers in your case are essential and you might want to try another meshing tool. As far as I know cfMesh is included in ESI OpenFOAM but not in Foundation OpenFOAM.

See also this thread for the problem: How to deal with non-orthogonal and skew faces