CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   OpenFOAM Meshing & Mesh Conversion (https://www.cfd-online.com/Forums/openfoam-meshing/)
-   -   [snappyHexMesh] problems resolving small mesh (https://www.cfd-online.com/Forums/openfoam-meshing/230386-problems-resolving-small-mesh.html)

otaolafr September 21, 2020 12:05

problems resolving small mesh
 
3 Attachment(s)
hello,
I am looking for some help with snappyHex or if someone can see another problem from a previous step.

I need to mesh really small geometries with small repetitive internal structures (3mm3) ( the total bounding box is (-1.0842e-19 -9.2767e-19 -1.38778e-17) (0.00505 0.00505 0.201133))

for the background mesh, I use blockMesh with cubic cells and a cell size of 0.002 (also tried 0.001/0.0015/0.0025). For the geometry itself, I triangulate it with salome with NETGEN 2D and wire discretization and local lenght of 5e-05 (i tried also 2.5e-5).

for the snappyHex
the castelated controls where:
maxLocalCells 2000000
maxGlobalCells 13000000
level ref (1 1) (tried 2 2, 3 3, 4 4, 5 5 and 6 6) for the howl region.
for the snapControls:
nSmoothPatch 5 (also 7)
tolerance 2
nSolveIter 300 (also 350)
nRelaxIter 4 (also 8)
nFeatureSnapIter 30 (also 100)

and with explicitFeatureSnap on.
snappyHexDict:
Attachment 80260
the feature detection works correctly see attached pictures...
the issues I am having is that the snapp of the geometry to go at least similar in a "look" i need to go at least up to level 5 5 of refiment, but when i do this the checking mesh bigings to give errors (higher the level the more face error i am facing):
non-orthogonality > 65 degrees :170 and faces with face pyramid volume < 1e-13: around 20000000 for the best results.

I am adding the results from the surfaceCheck of the stl and also the checkMesh -allGeometry -allTopology
Attachment 80258

Attachment 80259


I know that it is a proble of size of the geometry itself.. but I need really need to mesh this geometry as they idea is to compare it with experimental results so any help would be apreciated :)
image link:

https://i.ibb.co/5s6wtKz/text885.png

any guidance/help or pointing in a direction would be appreciate it.


All times are GMT -4. The time now is 02:18.