[VLima]

Dear all,

I have been struggling for quite some time with a relatively simple simulation, using a dem file which I converted into a stl file, as advised before by other users. In fact I have circled around those and other topics a lot, as I had many difficulties with snappyHexMesh and several other issues. I searched a lot but finally had to write here my questions and doubts and know what you think about this, so I'll get the insight to correct the errors that I am doing in this case.


1) First of all, the quality of the produced stl was at first my major concern but following the use of some tools in Blender I did not find any problems. However, as I have a very basic knowledge and use of Blender, I am still not totally sure about the stl, though it seems ok...
So I related the quality of the stl with what can be seen in figure1 attached. What do you think about those lateral wall spikes? Any hints are more than welcome.


2) As the simulations were getting too slow, I reduced the number of cells in blockMeshDict. Doing a surfaceMeshTriangulate of the stl some stripes appeared along the length, which I cannot understand where they are coming from (figure2). They appear in the simulation as well (figure3). Also strange are those steps in the bathymetry floor, which I relate with the decrease in the number of cells in blockMesh (right?). Please let me know all your thoughts.


3) Also I am not sure about what I see in figures 2/3 attached. There are gaps between the bathymetry and the background mesh, laterally and in the beginning and in the end. I was wondering if this can bring problems to the simulation, although the boundary conditions take care of the fluid flow.


4) In the end of the snappyHexMesh log :


Writing mesh to time constant
Wrote mesh in = 15.15 s.
Mesh snapped in = 593.3 s.
Checking final mesh ...
Checking faces in error :
non-orthogonality > 65 degrees : 57412
faces with face pyramid volume < 1e-13 : 0
faces with face-decomposition tet quality < 1e-30 : 0
faces with concavity > 80 degrees : 0
faces with skewness > 4 (internal) or 20 (boundary) : 0
faces with interpolation weights (0..1) < 0.02 : 0
faces with volume ratio of neighbour cells < 0.01 : 0
faces with face twist < 0.02 : 0
faces on cells with determinant < 0.001 : 0
Finished meshing with 57412 illegal faces (concave, zero area or negative cell pyramid volume)
Finished meshing in = 783.35 s.
End




From checkMesh I get :


Checking geometry...
Overall domain bounding box (-8 -2 -0.5) (12 2 1)
Mesh has 3 geometric (non-empty/wedge) directions (1 1 1)
Mesh has 3 solution (non-empty) directions (1 1 1)
Boundary openness (1.97036e-17 5.43299e-15 -1.259e-15) OK.
Max cell openness = 3.24855e-16 OK.
Max aspect ratio = 125.481 OK.
Minimum face area = 1.23242e-05. Maximum face area = 0.0339871. Face area magnitudes OK.
Min volume = 1.62566e-06. Max volume = 0.000758456. Total volume = 119.971. Cell volumes OK.
Mesh non-orthogonality Max: 85.9832 average: 20.7159
*Number of severely non-orthogonal (> 70 degrees) faces: 57364.
Non-orthogonality check OK.
<<Writing 57364 non-orthogonal faces to set nonOrthoFaces
Face pyramids OK.
***Max skewness = 12.6642, 3636 highly skew faces detected which may impair the quality of the results
<<Writing 3636 skew faces to set skewFaces
Coupled point location match (average 0) OK.


Failed 1 mesh checks.


So how bad is this ? What is acceptable or not in terms of the mesh quality?


Please let me learn your thoughts about this and my apologies for these basic questions and for mixing some concepts too I guess. I need of course to be sure that what I am doing is correct, to achieve the best and most precise results possible in the simulations.


Many thanks in advance for the time you might take with this!