Hi,
I have a viscous layer creation issue for both 3D and 2D geometries. Attached sol1.hdf file. Is there any workaround to make it work?
To elaborate, I extruded top and bottom surface mesh separately and all-together.
Is there any workaround if I create 3D separate VL cells (groups) and merge them with the main?
Thanks a lot.
sol1.hdf (950.7 KB)
hello,
the issue with your geometry is that you have extra edges in the faces that will not have the VL.
if you work around and remove this extra edges you can solve the issue.
you will not be able to directly create the VL with your geometry without modifications.
hope it helps.
sol1.hdf (931,2 KB)
Thank you very much Franco. Those edges were coming from revolve cut operation. That’s the reason surfaces are twisted. I was expecting that Salome would fail regenerating those faces. You made me wrong.
Hi Franco, I noticed that viscous layers are distorted nearby sharp corner edge. Do you know a remedy? GMSH also gives the same. Indexing issue, similar to the other 2D viscous layer post?
hello,
you have two options, you could ‘make’ the VL yourself by cutting the geometry and using partition, search in the forum i explained this already, or you can check the VL python api, that modifies the geometry to shirnk it and then create the VL, for you case it might work out of the box. the VL/BL python api, it was annonced in one of the last releases can not remember which, check the releases notes of the past few versions, when i tested myself i did not got good results with it. good luck.
in regards to gmsh, i did not understand what you meant, please keep in mind that one is not in the subject, so you need to explain yourself more detailed so one understands better and can help you 
(i meant it as a constructive criticism). in any case, eventhought i did not undrestood you what you meant exactly, i know that the gmsh algorithms are not fully integrated with everything else in salome, think them as some external mesher (which actually IS the case), instead of being coded inside salome, therefore there are some miss communication when connect. i know that gmsh 2D is not compatible with VL generation, [bug/issue] gmsh 2D not working correctly with viscous layer addition quasi structured mesh · Issue #41 · SalomePlatform/smesh · GitHub for 3D nooormally it is already compatible, but i can imagine it being still a little bit buggy, sincerently i never took too much the time with it, as i was able to generate correct meshes with netgen, and this ‘miss comunication’ between the algorithms are not a thing i can deal with, with my cyrrent workflows
