Hello, I am trying to use the viscous layer hypothesis in the meshing module to create boundary layer for my openFOAM heat transfer simulation study.
I first created a partition geom in the geometry module and created groups for fluid and solid regions as well as the fluid-solid interface that I want to create the boundary layer on. And then I switched to the meshing module. I used 1-3D NetGen algorithm for the overall meshing, and used the wall option for the viscous layer hypothesis.
The mesh generated successfully, but when carefully examining the mesh, I found that the viscous layer grows on both the fluid and solid region. Is it possible to generate the viscous layer only on fluid group of the partition geometry?
Here is more detailed info about the problem. The screen shot is showing the volume mesh of the fluid and fin regions generated for a heat sink heat transfer problem. The blue region is the solid fin region and the yellow region is the fluid region. If you zoom in close enough, you will find out that the viscous layer is generated on both side of the fluid-solid interface. However, I wanted to generate the viscous layer only on the fluid side, no the solid side. I tried a couple thing, but so far I haven’t found a way to suppress the viscous layer generate on the solid side. Thanks for your attention in advance. Any help is appreciated. Regards!
The viscous layer hypothesis lets you select which walls should have layers and which shouldn’t. It is usually a straightforward process.
As an alternative, you can generate a Python dump and manually edit the script to include the specific face/surface IDs where you want the layers applied.
I have not worked with compound meshing (ie., muti body). but i guess salome has dificulties to recognize where you want to place the BL as you are using a general algorithm for the complete body?
test this:
you have body A and B (B needs layers)
in geom, you create a compound of the two (if you dont have it already), then use repair/glue faces
on the result create two groups of solids, one for A and one for B
create a 3D mesh, with empty algorithm (or the one that you will mesh body A), define your hole thing as you did now, but create a sub mesh for body B, and in that add the layer hypothesis. this should work.
Thanks for your response. It seem like how the geometry was handle matters. I followed franco’s suggestion in the below response, and it works. Previously, I was using partitioned geometry instead of compound, and it generate viscous layers on both sides. To generate viscous layer in the desired region, one has to use compound instead partition, and the geometry has to be glued.
Hello franco, I followed your suggestion, and it worked for the tested geometry! Really appreciate your help. Now I am going to work on my real case geometry for my heat transfer simulation. Fingers cross. Regards.
no problem, just a small comment, it CAN work with partition, actually if you do that way, you dont need to glue faces, but neverhtless the important part is to create a submesh for each solid, and define the BL in that sub mesh and not globally(<= your issue whas there.)
