Non-conformal mesh between fluid zone without interface
Hello everyone,
I am doing a simulation of which only a small area is important. To avoid too many cells or large aspect ratio. I would like to generate a non-conformal mesh. I have a try using designModeler to build a test case(2d case of 2 rectangular) and form it into 2 bodies. Thus I can define the grid number at edge of each rectangular in meshing. I did it and defined an interface of matching before calculating it in fluent, but seems the result will be a little discontinuity at the non-conformal boundary. https://s2.doveoss.com/i/2022/07/29/svsn0x.png https://s2.doveoss.com/i/2022/07/29/svsihf.png I actually have a case file which is also using non-conformal mesh even including 2 interface there. But from the boundary I could not see there are any interface. And from the results I could not see any discontinuity. I am really confused how they did it. And I believe it could be a better way since if I have multi non conformal interface, the interface definition could be troublesome! I checked the threads related and found that it seems we have to define an interface between non-conformal mesh in fluent. So, I am wondering if anyone knows how that can be done? Like just treat the non-conformal interface as interior since it is all fluid. Any comments will be highly appreciated! Thanks in advance! |
Shouldn't you use Fluent Meshing here with Cutcell-mesh-approach, with hanging nodes?
|
Quote:
I am so confused before like with no interface how they did it. Now it is clear! Fluent can adapt to refine structure mesh after you give a base structure mesh and then choose region you wanna to refine and click adapt! Thanks for your help! |
No, that is not what I mean. In the latest versions of Fluent, there is a new meshing GUI, called fluent meshing. There you can generate your mesh as well, instead of ANSYS-meshing.
This fluent meshing allows you to create meshes based on cut-cell-technology, and that is more ore less what you desire, if I look to your non-corformal meshes in Post. Nevertheless, if fluent can adapt your currnet ANSYS-meshing mesh to something what you want, then that is fine as well. |
Quote:
Hello Jan, I see. Sounds great! The refinement I need just like cut cell. I think I will try fluent meshing next time. Before I think out of no way. Now there are two solutions. Cool! Thanks again! |
All times are GMT -4. The time now is 08:44. |