I've been trying for a week to eliminate the negative volumes from the mesh I created with Ansys CFX, but it doesn't work. I modeled in cad a 3D geometry of some buildings and created a mesh with tetrahedral + prism elements for the boundaries. This is not a Fluid Structure Interaction simulation, so I have no idea what is happening to have deformed elements, since I've adopted the same procedures I adopted for previous successful simulations. I have already recreated the mesh, reduced the timestep and even simplified the geometry in some problematic points. I observed in CFX-pre that some vectors from the opening boundary are not normal to the surface. Everytime this happens, I get an error message from CFX-solver saying I have negative elements....
Could someone help me?
In my point of view, the problem is in your mesh which has been generated by the grid generator (ICEM CFD, Ansys Meshing...etc.). There are some problems in the accurate definition for the volume that should contain the fluid (your buildings are closed volumes and can lead to real problems during the meshing process if not defined well). If possible could you please give more information about your geometry and the mesh software that you are using.
Yasser Ahmed, thank you for your answer. I'm using Ansys CFX meshing. Here are some information about it.
Mesh statistics:
Orthogonal Quality: 0,20
Skewness: 0,84
Aspect ratio: 660
I used edge sizing to improve the mesh quality.
Something weird: solver gives me the location of the first negative volume (x, y, z), but Cad indicates this location is far away from the geometry I exported.
Seems to me the geometry is not ok.
Yasser Ahmed, thank you for your answer. I'm using Ansys CFX meshing. Here are some information about it.
Mesh statistics:
Orthogonal Quality: 0,20
Skewness: 0,84
Aspect ratio: 660
I used edge sizing to improve the mesh quality.
Something weird: solver gives me the location of the first negative volume (x, y, z), but Cad indicates this location is far away from the geometry I exported.
Seems to me the geometry is not ok.
It is better in my point of view to use ICEM CFD for building your mesh instead of using Ansys meshing. This will given you full control on your mesh. In addition, try to check at the location that contains the negative volume, you may have small part (it must be deleted) which make the definition wrong for the solver. If we make sure that mesh is OK and the problem is still existed, you should check carefully your def file.
If you wish you can send to me your geometry file in DesignModeler Database format (agdb) and I will try build the mesh using ICEM CFD and you can check what will be the results.
In ANSYS, negative area (so is volume) were generated in case of direct meshing, i.e., meshing is done manually choosing nodes in clockwise directions.
- Badges
- Science topic