ansys icem tetra mesh to fluent_v6 conversion problems
 

Hi all,

I have a tetra mesh with prism layers generated in Ansys ICEM, I've attached a picture below.

Now when I use fluent_v6 solver to export the mesh so I can use it for my simulation in OpenFOAM. It converts the mesh, but skips the parts, INLET, OUTLET and ROTSYMM2. Also it creates shadow faces instead of the ROTSYMM2 part. Basically ROTSYMM1 and ROTSYMM2 are the sides of the grid that are periodic.

To explain a bit more, the grid is periodic and axisymmetric and it is a wedge of 5 degrees part of a cylindrical set up. I have set the rotational periodicity in the global mesh set up part.

Any help on how to get all the parts required for my boundary definition would be very welcome!

I've attached another image of the grid, just in case the other one is too small to view.

Cheers.

Bad geom, but should have been swept.
 

Hey Vasudha,

First things first... Your geometry was sloppy. In this zoom in, you can see that the extra pieces you added didn't fit well. This probably led to the problems.

Attachment 3282

On the other hand, such a simple model really should be swept (not tetra meshed). After fixing the geometry problems and building the topology, I set sizes on the curves on the lower surface and used patch conforming mesh to create a quad mesh on the lower surface. Since I had set the curve names up as separate parts, the 2D mesh had line elements in the separate parts. Then I used the extrude option with "inherited" for the sides, so that the side parts took on the names of the curve/line parts. I also set the "top" to the second periodic side part and set the volume to Fluid.

Attachment 3283

Attachment 3284

I will email you the files separately.

ICEM CFD Hexa...
 

If it were me, I would use ICEM CFD Hexa for this model...

Attachment 3288

It is very easy to hexa mesh this in 2D and then use the 2D to 3D Rotate option to get the 3D model which I could use to get exactly the mesh I wanted.

Attachment 3285

This would also convert the geometry from 2D to 3D, but since it didn't give me the option to change the "top" part, I simply needed to create a new part and put the created "top" surface into that part.

Attachment 3286

You could use OGrid to get an ideal boundary layer... This is just one way, but we could handle it differently depending on the physics...

Attachment 3288