CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   OpenFOAM Meshing & Mesh Conversion (https://www.cfd-online.com/Forums/openfoam-meshing/)
-   -   [Technical] Unspecified boundary types in the grids created employing Pointwise (https://www.cfd-online.com/Forums/openfoam-meshing/72311-unspecified-boundary-types-grids-created-employing-pointwise.html)

arash February 2, 2010 05:30

Unspecified boundary types in the grids created employing Pointwise
 
When using the revolve or extrude mesh in Pointwise, often some internal faces are created in the internal field which at the end remain as unspecified boundary types. These types of boundaries are problematic in OpenFOAM. Removing these unspecified boundaries results in removing the blocks sharing any of these certain boundaries (patches).

According to this, would you please let me know whether any of you have had similar problem or not? If so, how did you manage to fix that?

BTW, its is possible to remove such unused patches in GAMBIT without causing any other changes in the created grid; but, how should fix this just using Pointwise?

Regards,
Arash

rmatus February 2, 2010 18:09

Try Grid, Merge... Before Extruding
 
If you have unspecified internal faces in an extruded block, it most likely means the domains you started the extrusion from had some gaps between them. (Unmerged connectors in Pointwise-speak.)

You can easily check this by going to the Grid, Merge... panel. If there are gaps they will be outlined in red, and you can use the tools in Grid, Merge... to heal them.

Once they are merged, you can do the extrusion and there should be no unspecified internal faces.

arash February 2, 2010 20:16

Hi Rick,
Of-course, that could be a reason but the case I'm talking about is rather different. To be more clear, just imagine a meshed rectangle. If you rotate this rectangle about one of its axis 360 degrees to create a cylinder, you'll find that finally there is a face inside the field which is exactly located on the axis of rotation (I don't know if it is right to call that a face!). This is the unspecified boundary type which I mean. If you remove that face the entire block would be disappeared.

rmatus February 3, 2010 09:31

Try setting a cyclic boundary condition
 
Sorry, I was imagining the problem incorrectly. There is some discussion in this thread about the boundary conditions to use on a pole (singularity) as you have in this axisymmetric case. It looks like you should use a cyclic bc on the pole face.

To do that in Pointwise, go to the CAE, Set Boundary Conditions... panel. Click on New and then double click in the Name field of the new entry in the Boundary Conditions table to type whatever name you want to give that BC. Double click in the Type field and when the pull-down activates choose Cyclic for the BC type. Finally, select the face you want to apply this BC to from the display window and then click on the box in the Set column of the table to apply this BC to that face (or faces if you picked more than one).

Hope this helps.

Quote:

Originally Posted by arash (Post 244687)
Hi Rick,
Of-course, that could be a reason but the case I'm talking about is rather different. To be more clear, just imagine a meshed rectangle. If you rotate this rectangle about one of its axis 360 degrees to create a cylinder, you'll find that finally there is a face inside the field which is exactly located on the axis of rotation (I don't know if it is right to call that a face!). This is the unspecified boundary type which I mean. If you remove that face the entire block would be disappeared.


arash February 9, 2010 09:35

Rick, thanks for your instructions; now it's working.

rmatus February 9, 2010 09:56

Arash, happy to hear that worked.


All times are GMT -4. The time now is 10:59.