CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   ANSYS Meshing & Geometry (http://www.cfd-online.com/Forums/ansys-meshing/)
-   -   non-convex cells in ICEM 12.0 (http://www.cfd-online.com/Forums/ansys-meshing/85342-non-convex-cells-icem-12-0-a.html)

DaIN February 23, 2011 12:16

non-convex cells in ICEM 12.0
 
1 Attachment(s)
Hi there,

I'm trying to create a mesh to be converted into KIVA 3V mesh file.
After creating the geometry, blocking, association and meshing I finally got a mesh which looks apparently fine. Please look at the picture.
I checked the Quality Histograms and I obtained volume > 0 for all cells which I guess it's the mean condition for a mesh to be used in KIVA.
But, after running the code I got an error message saying I have non-convex cells.
I could find out though, these non-convex cells are located around the axis, but there's no way for me to get rid of them.

Does anyone know how to do that??

Thanks a lot,

daniel

Ludvik February 23, 2011 12:23

http://homepages.cae.wisc.edu/~hesse...rialIntro.html

DaIN February 23, 2011 12:35

Thanks a lot for your answer, but that's the tutorial I've been mainly following.
My geomtry is also different in a sense that I don't have periodic faces, but I do have an axis which must be setup as periodic.

It does not say anything about non-convex cells... :(

daniel

PSYMN February 24, 2011 12:57

Wedges or collapsed Hexas?
 
Are those wedges along the center axis or are they collapsed hexas (perhaps that is your problem). You can tell quickly with a mesh summary or simply by looking at the mesh/volume branch of the tree. Do you see any penta-6 or prism elements?

I am not sure what a periodic axis is, but have you considered an Ogird instead of wedges in the middle?

DaIN February 24, 2011 13:26

Thanks a lot for your answer :)

The non-convex cells (wedges??) are not along the axis, are near the axis, actually around it (I didn't check them all, there are 144) and i reckon they might have collapsed somehow. I can't see those kinds of cells in the mesh, but honestly I'm not sure if i'd be able to recognise them if the exist, becase I don't know how they look like. Is there anyway I can fix them?

About periodic axes I meant that i defined the bottom and top points of the axis with 360 degrees of periodicity, but I did not use more periodic conditions in the mesh.

Thanks a lot indeed for your help on this,

daniel

DaIN February 24, 2011 16:51

Oh sorry I forgot to mention it.

I did use O-grids, actually I used two. The first one is for the annular outer region and the second one is for the circular inner region. The problem is that as i will export the file into KIVA format, I think I canīt use O-grids due to the local axis of the blocks. For that reason, i always remove the inner block which never can get aligned with the others.

Thanks a lot,

daniel


All times are GMT -4. The time now is 11:11.