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/)
-   -   [ICEM] Pre inflation in ICEMCFD (http://www.cfd-online.com/Forums/ansys-meshing/109156-pre-inflation-icemcfd.html)

saisanthoshm88 November 10, 2012 11:02

Pre inflation in ICEMCFD
 
I was trying to some how realize the pre inflation for prism generation in ICEM. I tried to generate the tetrahedral mesh with octree and then the prism mesh. Later I deleted the tetrahedral mesh but retained the prisms. The resultant mesh is inputed now in the delaunay method to fill in the tetras.

But this doesn't work.

I did this to actually avoid the tetras squeezing up for the prism insertion instead I wanted them to be built in accordance with the existing prisms like the pre inflation does it in ansys meshing.

Could some one please explain why it doesn't work (or) suggest a better way of realizing what I intended.

PSYMN November 15, 2012 18:43

The Pre-prism method is not always best because you need to generate the mesh from the prism cap (inside surface of the prisms), which is not always good enough quality for the delaunay algorithm...

But every time you have a corner, the prism cap is more distorted than the original surface. For convex situations, it may be stretched, for concave situations, it may be squished...

You should also check to see if you have any overlapping elements, penetrations, etc.

The delaunay algorithm has some strict requirements that must be met.

One way to figure it out is to run the uncovered faces check...

This will find the tops of all your prisms... Put those in a new part call PRISMCAP or something like that...

Then run your mesh checks... It will find overlapping elements, etc. Run your quality checks to look for poor aspect ratio triangles...

PSYMN November 15, 2012 18:43

It could also be your delaunay settings. You want it to fill voids in the mesh, etc.


All times are GMT -4. The time now is 10:35.