CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   ANSYS Meshing & Geometry (https://www.cfd-online.com/Forums/ansys-meshing/)
-   -   [ICEM] Similar tetra refinement (https://www.cfd-online.com/Forums/ansys-meshing/89791-similar-tetra-refinement.html)

andreasp June 22, 2011 10:37

Similar tetra refinement
 
Hi all!

Somehow I can't believe I need to ask this because it seems trivial, but we are not getting the solution we want, so here's the question:

We have a 3D mesh consisting purely of tetra elements.
Now we want to refine this complete mesh with a similarity approach. That means, each tetra of the old mesh should be subdivided into 8 new ones, which are all similar to their "father". Thus, each edge should be split at its center into two new ones. The newly created nodes shall NOT be projected to the geometry surfaces. They should just stay at the same place where the former edge center used to be.

To me this really seems to be the simplest mesh refinement one could do. But how is it possible with ICEM???

Any help is appreciated!
Andreas

andreasp June 22, 2011 11:21

2 Attachment(s)
Maybe to shed a little more light on that issue:

We have tried the "Pure mesh refinement" method from the refine/coarsen menu. See the attached pngs for an example.

For the element marked in blue, the method seems to do what we want. At least the marked triangle is split into 4 as we desire.

However, the red element is refined awkwardly. Why? Do we have to explicitly tell ICEM to use the most simple dividing algorithm? If so, how can we do it?

Again, thanks in advance for any help!

PSYMN June 24, 2011 12:35

Different way to get there...
 
Under change mesh type, you could "add midside nodes". Turn off projection so they won't snap to surface.

Then "remove midside nodes" and use the option to convert to multiple linear elements...

It should give you exactly what you want.

andreasp June 27, 2011 04:22

Perfect, that did the job!

Thanks!


All times are GMT -4. The time now is 11:57.