CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CD-adapco (
-   -   Problems with Merge (

Sabe February 17, 2004 09:04

Problems with Merge

There is something curious going on when I merge vertices that do belong to a spherical shape!? The merging operation is successful but the cells can't be seen or selected anymore(there are still in cell set). The merging operation deletes way more then the model would need, only the half of the needed points remain! How come? I tried: -small sets of points, did not work out -i build a cylindrical model, that has the same number of vertices to merge and it works. so I assume it has something to with the spherical shape... Any advice is welcome,thanks in advance. Greatings Sabe

Brian February 17, 2004 14:50

Re: Problems with Merge
Polar meshes (those created in cylindrical or sperical coordinate systems) tend to concentrate a lot of cells (and vertices) at the centre of the circle. These cells usually end up as prism cells (or perhaps pyramids in a sperical system). As the number of cells in the centre is the same as on the circumference in the tangential direction, these cells can get very small, and the vertices that define them are therefore very close together.

VMERGE has an input variable, "tolerance". This value, by default 0.1mm, can easily be bigger than the distance between adjacent vertices in meshes created in this way.

This long story simply serves to tell you that you need to reduce the tolerance in the VMERGE command.

All times are GMT -4. The time now is 01:18.