Robinho4587
Mechanical
- Jan 27, 2012
- 2
Hi,
following problem: My task is to optimize the fe-mesh of a certain part. The boundary-nodes need to stay unchanged, so the optization must be accomplished by first changing the coordinates of the inner nodes and the generating a new delaunay-triangulation.
I wrote a function with Matlab for the optimization of the inner nodes, which works fine:
-Improvement of the "Average Shape Factor" from 0.685516 to 0.740928
-Improvement of the "Average Aspect Ratio" from 1.59 to 1.51
The next step is the delaunay-triangulation of the new coordinates, which replaces the old elements with better ones.
The delaunay-triangulation also improves the quality of the mesh as followed:
-Improvement of the "Average Shape Factor" from 0.740928 to 0.763868
-Improvement of the "Average Aspect Ratio" from 1.51 to 1.48
Unfortunately the delaunay-triangulation generates some critical elements, which are not acceptable.
Please take a look on the file i attached to this thread.
As you can see, all four nodes of the critical elements are in one plane and some elements contain nodes, which are not directly neighboured.
Does anybody know how to solve this problem?
Best regards!
following problem: My task is to optimize the fe-mesh of a certain part. The boundary-nodes need to stay unchanged, so the optization must be accomplished by first changing the coordinates of the inner nodes and the generating a new delaunay-triangulation.
I wrote a function with Matlab for the optimization of the inner nodes, which works fine:
-Improvement of the "Average Shape Factor" from 0.685516 to 0.740928
-Improvement of the "Average Aspect Ratio" from 1.59 to 1.51
The next step is the delaunay-triangulation of the new coordinates, which replaces the old elements with better ones.
The delaunay-triangulation also improves the quality of the mesh as followed:
-Improvement of the "Average Shape Factor" from 0.740928 to 0.763868
-Improvement of the "Average Aspect Ratio" from 1.51 to 1.48
Unfortunately the delaunay-triangulation generates some critical elements, which are not acceptable.
Please take a look on the file i attached to this thread.
As you can see, all four nodes of the critical elements are in one plane and some elements contain nodes, which are not directly neighboured.
Does anybody know how to solve this problem?
Best regards!
