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] Low quality mesh (http://www.cfd-online.com/Forums/ansys-meshing/102070-low-quality-mesh.html)

Memme May 21, 2012 11:59

Low quality mesh
 
I have to mesh an object and some of its changes. The critical area is represented by two surfaces with an angle of 30 degrees.
After this process I have to import the mesh in fluent.

Until the creation of prism, the quality of the mesh appears very good (> 0.1-0.2).

I explain what I do step by step.

I import a acis format, built topology (all lines are red).
I create a Octree mesh, delete all the volume elements and proceed with a smoothing operation:
- 50 iterations up to 0.6 (Laplace with allow node merging activeted);
- 10 iterations up to 0.4 (no Laplace).

After I create a Delaunay mesh to fill the volume.

Before generating prism:
- 10 iterations up to 0.6 (Laplace);
- 10 iterations up to 0.6 (no Laplace);
- 10 iterations up to 0.6 (Laplace);
- 10 iterations up to 0.6 (no Laplace).


I extracted a good mesh with these prism settings, but I have to improve it:
Initial height: 0.0002 (this size is necessary).
Height ratio: 1.2.
Number of layers: 5.
Fix marching not activated.
Min prism quality: 0.01.
Ortho weight: 0.2.
Fillet ratio: 0.1.
Max prism angle: 180.
Smoothing options:
Surface smoothing steps: 0.
Volume smoothing steps: 10.
Max directional smoothing steps: 5.
First layer smoothing steps: 1.
Auto reduction activated.
Use existing quad layers.

I noticed that there are a lots of problematic tetra_4 elements that not disapperead after a smoothing.
Also often after the smoothing, the export fails.

I've try also different settings to limit the pyramids, but I didn't got good results:
Prism settings:
Exponential.
Initial height: 0.0002.
Height ratio: 1.2.
Number of layers: 5.
Fix marching not activated.
Min prism quality: 0.000001.
Ortho weight: 0.2.
Fillet ratio: 0.1.
Max prism angle: 180.
Smoothing options:
Surface smoothing steps: 0.
Volume smoothing steps: 10.
Max directional smoothing steps: 5.
First layer smoothing steps: 1.
Auto reduction activated.

Finally:
- 50 iterations up to 0.6, penta_6 freeze (no Laplace).
- 5 iteration up to 0.01 (no Laplace).

I do not know what to improve.
I used also thin cuts and edge criterion options.
Before the prism, the histogram appears clear until 0.2.
Thanks in advance and sorry for my english.
Best regards.

Ralen May 25, 2012 02:02

Attach *.tin.


All times are GMT -4. The time now is 17:07.