CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   CFX (https://www.cfd-online.com/Forums/cfx/)
-   -   mesh adaptation for hexa mesh (https://www.cfd-online.com/Forums/cfx/21130-mesh-adaptation-hexa-mesh.html)

Pete March 31, 2005 04:41

mesh adaptation for hexa mesh
 
Hi,

I generated a hexa mesh from ICEM & applied the mesh adaptation in CFX 5.7.1. I used 2 adaptation steps. However, I get lots of prism & tetra elements after the adaptation and the solution is getting diverged (work fine without adaptation). Any remedy for this? The mesh adaptation doesn't seem to be designed for the hexa mesh anyway. Thanks for help. Pete

Glenn Horrocks March 31, 2005 18:26

Re: mesh adaptation for hexa mesh
 
Hi Pete,

The refiner does not refine hex meshes to hexes, rather they are refined to prisms and tets. Hex meshes are more tolerant of high aspect ratioes and other mesh nasties so if you refine a high aspect ratio hex mesh you will get a high aspect ratio tet mesh and it will be harder to converge.

Glenn Horrocks

caty April 5, 2005 21:07

Re: mesh adaptation for hexa mesh
 
Hi,

I have a hexa mesh and I need a finer mesh (indeed i need to remesh it), but i am thinking if I can use mesh adapt.

The file is transient, but it is posible to run a steady state with mesh adapt and after that to use that res file asd initial condition for the transient file using the interapolation?

Where can i see how my new mesh looks like?

Glenn Horrocks April 6, 2005 18:38

Re: mesh adaptation for hexa mesh
 
Hi,

Yes this is possible. You can load the mesh from a steady state mesh adaption results file into CFX-Pre and it will then be used as the mesh for whatever run you set up. As long as you keep the topology the same there should be no need for interpolation, just using the previous results file as an initial condition should work.

Glenn Horrocks

caty April 6, 2005 18:43

Re: mesh adaptation for hexa mesh
 
Hello Glenn,

thank you for the respone.


All times are GMT -4. The time now is 00:49.