CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   need to get sufficient results in 3D with TET (http://www.cfd-online.com/Forums/cfx/115980-need-get-sufficient-results-3d-tet.html)

imnull April 10, 2013 10:32

need to get sufficient results in 3D with TET
 
I need your advice.. My 2D HEX model and experiment are in agreement.
Nevertheless I need 3D. I am not good in ICEM to make 3D HEX (I have lots of freaking steps in geometry..). So I did 3D TET, results in numbers are not matching with 2D HEX.. but I would say If to refine the mesh, the results 3D TET will match to 2D HEX (I did try it already, with an increase of elements the results getting closer to 2D HEX). But, the way to refine TET is not good for me I already have 2.1M TET.. so if I put in *.def mesh adoption option.. will it helps to safe computing time to get results close to 2D HEX..? any advice are welcome thanks for your time. Thanks.
P.S. btw: i do model supersonic flow, with max M~2.5.

oj.bulmer April 10, 2013 11:43

21 million mesh!!? Are you sure?

You don't have to refine the mesh at every corner. Have a look at your 2D results and understand where you have larger gradients, and try to refine only there. Or, try mesh adaptation.

OJ

imnull April 10, 2013 12:21

1 Attachment(s)
Quote:

Originally Posted by oj.bulmer (Post 419674)
21 million mesh!!? Are you sure?

You don't have to refine the mesh at every corner. Have a look at your 2D results and understand where you have larger gradients, and try to refine only there. Or, try mesh adaptation.

OJ

yes my 3d tet has around that value.. i do have a blade with 16 cores and 36GB.

pls take a look on results of 2D HEX

Attachment 20643

as seen I have 5 areas... so, at each freakin corner.. hope you see the problem with using TET now..:rolleyes:

my run with mesh adoption is still on (2 days already).. converging process is ok (i do control it manually). will report on results.. do you have an experience to work with supersonic flow (I have some more @stupid@ questions)?

imnull April 11, 2013 09:57

calculation with mesh adaptation is done... yes, adaptation helps.


All times are GMT -4. The time now is 06:00.