CFD Online Discussion Forums

CFD Online Discussion Forums (
-   ANSYS Meshing & Geometry (
-   -   [ICEM] Patch dependent surface mesh problem! (

hadikhayyamian May 8, 2011 17:10

Patch dependent surface mesh problem!
2 Attachment(s)
hi guys;

I am trying to analyse a car (2D) in a wind tunnel test by ICEM and then Fluent. I sketched the car in silidworks then I expoerted in ICEM CFD in IGES format. When I define some surface and curve mesh parameters and use All Tri Patch Dependent Surface Mesh, my result is a bit weired (See the first atached photo). the interior surface have low density mesh! I even changed tetra size, but still has got problem.

When I use All Tri Patch Independent Surface Mesh, the result is quiet sensible (look at the second attached picture!).Attachment 7605

Attachment 7606

can any body help me!
Thanks Regards.

BrolY May 9, 2011 03:42

Your first picture means that the mesh didn't work. The patch dependent method requires a very good geometry. Your curves must be well defined. So, do a Build Topology, and check if it's right.

hadikhayyamian May 9, 2011 14:26

At Alexander;

Thanks for your reply.
Although I tried build topology, I am going to use it again. I will let u informed of the result.

zomayabssa February 24, 2014 12:01

Patch independant with inflation
2 Attachment(s)
Hi Guys,
I had quite the same problem when modelling a wind turbine blade profile and I fixed it by increasing the number of nodes in the curve mesh setup. But now I'm having two more problems:
1- My mesh is overlapping my trainling edge when using a patch dependant Tetra mesh with inflation around the blade profile. I tried to buid topology but it changed nothing. (see pictures)
2- I've noticed that when using Patch Independant the overlaping problem disapears but the inflation around the blade is not respected. So, is it possible to have an inflation with "curve mesh setup" when using patch Independant?
Thanks a lot in advance

juliom March 7, 2014 11:19

Yes you can remember that every time you use patch dependent you ened to use curve function, I mena curve size, in constrast with patch independent you need surface function... I do not see anything wring in your statement,.. If you solved your problema plesae share with us your approach!!

zomayabssa March 9, 2014 11:36

Hi Juloom,

Thanks for your reply,
Do you mean I need to define curve parameters like number of nodes, height and height ratio etc..? Beacause I've already defined these params. I also tried to create a curve topology but the problem still persist. :(



juliom March 10, 2014 10:03

Dear friend, My apologise for my grammar but I was writing very fast.
Did you check your geometry? this is very weird...
PLease try with patch independent and see if the problems is overcame.

zomayabssa March 11, 2014 12:38

Hi Juliom,

Thanks again for your interest.
Actually, Yes the trailing edge meshing problem disapears with the patch independant but the thing is that with patch independant I loose the inflation around the blade :( !



juliom March 11, 2014 15:03

This is only because you need to define the región on the airfoil as a regiuon to use prism and the far field tetra... this is the answer!!

zomayabssa March 12, 2014 20:06

I understand Julio but I've defined the region with "curve Mesh Setup" because I'm working in 2d. Prism is considered only in volume parts.

juliom March 13, 2014 08:35

Wow I see, you are right my colleague. So there might be a mistake or gap in your domain.
Please try to use a smlaler mesh size and a smaller tolerance while you are performing the domain check...

zomayabssa March 17, 2014 14:52

1 Attachment(s)
Thanks a lot Julio,
I don't know if you were refering to "Ignore size" in General/Global_mesh_setup when you said "smaller tolerance" but it seems to work perfectly when I put a much smaller value! (See attached figure).
Thanks again ;) !

juliom March 17, 2014 15:04

Nice, this is very rewarding for me, I wish to help everubody ask a question in this fórum and I will live 100 years helping colleagues and students, while I am learning from other experiences.
Waht you did is exactly what I mentioned bu tfrom another approach, at the end was the same!!!

All times are GMT -4. The time now is 15:26.