CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   ANSYS Meshing & Geometry (https://www.cfd-online.com/Forums/ansys-meshing/)
-   -   [ICEM] 2D Mesh of a Venturi : Errors during importation under FLUENT_v6 (https://www.cfd-online.com/Forums/ansys-meshing/84737-2d-mesh-venturi-errors-during-importation-under-fluent_v6.html)

 eeesprit February 7, 2011 17:33

2D Mesh of a Venturi : Errors during importation under FLUENT_v6

1 Attachment(s)
NB - Sorry for the english mistakes that would certainly follow...

First of all, good evening !

I'm working on a project, which aim is to study the possibility or not to cool the fuel cell of my school's car prototype, using the depression created by a Venturi which would be integrated on the structure of the vehicle.

The thermic part is over - now i wanna to see the impacts on the car's aerodynamics, especially drag coefficient Cd.

I started with very simple shapes : cylinder in an air flow - i am satisfied with the results which are very close to the theory.

Before trying to mesh complex shapes (i.e. the shape of the car) i would like to try with this :

a Venturi inluded in a Sphere (see the attached picture)

Because i'm beginner with ICEM & Fluent, i don't know how to mesh such a situation. Indeed, there are some "isolated walls" just in the middle of our an air flow.

Anyway, i tried to do something like that :

- definiting 3 surfaces for the Fluid (air)
- meshing these surfaces (i tried both TRI & HEXA unsuccessfully)
- setting boundary conditions
(by the way... what's the appropriate boundary condition for separation between my "fluid" surfaces ? interior ? interface ? thanks)

BUT

when i try to read the mesh with Fluent :

Quote:
 ... Warning: Invalid face type 1. Skipping ... Warning: cell 0 of thread 46 has NULL face pointer 2. Clearing partially read Error: Build Grid: Aborted due to critical error.

I read some posts on this forum with users who had the same problem : it was often related to "Uncovered Faces".
But i check the mesh, and there aren't any "Uncovered Faces".

Any ideas where the problem could come from ?? :confused:

Thank you very much.

Thibault

 PSYMN February 8, 2011 14:12

If this is a 2D model, how did you model the 4 regions? Are they separate parts (zones)? Do you have line elements between the zones (check this by turning on all parts, turn off shells, turn on lines... Do you see line elements between all the zones?

You should, other wise fluent can't keep the fluids apart.

Why do you even have 3 fluid regions? Can you put them all in the same part/zone? Then you wouldn't need line elements between them.

 eeesprit February 9, 2011 12:20

1 Attachment(s)

Quote:
 If this is a 2D model, how did you model the 4 regions? Are they separate parts (zones)? Do you have line elements between the zones (check this by turning on all parts, turn off shells, turn on lines... Do you see line elements between all the zones?
Yes i think i did this correctly ; indeed i have 3 surfaces, separated by lines.

Quote:
 Why do you even have 3 fluid regions? Can you put them all in the same part/zone? Then you wouldn't need line elements between them.
Actually, because i'm not used to ICEM, i thought it was the easiest way to do for me.
If i don't make these separated zones, how can i mesh the volume properly without taking in consideration the "wall elements" (i.e. the sphere) ?
That was a kind of "tip" to mesh the volume part after part, but maybe this is the reason of the errors under Fluent...

Basically, what are the main steps i need to do to mesh such a structure ?

Thanks,

Good evening

Thibault

 PSYMN February 9, 2011 14:13

What mesher are you using? Hexa? Patch Conforming quad?

 eeesprit February 9, 2011 14:15

Interesting.... :rolleyes:

Instead of setting 3 different surfaces with 3 names (SURF1, SURF2 & SURF3) i try to set only one name : SURF.

Then, i mesh the 3 surfaces (corresponding to one part only) together, and after creating a file.msh, i have a new error :

Quote:
 Error : FLUENT received a fatal signal (SEGMENTATION VIOLATION)
That's still doesn't work, but i'm going on ...
There are some other topics relating this error, certainly caused by a lack of memory ...

what should i do to avoid this error ?

 eeesprit February 9, 2011 14:16

Quote:
 What mesher are you using? Hexa? Patch Conforming quad?
That's it : Patch Conforming Quad !

 All times are GMT -4. The time now is 16:27.