|
[Sponsors] |
March 31, 2005, 10:16 |
Gambit Meshing Problem
|
#1 |
Guest
Posts: n/a
|
I have a gambit model which essentially consists of various hemispheres (generated by creating an arc and a line, joining these into a face and then revolving the face by 180 degrees to form a volume) of different sizes. They all originate in the same place and the smaller ones are subtracted from the larger ones (and retained) so they do not overlap.
When I try to mesh this I get inverted faces rendering it useless to Fluent (causes divergence). It's hard to describe this using words so if the actual .dbs file would help I'd be more than happy to e-mail it to you. It is 1024 kb. Thanks, Andrew. |
|
March 31, 2005, 13:54 |
Re: Gambit Meshing Problem
|
#2 |
Guest
Posts: n/a
|
When you say they originate in the same place, I'm assuming you mean they are concentric (tangent would be a HUGE problem in meshing).
If you're trying to subdivide the volume, then I would recommend using the split command instead of the subtract and retain command. If you subtract and retain, then you have to be sure to connect the adjacent faces or the mesh is discontinuous. Split command can keep the adjacent faces connectd. As far as the meshing problem, you need to use the mesh check tool in gambit to find the problem area. Once you know the problem area you can start to investigate what went wrong. Usually you get inverted volume meshes when you have highly skewed surface meshes. That, or you have two surfaces coming too close together for the mesh size chosen. There could be other problems, but those are the most common. Hope this helps, and good luck, Jason |
|
April 1, 2005, 11:31 |
Re: Gambit Meshing Problem
|
#3 |
Guest
Posts: n/a
|
I tried splitting the largest volume into separate volumes by generating hemispherical faces (without a base, obviously) but still get the same problem.
By meshing the faces separately I found that it was the faces on the base of the hemispheres that were causing the problems. Having found this though, I'm no closer to solving the problem. Is there a certain type of element that copes better with circular geometries? |
|
April 1, 2005, 11:50 |
Re: Gambit Meshing Problem
|
#4 |
Guest
Posts: n/a
|
I changed the element type from Hex to Tet/Hybrid. This, along with splitting as opposed to subtract and retain, seems to have sorted it out.
I checked all the meshes and none of them had inverted faces. Fluent doesnt pick up anything untoward either. Thankyou for your help. |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
GAMBIT meshing problem for boundary layer | Falah | FLUENT | 2 | November 30, 2020 15:16 |
Problem with Ansys meshing | Asghari_M | ANSYS Meshing & Geometry | 1 | June 8, 2011 11:38 |
[ANSYS Meshing] Migrating from GAMBIT to ANSYS Meshing | David-CFD | ANSYS Meshing & Geometry | 1 | April 1, 2011 06:22 |
Gambit meshing problem | David Banks | Main CFD Forum | 0 | July 19, 2007 12:48 |
Problem starting Gambit | ViHAR MALViYA | FLUENT | 0 | November 10, 2006 11:26 |