Folded faces in Gridgen.
I have been trying to extrude an unstructured tertahedral grid off a wing using Gridgen. I am using the COBALT package to run simulations on the resultant grids.
However, COBALT pre-processing (Blacksmith) keeps returning errors that the grid contains folded faces. I was wondering if there are any tools in Gridgen that I could use to check for and/or stop the creation of folded faces during the extrusion.
Additionally, any tips on how to build more robust grids that can capture boundary layer flows over complex objects (emphasis on using Gridgen) would be much appreciated. I have been having other issues extruding blocks from surface domains. Specifically, one of the problems that I have been having is that the block tends to split at corners or edges where the surface domains connect. This creates crevices in the block itself that are impossible to 'fill.' Also I was looking for a better method of estimating initial spacing of the extrusion to maintain low Y+ values.
Examine tools in Gridgen, etc.
There are several tetrahedral mesh quality measures available in Gridgen's Block Examine menu including volume, minimum included angle, maximum included angle, equiangle skewness, equivolume skewness, and centroid skewness. I am not sure what is meant by folded faces in Cobalt, so I don't know if any of these checks would indicate folded faces, but they will at least give you some measure of grid quality.
If you end up with gaps between blocks of extruded grid, it means the original domains were not connected to each other. You can check that quickly in the Connector, Merge Entities menu. If you see a red connector between the two domains when you are in this menu it means they are not connected. There are several quick cleanup tools in Merge Entities to correct any problems you find.
Have you looked at T-Rex meshing in Gridgen? It extrudes anisotropic tetrahedral grids and automatically detects interferences, sharp corners, and poor quality grids. Try running Gridgen's DLR-F6 tutorial to see how it works.
Finally, there is a y+ estimating script available for free download at the Glyph Script Exchange.
Hope this helps.
The T-Rex meshing did help a lot. Thanks. I feel like an idiot for not having tried that sooner. I just thought it would be more trouble than doing things like I have usually done (foolish pride and arrogance combined with ignorance). Thing is that I haven't been using the software for very long and I have no peers at hand who have ever used it before.
Gridgen has lots of hidden gems
Gridgen and Pointwise are continuously updated, so there are many features that long time users may not be aware of. We try to keep people aware of them through The Connector email newsletter, Pointwise Pointer, Focus On, and Focal Point print mailers, and our website, but most people are too busy to keep up with it all.
Don't hesitate to email email@example.com for help or even what-if questions to find out if there is an easier way to do things.
|All times are GMT -4. The time now is 18:49.|