Gambit: skewed mesh
I´m trying to mesh a symmetric halfmodel of a fuselage, which was imported as a step-file from Solid Works and always have some nodes, who fail the skewness check. I use triangular meshelements with a spacing of 25 on the fuselage and with a spacing of 200 on the symmetry plane. The volume is meshed with Tet/TGrid and a spacing of 200 as well - any suggestions?
Re: Gambit: skewed mesh
Don't used fixed spacing on your volume and symmetry plane. Use sizing functions to control the growth of the mesh from the fuselage out (they're available under tools... looks like a yellow radar screen).
Another thing though. If you're looking for a viscid solution, then your going to have huge problems if your first cell height is 200 (and probably just as bad if your first cell height is 25). The first cell height controls your boundary layer. For most of the turbulence models in Fluent, you can either shoot for a y+ value of less than 1 (which is a really refined mesh) or between 30 and 300 (y+ is related to wall shear stress and cell center height... you can use http://geolab.larc.nasa.gov/APPS/YPlus/ to estimate the cell height you need to get the right y+ values). For a simple fuselage, I would recommend the Spallart-Almaras model, and you'll get decent values for the boundary layer and wall shear stress (which is a big player in your drag estimations) if you shoot for y+ values between 30 and 300. For relatively simple external aero models using the SA model, I haven't seen much of a benefit for shooting for a y+ of 1 (considering how much more refined the mesh has to be). You can save some elements by using a Boundary Layer mesh, but this doesn't always work in Gambit, and you may have to go to TGRID to accomplish this.
I recommend looking into the Gambit manual for more information on Boundarly Layer meshes and Sizing Functions.
Hope this helps, and good luck, Jason
|All times are GMT -4. The time now is 00:26.|