CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CFX (
-   -   Icem - solver (

Jones October 14, 2006 08:37

Icem - solver
I have been trying to make a grid around an aerofoil using Icem. I have had a few problems when I try to run it through the Multiblock-info solver. I think it would be helpful if i knew how this solver generated the numbers. For now, it is jumping from -1 to -2 then back again...when it is meant to go smoothly till about -5. I also used bigeometric and other curves with points close to the aerofoil having a spacing of 10^-5. My tolerances for the curves were 10^-6 and for surfaces 10^-5. Also when I create a point using the curve intersection tool, the curve moves slightly. This is only visible if you have zoomed into the area quite alot. This is with setting the gap tolerance to 10^-6. SO after you try to create a surface with it, the curves dont join and it gives errors. I hope it is all understandable. Please let me know if it is not. Your help as soon as is possible would be really appreciated

myron October 16, 2006 07:57

Re: Icem - solver
With such tight tolerances, you'll need to reduce the Triangle Tolerance value in the Global Mesh Parameters. Probably down to 10-5 or 10-6. This will make the geometry more 'true' to reality. Change the setting then save and reload the geometry. (You'll want to do any geometry manipulations with the smaller value.) This should help the meshing to be closer to the true geometry.

Jones October 17, 2006 17:56

Re: Icem - solver
not sure what you mean by triangular tolerences. What does that mean? How does it affect everything? Where do you do that?

myron October 19, 2006 05:39

Re: Icem - solver
Look in the Global Mesh Parameters settings for 'Triange Tolerance'. When a tetin file is loaded, it's actually facetized for memory purposes. When you lower the 'triangle tolerance' and re-load the tetin file, those facets are smaller so the geometry is closer to the true shape. This only has a large effect if you're using very tight mesh spacings - and/or you have large differences in geometry sizes.

All times are GMT -4. The time now is 18:04.