Hi

I believe i have finally got my simulation working on a 3d airfoil, it converes well in under 40 interations and I think the mesh is of good quality (inflation layers etc).

The problem is that after a certain angle of attack the solution will not converge and oscilates periodically. In post processing it seems the airfoil is begining to stall ( ie seperation is visable). Would you all agree this has become a transient simulation?.

If so should i use the previous AOA results file for the inital values, and is there any way of working out what the timescale should be and also therefore the timesteps. Im completly new to transient solutions and would have prefered to avoid them!

Thanks

Steve

Thought id add comment...

My primary aim is to simply get Cl and Cd slopes vs AOA thats all, when the airfoil is stalling and wont converge, the peak of the oscillation is higher than the previous lift value and the trough is lower. Can i just assume that the value at the bottom of the trough is the correct value?

Thanks Steve

Hi,

As it says here: http://www.cfd-online.com/Wiki/Ansys...gence_criteria

Do a transient simulations. Use adaptive timestepping to quickly find the appropriate time scale.

Whether the average value or the trough depends on the application. If the frequency of the oscillation is high then the average will probably be most appropriate. If it is low then the trough might be better.

Glenn Horrocks