Hi, I need to compute flow calculations for a channel which contain a winglet on one of its walls. I would like to compare results given by DNS and RANS. I have access to Fluent, and to a finite-differences code based on Orlando book, but I am affraid a configuration which include a winglet will not be easy to model with the later.

What should I do ? Are there some tricks to use finite differences anyway ? Or is there some software which could do DNS with a finite element discretization ?

I hope that the software CFDesign which is discretising by fine element method may be suitable for your DNS and RANS.