BC for a wind tunnel

spepsicho

hi all

I'd like to simulate a typical problem: a car in a wind tunnel, to get correlation with real tests.

I've read many articles that say the BC type to apply for the top and external side walls is the symmetry one. But why??? In a real wind tunnel, the top and side walls are not moving and they (unfortunately) interfere with the tested model: the properties of symmetry BC (all gradients = 0, I guess) are thus not always verified. Shouldn't be more appropriate to simply use stationnary walls?

Please give me your ideas.

regards