Hi.

We are considering the use of CFD. We have a case (an airdistributor) with an outlet Re=10000 at a velocity of 8m/s. The air is guided through the distributor where it changes direction 2 times (2x90° bends).

Can anyone give an estimate of the expected accuracy of a simulation?

Thank you in advance.

Christian

Your geometry sounds pretty straightforward. The velocity is low enough that you're practically incompressible. Reynolds number isn't in the transition regime. It should be fairly easy to get guite good accuracy. To be honest though, it sounds simple enough (unless there is more to the geometry than a couple of bends) to do by hand using Idel'Chik or similar.