[allenfieldin]

Hello, foamers

Recently I want to use the dynamic Smagorinsky model for my LES simulation on a incompressible channel flow over a wavy bottom (like a sine-shaped hill), with horizontal planes being cyclic. I am a freshman using turbulence models, even conceptually not clear about that.

In openfoam, it is homogeneousDynSmagorinsky. But I have read a thread (actually quite long) talking about different models, it is quite confusing because the thread has covered different versions of foam.

1. So the first question is: in OpenFOAM2.3.0, the homogeneousDynSmagorinsky still calculate the Cs coefficient globally(based on the averaging of the whole computational domain)? or it now can account Cs locally?

2. I found there is a modified model proposed and uploaded by @alberto:
See: Improved implementation of dynamic Smagorinsky
where @alberto claimed that it can calculate the Cs locally. I want to know how does it work? have some one tried that?

3. I read Eugene's thesis, where he used oneEqEddy model and claimed that the dynSmagorinsky model will not deliver good performance around the wall boundaries which is important in my study. But I recently read a paper by Yang&Shen(2013), they actually uses dynamic Smagorinsky model for a wavy-bottom channel flow with high Re_tao=4096.
So, can someone give some opinion on that? I want to perform LES with both models at a low Re_tao and compare the results with DNS results.

4.Has someone got a tutorial using dynSmagorinsky model? like the numerical setup especially the wall function and the initial conditions. I think a tutorial would be of great use for beginners like me.

Thanks, any comment or suggestion would be quite really helpful.

[allenfieldin]

Oops, nobody comments on this, heart break