[bmikuz]

Hi,

I've prepared a LES simulation of turbulent flow in a fuel bundle using OpenFOAM version 2.3.0. Firstly, I used a Smagorinsky model with vanDriest damping at walls (fully hexa mesh with about 80 milion cells, average yPlus is 2.1, 2nd order schemes) and obtained results were reasonable. Now I want to test also a dynamic version of LES, namely homogeneousDynSmagorinsky model. Therefore, the vanDriest damping was switched off (in LESProperties) and test filter had to be selected (simple, laplace or anisotropic). Specifically, my first choice was the simple filter, which does not require any additional parameter. The problem is that the calculated nuSgs values are almost 1000-times smaller than molecular viscosity nu! Something must be wrong with the turbulence modeling!? Any suggestion or comment is welcome.

Since I don't know what could be the problem I decided to change test filter with the anisotropic one. Here I have to specify widthCoeff, which is not documented well and I don't understand what is its effect. For now I set it to be equal to 6 and the simulations is running nicely, but it would be really nice if somebody could explain what is widthCoeff? I couldn't find the answer from previous posts on the forum...

Thanks,
Blaz