Hi..

I work with turbulent natural convection flow of air in the square cavity with differentially heated vertical side walls. I use the following wall fonction for turbulent variables: K=0. at the wall Epsilon=Cte (K^(3/2))/l at the first inner grid point with l the closest distance to wall

I use Cte=0.4 is it true??? or may I use an other value???

Thanks for your help

Hi,

The constants in K-epsilon model are not the same for all flows.I suggest you look for relacted flows in square cavities.

Valdemir.

Hi, I consider well known van Leer class of the k-scheme: phi_f=phi_C+(1/4){(1+k)(phi_D - phi_C)+(1-k)(phi_C-phi_U)}. In many papers it is written that setting k = 1/3 we receive the scheme of the 3-rd order (so-called CUI scheme), but I receive the 3-rd order only at k = 1/2 (i.e. QUICK-scheme). Where can be an error? Grigory.