Hey all, for a turbulence s
 

Hey all,

for a turbulence simulation using the k-epsilon model I also have to consider wall functions. Comparing the pitzDaily example with simulations of cfx a strange difference at the upper wall (where the wall distance is reducing again) appeared. In the simulation with openFoam the the velocity near the wall was about 50% of the inflow, in contrast to cfx where the velocity was almost zero in that area.

One point I recognized is that the steady state solutions of cfx and of differ and cfx seems to stop much earlier. But after all I think the difference at the upper wall is also due to the wall functions. I've read that OF uses the nu_T wall function, in contrast to cfx which uses scalable wall functions. First, is there a big difference between both and second can I find more information about the wall functions used in OF? Is there a way to use almost the same wall functions?


Cheers
Florian