turbulent kinetic energy condition at wall

johndeas

Hi,

a very simple question. I always thought that, since in DNS the trace of the reynolds stress tensor is null at wall, a suitable condition for RANS calculation would be to fix the value of k=0 there. However, I recently came across a thesis in which a guy using Fluent used a zero derivative for k at the wall, and also an OpenFOAM tutorial with the same condition. Boths use wall function. I wanted to know the logic behind this.

Thanks,

JD