Dirichlet B.C. for k-epsilon

Jens Wyrwa

Hi,

For a scientific comparison, i try to simulate a grid tank with fluent. There is no mean flow in this experiment. Turbulence is produced by an oscillating grid at the bottom of the domain of interest. Turbulence performs diffusion and dissipation of itself in the fluid.

Experiments and analytical solutions are available for the case.

I want to give a Dirichlet (fixed value) boundary condition for k and epsilon on the bottom of the tank. (side walls - symmetry; top - pressure outlet) Is it a mistake to specify a " velocity inflow" boundary, is it impossible to specify such kind of boundary with Flunet 5.4 , or is there a bug in Fluent's standart k-epsilon discretisation?

Thanks for help in advance Jens