Hi CFX-user,

I have a question about the near wall region when using the standard k-epsilon model. The velocities in-between the wall an the center of the first control volume adjacent to the wall are modeled using the universal law of the wall. The epsilon in this region is set to epsilon=cmu^(3/4)*k^(3/2)/(kappa*d-y). What is the boundary condition for epsilon in the stand k-e model? And what is about the model of the k-equation in the region from the wall to the center of the first volume adjacent to the wall? Are there any changes in the source terms? Cheers

Fabian

Fabian,

Epsilon is impossible to define at a solid surface, its near-wall assymtotic solution is unknown! That is one of the reasons for using wall functions. Instead of apllying BC's at the wall, wall functions apply the BC's at the first grid point out from the wall (which must be in the log-layer) and the near-wall region is not modelled. The epsilon BC at this grid point is calculated with the equation you gave in your message. There are similar equations for k, and the u in the log-layer region (they'll be in the manual). So effectively the outer solution is made to assymtote to the log-law of the wall. NOTE:wall functions often lead to poor results as the law of the wall does not always hold true. Especially in adverse pressure gradients and separated flow regions. So be careful.

Regards Steve