I am trying to impose an epsilon boundary condition to the near wall node I that is equal to the one proposed by Chien and Launder (1980) paper. The approach is to tweak the discretised equation of the near wall node by using a big source term and an imposed value equivalent to the equilibrium lenght scale if y+>5.1. The implementation seems to be ok but I am testing the approach on a rectangular cavity and it seems that something goes wrong (
code attached). A few questions:
1) Does the imposed value need to be assigned to the pointer cvar_ep(iel) or to a new dynamic array ?
2) Is there any examples on how to do a weighting when there are 2 boundary faces for a single cell volume?
3) Is epsilon equation solved before k when ikecou=0?
Could you be more specific in "something goes wrong" ?
1) If you want to impose a value, you can either use source terms (as seems to be the case here), or boundary conditions. Neither is directly in cvar_ep(iel)
2) not to my knowledge. Cells with multiple boundary faces are mostly in ridges and corners, in which wall models are usually not so good anyways.
3) I am not sure. you would need to check in the source code, or simply use a higher verbosity in the numerical options so as to see which is solved first.