What values/expressions does CFX use at the no-slip wall boundary to solve the k and epsilon/omega equations in the case of two equation turbulence models?
In CFX a flux boundary condition is specified for the k-equation: F_k = 0. So you see the conservative value in CFX-Post.
We specify finite values for epsilon and omega. For epsilon this is the log-law value, whereas for omega we have the automatic wall treatment (blending between viscous sublayer and log-law relation).
eps = Cmu^(3/4)*k^(3/2)/(kappa*y)
where kappa is the Von Karman constant for smooth walls.
Please note that epsilon at the first interior node is set equal to this value. The boundary value for epsilon is not used anywhere.
Please see also the Documentation (Solver Theory Guide => Turbulence and Wall Function Theory => Modeling Flow Near the Wall => Automatic Near-Wall Treatment for Omega-Based Models) for a detailed description.