Multi-element airfoil problems.Scaling turbulence?
 

I'm doing a multi-element airfoil study. The conditions are: Re=25E6, Incompressible. As the Reynolds is relativly high, I'm using a k-eps High-Reynolds with SW157 (NON-EQUIL Wall Funct). I'm using a cartesian mesh (with a sublayer envolving the airfoil) for better control on yplus. A usual mesh from pro*am tool.

After a few runs I've seen that yplus is always bellow, say 10-15 in most of the airfoil. Comparing the results with EARSM turb model (not in Star) the separation in EARSM comes at 19 degrees while in k-eps we are at 24 degrees and still no separation. Could the small value of yplus be retarding the separation?

Lift at k-eps is always a bit smaller (until separation is reached) and Drag always bigger (until separation). Could it be due to bigger friction?

As there are recirculating zones (behind slat for example), is this model any good? Should I use non-linear models??? Cause off the Non-Equilibrium zones far away from the wall?

By the way... I'm scaling my model to unitary chord. To make the study at the same Reynolds as the real chord, should I act on viscosity or speed? It should be the same.... But..... This doubt is because of the turbulence and lenght apllied to the inlets?

HOW DO I SCALE TURBULENCE? This is confusing!!!!