turbulent Prandtl number
 

Hello!
I'm trying to verify turbulence models for heat exchange in flat plate boundary layer (air, Re about 10^5). Spalart-Allmaras results are good with conventional constant turbulent Prandtl number 0.9. However, Baldwin-Lomax is better for Pr_turb=1.2. I tried Kays-Crawford variable Pr_turb formula and realized that it is somewhat paradoxical for me: if turbulent viscosity is small, as, I suppose, it is far from the wall, then Pe_turb=0 and Kays-Crawford model gives Pr_turb=2*0.85=1.7 (just as at the wall surface). Is it Ok far from the wall?