k-Omega-SST: y+ values explode
 

Hello everyone,

I am calculating the cooling of a hot wall. My case is a 2D channel filled with water. My velocity is ramped and starts at 0 m/s and rises to about 5 m/s. The temperature is 1000 K at the wall (coupled, decreasing with time). The temperature of the water at the inlet is 300 K. I pre-estimated my yPlus for 5 m/s and a value of y+=1. My model is k-omega-SST with low-Re correlation.

After my first calculation was done, I evalueted y+ and it was around 55 instead of 1. So I reduced the distance of the first cell to 1/10 of the original value. I my view this should decrease y+ from 55 to about 5.5. To my surprise y+ went up to 55 again. Where am I mistaken? Is my model still valid, even though I am not resolving the viscous sublayer? How can I decrease y+?

I plot turbulence -> yPlus -> direction along my boundary.