grid adaption problem

justin

Now I am studying wind environemnt around buildings. The used turbulence model and wall function are rng k-eps and standard function. the domain and model are boxes. The separation really occured, but the velocity of some points on windward is larger than that at inlet boundary.

The problem is that if I set the lower y+ max allowed, the size of cas and dat files will increase from 150Mb to 800Mb and it seriously affect the speed of calculation. also the increased cells mainly are near windward face.

What is the problem? Can I just make the y+ of near the model wall lower than 150 and keep the other wall about 400?

Thanks