Question regarding Y Plus (y+)
 

I am using an SST k-w model to simulate airflow through an awkwardly shaped duct. I was told that if 95%+ of the y+ is below 5, then it is sufficiently accurate.

However, I found that if I simply add 'cells across gaps' and 'inflation layers', I get mostly sufficient y+ values (before I added more cells and layers, about 70% was below 5. After, it is about 98-99%). This seems like an inaccurate method to get the majority y+ values<5, but I am not sure.

So basically, can I assume my simulation is accurate if 98-99% of it is y+<5? Even if all I did was add cells and layers?

Simple answer: No.

This is just one of several criteria you should check. You may have sufficient points in your viscous and log layer region, but that doesn't say much about your other mesh qualities. Or about the mesh independency, choice of turbulence models, inlet conditions etc. For instance, you should check that your transition between these inflation layers and the 'main' flow happens smoothly. A large expansion factor can mess up your results.

So you are interested in what is going on near walls rather than in the middle of flow domain, right?

with new wall treatment even you can go up to yplus 10