[Pranjal193]

Hi,
I am simulating a HAWT (horizontal axis wind turbine) , specifically NREL Phase VI blade with a slot to help in passive boundary layyer control . I am simulating just one blade with a moving reference frame as was used in my primary article of reference. As I keep adapting mesh count based on y+ values the torque value (obtained from moments menu in reports) keeps going down, and significantly, moving further from the desired value. It was ~ 120 Nm at the count of 1.5M cells, and when increased to 9.2 M, it was 90 Nm. I dont understand how refining the mesh can lead to less accurate results.

I am using a tet mesh with no inflation layers as ANSYS meshing was creating bad quality elements when I tried to. I am using the SST k-omega model, and by refining the mesh I expect to get more accurate results. Is the lack of an inflation layer the problem here? Or, is it the slotted geometry(which might need a very refined mesh?)?

Considering that I do not have ICEM CFD, is there a way to make a better mesh if that is the problem?

[AlexanderZ]

i can hardly imagine you are able to control y+ value without inflation layer.

you may use mesh without inflation layer in case you are using wall treatment functions

on the other hand as you are focused on boundary layer simulation it sounds essential to have good mesh in that region.