Someone asked this same thing recently
Start with a 2d analysis and spend time there on mesh and comp domain size
I read all the airfoil related posts before I post. I did 2D analysis, local meshing, played with ratios. I do something wrong but I don't know what it is. Maybe missing a simple thing. That is why I have explained everything step by step.
you'll see in several other posts recommendations to do both global goals and also surface goals
what does the final mesh look like? how sensitive is the solution to mesh? what about comp domain size?
what are you comparing the result against? hand calc? known values? test? how close do you need the result to be?
what do you know about the turbulence developed by your wing? is it possible it is outside the scope of KE?