Hey Eric,
first guess: 100 is not a really fine resolution for a Reynoldys number of 4000.
I need some additional information to evaluate possible error sources:
– boundary condition at the cylinder
– used turbulence modell
– max Mach number
Best Marc