2d axisymmetric: nozzle--> massflow too low?!

Friederike

I am investigating annular inlet nozzles for different compressible flow regimes. For simplicity I am using a 2d axisymmetric approach. But for different geometries that I tried, the nozzle chokes at much lower massflows (-30%) than it should be (ideal gas). Does anyone have similar experiences? A Fluent guy mentioned to me once that the massflow predictions of the 2d axisymmetric solver are less reliable. But with 30% discrepancy - do I have to go 3d then? Any help is much appreciated! Thanks!