While doing 2D simulation of underexposed and over-expanded supersonic jets my results is getting converged with low NPRs (2-4) with choosing Zero Supersonic/Initial Gauge Pressure at the inlet of nozzle in the Fluent solver. But it get diverged with high NPRs (4-10) value. However results are converging when putting nearer value of Supersonic/Initial Gauge Pressure to inlet nozzle pressure. Kindly help me to resolve this problem ?