Your problem cries out for an empirical solution because of the chaotic nature of the gas flow. Nozzle efficiencies and pressure drops thru the piping makes this a difficult problem to establish precise boundary conditions for. It would probably would be simpler to construct your device and hook it to a welding tank of nitrogen with a regulator. Vary the pressure to establish this requirement and put a flow meter in the line to calculate the required volume of your tank. It would also be easy (if your design allowed) to vary the nozzle diameter and optimize it against the input pressure by experimentation.
just say'n,
Timelord