Hi all, about the choice of the inlet-outlet BCs for a subsonic compressible nozzle flow is a difficulty puzzling me. And the non reflecting conditions presented by David H. Rudy and the NSCBC presented by S.K. Lele appear not obviously suitable for the probleme of this type.Any help is greatly appreciated. Best regards

At the inflow, characteristic analysis dictates that the transverse velocities be specified (usually zero). Two variables from the group density, temp, and longitudinal velocity must also be specified, and one variable (usually longitudinal velocity) must be allowed to float.

At the outflow, the transverse and longitudinal velocities, as well as density, are allowed to float. Temperature is specified. Often, pressure is specified, which in turn leads to specifying temperature after calculating density.

These conditions will reflect accoustic waves. For non-reflecting conditions, your best bet is to follow the work of Kevin Thompson and Poinsot and Lele.

Hi, You can also follow Francois Dubois recipes according which the problem is well posed if: 1) for inlet boundary the density and temperature is given; 2) for outlet boundary there are two cases: 2.1) subsonic outlet BC => static pressure has to be given; 2.2) supersonic outlet BC => no datum has to be prescribed. All above is related to Riemann problem solution between boundary manifold and state. Happy coding!

Also, be advised - you have to solve linearised Riemann problem, so, in case of subsonic outlet, the prescribed static pressure has to be close enough to that one in the last cell...