[cfdnoob]

Can some one explain the riemann invariants in relation to the boundary condition and the isentropic relations.

[FMDenaro]

Originally Posted by cfdnoob

It was a while back, but i can remember my teacher saying something about
that an inflow needs 3 parameters going in and 1 going out and for an outflow the other way around. Can someone explain what he meant again (i know this can sound kind of vague)?
Is it that I need to fix 3 values at the boundary condition and it gives me 1 value back and then at the outflow you get in one fixed value and get three back.

The analysis comes from the sign of the eigenvalues of the Euler equation for compressible flow. The distinction in the sign appears for subsonic and supersonic flows.
Any book on compressible flow can help you

[cfdnoob]

Originally Posted by FMDenaro

The analysis comes from the sign of the eigenvalues of the Euler equation for compressible flow. The distinction in the sign appears for subsonic and supersonic flows.
Any book on compressible flow can help you

Do you mean that for pressure, temperature and density you get a machnumber and the other way around?

[FMDenaro]

Consider the case of 1D Euler (mass, momentum, energy) equations.
You have 3 eigenvalues u-a, u, u+a, therefore if u<a the first characteristic curve dx/dt=u-a produces an information coming from inner to inflow, as well as produces an information coming from outflow to inner.
In case u>a you have a different framework, no information from inner to inflow and no information from outflow to the inner.

[FMDenaro]

of course a = sqrt(gamma*R*T) = sqrt(gamma*p/rho)

[cfdnoob]

that sounds like the riemann invariants
In The programm i use it says i need to set 2 (total pressure and temperature) and M is allowed to float. So why are there only 2 things 'coming in'