[Ozzy]

Hello all,

I am at the moment writing a code for a uni project using a second order Lax-Wendroff scheme to model unsteady flow.

I am trying to implement boundary conditions using characteristics but am struggling....the method described by Benson for a 'closed' boundary does not produce reflections but results in big oscillations in my program:

1) Do I not need to set the temperature at the boundary to be a constant?

2) Will the homoentropic method at the boundary suffice?

I would really appreciate any help

Thanks,

Ozzy

[ignat]

Originally Posted by Ozzy

:

1) Do I not need to set the temperature at the boundary to be a constant?

It depends on flow type at the boundary.

if it's subsonic INflow (M<1) then 2 constant values should be specified, usually total Entalpy and entropy





and 1 parameter should be determined by closest cell inside the domain, ususlly velocity u.

if bounday is subsonic OUTflow then 1 constant parameter
is specified - ambient pressure and 2 others are dtermined by closest cell inside the domain.