[harry123]

Hi,

I happened to read that, use of the conventional outflow boundary condition in conjunction with specified pressure boundary constitutes an indeterminate problem. But I'm not able to understand why.

I will be thankful if someone could comment on it or provide some sources from which I can get to know about this topic.

Thanks in advance

Regards,
Harry

[saurya]

Hi Harry,
The effect of boundary condition on stability, convergence and accuracy of the numerical schemes is very much case dependent (exit flow conditions; subsonic/supersonic/mixed flow regime, and the size of the computational domain etc). The conventional exit boundary condition with imposition of exit static pressure may not be problematic for a case when it doesn't un-physically (severely to blow up the computations) obstruct the passage of acoustic and material waves between the computational domain and free stream surroundings.
The conventional method can be reliably applied to the problems in external aerodynamics where the exit boundary is far (around 20/30 chord distance) from the lifting bodies. In contrary, limited domain size in internal aerodynamics causes significant influence of the developed flow structures on the exit boundary. This would lead to convergence and accuracy issues, unless the direction of propagation of the waves are properly considered during boundary treatment. In order to avoid unphysical reflection of the waves at the boundaries, the method of non-reflecting boundary conditions based on characteristics can be applied.
For more information look for Non-reflecting Boundary condition by Giles.
Have a look at the paper in the link below, which might help.

http://turbomachinery.asmedigitalcol...icleID=1896640


Regards,
Saurya