Hello all; Is there any special numerical treatement of farfield boundary condition in the case of incompressible flow around airfoil?

Dear Bounecer,

If you are using a freestream BC at the farfiled, and the farfield is really far, then no special treatment is required. If you are however looking for a near farfield then you need to use NRBCs (Non Reflecting Boundary Conditions) which can either be charcateristic based or you can simply make use of the same flux scheme in the interior to calculate the flux at the farfield, with the exception that one of the states will be freestream. This absorbing BC has been widely used from the 1990s esp. in turbomachinery computations.

Hope this helps

Regards,

Ganesh

Farfield is situated at 15xcord, is that really far? at farfield i've u=1 and v=0, how can i implement farfield BC?

Dear Bounecer,

A farfield of 15chords may not still be good if you use freestream conditions in the farfield. To use a simple absorbing typr BC, as I mentioned in the last post, use the same flux formula at the farfield as you have used in the interior. The flux formula would require two states: at the farfield, one comes from the interior cell, the other is the freestream values(ie u=1,v=0 etc...). You can also try imposing the freestream BC and check out its effects, esp. in the convergence; all you need to do is to compute the flux with the freestream values. You can then see for yourself the disticntion between a RBC and a NRBC.

Hope this helps

Regards,

Ganesh

I recommend using an asymptotic treatment based on a point vortex model. See

Far-field boundary conditions for transonic lifting solutions to the Euler equations

THOMAS, J. L.;SALAS, M. D. (NASA, Langley Research Center, Hampton, VA) AIAA Journal 1986 0001-1452 vol.24 no.7 (1074-1080)