i am studying shock interactions in supersonic intake (2D)... so, basically to set up a a supersonic flow in fluent, what are the boundary conditions for compressible flow problems ?? i tried various combinations, and my solution is not converging to more than 1.0E-2... its slowly converging and then starts oscillating and diverging after reaching 1.0E-2... I hav to fix back pressure at one particular boundary... how different is the setting up of a problem for supersonic flow different from that of a subsonic flow...?? what might be the reason for the divergence ??? I hav even tried lott of grids !! starting from very course grids to very very fine grids ! and I am not even able to ger convergence for Inviscid Flow ! pls help cheers

Hi,

Supersonic flow and subsonic flow as of cfd is concern depends upon the boundary conditions. Like if we see the characteristics of the flow in subsonic flow one will come out of the domain and two will go inside domain at the outlet two will go out and one will come in so accordingly we need to give information at the boundary as two variables have to specified for subsonic flow at inlet and one variable will be extrapolated and similarly at out let u need to specify one and the remaining two will be extrapolated. But the case of supersonic case is different all the three characteristics are in the same direction moving into the domain at inlet and out at the exit so all the three variables are to be specified at the inlet and no need to specify at the outlet. briefly this is the boundary conditions for the two types of flows but when u implement them in fluent for supersonic case take the pressure inlet at inlet and pressure outlet or far field pressure at the exit according to the boundary position. and while specifying the pressure specify the supersonic gauge pressure if u have the information that is if ur flow is supersonic at the outlet then the pressure at th out let will become ur supersonic gauge pressure.

hope this may help u,

you try lower(first order scheme)order scheme, it will work.

thanks for ur replies... i am trying first order scheme only... residuals start oscillating and diverge after 1.0E-2 ... Gambit is also giving me troubles when i try a boundary layer mesh (to run viscous flow).. i mean, i am not able to mesh the surrounding areas,as the first grid is very very close to the surface ( approx 8.0E-6 ).. and hence i am UNABLE to use any boundary layer mesh... how do i tackle this problem ?