[mruda]

Hey everyone,

I'm currently using SU2 to investigate a serpentine duct subject to inlet flow of Mach 0.5. More specifically, I'm interested in the pressure loss at the outlet portion.

Herein lies my question. Ideally, I'd like to apply a mass flow boundary at the outlet, like what is in Fluent and some other codes. However, the only subsonic outlet I see requires setting a specific back pressure, which defeats the purpose of my study.

The supersonic outlet boundary setting extrapolates from within the domain, but would effectively ignore the incoming characteristic, which makes me uneasy. Is this an acceptable solution to my problem? I am only interested in the internal flow, and my domain terminates at the outlet. My gut says simple extrapolation would be fine, but I'd love to get some input from those with more knowledge.

Thanks in advance!

[RcktMan77]

You don't want an extrapolation BC at the exit. The exit boundary is the one driving the flow in this case driven by the difference in pressure between free stream and and the control volume exit. The mass flow boundary condition you mention, while convenient, is basically just changing the back pressure at the exit until it converges at the target mass flow value. You can do this iteratively yourself using the subsonic back pressure boundary condition available.

The extrapolation boundary condition is typically used in situations where you have a situation where a boundary is present, but you don't want to have it exert any influence on the surrounding flowfield. For example if you're simulating the external flow over a wing-body-nacelle configuration, and you're not modeling flow through the nacelle, then you might specify an extrapolation boundary condition on a plane normal to the flow direction that is located just slightly downstream of the nacelle's inlet lip and inside of the placeholder duct instead of placing a wall boundary there.