Learning Forum

[manjuk]

Hey everyone,
      As the title says, I'm currently trying to simulate a steady, 2D flow through a channel in FLUENT. The flow at the inlet is supersonic, I have the solver set to density-based, and the flow is inviscid. Also, the air is an ideal-gas.
      I'm working on setting my boundary conditions at the inlet and outlet (the other edges are just walls). At the inlet, I'm struggling to find a boundary condition that would allow me to define the Mach number (since a velocity-inlet won't work with compressible flows). Also, I'm not sure what to do for the outlet. I've tried a mass-flow-outlet, a pressure-outlet, and a pressure-far-field, but none of them have worked for me. The residuals don't converge, and typically when I view a contour of the static pressure, I see extremely low pressure everywhere except for a very small area near the outlet, where the pressure is extremely high.
      Are there any specific boundary conditions I should use? Or could the problem be elsewhere in my settings? Any help would be greatly appreciated.

[DrAmine]

You can use pressure inlet and provide the supersonic static pressure + total pressure based on the isentropic relationships.

[manjuk]

Thank you Amine. Would the pressure inlet allow me to enter the Mach number as, say, 2?

 

Also, what would you suggest for the outlet?

 

[Rob]

Yes, review the boundary settings. However, the solution will be fairly stiff so you may struggle to converge it. Think through how the flow should behave in the model: why is the inlet supersonic? 

[manjuk]

Thank you rwoolhou.

The flow at the inlet is supersonic just because it's part of the problem that I'm trying to solve. I designed the geometry for a Mach number of 2 at the inlet, so I need to be able to define the Mach number as 2 for the inlet boundary condition. Is there an inlet I could use that would let me do that?

[Rob]

You've got the mass flow boundary, and that also gives an upstream pressure. The biggest problem you'll have is stopping the flow bouncing shockwaves through the domain and being generally difficult to converge. Also making sure you get the downstream conditions correct will be difficult. 

[manjuk]

Okay, thank you!

[DrAmine]

Compressible flow and supersonic I will always favorize pressure I let as it contains the whole information. But here you need to know total and static pressure. Mass flow inlet is a bit weaker as it let the stagnation pressure floats but is ok.

[manjuk]

Thank you Amine!