[mkhm]

I have a diffuser where the flow is supersonic in the converging part and subsonic in the divergent part. I am using my own 1D solver built based on rhoCentralFoam, which has been verified and tested for supersonic nozzles, however this is time I'll use it for a diffuser (geometry the same, different inlet and outlet boundary conditions are required). For the supersonic inlet, I specify three boundary conditions: fixedValue for pressure (p), temperature (T), and velocity (U). However, for the subsonic outlet, I have encountered issues as the solver does not converge towards the correct solution. According to theory, at the subsonic outlet, only one boundary condition needs to be fixed. Therefore, I have tried various combinations for the subsonic outlet, such as:

1) did not work
U_outlet fixedValue
p_outlet zeroGradient;
T_outlet zeroGradient;

T_inlet fixedValue
p_inlet fixedValue
U_inlet fixedValue

2) wrong solution
p_outlet fixedValue;
T_outlet zeroGradient;
U_outlet zeroGradient;

T_inlet fixedValue
p_inlet fixedValue
U_inlet fixedValue

8) wrong solution
p_outlet & U_outlet zeroGradient;
T_outlet fixedValue;

T_inlet fixedValue
p_inlet fixedValue
U_inlet fixedValue

37) U < 0
T_outlet zeroGradient
p_outlet fixedValue
U_outlet pressureInletOutletVelocity

T_inlet fixedValue
p_inlet fixedValue
U_inlet fixedValue

38) U < 0
T_outlet zeroGradient
p_outlet freestream
U_outlet zeroGradient

T_inlet fixedValue
p_inlet fixedValue
U_inlet fixedValue

39) crashed very early
T_outlet freestream
p_outlet freestream
U_outlet zeroGradient

T_inlet fixedValue
p_inlet fixedValue
U_inlet fixedValue

40) crashed very early
T_outlet freestream
p_outlet freestream
U_outlet freestream

T_inlet fixedValue
p_inlet fixedValue
U_inlet fixedValue

41) crashed very early
T_outlet zeroGradient
p_outlet freestream
U_outlet freestream

T_inlet fixedValue
p_inlet fixedValue
U_inlet fixedValue

41) converges towards supersonic and U <0
T_outlet inletOutlet
p_outlet zeroGradient
U_outlet inletOutlet

T_inlet fixedValue
p_inlet fixedValue
U_inlet fixedValue
etc.

Could someone please tell me what the appropriate subsonic outlet boundary condition to impose is?


P.S.: some of the combinations lead to negative values for velocity. I tried to use inletOutlet and freestream BCs, but it did not help.

[mkhm]

I think the problem for OpenFOAM is how to go from a low pressure to high pressure. While I am expecting it to follow the isentropic relationship, the fact of having higher pressure at output than the one at input seems problematic for OpenFOAM. That is the reason why velocity becomes negative (flow goes from high pressure to low pressure).



Has anyone encountered this problem before?