[masimasiqi]

I am running twoPhaseEulerFoam in the case of bubbleColumn and I changed the inlet to the top and outlet to the bottom which can be approximated to a standing pipe with air injected from top

- in blockMesh I exchanged the position of the patch inlet and patch outlet from the sample case, so the air bubbles are injected from the top and the mixture fluid flows out from bottom.

code from file blockMeshDict

patches
(
    patch inlet
    (
        (3 7 6 2)
    )
    patch outlet
    (
        (1 5 4 0)
    )

- in setFields I set the whole fluid domain with water(alpha.air=0)

code from file alpha.air

dimensions      [0 0 0 0 0 0 0];

internalField   uniform 0;

- because the inlet is changed to the top so the velocity of injected air should change direction

code from U.air

dimensions      [0 1 -1 0 0 0 0];

internalField   uniform (0 -0.1 0);

- in thermophysicalProperties, I change the equationOfState from perfectGas/perfectFluid to rhoConst in thermophysicalProperties to obtain incompressible flow.

code from thermophysicalProperties.air

thermoType
{
    type             heRhoThermo;
    mixture         pureMixture;
    transport       const;
    thermo          hConst;
    equationOfState rhoConst;
    specie           specie;
    energy          sensibleInternalEnergy;
}
mixture
{
    specie
    {
        molWeight   28.9;
    }
    equationOfState
    {
        R           3000;
        rho         1164;
    }
    thermodynamics
    {
        Cp          1007;
        Hf          0;
    }
    transport
    {
        mu          1.84e-05;
        Pr          0.7;
    }
}

the result seems to be right, air injects into water

alpha.air


p_rgh


but if i deal with the original thermophysicalProperties (equationOfState : perfectGas) which means compressible flow. The outcome is quite unrealistic... The upper part is filled with complete air and pushing down the fluid, there is no "injection" into the water

alpha.air

p_rgh


I think the unrealistic behavior in compressible flow (perfectGas/perfectFluid) is because of the density change affects the pressure.

the solver twoPhaseEulerFoam use p_rgh rather than p to solve the result.





(btw all the other boundary conditions remain exactly the same as original case)

code of file p_rgh

dimensions          [1 -1 -2 0 0 0 0];

internalField       uniform 1e5;

boundaryField
{
    inlet
    {
        type            fixedFluxPressure;
        value           $internalField;
    }
    outlet
    {
        type            prghPressure;
        p               uniform 110000;
        value           $internalField;
    }
    walls
    {
        type            fixedFluxPressure;
        value           $internalField;
    }
}

code of file p

dimensions          [1 -1 -2 0 0 0 0];

internalField       uniform 1e5;

boundaryField
{
    inlet
    {
        type               calculated;
        value              $internalField;
    }
    outlet
    {
        type               calculated;
        value              $internalField;
    }
    walls
    {
        type               calculated;
        value              $internalField;
    }
}

I also have tried to change several boundary conditions of p_rgh, p, U.air and U.water, nothing works out.



Does anybody has any idea or relative thought about how could I deal with the compressible case?

- about boundary condition especially of p_rgh?
- about part of the solver code about pressure to dig into?

thanks a lot, even some hints will be really helpful.