[kaaja]

I struggle understanding the meaning of this boundary condition.
I have two fluids.
I use the following BCs for the velocity

internalField   uniform ($mUmean  0 0); 
 
boundaryField 
{ 
    outlet 
    { 
        type            outletPhaseMeanVelocity; 
        alpha           alpha.water; 
        Umean           $Umean; 
        value           $internalField; 
 
    } 
    inlet 
    { 
        type                    fixedValue; 
        value                   $internalField; 
    }

Am I right when I am saying that the alpha.water at the outlet is set "Umean" and that the velocity of the other phase adjusts so that there is mass conservation of both phases?

In case this is important information, I use the following BCs for the alpha:

internalField   uniform 0;

    outlet
    {
        type            variableHeightFlowRate;
        lowerBound      0;
        upperBound      1;
        value           $internalField;
    }
    inlet
    {
        type            fixedValue;
        value           $internalField;
    }

[clapointe]

Searching the code for the corresponding file (https://github.com/OpenFOAM/OpenFOAM...VectorField.H)) is a good place to check. Quoting the provided information :

"This boundary condition adjusts the velocity for the given phase to achieve
the specified mean thus causing the phase-fraction to adjust according to
the mass flow rate.".

Caelan

[kaaja]

I have seen that definition, but I am not fully sure what it means.

As far as I understand, the velocity for alpha.water becomes the value one presicribes, "Umean", correct?

Then it says "... thus causing the phase-fraction to adjust according to
the mass flow rate."
Does this mean that the value of alpha.water at the outlet changes so that the mass flow rate over the whole outlet equals the mass flow rate over the whole inlet?

[clapointe]

There is nothing about the other patches -- if the velocity is changed on the patch, the phase fraction is also adjusted on the patch to conserve mass :

    // Set the refValue and valueFraction to adjust the boundary field
    // such that the phase mean is Umean_
    if (Uzgmean >= Umean_)
    {
        refValue() = Zero;
        valueFraction() = 1.0 - Umean_/Uzgmean;
    }
    else
    {
        refValue() = (Umean_ + Uzgmean)*patch().nf();
        valueFraction() = 1.0 - Uzgmean/Umean_;
    }

[kaaja]

Am I right when I say

1)that the BC sets U = Umean everywhere on the outlet, and that alpha.water is adjusted so that mass conservation is obtained?

or

2) A velocity assuming zeroGradient is calculated. This velocity is adjusted by a factor so that the mean of this velocity equals Umean. Finally the alpha.water is adjusted so that mass conservation is maintained.

edit: Added guess 2.

[clapointe]

Scanning the code, it looks like the latter. I would encourage you to run a test using this bc and visualize the velocity on the patch (eg in paraview). If it's non-uniform, we'll know it's the latter. You could also use the patchAverage function object for further reassurance.

Caelan

[kaaja]

I will try your suggestion.
Thanks for helping, Caelan!