CFD Online Discussion Forums - Is MULES the right choice for modified transport equation?

Hi everybody,
I want to solve the following modified transport equation for the calculation of sensitivities:

$\frac{\partial}{\partial t} (\delta \alpha) + \nabla \cdot (\vec{u} \ \delta \alpha) + \nabla \cdot (\delta \vec{u} \ \alpha)$

Within interFoam, I implemented the the following code right after solving alphaEqn.H (along with initial fields dAlpha and dU, calculation of dPhi and other modifications):

Code:

{

    word dAlphaScheme("div(phi,dAlpha)");

    word dPhiScheme("div(dPhi,alpha)");

    

    #include "alphaSuSp.H"

    

    surfaceScalarField dAlphaPhi

    (

    fvc::flux(phi, dAlpha, dAlphaScheme)

    + fvc::flux(dPhi, alpha1, dPhiScheme)

    );

    

    MULES::explicitSolve

    (

    geometricOneField(),

    dAlpha,

    dAlphaPhi,

    Sp,

    Su

    );

}

It works, but I'm not sure if MULES::explicitSolve is the right way to solve this equation. After a few time steps the solution smears heavily. What would you recommend?