[mshehata85]

Dear Foamers,

My question is regarding the "phasePressureCoeffs" usually specified in the turbulenceProperties.particle file in OpenFoam (for example in the fluidized bed twoPhaseEulerFoam tutorial)

I have done an extensive online research and I concluded it is a way to compute the derivative of inter-particle stress w.r.t volume fraction in an attempt to avoid exceeding the specified maximum volume fraction.

However, I do not really understand the purpose of it as I thought that the "frictionalStressModel" is the one responsible to avoid the over-packing by incorporating the frictional component after exceeding the specified threshold.

So my questions are:

1- Why and when I need to activate the phase pressure model (by setting g0 to a positive value)? is there any practical reason to activate it?

2- Does anybody know the reference to this exponential model? No reference is available in the " phasePressureModel.H" or any where else on the documentation.

3- is the default parameters ( g0, preAlphaExp, expMax, and alphaMax) valid for all two-phase cases. if not, how to estimate these parameters? Do they affect the resulting accuracy?

4- Why "alphaMax" is specified twice in the same file, once in the "kineticTheoryCoeffs" and another time in the "PhasePressureCoeffs"? I thought both of them refer to the same criteria!!!

I appreciate any help because I have been looking for an answer for more than 4 months. So, this Forum is my last resort.

Best Regards.

Mahmoud

[mshehata85]

Can anyone please help? My thought right now is that it is important only in case of compressible two-phase system. In my simulation, I deactivated the energy equation and described the density of both phases as constant (i.e. incompressible flow). So, I think I need to deactivate the phasePressureCoeffs. Am I right?

[hooman.4028]

Originally Posted by mshehata85

Can anyone please help? My thought right now is that it is important only in case of compressible two-phase system. In my simulation, I deactivated the energy equation and described the density of both phases as constant (i.e. incompressible flow). So, I think I need to deactivate the phasePressureCoeffs. Am I right?

Hello Mahmoud,

Have you had any luck answering these questions?

[mshehata85]

Originally Posted by hooman.4028

Hello Mahmoud,

Have you had any luck answering these questions?

Not yet unfortunately.

[mnikku]

I don't have a reference, but essentially it seems just an optional way to compute pPrime.

In kineticTheoryModel.C

Foam::RASModels::kineticTheoryModel::pPrime() const
{
...
    tmp<volScalarField> tpPrime
    (
        Theta_
       *granularPressureModel_->granularPressureCoeffPrime
        (
            alpha_,
            radialModel_->g0(alpha_, alphaMinFriction_, alphaMax_),
            radialModel_->g0prime(alpha_, alphaMinFriction_, alphaMax_),
            rho,
            e_
        )
     +  frictionalStressModel_->frictionalPressurePrime
        (
            phase_,
            alphaMinFriction_,
            alphaMax_
        )
    );
...
return tpPrime;
}

While in phasePressureModel.C

Foam::RASModels::phasePressureModel::pPrime() const
{
    tmp<volScalarField> tpPrime
    (
        g0_
       *min
        (
            exp(preAlphaExp_*(alpha_ - alphaMax_)),
            expMax_
        )
    );
...
return tpPrime;
}

In file phasePressureModel.h the functionality is descripted as: "Particle-particle phase-pressure RAS model. The derivative of the phase-pressure with respect to the phase-fraction is evaluated as..." .

So it seems to be a simpler way to model the solids pressure, essentially preventing overpacking.

That's what I was able to figure out (so far), experts will hopefully correct me if I am wrong. I have used this model successfully, while it is simpler, it seems to work just as nicely (or better) and faster(?) than kinetic theory of granular flows, at least in the limited amount of cases I have tested (gas-solid fluidized bed simulations).

[Roham..Seif]

Originally Posted by mnikku

I don't have a reference, but essentially it seems just an optional way to compute pPrime.

In kineticTheoryModel.C

Foam::RASModels::kineticTheoryModel::pPrime() const
{
...
    tmp<volScalarField> tpPrime
    (
        Theta_
       *granularPressureModel_->granularPressureCoeffPrime
        (
            alpha_,
            radialModel_->g0(alpha_, alphaMinFriction_, alphaMax_),
            radialModel_->g0prime(alpha_, alphaMinFriction_, alphaMax_),
            rho,
            e_
        )
     +  frictionalStressModel_->frictionalPressurePrime
        (
            phase_,
            alphaMinFriction_,
            alphaMax_
        )
    );
...
return tpPrime;
}

While in phasePressureModel.C

Foam::RASModels::phasePressureModel::pPrime() const
{
    tmp<volScalarField> tpPrime
    (
        g0_
       *min
        (
            exp(preAlphaExp_*(alpha_ - alphaMax_)),
            expMax_
        )
    );
...
return tpPrime;
}

In file phasePressureModel.h the functionality is descripted as: "Particle-particle phase-pressure RAS model. The derivative of the phase-pressure with respect to the phase-fraction is evaluated as..." .

So it seems to be a simpler way to model the solids pressure, essentially preventing overpacking.

That's what I was able to figure out (so far), experts will hopefully correct me if I am wrong. I have used this model successfully, while it is simpler, it seems to work just as nicely (or better) and faster(?) than kinetic theory of granular flows, at least in the limited amount of cases I have tested (gas-solid fluidized bed simulations).

Dear Mnikuu,

Years have passed from your comment on phase pressure model. What happened after that? can you provide some information about when it is fine to use this model?
In my case, its results are much closer to experiments rather than kinetic theory of granular flow.

Thank you very much.
Sincerely,
Roham