[Bokse]

I'm trying to understand which version of Menter's k-omega SST turbulence model OpenFOAM 2.2.2 is using.

Looking at the incompressible k-omega SST model, the production limiter is only added to the k equation.

// Turbulent kinetic energy equation
    tmp<fvScalarMatrix> kEqn
    (
        fvm::ddt(k_)
      + fvm::div(phi_, k_)
      - fvm::laplacian(DkEff(F1), k_)
     ==
        min(G, c1_*betaStar_*k_*omega_) // Production limiter
      - fvm::Sp(betaStar_*omega_, k_)
    );

This corresponds with Menter's suggestion from 1993 (http://turbmodels.larc.nasa.gov/sst.html). Menter suggests setting c1 to 20 but c1 is 10. This corresponds to menters suggestion from 2003. Here it's suggested to use 10 and use the production limiter for both k and omega. Looking at omega the limiter has not been used:

 
// Turbulent frequency equation
    tmp<fvScalarMatrix> omegaEqn
    (
        fvm::ddt(omega_)
      + fvm::div(phi_, omega_)
      - fvm::laplacian(DomegaEff(F1), omega_)
     ==
        gamma(F1)*S2                               // No limiter
      - fvm::Sp(beta(F1)*omega_, omega_)
      - fvm::SuSp
        (
            (F1 - scalar(1))*CDkOmega/omega_,
            omega_
        )
    );

So my question is why the production limiter is only used for k?

I found a bug report on the same topic, but I don't think it provides a answer to the question. http://openfoam.org/mantisbt/view.php?id=286

Thanks in advance.

[msuaeronautics]

...........

[msuaeronautics]

The NASA resource states that Menter required production limiters on both equations, but his actual 2003 paper only seems to call for the limiter on k. So, I suppose that the code only does what's been called for.

[mechy]

I have compared Fluent's and OFs' kw-sst model. and found always the OF value of w is higher than Fluent' s w.(between 10 to 100 times).
but other fields (k U p) is very close to Fluent ones.
I think this thread is the reason of higher w in OF

[msuaeronautics]

Originally Posted by mechy

I have compared Fluent's and OFs' kw-sst model. and found always the OF value of w is higher than Fluent' s w.(between 10 to 100 times).
but other fields (k U p) is very close to Fluent ones.
I think this thread is the reason of higher w in OF

Can you provide some data showing this?

[mechy]

OK
I will provide some examples,as soon as possible

[mjb28]

I'm also interested in why the production limiter is not used in the omega equation, or indeed if it should be.

Does anyone know where the correction referred to on the NASA website comes from? Could it be the following reference (I don't have access to this so cannot check, does anybody else?), or is it just knowledge that Chris Rumsey has? (who is curator of the NASA page).

Menter, Florian R. "Review of the shear-stress transport turbulence model experience from an industrial perspective." International Journal of Computational Fluid Dynamics 23.4 (2009): 305-316.

There is another bug report about the same issue and it seems the developer (Sergio) decided there is not a published paper to base the correction on:

http://openfoam.org/mantisbt/view.php?id=486

[msuaeronautics]

I have personally tried using a custom limiter on omega, which appears to keep omega in check, but introduces some non-physical elements (like hotspots in TKE). Will have to do further testing.

[KateEisenhower]

Hello all,

the omega-equation has changed with 2.3.x:

    // Turbulent frequency equation
    tmp<fvScalarMatrix> omegaEqn
    (
        fvm::ddt(omega_)
      + fvm::div(phi_, omega_)
      - fvm::laplacian(DomegaEff(F1), omega_)
     ==
        gamma(F1)
       *min(S2, (c1_/a1_)*betaStar_*omega_*max(a1_*omega_, b1_*F23()*sqrt(S2)))
      - fvm::Sp(beta(F1)*omega_, omega_)
      - fvm::SuSp
        (
            (F1 - scalar(1))*CDkOmega/omega_,
            omega_
        )
    );

But this implementation doesn't correspond with Menter's 2003 paper. Is there an explanation for this? Or is there a more recent publication?

Best regards,

Kate

Originally Posted by Bokse

I'm trying to understand which version of Menter's k-omega SST turbulence model OpenFOAM 2.2.2 is using.

Looking at the incompressible k-omega SST model, the production limiter is only added to the k equation.

// Turbulent kinetic energy equation
    tmp<fvScalarMatrix> kEqn
    (
        fvm::ddt(k_)
      + fvm::div(phi_, k_)
      - fvm::laplacian(DkEff(F1), k_)
     ==
        min(G, c1_*betaStar_*k_*omega_) // Production limiter
      - fvm::Sp(betaStar_*omega_, k_)
    );

This corresponds with Menter's suggestion from 1993 (http://turbmodels.larc.nasa.gov/sst.html). Menter suggests setting c1 to 20 but c1 is 10. This corresponds to menters suggestion from 2003. Here it's suggested to use 10 and use the production limiter for both k and omega. Looking at omega the limiter has not been used:

 
// Turbulent frequency equation
    tmp<fvScalarMatrix> omegaEqn
    (
        fvm::ddt(omega_)
      + fvm::div(phi_, omega_)
      - fvm::laplacian(DomegaEff(F1), omega_)
     ==
        gamma(F1)*S2                               // No limiter
      - fvm::Sp(beta(F1)*omega_, omega_)
      - fvm::SuSp
        (
            (F1 - scalar(1))*CDkOmega/omega_,
            omega_
        )
    );

So my question is why the production limiter is only used for k?

I found a bug report on the same topic, but I don't think it provides a answer to the question. http://openfoam.org/mantisbt/view.php?id=286

Thanks in advance.