[khedar]

Hi all,
I get the following error after I modified my Turbulence model.

keyword div((thermo:mu*dev2(T(grad(U))))) is undefined in dictionary

This error comes only when I add RST terms to devRhoReff() and divDevRhoReff(U) terms.

tmp<volSymmTensorField> sBSLEARSMNew::devRhoReff() const
{
    return tmp<volSymmTensorField>
    (
        new volSymmTensorField
        (
            IOobject
            (
                "devRhoReff",
                runTime_.timeName(),
                mesh_,
                IOobject::NO_READ,
                IOobject::NO_WRITE
            ),
            rho_*R_+mu()*dev(twoSymm(fvc::grad(U_)))
        )
    );
}


tmp<fvVectorMatrix> sBSLEARSMNew::divDevRhoReff(volVectorField& U) const
{
    return
    (
        fvc::div(rho_*R_)
      - fvm::laplacian(muEff(), U) 
      - fvc::div(mu()*dev2(T(fvc::grad(U))))
    );
}

I have searched a lot but am not able to find the solution. My doubt is the "thermo:" before the mu term. I have never seen div scheme to contain such term. But when I add this to the div scheme i get a different error.

--> FOAM FATAL IO ERROR: 
attempt to read beyond EOF

file: /home/i6070205_/OpenFOAM/i6070205_-2.3.0/run/simple/2Round/4test/system/fvSchemes.divSchemes.div((rho*R)) at line 38.

my div schemes are:

divSchemes
{
    default            none;
    div(phi,T)         bounded Gauss upwind;
    div(U)          Gauss linear;
    div(phi,U)        bounded Gauss upwind;
    div(phi,k)         bounded Gauss upwind;
    div(phi,K)         bounded Gauss upwind;
    div(phi,omega)      bounded Gauss upwind;
    div((rho*R))        Gauss upwind;
    div(R)          Gauss upwind;
    div(phi,R)          Gauss upwind;
    div(phi,h)          bounded Gauss upwind;
    div((muEff*dev2(grad(U).T()))) Gauss linear;
    div(phi,Ekp)        bounded Gauss upwind;
    div(phi,e)          bounded Gauss upwind;
    div((mu*dev2(T(grad(U))))) Gauss upwind;
    div((nu*dev2(T(grad(U))))) Gauss upwind;
    div((muEff*dev2(T(grad(U)))))   Gauss linear;
    div((thermo:mu*dev2(T(grad(U))))) Gauss linear;

}

I am using OF2.3.1 and this turbulence model is based on kOmegaSSL.

It would be great if someone could help me out.

khedar

[Tobi]

Hi,

I am not sure about your implementation and what you did but try,...

div((rho*R))    Gauss upwind grad(rho);

If it is like that, I am not sure if you did everything correct in the mathematical point of view. The thermo: keyword is fine because you used it in your function.

[khedar]

Hi Tobi,
thanks for your reply. When i changed as you suggested i got a new error.

--> FOAM FATAL ERROR: 

    request for surfaceScalarField grad(rho) from objectRegistry region0 failed
    available objects of type surfaceScalarField are

2
(
ghf
phi
)

But when i tried to change to "Gauss linear", it worked. So probably it has something to do with type of scheme. But I don't know OF that much so can't understand why it would be so.

khedar

[Tobi]

Hi,

Then use grad(phi).
However, this is not a good stuff. Just one question. What is R_ in your code?

Sent from my HTC One mini using CFD Online Forum mobile app

[khedar]

Hi, I tried using grad(Phi) also but then I get some other error. But with Gauss linear it works so I am sticking to that for the Moment..
R is the Reynolds Stress Tensor

[khedar]

I have one more doubt regarding a term in the kOmegaSST model as implemented in OF.

// Turbulent frequency equation
    tmp<fvScalarMatrix> omegaEqn
    (
        fvm::ddt(rho_, omega_)
      + fvm::div(phi_, omega_)
      - fvm::laplacian(DomegaEff(F1), omega_)
     ==
        rhoGammaF1
       *min
        (
            GbyMu,
            (c1_/a1_)*betaStar_*omega_*max(a1_*omega_, b1_*F23()*sqrt(S2))
        )
      - fvm::SuSp((2.0/3.0)*rhoGammaF1*divU, omega_)
      - fvm::Sp(rho_*beta(F1)*omega_, omega_)
      - fvm::SuSp
        (
            rho_*(F1 - scalar(1))*CDkOmega/omega_,
            omega_
        )
    );

When I compare it to the original model, I don't understand from where the following term comes from.

      - fvm::SuSp((2.0/3.0)*rhoGammaF1*divU, omega_)

[khedar]

I think I understand where the source term with 2/3 in the previous post comes from. It is from the production term.



When the production term is evaluated we get three terms and the last term with k turns out to be the source term.

But when we do this and seeing by the way Gbymu is calculated,

    volScalarField GbyMu((tgradU() && dev(twoSymm(tgradU()))));

I think Production is not correctly estimated, since it does not contain all the terms (Only the first two terms in the stress tensor). And applying the production limiter to this term seems incorrect also.

[Tobi]

The term 2/3 rho k is not there because it goes to the pressure.

Sent from my HTC One mini using CFD Online Forum mobile app

[khedar]

But then what is this following term in the omega and k equation.?

Originally Posted by khedar

When I compare it to the original model, I don't understand from where the following term comes from.

      - fvm::SuSp((2.0/3.0)*rhoGammaF1*divU, omega_)

[Tobi]

Oh, I was talking about the buossinesq eddy viscosity approximation and the momentum equations, not about the turbulence models itself.

Sent from my HTC One mini using CFD Online Forum mobile app

[khedar]

The term which i referred to in my previous post comes indeed from the Production term.

See here:
http://caefn.com/openfoam/keqn-sgs-model

In this blog post(although explained for k-sgs model but k equation treatment is more or less same.), the author derives how the 2/3 term comes from production. The reason for doing this way is answered on the same post. I quote here.

My question:

Why the source code contains two terms instead of one for the production. In other words why do we need to separate it into two? (Refer your equation 6.)

Authors answer:

Hi Khedar,

Thanks for your question.
I think the reason why the production term is divided into two part is that the SGS kinetic energy ksgs is only in the 1st term of the R.H.S. of Eq. (6). This term can take both positive and negative values so the SuSp operator is used in order to create diagonally dominant matrix.