[cjc96]

As an update for those looking for a "cleaner" or just more general solution, I've since implemented the previously discussed codedFunctionObject as a traditional functionObject following the same format as other fieldExpression derived objects.

I've included the .C and .H files, to be inserted and compiled in either the main or a user-defined src/functionObjects/fields directory, alongside an example of how to load the function within a simulation (by adding to controlDict or as a file within the system directory)

Don't forget to make the necessary additions to Make/files prior to compiling!

[SonnyD]

Hi guys,

I know the discussiun ended a year ago. But this is my way to to qualify my LES mesh resolution.

I usually just calculate the UPrime2Mean of course, as you explained before, to calculate my TKE via diagonal components of the Reynolds Stress Tensor. The SGS-part of the TKE is modelled anyway and printed (in e.g. kEquation and dynamicKEquation LES-Model) automatically. If it is not, one can use the tubulenceFields functionObject in controlDict.
To finally quantify my LES-mesh quality I use the Paraview calculator:
(0.5*(UPrime2Mean_XX+UPrime2Mean_YY+UPrime2Mean_ZZ ))/((0.5*(UPrime2Mean_XX+UPrime2Mean_YY+UPrime2Mean_Z Z))+k)

I guess this is a pretty easy way get your results and it might work for DES aswell.

David

[FluentStarter]

Originally Posted by cjc96

As an update for those looking for a "cleaner" or just more general solution, I've since implemented the previously discussed codedFunctionObject as a traditional functionObject following the same format as other fieldExpression derived objects.

I've included the .C and .H files, to be inserted and compiled in either the main or a user-defined src/functionObjects/fields directory, alongside an example of how to load the function within a simulation (by adding to controlDict or as a file within the system directory)

Don't forget to make the necessary additions to Make/files prior to compiling!

Sorry but which are the necessary changes in the Make/files?

[David*]

Originally Posted by SonnyD

Hi guys,

I know the discussiun ended a year ago. But this is my way to to qualify my LES mesh resolution.

I usually just calculate the UPrime2Mean of course, as you explained before, to calculate my TKE via diagonal components of the Reynolds Stress Tensor. The SGS-part of the TKE is modelled anyway and printed (in e.g. kEquation and dynamicKEquation LES-Model) automatically. If it is not, one can use the tubulenceFields functionObject in controlDict.
To finally quantify my LES-mesh quality I use the Paraview calculator:
(0.5*(UPrime2Mean_XX+UPrime2Mean_YY+UPrime2Mean_ZZ ))/((0.5*(UPrime2Mean_XX+UPrime2Mean_YY+UPrime2Mean_Z Z))+k)

I guess this is a pretty easy way get your results and it might work for DES aswell.

David

Note that in this version the denominator trace does NOT include k, while in cjc96's version it does! What's right?


/EDIT: I mixed up R and k, they are the difference!

[salehi144]

Hi,

I think Conor's code is implemented nicely. However, in my opinion, there is one small issue which is here, the mean resolved turbulent kinetic is compared energy with the instantaneous modeled energy. I think the correct comparison, would be to compare mean resolved and mean modeled kinetic energies. So, first, one needs to average both U (with primes) and R (without primes) during the simulation and then run the functionObject script.

Additionally, I rewrote the whole script to make it more aligned with the current OpenFOAM coding style. It is assumed that UPrime2Mean and RMean are already available.

readFields
{
    type            readFields;
    libs            (fieldFunctionObjects);
    fields          (UPrime2Mean turbulenceProperties:RMean);
}

LESIndex
{
    type            coded;
    libs            (utilityFunctionObjects);
    name            LESIndex;

    codeExecute
    #{
        auto* LESIndexPtr =
            mesh().getObjectPtr<volScalarField>("LESIndex");

        if (!LESIndexPtr)
        {
            Info<< "Create total turbulent kinetik energy field" << nl;
            LESIndexPtr = new volScalarField
            (
                IOobject
                (
                    "LESIndex",
                    mesh().time().timeName(),
                    mesh(),
                    IOobject::NO_READ,
                    IOobject::AUTO_WRITE
                ),
                mesh(),
                dimless
            );

            regIOobject::store(LESIndexPtr);
        }
        auto& LESIndex = *LESIndexPtr;

        Info<< "Computing total turbulent kinetik energy field\n" << endl;
        
        auto* RPtr = 
            mesh().getObjectPtr<volSymmTensorField>("turbulenceProperties:RMean");
        auto* UPrime2MeanPtr = 
            mesh().getObjectPtr<volSymmTensorField>("UPrime2Mean");

        if (RPtr && UPrime2MeanPtr)
        {
            auto& R = *RPtr;
            auto& UPrime2Mean = *UPrime2MeanPtr;

            volScalarField resolvedTKE = 0.5*tr(UPrime2Mean); 
            volScalarField totalTKE = 0.5*tr(R) + 0.5*tr(UPrime2Mean);
            LESIndex = resolvedTKE/(totalTKE +
                    dimensionedScalar("small", totalTKE.dimensions(), SMALL));
            LESIndex.write(); // Not necessary if the funcitonObject is used during the simulation
        }
        else
        {
            Warning << endl
                    << "    Unable to compute total turbulent kinetik energy" << endl
                    << "    UPrime2Mean and/or RMean Unavailable" << endl;
        }
    #};
}

Saeed

[YongCLee]

Originally Posted by Andrea1984

Hi Conor,

an alternative method could be to hardcode resolved TKE, epsilon and all the other variables you need in a modified version of the solver. I gave it a go here: pimpleTKEBudgetFoam, which also includes a LESResolutionIndex = k_res/(k_modelled + k_res)

Andrea

When you calculate UPrime, it seems that you use U from current timestep, but use Umean from last timestep. Because the function to calculate Umean is done after "tkeBudget.H".