[FoamingSimon]

Hello,

I am having troubles with extracting the Reynolds stress from a simulation. Wether I use

"simpleFoam -postProcess -func R"

or (in the controlDict)

 functions
{
    turbulenceFields1
    {
        type        turbulenceFields;
        libs        ("libfieldFunctionObjects.so");
        field		R;
        writeControl        onEnd;
    }
}

The result entire simulation executes well, but the extracted Reynolds stresses is the same:

No MRF models present

No finite volume options present
    functionObjects::turbulenceFields R writing field: turbulenceProperties:R

Which seems fine, except there is not output. Nothing to be found neither in the folder nor when viewing results via Paraview.

Running more postproccessing together with the Reynolds stress via the controlDict, for instance wallshear stress, works fine and produces:

    functionObjects::turbulenceFields turbulenceFields1 writing field: turbulenceProperties:R
wallShearStress wallShearStress1 write:
    writing field wallShearStress
    min/max(ground) = (-0.119511319929 -1.42561710357e-18 -2.02731918065e-05), (-0.115200085848 2.42927606607e-18 1.24279923199e-05)

But again, no Reynolds stresses. Which I need.

I am using the "Windows Binary v1812" version of OpenFOAM.

Does anyone have suggestions on what to do, to get the Reynolds stress?
Or a way of fixing my problem?

[FoamingSimon]

Still no solutions in hand, but more information gathered:

When running either

postProcess -func R

or

postProcess -funcs '(turbulenceFields)'

The result is:

Executing functionObjects


--> FOAM FATAL ERROR:
Turbulence model not found in database, deactivating

    From function bool Foam::functionObjects::turbulenceFields::compressible()
    in file turbulenceFields/turbulenceFields.C at line 107.

FOAM exiting

[blackpingu]

Try this maybe :

 YourSolverHere -postProcess -func R

[FoamingSimon]

Hello Blackpingu

Unfortunately, it does not make any difference. An excerpt of the result from running above code is:

Time = 450
Reading field p

Reading field U

Reading/calculating face flux field phi

Selecting incompressible transport model Newtonian
Selecting turbulence model type RAS
Selecting RAS turbulence model kEpsilon
kEpsilonCoeffs
{
    Cmu             0.03;
    C1              1.21;
    C2              1.92;
    sigmaEps        1.3;
    C3              0;
    sigmak          1;
}

No MRF models present

No finite volume options present
    functionObjects::turbulenceFields R writing field: turbulenceProperties:R

Also, still no "R" in the postprocess folder

[blackpingu]

Originally Posted by FoamingSimon

Hello Blackpingu

Unfortunately, it does not make any difference. An excerpt of the result from running above code is:

Time = 450
Reading field p

Reading field U

Reading/calculating face flux field phi

Selecting incompressible transport model Newtonian
Selecting turbulence model type RAS
Selecting RAS turbulence model kEpsilon
kEpsilonCoeffs
{
    Cmu             0.03;
    C1              1.21;
    C2              1.92;
    sigmaEps        1.3;
    C3              0;
    sigmak          1;
}

No MRF models present

No finite volume options present
    functionObjects::turbulenceFields R writing field: turbulenceProperties:R

Also, still no "R" in the postprocess folder

Check your time folders for a file named "turbulenceProperties:R", in folder "450" for example.

[FoamingSimon]

No such file nor anything similar.
If I postprocess for yplus for example, I do get a "yPlus"-file with the results.

[blackpingu]

That is strange, what case are you running?

[FoamingSimon]

Indeed.

It does not matter what I run, it seems.
But currently, I run a modified "turbine siting tutorial". Basically a square domain, with ABL inflow.

I have attached a 2D-ABL flow study I have made, where I cannot extract the Reynold's stresses as well.

[blackpingu]

Try to change to this in the controldict file instead of "onEnd"

turbulenceFields1
    {
        type        turbulenceFields;
        libs        ("libfieldFunctionObjects.so");
        field		R;
        writeControl        writeTime;
    }

or

turbulenceFields1
    {
        type        turbulenceFields;
        libs        ("libfieldFunctionObjects.so");
        field		R;
        writeControl        timeStep;
    }

[FoamingSimon]

Thank you for your perseverance, Blackpingu, but no - changing the writeControl to either writeTime or timeStep did not result in any Reynolds stress outputs. Neither in the postProcess folder, nor in any of the arbitrary timestep folders.

Although, the writeControl "timeStep" did result in a lot of empty folders - which makes sense, since an output is written for every time increment.

Do you have any other ideas?

[blackpingu]

Originally Posted by FoamingSimon

Thank you for your perseverance, Blackpingu, but no - changing the writeControl to either writeTime or timeStep did not result in any Reynolds stress outputs. Neither in the postProcess folder, nor in any of the arbitrary timestep folders.

Although, the writeControl "timeStep" did result in a lot of empty folders - which makes sense, since an output is written for every time increment.

Do you have any other ideas?

Not that i can think of, but I wasn't able to run your case properly, didn't get files for U or P or anything, do you get decent results for the velocity? Which OF version are you running?

[FoamingSimon]

That sounds weird, as I get reasonable results, as can be seen on the attached .png-file, showing the velocity profiles through the domain.

- Did you run it via the Allrun-file?
Otherwise, you would have to copy the 0.orig and rename it 0 before running it.

I am using the "Windows Binary v1812" version of OpenFOAM.

[blackpingu]

Originally Posted by FoamingSimon

That sounds weird, as I get reasonable results, as can be seen on the attached .png-file, showing the velocity profiles through the domain.

- Did you run it via the Allrun-file?
Otherwise, you would have to copy the 0.orig and rename it 0 before running it.

I am using the "Windows Binary v1812" version of OpenFOAM.

Yep I did but no luck

[FoamingSimon]

Originally Posted by blackpingu

Yep I did but no luck

Maybe this is due to some differences in versions of OF.
What version do you use?
And is it normally possible for you to extract the Reynolds stresses?

[FoamingSimon]

New error type..
Using:

postProcess -func 'components(turbulenceProperties:R)'

The result is:

Time = 0

Reading fields:

Executing functionObjects
--> FOAM Warning :     functionObjects::components components(turbulenceProperties:R) failed to execute.

For each time..

[blackpingu]

I use OpenFOAM 5, and yes works everytime.

[FoamingSimon]

Switched OF version from the windows binary docker version to ubuntu via windows, and everything works like a charm now.