[vut]

Hi everybody,

I have a question which is tricky for months. I would be really appreciate your shares and your answers.

I want to simulate an internal flow through an injector. I use pisoFoam, LES and I tried oneEqEddy and mixed Smagorinsky sub-grid models.

The trouble I have is the flow is not at all turbulent which is the situation I waited for. Here you have my input files for mixed Smagorinsky:
=======================
In folder 0
=======================
=== p
dimensions [0 2 -2 0 0 0 0];

internalField uniform 0;

boundaryField
{
Inlet
{
type zeroGradient;
}

Outlet
{
type fixedValue;
value uniform 0;
}

Wall
{
type zeroGradient;
}
}



=== U

dimensions [0 1 -1 0 0 0 0];

internalField uniform (0 0 0);

boundaryField
{
Inlet
{
type turbulentInlet;
referenceField uniform (0 0 7.1);
fluctuationScale (0.02 0.02 0.02);
value uniform (0 0 7.1);
alpha 0.5;
}

Outlet
{
type zeroGradient;
}

Wall
{
type fixedValue;
value uniform (0 0 0);
}
}



=======================
In folder constant
=======================

=== LESProperties

LESModel mixedSmagorinsky;

delta cubeRootVol;

printCoeffs on;

mixedSmagorinskyCoeffs
{
filter simple;
ce 1.05;
}

cubeRootVolCoeffs
{
deltaCoeff 1;
}

PrandtlCoeffs
{
delta cubeRootVol;
cubeRootVolCoeffs
{
deltaCoeff 1;
}

smoothCoeffs
{
delta cubeRootVol;
cubeRootVolCoeffs
{
deltaCoeff 1;
}

maxDeltaRatio 1.1;
}

Cdelta 0.158;
}

vanDriestCoeffs
{
delta cubeRootVol;
cubeRootVolCoeffs
{
deltaCoeff 1;
}

smoothCoeffs
{
delta cubeRootVol;
cubeRootVolCoeffs
{
deltaCoeff 1;
}

maxDeltaRatio 1.1;
}

Aplus 26;
Cdelta 0.158;
}

smoothCoeffs
{
delta cubeRootVol;
cubeRootVolCoeffs
{
deltaCoeff 1;
}

maxDeltaRatio 1.1;
}



=== turbulenceProperties

simulationType LESModel;



=======================
In folder system
=======================

=== fvScheme

ddtSchemes
{
default backward;
}

gradSchemes
{
default Gauss linear;
}

divSchemes
{
default none;
div(phi,U) Gauss linear;
div((nuEff*dev(T(grad(U))))) Gauss linear;

div((dev((((surfaceSum((magSf*interpolate(sqr(U))) )|surfaceSum(magSf))-sqr((surfaceSum((magSf*interpolate(U)))|surfaceSum (magSf))))+(((0.666667)*((0.5*((surfaceSum((magSf* interpolate(magSqr(U))))|surfaceSum(magSf))-magSqr((surfaceSum((magSf*interpolate(U)))|surface Sum(magSf)))))+((((2*ck)|ce)*sqr(delta))*magSqr(de v(symm(grad(U)))))))-(nuSgs*twoSymm(grad(U))))))+(-nuEff*dev(twoSymm(grad(U)))))) Gauss linear;

// div((nu*dev(T(grad(U))))) Gauss linear;
div((grad(U)&U)) Gauss linear;
div(phi,k) Gauss limitedLinear 1;
div(phi,B) Gauss limitedLinear 1;
div(B) Gauss linear;
}

laplacianSchemes
{
default Gauss linear corrected;
}

interpolationSchemes
{
default linear;
}

snGradSchemes
{
default corrected;
}

fluxRequired
{
default no;
p ;
}



=== fvSolution

solvers
{
p
{
solver GAMG;
tolerance 1e-06;
relTol 0.1;
smoother GaussSeidel;
nPreSweeps 0;
nPostSweeps 2;
cacheAgglomeration on;
agglomerator faceAreaPair;
nCellsInCoarsestLevel 10;
mergeLevels 1;
}

pFinal
{
$p;
smoother DICGaussSeidel;
tolerance 1e-06;
relTol 0;
}

"(U|k|B|nuTilda)"
{
solver smoothSolver;
smoother GaussSeidel;
tolerance 1e-05;
relTol 0;
}
}

PISO
{
nCorrectors 2;
nNonOrthogonalCorrectors 3;
}



Once again, all your ideas are welcome and thank you in advance!

vut

[adambarfi]

did you find your answer? I have a same question
If you found it can you please let me know about it

Regards,
Mostafa

[vut]

Dear Mostafa,

I do not have the answer for now. I hope somebody knows how to solve it.

I myself try to change some parameters to obtain a more turbulent flow inside the injector.

Without doubt, I give you further information if the problem is fixed.

Best regards,

vut

[adambarfi]

hi,

I've found the solution.
at first I defined a variable deltaT in a way that the Courant number doesn't exceed Co=0.6.
Then, I let it solve for a long time to reach a fully-developed turbulent flow (in my case, I should reach a fully-developed condition). actually the c_k and c_e which lead to C_S are very important, you should specify them in a way that the turbulent dissipation decreases and eddies formed in flow don't decay!

hope it helps you

Regards,
Mostafa

[vut]

Thank you, Mostafa for replying me.

Can you give more details of the way you modify C_s (or C_k and C_e)? Did you mean we can do so by reducing Ce in LESProperties file?



mixedSmagorinskyCoeffs
{
filter simple;
ce 1.05;
}



May you attach your LESProperties, fvSolution and fvSystem files?

I am not yet so confident with OpenFOAM.

Thanks in advance,

vut

[Alish1984]

Originally Posted by adambarfi

hi,

I've found the solution.
at first I defined a variable deltaT in a way that the Courant number doesn't exceed Co=0.6.
Then, I let it solve for a long time to reach a fully-developed turbulent flow (in my case, I should reach a fully-developed condition). actually the c_k and c_e which lead to C_S are very important, you should specify them in a way that the turbulent dissipation decreases and eddies formed in flow don't decay!

hope it helps you

Regards,
Mostafa

Could you pls explain more
1. What is ur case? A channel or pipe?
2. An estimate dimension. And mesh size
3. What is ur boundary? Periodic? Or inlet, outlet and wall
4. And the most important, what do u mean by "long time"