View Issue Details

IDProjectCategoryView StatusLast Update
0000835OpenFOAMBugpublic2018-08-05 18:57
Reporteruser554Assigned Tohenry  
PriorityhighSeveritymajorReproducibilityalways
Status closedResolutionno change required 
Summary0000835: mistake equation in yPlusRAS command
Descriptionin the code of yPlusRAS command the equation that is used is for y-star not y-plus and this causes some troubles.
proposed to make another command for y-star and modify yPlusRAS command(I haven't tested for yPlusLES but probably it has the issue too)
TagsExcept buoyantKepsilon Model, I could not find it in any other model where Gb has been calculated., is there any other model involves Generation due to buoyancy

Activities

henry

2013-05-05 12:54

manager   ~0002159

The yPlusRAS command does not evaluate y+ directly but calls the yPlus function from the nut or mut BC.

user554

2013-05-05 13:34

  ~0002160

Last edited: 2013-05-05 13:39

in mutkWallFunctionFvPatchScalarField the equation uses for y-plus is related to y-star not y-plus

y-star=pow(0.09,0.25)*sqr(k)*nearDist()/mu
y-plus=rho*u_star*y/mu which u*=aqrt(taw_w/rho)
y-star is related to turbulent kinetic energy while y-plus relates to friction velocity(that relates to wall skin friction in its own)
 
in the code of mut
:
forAll(mutw, faceI)
    {
        label faceCellI = patch().faceCells()[faceI];

        scalar yPlus =
            Cmu25*y[faceI]*sqrt(k[faceCellI])/(muw[faceI]/rhow[faceI]);

        if (yPlus > yPlusLam_)
        {
            mutw[faceI] = muw[faceI]*(yPlus*kappa_/log(E_*yPlus) - 1);
        }
    }
thanks.

henry

2013-05-05 14:23

manager   ~0002161

Each nut/mut BC provides a consistent y+ function which takes into account the potentially extreme non-linear distribution of the near-wall velocity particularly for wall-functions. The simple nieve funcion for y+ using a simple estimate for the near-wall velocity gradient is appropriate for the laminar BC and for low-Re models in which the near-wall distributions are resolved.

user554

2013-05-05 14:47

  ~0002162

that you tell is true.but when having separation regions or high adverse pressure gradients in wall its more accurate that we use finer mesh near wall so that y-plus be around 1 to resolve viscous sub layer.this way we can include wall phenomenas exactly in return of have a coarser mesh and use wall-functions instead.
so,its good to have a real yPlus postProcessing tool as well as yStar for those that want to use coarser mesh and wall-functions and need it with true name.
best regards.

henry

2013-05-05 15:00

manager   ~0002163

Each nut/mut BC provides a consistent y+ function, for example compare the y+ functions in

nutkWallFunctionFvPatchScalarField

and

nutUWallFunctionFvPatchScalarField

Each provides a y+ function consistent with the wall-function formulation and hence the post-processing utility returns the y+ distribution used by and consistent with the wall-function.

user554

2013-05-05 16:19

  ~0002164

the view is understandable in genaral.but I want to know the true yPlus in a turbulent kOmegaSST with yplus<=1 to be sure that separation regions and adverse pressure gradients are calculated and included truely not through some "wall-function rough estimations".
maybe someone dosen't want to use wall functions at all when he has a low-Re grid mesh.
I want "y plus" exactly and only "y plus" not "y star" or any other thing instead of "y plus".

henry

2013-05-05 16:24

manager   ~0002165

> kOmegaSST with yplus<=1

You need a low-Re form of kOmegaSST which does not use wall-functions and the nut/mut BC of which will provide the consistent y+.

user554

2013-05-05 18:01

  ~0002166

so is there such version of kOmegaSST in OpenFOAM?
if i understand correctly then you say if anyone use yPlusRAS(or yPlusLES) he/she get an appropriate and correct value of y-plus regardless of using wall-functions or not(even if wall functions are not implemented it works well).is it so?
are you intending y-star values are a good estimate for y-plus values in all situations,or i'm wrong?

henry

2013-05-05 18:37

manager   ~0002167

We have not yet been sponsored to implement a low-Re version of kOmegaSST but we have implemented kkLOmega which is a low-Re k-omega model with support for transision. If you would like us to implement a low-Re version of kOmegaSST for you contact us via enquiries and we can discuss a contract to cover it. We would also implement appropriate BCs and the correspending y+ function for post-processing.

user554

2013-05-05 19:33

  ~0002168

is kkLOmega suitable for internal flow with a finer mesh near the wall and regions far the wall?
I did an enquiry.
thank you.

henry

2013-05-05 19:37

manager   ~0002169

Low Reynolds-number k-kl-omega turbulence model for
    incompressible flows.

    Turbulence model described in:
    \verbatim
        D. Keith Walters, Davor Cokljat
        "A Three-Equation Eddy-Viscosity Model for Reynold-Averaged
        Navier-Stokes Simulations of Transitional Flow"
    \endverbatim

    J. Fluids Eng. 130(12), 121401 (Oct 24, 2008)

Issue History

Date Modified Username Field Change
2013-05-05 12:44 user554 New Issue
2013-05-05 12:54 henry Note Added: 0002159
2013-05-05 12:54 henry Status new => closed
2013-05-05 12:54 henry Assigned To => henry
2013-05-05 12:54 henry Resolution open => no change required
2013-05-05 13:34 user554 Note Added: 0002160
2013-05-05 13:34 user554 Status closed => feedback
2013-05-05 13:34 user554 Resolution no change required => reopened
2013-05-05 13:39 user554 Note Edited: 0002160
2013-05-05 14:23 henry Note Added: 0002161
2013-05-05 14:23 henry Status feedback => closed
2013-05-05 14:23 henry Resolution reopened => no change required
2013-05-05 14:47 user554 Note Added: 0002162
2013-05-05 14:47 user554 Status closed => feedback
2013-05-05 14:47 user554 Resolution no change required => reopened
2013-05-05 15:00 henry Note Added: 0002163
2013-05-05 15:00 henry Status feedback => closed
2013-05-05 15:00 henry Resolution reopened => fixed
2013-05-05 16:19 user554 Note Added: 0002164
2013-05-05 16:19 user554 Status closed => feedback
2013-05-05 16:19 user554 Resolution fixed => reopened
2013-05-05 16:24 henry Note Added: 0002165
2013-05-05 16:24 henry Status feedback => closed
2013-05-05 16:24 henry Resolution reopened => no change required
2013-05-05 18:01 user554 Note Added: 0002166
2013-05-05 18:01 user554 Status closed => feedback
2013-05-05 18:01 user554 Resolution no change required => reopened
2013-05-05 18:37 henry Note Added: 0002167
2013-05-05 18:37 henry Status feedback => closed
2013-05-05 18:37 henry Resolution reopened => no change required
2013-05-05 19:33 user554 Note Added: 0002168
2013-05-05 19:33 user554 Status closed => feedback
2013-05-05 19:33 user554 Resolution no change required => reopened
2013-05-05 19:37 henry Note Added: 0002169
2013-05-05 19:37 henry Status feedback => closed
2013-05-05 19:37 henry Resolution reopened => no change required
2018-08-05 18:57 atulkjoy Tag Attached: Except buoyantKepsilon Model
2018-08-05 18:57 atulkjoy Tag Attached: I could not find it in any other model where Gb has been calculated.
2018-08-05 18:57 atulkjoy Tag Attached: is there any other model involves Generation due to buoyancy