CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > OpenFOAM

grid length scale in DES Spalart-Allmaras - definition of dTilda

Register Blogs Members List Search Today's Posts Mark Forums Read

Reply
 
LinkBack Thread Tools Display Modes
Old   February 2, 2011, 17:14
Default grid length scale in DES Spalart-Allmaras - definition of dTilda
  #1
New Member
 
Join Date: Mar 2009
Posts: 18
Rep Power: 8
fsaltara is on a distinguished road
Does someone know how is calculate the grid length scale used in dTilda? In the original model, Spalart proposed to use the maximum length of the cell, but in the lesdelta options I was unable to find something similar. I am using the cube root delta.
fsaltara is offline   Reply With Quote

Old   May 16, 2012, 18:50
Default
  #2
New Member
 
Join Date: Nov 2011
Posts: 13
Rep Power: 5
0.1 watts is on a distinguished road
Quote:
Originally Posted by fsaltara View Post
Does someone know how is calculate the grid length scale used in dTilda? In the original model, Spalart proposed to use the maximum length of the cell, but in the lesdelta options I was unable to find something similar. I am using the cube root delta.
Looking through the src it seems that the user is restricted to the usual LESDelta options.

One of the other options is maxDeltaxyz: 'Maximum of x, y and z; for structured hex cells only'.

Is this equivalent to the DES Delta parameter, and if so does it work with snappyHexMesh grids?
0.1 watts is offline   Reply With Quote

Old   May 17, 2012, 04:53
Default
  #3
Senior Member
 
Vesselin Krastev
Join Date: Jan 2010
Location: University of Tor Vergata, Rome
Posts: 361
Rep Power: 10
vkrastev is on a distinguished road
Until OF-2.1.0, the Spalart-Allmaras DES delta definition (max DeltaX, DeltaY, DeltaZ) was not available, so the only reasonable choice appeared to be the classical cubic root Delta definition (which, however, does not return the expected behavior in the RANS zone of the original S-A DES model). Now it seems that someone has detected (at last...) this deficiency, so the maxDeltaxyz is the consistent choice for the DES model in its original form (although "consistent" does not mean "best", especially in the LES region of the flow with highly anisotropic or unstructured grids)

Regards

V.
vkrastev is offline   Reply With Quote

Old   May 17, 2012, 06:35
Default
  #4
New Member
 
Join Date: Nov 2011
Posts: 13
Rep Power: 5
0.1 watts is on a distinguished road
Thanks for your advice, I'll try both with DDES and see what the differences are.
0.1 watts is offline   Reply With Quote

Old   July 17, 2012, 05:07
Default
  #5
Senior Member
 
Anne Gerdes
Join Date: Aug 2010
Location: Hamburg
Posts: 152
Rep Power: 7
Anne Lincke is on a distinguished road
Dear all,

I have a question concerning the computation of (max DeltaX, DeltaY, DeltaZ) in
src/turbulenceModels/LES/LESdeltas/maxDeltaxyz.

Here are some lines of the code

HTML Code:
 forAll(cells,cellI)
    {
        scalar deltaMaxTmp = 0.0;
        const labelList& cFaces = mesh().cells()[cellI];
        const point& centrevector = mesh().cellCentres()[cellI];

        forAll(cFaces, cFaceI)
        {
            label faceI = cFaces[cFaceI];
            const point& facevector = mesh().faceCentres()[faceI];
            scalar tmp = mag(facevector - centrevector);
            if (tmp > deltaMaxTmp)
            {
                deltaMaxTmp = tmp;
            }
        }
        hmax()[cellI] = deltaCoeff_*deltaMaxTmp;
    }
To my understanding, the method takes the maximum from faceCentre to cellCentre, for each cell.
But in order to obtain the desired (max DeltaX, DeltaY, DeltaZ), isn't it necessary to multiply the obtained value with a factor 2?
Is it necessary to set deltaCoeff to 2 in constant/LESProperties? As far as I can see the default value is 1.

I would appreciate any answers.

Kind Regards
Anne
Anne Lincke is offline   Reply With Quote

Old   July 24, 2012, 05:07
Default
  #6
Member
 
David GISEN
Join Date: Jul 2009
Location: Germany
Posts: 37
Rep Power: 8
David* is on a distinguished road
Hi Anne,
if you are using Hexamesh, you have to set deltaCoeff to 2 to obtain the correct value. As far as I understand, it is implemented in the present way to account for Polymeshes. There is a bug report (ID 0000474) at http://www.openfoam.org/bugs/ including this topic.
Regards,
David
David* is offline   Reply With Quote

Old   July 25, 2012, 10:23
Default
  #7
Senior Member
 
Anne Gerdes
Join Date: Aug 2010
Location: Hamburg
Posts: 152
Rep Power: 7
Anne Lincke is on a distinguished road
Hey David,

thank you for this hint. I found this thread a while ago, but could not find it anymore.

Kind Regards
Anne
Anne Lincke is offline   Reply With Quote

Old   August 1, 2012, 05:30
Default
  #8
Senior Member
 
Anne Gerdes
Join Date: Aug 2010
Location: Hamburg
Posts: 152
Rep Power: 7
Anne Lincke is on a distinguished road
Hello,

I have one small further question.

Citation from the Bug report site

Quote:
In case of polyhedral meshes it is easier to calculate the distance between the face centres and the cell centre. In the particular case of hexa you can just use the coeff for maxDeltaxyz to get the max(h) from face to face on the cell:

maxhxhyhzDeltaCoeffs
{
deltaCoeff 2;
}

The same with IDDESDeltaCoeffs, you can use deltaCoeff_

IDDESDeltaCoeffs
{
deltaCoeff 1;
cw 0.15;
}
Doesn't it has to be
Quote:
IDDESDeltaCoeffs
{
deltaCoeff 2;
cw 0.15;
}
when using hex meshes?

Kind Regards
Anne
Anne Lincke is offline   Reply With Quote

Old   May 27, 2013, 16:28
Default
  #9
Member
 
Join Date: Jul 2010
Posts: 36
Rep Power: 7
MaryBau is on a distinguished road
Hi everyone;

Does anyone know the answer to Ann's previous question about how to define deltaCoeff for hex meshes?

I believe that when using DES or DDES:

maxDeltaxyzCoeffs
{
deltaCoeff 2;
}

but for IDDES the deltaCoeff should be defined as 1. So:

IDDESDeltaCoeffs
{
deltaCoeff 1;
cw 0.15;
}

Is that correct?

Thanks for your help.

Last edited by MaryBau; May 29, 2013 at 11:48.
MaryBau is offline   Reply With Quote

Old   December 11, 2014, 05:48
Default
  #10
New Member
 
Join Date: Nov 2011
Posts: 28
Rep Power: 5
Andy_bm is on a distinguished road
This question is actual for me. I use LES turbulence model and snappyhexmesh. What is the delta should I choose 1 or 2? Because there is the problem of the flow past cylinder. If Deltacoeffs = 1 then Cd = 0.33, if deltaCoeffs = 2 then Cd = 1.2 as in the experiment. But I have coarse mesh y+ ~ 40, V=72m\s

Thanks.
Andy_bm is offline   Reply With Quote

Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
How to use DES well? Daniel Main CFD Forum 0 October 26, 2008 06:59
grid generation for Detached eddy simulation (DES) ROOZBEH Main CFD Forum 1 June 18, 2007 12:02
Kolomogorov length scale Bobby Main CFD Forum 8 November 23, 2002 04:54
Structured Grid Definition craig shores Main CFD Forum 3 March 21, 2001 15:48
Grid Independent Solution Chuck Leakeas Main CFD Forum 2 May 26, 2000 11:18


All times are GMT -4. The time now is 19:39.