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

About the non-orthogonal mesh and non-orthogonal corrector

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

Reply
 
LinkBack Thread Tools Display Modes
Old   January 26, 2013, 10:23
Default About the non-orthogonal mesh and non-orthogonal corrector
  #1
Member
 
jack
Join Date: Jul 2011
Posts: 52
Rep Power: 5
lg88 is on a distinguished road
Hello everyone
I used a modified solver to simulate a circular pipe flow.The mesh is non-orthogonal.So I think I need use non-orthogonal corrector in the fvSchemes. But the results is not good as the time developing.It seems that the mesh have great influence to the velocity.I have attached my velocity contours and mesh here.I hope you can help me to analysis the the fvSchemes I used. I think the problem maybe lie in the non-orthogonal corrector,but I don't have enough experience.I set the fvSchemes as following:
Code:
ddtSchemes
{
    default         Euler;
}

gradSchemes
{
      default        cellLimited leastSquares 1;
   // grad(p)         Gauss linear;
}

divSchemes
{
     default         none;
    div(phi,U)      Gauss linearUpwindV cellMDLimited leastSquares 1.0;
    div(phi)      Gauss linear 1;
    div(phiB,Dsig)  Gauss linear 1;
}

laplacianSchemes
{
    default        Gauss linear corrected;
    laplacian(nu,U) Gauss linear corrected;
    laplacian((1|A(U)),p) Gauss linear corrected;

    laplacian(DT,T)     Gauss linear corrected;
}

interpolationSchemes
{
    default         linear;
    interpolate(HbyA) linear;
}

snGradSchemes
{
    default         corrected;
}

fluxRequired
{
    default         no;
    p;
}
Thank you for your advice.
regards!


lg88
Attached Images
File Type: jpg 345.jpg (97.6 KB, 121 views)
File Type: jpg 346.jpg (30.7 KB, 94 views)
lg88 is offline   Reply With Quote

Old   January 27, 2013, 07:15
Default
  #2
Super Moderator
 
Bruno Santos
Join Date: Mar 2009
Location: Lisbon, Portugal
Posts: 8,258
Blog Entries: 34
Rep Power: 84
wyldckat is just really nicewyldckat is just really nicewyldckat is just really nicewyldckat is just really nice
Hi lg88,

What do you have in "system/fvSolution"? I'm asking this because the number of correctors is defined on that file: http://www.openfoam.org/docs/user/fvSolution.php

Best regards,
Bruno
wyldckat is offline   Reply With Quote

Old   January 27, 2013, 09:53
Default
  #3
Member
 
jack
Join Date: Jul 2011
Posts: 52
Rep Power: 5
lg88 is on a distinguished road
Thank you for your reply.I set my fvSolution as following:
Code:
solvers
{
    p
    {
        solver           PCG;
        preconditioner   DIC;
        tolerance        1e-06;
        relTol           0;
    };
    pFinal 
    { 
        solver           PCG;
        preconditioner   DIC;
        tolerance        1e-06;
        relTol           0;
    };

    ephi
    {
        solver          PCG;
        preconditioner  DIC;
        tolerance       1e-06;
        relTol          0;
    };

    U
    {
        solver           PBiCG;
        preconditioner   DILU;
        tolerance        1e-05;
        relTol           0;
    };
}

PISO
{
    nCorrectors     3;
    nNonOrthogonalCorrectors 3;
    pRefCell        0;
    pRefValue       0;
}

// ************************************************************************* //
I have set the number of correctors from 1 to 3,but the result is almost the same in the end.The velocity at the four corners has odd behavious as the attached picture show.


Best regards!

lg88
lg88 is offline   Reply With Quote

Old   January 27, 2013, 16:15
Default
  #4
Super Moderator
 
Bruno Santos
Join Date: Mar 2009
Location: Lisbon, Portugal
Posts: 8,258
Blog Entries: 34
Rep Power: 84
wyldckat is just really nicewyldckat is just really nicewyldckat is just really nicewyldckat is just really nice
Hi lg88,

I'm not experienced enough on this, but I suggest that you also provide the following details, so that someone with more experience can help you during this week:
  1. Have you checked the state of the mesh? What does checkMesh give you? And what about a full check, namely:
    Code:
    checkMesh -allGeometry -allTopology
  2. What solver (or based on which solver) are you using?
    • On a side note: have you confirmed that the solver you're using does apply this number?
  3. What's in play? More specifically: the fluid type, speeds that it reaches, the Reynolds number it can reach, and so on...
Best regards,
Bruno
wyldckat is offline   Reply With Quote

Old   January 28, 2013, 07:51
Default
  #5
Member
 
jack
Join Date: Jul 2011
Posts: 52
Rep Power: 5
lg88 is on a distinguished road
Hi Bruno
I run the command
Code:
checkMesh
and got the following message:
Code:
Checking geometry...
    This is a 3-D mesh
    Overall domain bounding box (-1 -1 0) (1 1 0.02)
    Mesh (non-empty, non-wedge) directions (1 1 1)
    Mesh (non-empty) directions (1 1 1)
    Mesh (non-empty, non-wedge) dimensions 3
    Boundary openness (4.2856998285e-20 5.9259059357e-20 -2.7446997359e-19) Threshold = 1e-06 OK.
    Max cell openness = 3.9590866047e-16 OK.
    Max aspect ratio = 159.52539546 OK.
    Minumum face area = 1.5253865084e-06. Maximum face area = 0.0014060260459.  Face area magnitudes OK.
    Min volume = 1.061859197e-07. Max volume = 2.8120520918e-05.  Total volume = 0.062779645967.  Cell volumes OK.
    Mesh non-orthogonality Max: 88.365490847 average: 7.2145668686 Threshold = 70
   *Number of severely non-orthogonal faces: 32.
    Non-orthogonality check OK.
  Writing 32 non-orthogonal faces to set nonOrthoFaces
    Face pyramids OK.
    Max skewness = 0.81894169959 OK.

Mesh OK
I think the mesh quality is good.
I modified my solver according icoFoam.
The fluid in my case is not a real material.I set its density to 1,kinematic viscosity 0.001,velocity 1 and Reynolds number 1000.

Then I modified the laplacianSchemes from
Code:
{
    default        Gauss linear corrected;
    laplacian(nu,U) Gauss linear corrected;
    laplacian((1|A(U)),p) Gauss linear corrected;

    laplacian(DT,T)     Gauss linear corrected;
}
to
Code:
{
    default        Gauss linear limited 0.5//or 0.333;
    laplacian(nu,U) Gauss linear limited 0.5//or 0.333;
    laplacian((1|A(U)),p) Gauss linear limited 0.5//or 0.333;

    laplacian(DT,T)     Gauss linear limited 0.5//or 0.333;
}
But I got worse result.
I have no ideal about it.

Thanks for your advice.

regards!

lg88
lg88 is offline   Reply With Quote

Old   January 29, 2013, 03:53
Default
  #6
Member
 
jack
Join Date: Jul 2011
Posts: 52
Rep Power: 5
lg88 is on a distinguished road
Hello everyone
When I set the laplacianSchemes from:
Code:
{
    default        Gauss linear corrected;
    laplacian(nu,U) Gauss linear corrected;
    laplacian((1|A(U)),p) Gauss linear corrected;

    laplacian(DT,T)     Gauss linear corrected;
}
to
Code:
{
    default        Gauss linear limited 0.5//or 0.333;
    laplacian(nu,U) Gauss linear limited 0.5//or 0.333;
    laplacian((1|A(U)),p) Gauss linear limited 0.5//or 0.333;

    laplacian(DT,T)     Gauss linear limited 0.5//or 0.333;
}
,the calculation will be divergence with few iterations.The velocity will be very large.I don't have any experience on this.Thank you for your advice.

regards!

lg88
lg88 is offline   Reply With Quote

Old   March 4, 2013, 11:56
Default
  #7
Senior Member
 
HECKMANN Frédéric
Join Date: Jul 2010
Posts: 236
Rep Power: 7
fredo490 is on a distinguished road
In my point of view your skewness looks a bit high (0.8). I had a similar problem that I solved by introducing a skewcorrector. You can look at my scheme here Error due to Unstructured Mesh with custom solver
fredo490 is offline   Reply With Quote

Old   March 5, 2013, 04:04
Default
  #8
Member
 
Join Date: Nov 2012
Posts: 58
Rep Power: 4
startingWithCFD is on a distinguished road
Sorry for the stupid question but I am curious:
If it is a simple, straight pipe, why do you not have a square block in the middle?
Why are its edges curved in this manner?
Maybe changing this may improve your mesh properties and therefore correct your solution.
startingWithCFD is offline   Reply With Quote

Old   March 5, 2013, 04:11
Default
  #9
Senior Member
 
HECKMANN Frédéric
Join Date: Jul 2010
Posts: 236
Rep Power: 7
fredo490 is on a distinguished road
I guess he used a "smoothing" function from his grid generator. Sometimes I get the same kind of restults with Pointwise.
fredo490 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



All times are GMT -4. The time now is 02:09.