[tayo]

Hello,
I'm puzzled about the way openfoam computes the gradient scheme. I have a temperature field, T that varies between 358 - 400K and I simply take the gradient using fvc::grad(T), with Gauss linear scheme. With a mesh size is approx. 0.3 mm, some of the results for the gradient are printed below:

const volScalarField& T = alpha1_.db().lookupObject<volScalarField>("T");
volVectorField gradT = fvc::grad(T);
Info << "gradT =" << gradT << endl;
(-164633 -164633 0)
(0 -164633 0)
(0 -164633 -2.98156e-10)
(0 -164633 -2.98156e-10)
(0 0 2.98156e-10)
(0 0 1.49078e-10)
(164633 -164633 0)
(0 0 0)
(0 0 0)
(164633 0 0)
(0 0 0)
 :  :  :
 :  :  :

Some of my basic questions are:
1.) Why does it print out negative numbers?
2.) How does it get gradient values of (0 0 0) when the T values vary between 358 - 400K.
3.) Why is the z- gradient values so low (2.98156e-10) considering the small mesh size?
Kindly help explain what's happening here. Thanks

[alexeym]

Quite usual questions:

1. What type of BCs do you use?
2. Can you post checkMesh output?
3. Can you post your fvSchemes?
4. Can you post your case files?

[akidess]

Post plots of T and gradT from paraview as well.

[tayo]

Thanks for the response. Below is my boundary condition
Temp.: fixedGradient on wall, fixedValue inlet, zeroGradient outlet
Pressure: zeroGradient inlet and wall, fixedValue outlet
Velocity: no slip wall, fixedValue inlet, zeroGradient outlet

Here is my fvScheme:

ddtSchemes
{
    default         Euler;
}
gradSchemes
{
    default         Gauss linear;
}
gradSchemes
{
    default         Gauss linear;
}
divSchemes
{
    div(rho*phi,U)  Gauss limitedLinearV 1;
    div(phi,T)  Gauss upwind;
    div(phi,p_rgh)  Gauss upwind;
    div(phi,alpha)  Gauss vanLeer01;
    div(phirb,alpha) Gauss interfaceCompression;
: : :
}

The checkMesh also seems ok:

Create time

Create polyMesh for time = 0

Time = 0

Mesh stats
    points:           93636
    faces:            267729
    internal faces:   254991
    cells:            87120
    boundary patches: 3
    point zones:      0
    face zones:       0
    cell zones:       1

Overall number of cells of each type:
    hexahedra:     87120
    prisms:        0
    wedges:        0
    pyramids:      0
    tet wedges:    0
    tetrahedra:    0
    polyhedra:     0

Checking topology...
    Boundary definition OK.
    Cell to face addressing OK.
    Point usage OK.
    Upper triangular ordering OK.
    Face vertices OK.
    Number of regions: 1 (OK).

Checking patch topology for multiply connected surfaces ...
    Patch               Faces    Points   Surface topology                  
    inlet               2640     2754     ok (non-closed singly connected)  
    outlet              2640     2754     ok (non-closed singly connected)  
    heatWall            7458     7684     ok (non-closed singly connected)  

Checking geometry...
    Overall domain bounding box (0 0 0) (0.02 0.01 0.01)
    Mesh (non-empty, non-wedge) directions (1 1 1)
    Mesh (non-empty) directions (1 1 1)
    Boundary openness (3.88444e-16 1.23164e-16 -3.56799e-15) OK.
    Max cell openness = 8.82326e-17 OK.
    Max aspect ratio = 1.33333 OK.
    Minumum face area = 7.5e-08. Maximum face area = 1.0101e-07.  Face area magnitudes OK.
    Min volume = 2.27273e-11. Max volume = 2.52525e-11.  Total volume = 2e-06.  Cell volumes OK.
    Mesh non-orthogonality Max: 0 average: 0
    Non-orthogonality check OK.
    Face pyramids OK.
    Max skewness = 0.000109995 OK.
    Coupled point location match (average 0) OK.

Mesh OK.

Time = 0.0005

Mesh stats
    points:           95614
    faces:            272670
    internal faces:   259779
    cells:            88618
    boundary patches: 3
    point zones:      0
    face zones:       0
    cell zones:       1

Overall number of cells of each type:
    hexahedra:     88339
    prisms:        0
    wedges:        0
    pyramids:      0
    tet wedges:    0
    tetrahedra:    0
    polyhedra:     279

Checking topology...
    Boundary definition OK.
    Cell to face addressing OK.
    Point usage OK.
    Upper triangular ordering OK.
    Face vertices OK.
    Number of regions: 1 (OK).

Checking patch topology for multiply connected surfaces ...
    Patch               Faces    Points   Surface topology                  
    inlet               2640     2754     ok (non-closed singly connected)  
    outlet              2640     2754     ok (non-closed singly connected)  
    heatWall            7611     7859     ok (non-closed singly connected)  

Checking geometry...
    Overall domain bounding box (0 0 0) (0.02 0.01 0.01)
    Mesh (non-empty, non-wedge) directions (1 1 1)
    Mesh (non-empty) directions (1 1 1)
    Boundary openness (-4.07224e-16 3.9387e-17 -3.50976e-15) OK.
    Max cell openness = 1.76465e-16 OK.
    Max aspect ratio = 1.33336 OK.
    Minumum face area = 4.6875e-09. Maximum face area = 1.01015e-07.  Face area magnitudes OK.
    Min volume = 3.55078e-13. Max volume = 2.52538e-11.  Total volume = 2e-06.  Cell volumes OK.
    Mesh non-orthogonality Max: 29.6211 average: 1.53168
    Non-orthogonality check OK.
    Face pyramids OK.
    Max skewness = 0.334633 OK.
    Coupled point location match (average 0) OK.

Mesh OK.

The issue seems to be with the gradScheme. It gives fairly the same result when for fvc::grad(T) and fvc::grad(T-TSat) where TSat is a dimensionedScalar at 371K while T is a volScalarField defined in previous post.

volScalarField gradT = mag(fvc::grad(T));
volScalarField gradTt = mag(fvc::grad(T-TSat));
Info << min(gradT) << max(gradT) << min(gradTt) << max(gradTt) << endl;

Checking the min. & max. values for mag(grad(T)) and mag(grad(T-TSat) shows that they are exactly the same. This is clearly wrong but I don't understand why. However, when I compute these values by postprocessing with funkySetField, I get more reasonable result as shown in the plots. So my question is this: why are these two grad(T-TSat) results significantly different? How do I best define grad(T-TSat) to give the expected result? Thanks

[alexeym]

Hi,

Concerning grad(T) and grad(T - Tsat), as Tsat is constant, grad(Tsat) = 0 and grad(T) = grad(T - Tsat).

From you graphs I can see the areas of constant temperature, so in these areas grad(T) will be 0.

And finally negative numbers - we need to continue guessing what are the real initial and boundary conditions.