[ari003]

Hello Foamers,
Hope that all is good with you people amid this situation. My problem might be very basic to some people but I'm facing some issues while troubleshooting this. The problem is as :-

I m using some set of velocity datas collected at every 0.5 time step(setting up in constant/boundaryData) as my timeVaryingMappedField at the inlet of my domain and I m wishing to use some kind of wall-function for this simulation. While using the Spalding wall function and setting the required b.c in the nut file I m getting some abnormal Courant number at some midway of my simulation. I ran the same simulation without wall function and it was running perfect but the aim is to run with the wall function. The nut file is shown below and I believe the b.c condition at the inlet might be something else.

/*--------------------------------*- C++ -*----------------------------------*\
| ========= | |
| \\ / F ield | OpenFOAM: The Open Source CFD Toolbox |
| \\ / O peration | Version: splitCyclic |
| \\ / A nd | Web: www.OpenFOAM.org |
| \\/ M anipulation | |
\*---------------------------------------------------------------------------*/
FoamFile
{
version 2.0;
format ascii;
class volScalarField;
location "0";
object nut;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //

dimensions [ 0 2 -1 0 0 0 0 ];

internalField uniform 0;
boundaryField
{
bottomWall
{
type nutUSpaldingWallFunction;
value uniform 0;
}
topWall
{
type nutUSpaldingWallFunction;
value uniform 0;
}
sides1_half0
{
type cyclic;
}
sides2_half0
{
type cyclic;
}
inlet
{
type calculated;
value uniform 0;
}
outlet
{
type calculated;
value uniform 0;
}
sides2_half1
{
type cyclic;
}
sides1_half1
{
type cyclic;
}

}

Will be really helpful if anyone can provide some solution for this problem.

[HPE]

some ideas:

- checkMesh output (particularly nearby wall patches)?
- reduce max Co (by reducing the time-step)?
- when you don't you nutUSpalding, what do you put instead, so that it starts working?
- double verification of the U inlet dataset?
- don't see any problem in nut setup.

[ari003]

Originally Posted by HPE

some ideas:

- checkMesh output (particularly nearby wall patches)?
- reduce max Co (by reducing the time-step)?
- when you don't you nutUSpalding, what do you put instead, so that it starts working?
- double verification of the U inlet dataset?
- don't see any problem in nut setup.

Thank you Sir for the reply. I have ran the setup with 0.05 sec time step even that reaches such abnormal Co at course. In normal case without wall func I m using 0.5 sec.

The checkMesh seems fine to me and here it is as follows:

Checking geometry...
Overall domain bounding box (0 0 -65) (1200 130 65)
Mesh has 3 geometric (non-empty/wedge) directions (1 1 1)
Mesh has 3 solution (non-empty) directions (1 1 1)
Boundary openness (-1.04754e-16 8.29579e-18 6.81378e-17) OK.
Max cell openness = 2.09998e-16 OK.
Max aspect ratio = 33.3682 OK.
Minimum face area = 0.649321. Maximum face area = 22.499. Face area magnitudes OK.
Min volume = 3.2466. Max volume = 97.4956. Total volume = 2.028e+07. Cell volumes OK.
Mesh non-orthogonality Max: 0 average: 0
Non-orthogonality check OK.
Face pyramids OK.
Max skewness = 2.72848e-13 OK.
Coupled point location match (average 4.62593e-19) OK.

Mesh OK.

End

The nut setup when I dont use nutUSpalding is as follow:

internalField uniform 0;
boundaryField
{
bottomWall
{
type zeroGradient;
}
topWall
{
type zeroGradient;
}
sides1_half0
{
type cyclic;
}
sides2_half0
{
type cyclic;
}
inlet
{
type calculated;
value uniform 0;
}
outlet
{
type calculated;
value uniform 0;
}
sides2_half1
{
type cyclic;
}
sides1_half1
{
type cyclic;
}

}

I m just wondering with the inlet bc when I use the nutSpalding function. Is it fine to use Calculated value 0 when I m feeding vel field at inlet?

[HPE]

yes, it should be ok.
- can you execute `checkMesh -allTopology -allGeometry`?
- what does yPlus estimation suggest on the walls?

[ari003]

After executing the first suggestion it is as follow:

Checking patch topology for multiply connected surfaces...
Patch Faces Points Surface topology Bounding box
bottomWall 7200 7471 ok (non-closed singly connected) (0 0 -65) (1200 0 65)
topWall 7200 7471 ok (non-closed singly connected) (0 130 -65) (1200 130 65)
sides1_half0 12000 12291 ok (non-closed singly connected) (0 0 -65) (1200 65 -65)
sides1_half1 12000 12291 ok (non-closed singly connected) (0 0 65) (1200 65 65)
sides2_half0 12000 12291 ok (non-closed singly connected) (0 65 -65) (1200 130 -65)
sides2_half1 12000 12291 ok (non-closed singly connected) (0 65 65) (1200 130 65)
outlet 3000 3131 ok (non-closed singly connected) (1200 0 -65) (1200 130 65)
inlet 3000 3131 ok (non-closed singly connected) (0 0 -65) (0 130 65)

Checking geometry...
Overall domain bounding box (0 0 -65) (1200 130 65)
Mesh has 3 geometric (non-empty/wedge) directions (1 1 1)
Mesh has 3 solution (non-empty) directions (1 1 1)
Boundary openness (-1.04754e-16 8.29579e-18 6.81378e-17) OK.
Max cell openness = 2.09998e-16 OK.
Max aspect ratio = 33.3682 OK.
Minimum face area = 0.649321. Maximum face area = 22.499. Face area magnitudes OK.
Min volume = 3.2466. Max volume = 97.4956. Total volume = 2.028e+07. Cell volumes OK.
Mesh non-orthogonality Max: 0 average: 0
Non-orthogonality check OK.
Face pyramids OK.
Max skewness = 2.72848e-13 OK.
Coupled point location match (average 4.62593e-19) OK.
Face tets OK.
Min/max edge length = 0.149843 5 OK.
All angles in faces OK.
Face flatness (1 = flat, 0 = butterfly) : min = 1 average = 1
All face flatness OK.
Cell determinant (wellposedness) : minimum: 0.00101484 average: 2.39035
Cell determinant check OK.
Concave cell check OK.
Face interpolation weight : minimum: 0.482649 average: 0.494306
Face interpolation weight check OK.
Face volume ratio : minimum: 0.932922 average: 0.977987
Face volume ratio check OK.

Mesh OK.

End

Seems same as the checkMesh option.
The yPlus looks like this

# y+ ()
# Time patch min max average

1000.00 bottomWall 5.176402e+01 4.951380e+02 1.560630e+02
1000.00 topWall 4.917430e+01 2.224669e+02 1.550358e+02

When the Co overshoot takes place the yPlus value also overshoots correspondingly.But the above one is from the case which I ran without wall function.

[ari003]

Originally Posted by HPE

yes, it should be ok.
- can you execute `checkMesh -allTopology -allGeometry`?
- what does yPlus estimation suggest on the walls?

Sir, I just want to let you know that I was using WALE model for this case. Is that can be a reason for this?