[PorscheGTIII]

Greetings all!

I come to you very frustrated at the moment. I'm simulating a propeller-type hydrokinetic turbine in a large channel as shown bellow.


I'm using the frozen rotor approach to solve this problem using simpleFoam with the MRF option and the k-omegaSST turbulence model. I create an unstructured tetrahedral and prism mesh using ANSYS Meshing and then import the mesh to OpenFOAM format using fluent3DMeshToFoam. Once the file structure is set up, I check the mesh quality with checkMesh, decompose the mesh for a parallel run using the scotch method, and renumber the mesh to reduce the bandwidth. The simulations are initialized with potentialFoam for both the velocity and pressure fields.

For a very coarse mesh (~2 million cells), the simulation behaves very well. Values for torque and thrust (drag) on the turbine converge very nicely on firm values and the residuals drop very well as shown bellow.



However, when I refine the mesh (~7.5 million cells), the simulation does not behave as nicely. The residual for pressure only drops about 2 orders of magnitude and values for torque and thrust are not even resolved to one decimal place. (See below)



So I thought maybe because I was capturing more flow features was the reason the residuals were hanging, so I created a 47 million cell mesh and thought the residuals would improve. Nope, the behaved much of the same as the 7.5 million cell mesh.

This is the results from my checkMesh runs between the 2 and 7.5 million cell mesh:

2 Million Cell Mesh Stats

/*---------------------------------------------------------------------------*\
| =========                 |                                                 |
| \\      /  F ield         | OpenFOAM: The Open Source CFD Toolbox           |
|  \\    /   O peration     | Version:  2.2.2                                 |
|   \\  /    A nd           | Web:      www.OpenFOAM.org                      |
|    \\/     M anipulation  |                                                 |
\*---------------------------------------------------------------------------*/
Build  : 2.2.2-9240f8b967db
Exec   : checkMesh -case testCase
Date   : Nov 01 2013
Time   : 01:15:37
Host   : "Ranka"
PID    : 26751
Case   : ./testCase
nProcs : 1
sigFpe : Enabling floating point exception trapping (FOAM_SIGFPE).
fileModificationChecking : Monitoring run-time modified files using timeStampMaster
allowSystemOperations : Disallowing user-supplied system call operations

// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //
Create time

Create polyMesh for time = 0

Time = 0

Mesh stats
    points:           146558
    faces:            1280820
    internal faces:   1258192
    cells:            605624
    faces per cell:   4.19239
    boundary patches: 5
    point zones:      0
    face zones:       3
    cell zones:       2

Overall number of cells of each type:
    hexahedra:     0
    prisms:        116496
    wedges:        0
    pyramids:      20
    tet wedges:    0
    tetrahedra:    489108
    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               2364     1749     ok (non-closed singly connected)  
    outlet              2400     1767     ok (non-closed singly connected)  
    river_bed           7592     3913     ok (non-closed singly connected)  
    free_surface        5144     3315     ok (non-closed singly connected)  
    turbine             5128     2566     ok (closed singly connected)      

Checking geometry...
    Overall domain bounding box (-3.048 -1.8288 -3.048) (3.048 1.2192 3.048)
    Mesh (non-empty, non-wedge) directions (1 1 1)
    Mesh (non-empty) directions (1 1 1)
    Boundary openness (6.72913e-17 -2.14284e-16 -2.24694e-17) OK.
    Max cell openness = 9.18905e-16 OK.
    Max aspect ratio = 26.5342 OK.
    Minimum face area = 2.56528e-07. Maximum face area = 0.0276275.  Face area magnitudes OK.
    Min volume = 7.93254e-10. Max volume = 0.00137809.  Total volume = 88.9365.  Cell volumes OK.
    Mesh non-orthogonality Max: 82.8675 average: 14.8611
   *Number of severely non-orthogonal (> 70 degrees) faces: 34.
    Non-orthogonality check OK.
  <<Writing 34 non-orthogonal faces to set nonOrthoFaces
    Face pyramids OK.
    Max skewness = 2.38013 OK.
    Coupled point location match (average 0) OK.

Mesh OK.

End

7.5 Million Cell Mesh Stats

/*---------------------------------------------------------------------------*\
| =========                 |                                                 |
| \\      /  F ield         | OpenFOAM: The Open Source CFD Toolbox           |
|  \\    /   O peration     | Version:  2.2.2                                 |
|   \\  /    A nd           | Web:      www.OpenFOAM.org                      |
|    \\/     M anipulation  |                                                 |
\*---------------------------------------------------------------------------*/
Build  : 2.2.2-9240f8b967db
Exec   : checkMesh -case 7357266_Cells
Date   : Oct 28 2013
Time   : 14:19:03
Host   : "Ranka"
PID    : 19244
Case   : ./7357266_Cells
nProcs : 1
sigFpe : Enabling floating point exception trapping (FOAM_SIGFPE).
fileModificationChecking : Monitoring run-time modified files using timeStampMaster
allowSystemOperations : Disallowing user-supplied system call operations

// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //
Create time

Create polyMesh for time = 0

Time = 0

Mesh stats
    points:           1771385
    faces:            15543560
    internal faces:   15383084
    cells:            7357266
    faces per cell:   4.20355
    boundary patches: 5
    point zones:      0
    face zones:       3
    cell zones:       2

Overall number of cells of each type:
    hexahedra:     0
    prisms:        1497580
    wedges:        0
    pyramids:      0
    tet wedges:    0
    tetrahedra:    5859686
    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               15540    9823     ok (non-closed singly connected)  
    outlet              15958    10032    ok (non-closed singly connected)  
    free_surface        35872    20655    ok (non-closed singly connected)  
    river_bed           56652    28637    ok (non-closed singly connected)  
    turbine             36454    18229    ok (closed singly connected)      

Checking geometry...
    Overall domain bounding box (-3.048 -1.8288 -3.048) (3.048 1.2192 3.048)
    Mesh (non-empty, non-wedge) directions (1 1 1)
    Mesh (non-empty) directions (1 1 1)
    Boundary openness (3.19358e-16 -5.6173e-15 4.75543e-16) OK.
    Max cell openness = 1.7247e-15 OK.
    Max aspect ratio = 31.2746 OK.
    Minimum face area = 1.48502e-07. Maximum face area = 0.0035493.  Face area magnitudes OK.
    Min volume = 3.97334e-11. Max volume = 6.47112e-05.  Total volume = 88.9538.  Cell volumes OK.
    Mesh non-orthogonality Max: 80.6111 average: 14.8982
   *Number of severely non-orthogonal (> 70 degrees) faces: 165.
    Non-orthogonality check OK.
  <<Writing 165 non-orthogonal faces to set nonOrthoFaces
    Face pyramids OK.
    Max skewness = 2.48881 OK.
    Coupled point location match (average 0) OK.

Mesh OK.

End

System Files:
https://dl.dropboxusercontent.com/u/806642/controlDict
https://dl.dropboxusercontent.com/u/806642/fvSchemes
https://dl.dropboxusercontent.com/u/806642/fvSolution

BC's
https://dl.dropboxusercontent.com/u/806642/U
https://dl.dropboxusercontent.com/u/806642/k
https://dl.dropboxusercontent.com/u/806642/nut
https://dl.dropboxusercontent.com/u/806642/omega
https://dl.dropboxusercontent.com/u/806642/p

I'm a little frustrated right now. Is my issue a mesh issue? Or is there something that is flawed about my simulation?

Thanks for your help so far!

[pbachant]

I've had similar problems with pimpleDyMFoam that were alleviated by adding nonOrthogonalCorrectors (in fvSolution).

I set the max nonorthogonality in the mesh to 70 degrees (it was 65 by default with snappyHexMesh). It looks like you have some cells above that value.