[mankaran90]

Hello Everyone

I am trying to simulate the flow of flue gas in a cyclone seperator using OPENFOAM gui simflow. When I try to solve the simulation the residuals starts decreasing at first and after some time begin to increase and then the solving stops and I get the following error

FOAM FATAL IO ERROR: 
wrong token type - expected Scalar, found on line 0 the word 'nan'

file: /home/mankaran90/simflow/CycSep/CycSep/system/data.solverPerformance.p at line 0.

--> FOAM Warning : 
    From function Foam::Istream& Foam::operator>>(Foam::Istream&, Foam::doubleScalar&)
    in file lnInclude/Scalar.C at line 93.

FOAM exiting

Please help!! Its important!! I shall be thankful to you guys!!

Mankaran

[BlnPhoenix]

You received an error that indicates that your solution diverged. So probably the solver blew up because of an error in your setup. This could be mesh related, BC's related or an numerically issue.

[mankaran90]

Thanks BlnPhoenix!!

I am new to CFD and can you suggest how to proceed with finding out the problem (wether it is mesh related, BC related or a numerical issue). Any source where I can learn about CFD in Openfoam and its related errors comprehensively!!

[BlnPhoenix]

Ah yes. Try to do the following to provide enough information users on this forum can try to help you.

1) Post your geometry, mesh and the CheckMesh output to see if it can be mesh related.
2) Post your Boundary and Initial Conditions to see if this can cause issues.
3) Post your used solver, turbulence model and the numerical stuff like schemes, relaxation factors etc.


Points 1) and 2) are most relevant in my opinion, so get these informations first if possible.

[mankaran90]

Actually I am working as a research fellow in a university and my project involves burning of biomass bales in a combustor and for removing particulate matter from the flue gas, a cyclone seperator is installed. The following link provides step by step instructions for simulating a cyclone seperator in Simflow- an Openfoam GUI
https://sim-flow.com/docs/3.1/tutori...bystep:cyclone
I decided to complete this tutorial before simulating using our conditions. So, I followed all the steps given in this tutorial but I did not use the cyclone.stl.gz file used in the tutorial. Instead, I created my own cyclone seperator model in Catia v5, exported the model as .stp file instead of stl.gz or .stl and in simflow, I scaled the geometry (from mm to meter), seperated inlet, bottom and top faces in the geometry (this step is not performed in the tutorial in the upper link) and then meshed. I used the exact same setup as in the tutorial (even same values) and got the error.

[mankaran90]

And also in the hexmesh panel, under the base tab, I did not define minimum and maximum dimensions of the base box, so that inlet, bottom and top faces of the geometry fall slightly outside the box. Instead, I used the auto size option.

[BlnPhoenix]

Ok, can you evaluate the mesh quality inside your CFD suite? When you have openfoam installed you could also go with a terminal to the folder containing the "0", "system" and "constant" folders and run the command "checkMesh".

[mankaran90]

This is the result I got after applying mesh checking:

/*---------------------------------------------------------------------------*\
| =========                 |                                                 |
| \\      /  F ield         | OpenFOAM: The Open Source CFD Toolbox           |
|  \\    /   O peration     | Version:  4.1                                   |
|   \\  /    A nd           | Web:      www.OpenFOAM.org                      |
|    \\/     M anipulation  |                                                 |
\*---------------------------------------------------------------------------*/
Build  : 4.1
Exec   : /opt/openfoam4/platforms/linux64GccDPInt32Opt/bin/checkMesh -noFunctionObjects -constant
Date   : Jul 26 2017
Time   : 18:08:50
Host   : "Mankaran90-Inspiron-N5010"
PID    : 3529
Case   : /home/mankaran90/simflow/CycSep/CycSep
nProcs : 1
fileModificationChecking : Monitoring run-time modified files using timeStampMaster
allowSystemOperations : Allowing user-supplied system call operations

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

Create polyMesh for time = constant

Time = constant

Mesh stats
    points:           90131
    faces:            260907
    internal faces:   249326
    cells:            85949
    faces per cell:   5.936462
    boundary patches: 4
    point zones:      0
    face zones:       0
    cell zones:       0

Overall number of cells of each type:
    hexahedra:     81521
    prisms:        2903
    wedges:        0
    pyramids:      0
    tet wedges:    7
    tetrahedra:    0
    polyhedra:     1518
    Breakdown of polyhedra by number of faces:
        faces   number of cells
            4   1026
            5   492

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
                     top      335      352  ok (non-closed singly connected)
                   inlet      120      142  ok (non-closed singly connected)
                  bottom       69       78  ok (non-closed singly connected)
    Cyclone_seperator_fluid_domain_geometry_1    11057    12070  ok (non-closed singly connected)

Checking geometry...
    Overall domain bounding box (-0.1976016 -0.1982511 -1.67) (0.24 0.1974258 0.16)
    Mesh has 3 geometric (non-empty/wedge) directions (1 1 1)
    Mesh has 3 solution (non-empty) directions (1 1 1)
    Boundary openness (5.738213e-17 2.612785e-16 -6.240371e-17) OK.
    Max cell openness = 3.025423e-16 OK.
    Max aspect ratio = 8.903284 OK.
    Minimum face area = 1.399115e-07. Maximum face area = 0.000390403.  Face area magnitudes OK.
    Min volume = 9.414362e-08. Max volume = 3.433906e-06.  Total volume = 0.1447505.  Cell volumes OK.
    Mesh non-orthogonality Max: 62.06493 average: 5.023939
    Non-orthogonality check OK.
    Face pyramids OK.
 ***Max skewness = 5.681266, 44 highly skew faces detected which may impair the quality of the results
  <<Writing 44 skew faces to set skewFaces
    Coupled point location match (average 0) OK.

Failed 1 mesh checks.

End

A pop-up window says that there are problematic skew faces

[mankaran90]

And in hex-meshing controls these are the values under following tabs:
General:
merge tolerance 1e-06
Combine faces: unchecked
CastellatedMesh:
Max local cells: 1000000
Max global cells 10000000
Min refinement cells: 10
Max Load unbalance: 0.1
Cells between levels: 4
Gap level increment: 0
Feature angle: 30
Snap:
Tolerance: 1
Feature snapping:
Implicit: checked
Explicit: checked
Multi region: checked
Feature angle 120
Feature refinement: 0
Iterations:
Smoothing: 3
Displacement: 500
Relaxation: 5
Edge snapping: 10
Layers:
Relative sizes: checked
Size control:
Expansion ratio: 1.25
First layer thickness: 0.2
Min thickness: 0.1
Layer growth control
Feature angle: 180
Slip feature angle: 30
Concave angle: 90
Buffer cells no extrude: 0
Min median axis angle: 90
Max face thickness ratio: 0.5
Max thickness to medial ratio: 0.5
Iterations:
Mesh displacement: 100
Layer addition: 50
Snapping relaxation: 5
Surface smoothing: 5
Interior smoothing: 3
Smooth layer thickness: 10
Relaxed iterations: 20
Quality:
Min
Volume: 1e-14
Tet quality: 1e-20
Area: -1
Twist: 0.05
Determinant: 1e-02
Face weight: 0.05
Volume ratio: 1e-02
Triangle twist: -1
Max
Non-orthogonality: 65
Boundary skewness: 20
Internal skewness: 4
Concaveness: 80
Advanced:
Smooth iterations: 4
Error reduction: 0.75
Relaxed max non-ortho: 75

[BlnPhoenix]

Ok, there might be the problem. If you say Boundary Conditions are identical and the overall geometry did not change, i would assume the mesh error is the main problem. OpenFoam is pretty picky about mesh quality.

Try to get rid of the skewed cells, so that checkMesh runs without any errors.

Can you try to use the default geometry not your own, to see if it runs without crashing?

[mankaran90]

I also tried meshing with lesser divisions, it passed the checkmesh without any errors but still got the above mentioned error with increasing residuals
the mesh is fine, the problem is either in defining BC or numerical schemes?
Plus after getting the above mentioned error, the console read

MPI_ABORT was invoked on rank 1 in communicator MPI_COMM_WORLD 
with errorcode 1.

NOTE: invoking MPI_ABORT causes Open MPI to kill all MPI processes.
You may or may not see output from other processes, depending on
exactly when Open MPI kills them.
--------------------------------------------------------------------------
[Mankaran90-Inspiron-N5010:05138] 1 more process has sent help message help-mpi-api.txt / mpi-abort
[Mankaran90-Inspiron-N5010:05138] Set MCA parameter "orte_base_help_aggregate" to 0 to see all help / error messages

OK, I'll try with the default geometry and then post....