[andreaRocca]

Hello Folks


I'm working with foam-extend simulating a Francis turbine and I have problems to get a solution.



In the CAD the surfaces composing the interfaces perfectly match each other.

After meshed (discretization error) the differences between areas at the interfaces are:

Spiral-GuideVanes: Diff = -1.54116647556601e-05
Bulb-DraftTube: Diff = 2.25976241261805e-05


According to your experience, is it enough an area difference of O(e-05) for GGI (or mixing plane) interfaces?



Bests

[hjasak]

This should not be a problem - compare the error in area between the two surfaces with respect to the flux error.


Can you post the whole message?

[andreaRocca]

Thank you for reply Prof. Jasak.

The message is the following:



GGI pair (SP-GV, GV-SP)
Area: 0.254671269026831 0.254686680691586 Diff = -1.54116647556601e-05 or 0.00605122525991961 %
Flux: 0.949992608907554 0.950088338717345 Diff = -9.57298097908899e-05 or 0.010075885145599 %
GGI pair (BULB-DT, DT-BULB)
Area: 0.138815526804067 0.138792929179941 Diff = 2.25976241261805e-05 or 0.0162788879936147 %
Flux: 0.972494692318502 0.972510217271029 Diff = -1.55249525264356e-05 or 0.0015963793748101 %


Regarding the last interface (mixing-plane) the mixing plane check is not working in parallel (this can be verified also running the tutorial "axialTurbine_mixingPlane").



I made the grids using Numeca AG5 and Omnis. The checkMesh looks fine.




Create time

Create polyMesh for time = 0

Initializing the GGI interpolator between master/shadow patches: SP-GV/GV-SP
Initializing the GGI interpolator between master/shadow patches: BULB-DT/DT-BULB
Time = 0

Mesh stats
all points: 22230489
live points: 13966149
all faces: 46481093
live faces: 38339883
internal faces: 35820020
cells: 12196229
boundary patches: 27
point zones: 0
face zones: 6
cell zones: 4

Overall number of cells of each type:
hexahedra: 11986443
prisms: 0
wedges: 0
pyramids: 0
tet wedges: 0
tetrahedra: 0
polyhedra: 209786

Checking topology...
Boundary definition OK.
Point usage OK.
Upper triangular ordering OK.
Face vertices OK.
*Number of regions: 4
The mesh has multiple regions which are not connected by any face.
<<Writing region information to "0/cellToRegion"
Nuumber of cells per region:
0 330992
1 0
2 0
3 0


Checking geometry...
This is a 3-D mesh
Overall domain bounding box (-0.7879466659 -1.671217118 -0.2864486732) (0.8220821991 4.442373826 3.74207617)
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 (-7.36175726854743e-13 -2.65983934844024e-12 2.58673887309245e-12) Threshold = 1e-06 OK.
Max cell openness = 5.15135387404077e-08 OK.
Max aspect ratio = 267.233167283032 OK.
Minumum face area = 3.35187928762322e-09. Maximum face area = 0.011201607594839. Face area magnitudes OK.
Min volume = 3.13556677637343e-12. Max volume = 0.00116740721267175. Total volume = 3.08841375862224. Cell volumes OK.
Mesh non-orthogonality Max: 72.648044449205 average: 18.5441847118136 Threshold = 70
*Number of severely non-orthogonal faces: 22.
Non-orthogonality check OK.
Writing 22 non-orthogonal faces to set nonOrthoFaces
Face pyramids OK.
Max skewness = 2.77047715280581 OK.

Mesh OK.

End

[jennyjian]

Hi, did you use the ggi interface in parallel mode ? I find the initialization takes too much time when I use ggi interface in parallel? Do I need to decompose the mesh in another way ? Thanks!