[nskelly]

Hi all,

This is my first post in this forum, and let me firstly say everyone has helped me a lot already with all the problems I have come across so far.

Now here's my problem. I am aiming to do a LES of a simple backward facing step and I am using blockMesh to create the mesh. I have created the geometry using 5 blocks but I need to have a fine mesh around particular regions, namely at the step, and upper and lower walls. I have managed to get a fine mesh around those regions, but grading has made the mesh downstream larger. The cells end up getting quite large, because the computational domain is really long (replicating an experiment). If I do run simulations, my time stepping needs to be really small which is not ideal as I have lots of simulations to run. I have tried to run simulations and the Courant number just gets larger, and also takes a lot of time, this is with a timestep of 1e-6.

Any tips or advice would really help. I've attached some pics of the mesh I have made and the blockMeshDict if needed. If I have not said any important information please let me know, I'm new to all of this.

__________________________________________________ _
checkMesh
__________________________________________________ _
Create time

Create polyMesh for time = 0

Time = 0

Mesh stats
points: 3246903
faces: 7564600
internal faces: 5395400
cells: 2160000
faces per cell: 6
boundary patches: 7
point zones: 0
face zones: 0
cell zones: 0

Overall number of cells of each type:
hexahedra: 2160000
prisms: 0
wedges: 0
pyramids: 0
tet wedges: 0
tetrahedra: 0
polyhedra: 0

Checking topology...
Boundary definition OK.
***Total number of faces on empty patches is not divisible by the number of cells in the mesh. Hence this mesh is not 1D or 2D.
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
outlet 1600 2403 ok (non-closed singly connected)
inlet 800 1203 ok (non-closed singly connected)
lowerWall 3000 4506 ok (non-closed singly connected)
upperWall 3000 4503 ok (non-closed singly connected)
front 1080000 1082301 ok (non-closed singly connected)
back 1080000 1082301 ok (non-closed singly connected)
defaultFaces 800 1203 ok (non-closed singly connected)

Checking geometry...
Overall domain bounding box (-0.2 -0.004889 -5e-05) (0.5 0.0052 5e-05)
Mesh has 2 geometric (non-empty/wedge) directions (0 1 1)
Mesh has 2 solution (non-empty) directions (0 1 1)
All edges aligned with or perpendicular to non-empty directions.
Boundary openness (-4.47912e-20 -2.71433e-17 2.56684e-13) OK.
Max cell openness = 2.2044e-16 OK.
Max aspect ratio = 90.1814 OK.
Minimum face area = 1.36439e-12. Maximum face area = 2.05117e-07. Face area magnitudes OK.
Min volume = 6.82195e-17. Max volume = 1.02559e-11. Total volume = 6.0845e-07. Cell volumes OK.
Mesh non-orthogonality Max: 0 average: 0
Non-orthogonality check OK.
Face pyramids OK.
Max skewness = 5.28696e-12 OK.
Coupled point location match (average 2.76656e-17) OK.

Mesh OK.

[anuragm]

Perhaps you have not noticed this but there is an error in your mesh.

***Total number of faces on empty patches is not divisible by the number of cells in the mesh. Hence this mesh is not 1D or 2D.

The usual cause for the error blockMesh throws is that the blocks are not properly defined or are skewed. However, in your case this is simply because OpenFOAM wants a 2-D mesh in the XY plane with Z as the empty direction. If you go to the casedir and execute

> paraFoam -block &

you can see the blocks in Paraview. It just makes it easier to troubleshoot or add/edit blocks.

Now here's my problem. I am aiming to do a LES of a simple backward facing step and I am using blockMesh to create the mesh. I have created the geometry using 5 blocks but I need to have a fine mesh around particular regions, namely at the step, and upper and lower walls. I have managed to get a fine mesh around those regions, but grading has made the mesh downstream larger. The cells end up getting quite large, because the computational domain is really long (replicating an experiment). If I do run simulations, my time stepping needs to be really small which is not ideal as I have lots of simulations to run. I have tried to run simulations and the Courant number just gets larger, and also takes a lot of time, this is with a timestep of 1e-6.

Split the block(s) where this problem occurs in two halves. This way you have more control over the grading/mesh size.

[nskelly]

Hi

Thanks for your help. I actually made a silly mistake of not defining the boundary of the step, so the geometry was 'open'.

It is all resolved now. Thanks