Multi-Building Mesh
 

I am having difficulty getting snappyHexMesh to produce a "clean" -looking mesh when the block mesh is not parallel to the building walls (there is about a 30 degree angle in my case). If anyone can offer help regarding this matter, it would be greatly appreciated. I am attaching an image of the buildings (viewed from below) for reference. Thanks for any help/experience you can offer. Ultimately, the model is unstable at high windspeeds, but I think some of this may be a problem with the meshing.

Greetings Robert and welcome to the forum!

You're not making snappyHexMesh's life very easy :( Since the natural orientation of the base mesh is not aligned with the main feature orientations of those zones, you get those very bad adjustments :(

If you can't improve the alignment of the base mesh on that zone, you'll have to increase the resolution there, in order for the mesher to accommodate cells there. Otherwise, even if the snapping process did things exactly the way you want them to be, the resulting cells will be veeeery bad in those zones!

Best regards,
Bruno

Winds at boundary of finer mesh
 

I've rotated my domain and the results are much more stable. My current challenge involves a two difference in windspeed at the boundary between the finer mesh and the blockmesh, particularly near ground level and upstream of the buildings. I've attached a cross-section of 5 meter winds with mesh. Any idea how to resolve this apperently non-physical result? I appreciate any advice.

Robert

Hi Robert,

I'm not quite certain if I understand what the non-physical result is, so I'm going to have to suggest that you:

• Check the units for the mesh, all of boundaries and transport properties.
• Try using other schemes in fvSchemes: http://www.openfoam.org/docs/user/fvSchemes.php
• Try increasing the number of cells between each transition of mesh resolution.

Good luck!
Bruno

Boussinesq Solvers
 

Thanks for the help on this, Bruno.

Can anyone offer advice regarding solvers for Boussinesq flow? I plan to add a source term to one of the existing solvers. I was considering buoyantBoussinesqSimpleFoam vs buoyantBoussinesqPisoFoam. Currently, higher resolution flow around buildings with scalarSourceSimpleFoam yields unstable results. I hope that moving over to version 2.1 and one of the Boussinesq solvers will help this. The complex geometry of our buildings is the challenge here and we have already simplified the buildings as much as possible.