[Korimbo]

Hello,

I am trying to simulate the wind flow around two buildings represented by simple blocks. I am using trimmed mesh and my problem is that I get 3.3% of bad cells (i have activated cell quality remediation) and my solution diverges after around 100 iterations. My goal is to get the velocity between the buildings which is why the most refined is there.

I am using Thresholds to visualise where the bad cells are. I have some cells with a skewness angle higher than 85 on the edges of the buildings and the whole ground has a volume change of less than 0.01. These are the values that define the bad cells according to the documentation. There aren't any cells with cell quality less than 1e-5. I would like to know how i can improve my mesh quality since it seems to be the reason for the divergence.

I am attaching images of my mesh. I am using a symmetry plane between the buildings which is why only one building is visible.

[The]

can you give us some insights about the physical values? what kind of simulation you are doing: steady, unsteady, turbulence modell etc

[Korimbo]

It's a steady state simulation and I am actually testing different turbulence models - Spalart-Allmaras, k-w SST and Reynolds Stress Model. In this particular case I was using SA and it diverged and I guess it is due to the mesh quality.

[The]

Originally Posted by Korimbo

It's a steady state simulation and I am actually testing different turbulence models - Spalart-Allmaras, k-w SST and Reynolds Stress Model. In this particular case I was using SA and it diverged and I guess it is due to the mesh quality.

Which model are you currently using and what is your Y+ value? How many prism layers do you use?

[Korimbo]

I am using 12 prism layers and y+ value is less than 1 at most places. I am currently focused on using Spalart-allmaras.