error:
The quality of the mesh which is assigned to this simulation is below the recommended limits. We encourage you to investigate the mesh quality with the suggestions provided in the documentation. We do not recommend to proceed with this mesh as-is because the simulation may diverge quickly, or the results may be unreliable.
Hello,
The warning message contains a link to a documentation page with tips. Have you already gone through those tips? 
Cheers
Yes I tried to follow them the best I can, the mesh quality is still too low
Hi jwage,
I noticed in the Numerics area some of the settings were non-standard.
Can I suggest reset to the defaults, there is a button on the bottom left of the Numerics dialog box to reset to the defaults.
CFD simulation settings are very sensitive and using the defaults by Simscale is generally the best, well that has been my experience.
See this doc for help.
Hi again jwage,
You mention the mesh as an issue, but I noticed the mesh has some 4 million cells which should be ok, but if you wish to improve mesh quality, create a regional refinement and assign a geometric primitive box around your building with a finer mesh.
What do you mean by the mesh is ok? SimScale says that it is not recommended to use the mesh because its quality was too low or something. And when I tried to use the mesh in the simulation it would just have an error.
The mesh was still super low quality after I created a regional refinement
Hi jwage,
I now see the model & flow volume are very big, hence the low mesh resolution mentioned earlier.
Use the mesh clip if you wish to look at the meshing & also review the mesh log and review the acceptable/recommended limits.
Mesh log info,
Here is a mesh clip of your cells.
…and here is a slice through the mesh, showing the sphere and the increased resolution compared to the regional mesh.
Boxes rather than a spheres for regional refinements are typically selected, as a box provides a closer fit to the building… A couple of ways to go to increase mesh quality. Increase the overall regional mesh from possibly 5 to 6 or 7, change the sphere to a smaller closer fitting rectangular volume and choose a shorter edge length and/or a combination of both.
Simscale has most questions answered in the documentation area, if you do a search.
Just saw this, in another post might help.
An interesting project, I see someone is asking similar Forum question to you, that you may care to review. Look at Public projects to find working case, similar to your model/simulation, there are hundreds of simulations in public projects.
Hi jwage,
I just noticed this Simulation that is current underway, somewhat similar to your project, but wind is coming from east/north. Its a very large 15.4million cell simulation, but you can see its a large area & 2 regional refinements have been created. Interesting to see the parameters, note the ground plane settings in this model, as in real life and in accordance with wind codes wall (ground surface) friction impedes wind velocity close to the surface.
Note the settings for the ground plane
Boundary layer / wall friction info
