SimScale CAE Forum

Rhino Meshing Failure


Hello Everyone,

Rhino Meshing Failure(s): here’s a simscale newbie who needs to carry out an urban wind flow analysis (architecture/urban design sector). Prepared the urban blocks by Rhino5, tried different input strategies: the original blocks, the reciprocal of the blocks by BooleanDifference, and an exploded version of the reciprocal. Links below…

All failed to generate the necessary mesh…

Would be grateful if anyone could help…


Hi @clk!

We will have a look at that. In the meantime I am tagging my colleague @yosukegb4 who might help you identify modeling issues with Rhino.




Update: The model has intersections as well as regions where it is not closed. The geometry has to be watertight! See attached picture for more information.

I would suggest to check the geometry with Rhino which has nice features especially for such kind of issues. Let us know how things go!



Hello @jousefm and @yosukegb4 updates:

Thanks for your quick reply! I rebuilt the model with less details and gave each block a clean boolean and ran a check to eliminate all unnecessary curves and open surfaces…still couldn’t generate the mesh. See the second model (‘Simplified - Original’) of the link…



Hi @clk! Just gave it a quick spin to see if a change in the format does something - indeed it does.

I uploaded the Rhino geometry to OnShape and exported it as an IGES file which seems to work without an error. Of course some refinements need to be added but that was just to see if another format does the job.

Try it with the IGES file and let me know if you need any further help!



Aha, Rhino, I can help :smile:

Using your NO_BOX geometry of Post #4 project, I found and ‘fixed’ the watertight issues doing this;

  1. Boolean unioned all separate building complexes into individual closed solids.

  2. On each of the complexes that showed as single closed polysurfaces (or extrusions) after the union, I ran MergeAllFaces to minimize surface edges on all coplanar surfaces.

  3. Only one complex did not succeed the Boolean Union. It was the tallest structure made of 4 extrusion layers.

  4. To turn the tallest structure into a semi-nice closed polysurface (watertight), I had to do this:

    1. Level 3 was not a close polysurface (or extrusion, it was created by an extrusion of an OPEN curve), so I closed the curve with a zero radius fillet at the start point (start point was found by using CrvStart command)
    2. Then I extruded that closed curve of level 3 to the same height
    3. Then I had to overlap each level by 2 mm so that they would Boolean union.
    4. Then I ran MergeAllFaces on it and found that some of the exterior wall surfaces that should be coplanar between levels were NOT coplanar, so there are ‘steps’ in some exterior walls when the level changes.
    5. This is where I stopped, what should be done is to make sure the individual level closed curves that the 4 levels were made from are made from EXACTLY the same curve sections on sections that continue across the levels. I will leave that for you.

Here is a .3dm file of all the watertight objects (closed polysurfaces/extrusions with MergeAllFaces applied) for as far as I got.

I have not tried meshing it but it should as long as all ground level faces are truly at ground level.

Hope this helps.

I personally think it better that you use properly created native .3dm file for import to SimScale vs adding another level of bandaid fixes like converting to IGES.



Hi @clk,

I use Rhino 5 for Mac for modeling.
I always export STL files from Rhino for CFD simulations.

It’s very convenient if IGES, STEP, 3DM files can make good CFD meshes,
but in my cases only STL files make good CFD meshes.
I think the cause is that my model shapes are a bit complicated.

It may not be directly related to your case this time,
but if you can not create a mesh from a complicated model,
it may be good to try using a STL file.



Hello everyone,

Bingo. Rhino export as STL solved the issue.

Thanks everyone…