Quirk #97: FilletEdge?

Here is a circular array of 5 rectangular slabs who’s intersection is filleted with a 3.5 mm radius:

and my GH looks like this:

This looks fine :grinning:, but when I increase the fillet radius to 6.4 mm I get this:

and my GH looks like this:


Basically the FilletEdge results seem to be rather random based on the radius given and the settings for the B and M parameters. I’ve tried lots of combinations and there doesn’t seem to be any consistency. So I’m wondering if this could be the problem:


The result of the Brep Join and Solid Union appears to be a single closed BRep, but the Rhino display shows that the edges from each of the 5 slabs are still part of the Brep (maybe?) So I’m wondering if FilletEdge is somehow getting confused by the extra edges - if they are truly there - or, if not, is there some other trick to get FilletEdge to behave consistently?

Quirk#97.gh (11.5 KB)

1 Like

You need to merge the coplanar faces before doing the fillets.

merge_coplanar_faces.gh (5.5 KB)


Wow - spectacular. That really fixes it. Thanks.

Interestingly enough I did try doing a DeBrep and then merging the faces from that, but that did not help. I did that becauseI didn’t know the FMerge component was able to extract the Brep faces itself. But now I do.

I’m also thinking that doing that would be one way to help Rhino do a better job of converting it’s internal geometry to an STL file for 3D printing. An extra added benefit for sure.


1 Like

When you deconstruct, you need to join first to be able to merge.

Alternatively in some situations it might be more efficient to Region Union the curves and fillet sharp corners before extruding.

If the problem is solved, please click the image button, thanks.

1 Like