why I cannot join these surfaces.gh (59.3 KB)
The problematic surfaces’ edges are partly from the base surface isocurves. So, why cannot I join them in the final stage?
I’m not sure if this is a bug, but here’s a solution for your reference.
BrepJoinIssue.gh (108.9 KB)
why I cannot join these surfaces a.gh (35.2 KB)
Thank you!
I guess this is a bug. There is no reason for that. I didn’t rebuild the surface edge in the progress.
@jessesn
I forgot to mention that I upload the surfaces to the Fusion 360 and tried to stitch(join) the surfaces but also failed.
I have logged this issue to our development engineers. Thank you.
RH-96273 Join: Unable to join
Thank you, but it seems this has nothing to do with which tool you used to create the surfaces; it’s related to the position and geometric continuity of the adjacent surface edges.
The ripples are likely because the sections are not (all) perpendicular to the rails.
Got it, Thank you very much!
“Cannot be joined, why is that?”
Each segment edge curves away from its neighbour. To join the segments you have to split the edges where they diverge:
“also, why it has ripples?”
I suspect it is because the sweep algorithm generates a conflict between the rails pulling down and the shapes pulling up. This may be exacerbated by the shapes not lying on planes perpendicular to the two rails - although experiments with perpendicular shapes aren’t fully ripple free, particularly where the isocurves crowd together at the sharpest curvature.
It is also better, albeit not perfect, if you change one rail to follow the crown line and model a half segment. You can then rebuild the surface to, say 48 x 5 points to remove the ripples.
I don’t think the topic is about suggesting different methods but much rather why the surface segments can’t be joined.
I projected the sweep and one rail to the XY plane and split the surface with a circle.
Split within the linear part of the surface edge, the surfaces can be joined.
This scaled curvature graph might show the problem:
My suggestion would be to avoid splitting with isocurves. The resulting curve has more control points than the initial rebuilt curve use for the revolution. It is questionable why you even create a revolution first.
Finally I think it is irrelevant how you create the surface, whether you loft or sweep or whatever.
If you want to join an array of segments, the crucial part is that the edges are within the angle segment, 45° in your case…
why I cannot join these surfaces mrtn.gh (64.9 KB)
When I had a similar design it seemed to me to be a trimming issue. So this is the method I used:
Which made this using the original geometry from post #1 above:
trimmer.gh (109.4 KB)
Indeed, it got repaired! Thank you very much!
@Birk_Binnard Thank you!
R9 BETA cannot cap the tranformed surface v0.gh (22.2 KB)
I am not sure if this is a similar issue.
There is a rarely used component called Mobius Transform in the K2 category.
It can deform a surface around a circle. When Pi is 0.5, it forms a perfect half sphere. The resulting half sphere can be joined as a closed Brep, but cannot be capped; it returns an error.
