Rebuilding a corrupt, third party imported brep, turning it into a surface

Grasshopper
1

seat-rebuild
seat-rebuild.jpg724×532 45 KB

190524.gh (1.3 MB)

Dear all,

I want to share a Grasshopper problem with you, which I however solved but I am not happy with the surface quality, neither with the complexity of the GH-file. Maybe this could be done much easier.

I imported a test shape (knowing it was quite badly built) and made sections through it, extracted the edge curves and turned it into a surface with NetSurf (file attached above).

  • There are some problems with the section curves, sometimes they are not covering the whole brep, depending on where I position the sections. Is there a workaround?
  • Could the whole setup be more optimised, simpler?
  • Is there something I missed or is there a much better approach to this problem?

Sorry, this is a big file but yet I think the problem is super interesting.

Best regards,
SFS

2

I’m curious why you are using Grasshopper and not basic Rhino?

Can you upload a .3dm file with the input and final result?

3

I would say, because it’s more exact with Grasshopper. The 3dm file is internalised in the brep component.

4

How is it “more exact” with Grasshopper? There is no difference is the accuracy of the calculations.

5

“more exact” was perhaps not entirely correct.
I could do it in Rhinoceros but the ambition here is to create a parametric tool/method that can be reused. Also I want to choose e.g. points on curves and surfaces mathematically rather than with the mouse, which was my idea of “more exact”.
Additionally, the goal is to create a surface with perfect zebra pattern analysis. That’s a tough one.
But I’m open for a perfect surface made in this Rhinoceros anyday, so far I wasn’t happy with the results. :slight_smile:

6

If you want an untrimmed surface you could use something like that.
https://www.grasshopper3d.com/forum/topics/morph-from-source-grid-to-target-boundary-curve

7

I’ve been working on a GH tool to clean up a certain type of solid geometry construction with nearly coincident surfaces where parts are not joining. I can make centreline end points precisely identical and generate new surfaces. But I cannot necessarily get clean and complete surface intersections in GH - whilst I can in Rhino.

The difference is that I can adjust the document tolerance on the fly in Rhino. By creating the surfaces with a tighter than normal tolerance and finding the intersections with a looser than normal tolerance I can achieve an intersection that fails at normal tolerance. As far as I know (and I’m a GH newbie, so let me know if I am mistaken) it isn’t possible to have different tolerances at different stages in the GH process (I know there are some components where tolerance can be input, but not many) so I cannot achieve a result in GH that I can in Rhino.