I’m trying to generate a surface using relaxed points and edge curves after performing relaxation, as visible here:
The purpose of generating this surface is that I need to acquire normal vectors, so that I can separate the three (timber) layers - however I have very little experience in working with surfaces in GH (and maybe also just GH in general…), so I’m currently struggling to achieve a nice result. I’ve searched a bit on the forum and tried with the “patch surface” option using different setting for spans and flexibility, but again; the result is always a bit too messy. The problem occur around the holes, probably because the UV division in this “vertical” part of the structure becomes too coarse? Ideally, the surface shouldn’t try to wrap around these holes, but just leave them open.
I know that it is somehow possible, since the following GH-user has demonstrated a very nice result on this particular problem (https://www.grasshopper3d.com/photo/proceso-17-nurbs-e?context=user)
Do anyone have an idea on how to tackle this problem? Maybe my approach is wrong, and I should be lofting curves instead, using mesh, multiple surfaces or… ? Suggestions are very welcomed
Thanks in advance!
Surface_Generation_Problem.gh (544.3 KB)
This is not a problem about your knowledge of grasshopper, but in rhino or even more precisely in NURBS.
Nurbs object have continuity.
Having a jagged , nice and clean “cut” in the middle of a surface is like asking this:
to be a “single” curve object. It’s impossible.
What you can do is to greatly increase the control-points count so you can “cheat” it and make an hole in the classic way.
Like, for keepeing comparison, this:
Contuinuity is mantained (tengency).
Is there a particular reason you want to do it a single surface? If not, avoid that.
I would use multiple surfaces (and keeping them tangent each-other where needed) or a mesh.
The link you showed don’t make it clear if it is a single surface or a polysurface/brep.
Also, if you want to post your data, it would be more useful to work with the previous “steps”.
All your geometries have no data-structure… re-sorting them would be a trouble.
Thanks for the reply and suggestions! Makes sense about the continuity, and probably more suitable to go with a different approach instead of trying to brute force it with additional control points.
As long as I am somehow able to extract normal vectors to associated curves that lie on each surface, I dont mind it being multiple surfaces or a mesh, which I guess shouldn’t be a problem.
Attached is the full WIP - I hope it’s not too-too messy.
01_Miniproject.gh (120.0 KB)