I’m trying to create a pipe Delaunay structure to 3D print, ideally one in which the higher the gradient the smaller the triangles, to differentiate it from a regular triangle mesh and give it a more structural character. Recently my school changed to rhino 6 and I can’t find a way to translate a 2D Delaunay into a 3D double curvature surface.
In a past discussion forum I found a reference to a plugin called MeshEdit which contains a component called “3D Delaunay”, but it only seems available for Rhino 5
I also downloaded a plugin called SuperDelaunay and worked around one of their tutorial codes but I’ve not been successful in translating a 2D code to 3D
Hi @martinsiegrist, we only have Rhino 6 not Rhino 7 I don’t see TriRemesh in Rhino 6
besides this, I found an old discussion where someone brought up a similar topic and someone suggested “you are after a Del solver that could accept a Surface as “guide” (for instance your tube like facade, the roof etc etc)” However the discussion was never continued. Don’t know if this might be a dead end or not?
It says that the platform is Rhino 5, but it doesn’t mean is only avaliable for Rh5, almost everything that works for Rhino 5 works for Rhino 6 or 7 as well.
It sounds like maybe you are after a mesh on a curved surface rather than a volumetric mesh (instead of triangles covering a surface, a volumetric mesh would be tetrahedra filling a volume).
@DanielPiker Thanks for the reply! is there a way to control the density of the mesh according to the gradient? and/or is there a way to pipe the mesh? I ask because I’m only used to work with curves not meshes
Yes, you can define any gradient you like.
In that example I showed it with the gradient in the Y direction.
To use something different you need to create a value for each point of the input mesh according to its location.
Here’s a file showing examples of sizing by surface u parameter, or by proximity to a point sized_mesh2.gh (233.4 KB)
@martinsiegrist at this point just for curiosity where can I get the WIP version of Trimesh? I found NGON plugin, seems to be very good for what I am trying to do.
@Dani_Abalde you where right it does work in Rhino 6 as well, I guess now I know. on the other hand I am sure that I am doing something wrong, I feel the problem is establishing the surface as the boundary for joining the points, right know the Delaunay is connecting all points together. I know @DanielPiker already solved the issue with another method but I would still want to know if there is a solution with this approach?
Many thanks @DanielPiker! you really solved it in no time!
Not wanting to be someone to abuse here, but since I’m studying architecture I was interested in knowing if there is a way to link the tessellation or triangle division based on the forces, like a type of optimization, considering the form. Is there a plugin or tutorial that could help me with this? if you could point me in the right direction I would appreciate it infinitely!
Thanks @DanielPiker you’ve been so helpful! Great to know that, I will try it later in my personal laptop, don’t think my school will let me install a WIP lol
I’ve just acquainted myself with TriRemesh in the last few days and it is of itself hugely impressive. I am interested in something like the volumetric mesh that would be tetrahedra-like, filling a volume.
You mention this but I’m unclear if this is possible or not. As a beginner in Grasshopper, I may be missing something.
However, as a first or next step toward the goal of creating at least another layer of tetrahedra multi-piping, even if manually, would be possible somehow to accurately locate Points on the Nodes that are generated by TriReMesh (and multi-pipe)?
Or has that precision ability been lost with the conversion from NURBS to mesh?
Interesting to stumble on this today. In the past, every time I’ve ever researched this stuff for different projects, I’ve also wanted to do ‘volumetric’ style delaunay/voronoi geometry.
But all I ever see is ppl only doing it 2-dimensionally along a surface
I have some stuff I got from @DanielPiker , but I think there’s still improvements that can be made.
When I get back into this soon, this is close to where I will pick up: