# Connect 4 circles with straight line

@akilli I am working with ‘basic’ forms like a sphere (dome), a cylinder, a cone, and a hyperboloïd. Those are selfsupporting structures, that I start cutting through to connect them to each other. The other one there structure is ok, but the hyperboloïd needed the second structure underneath to be strong enough. If it is not possible that the turned hyperbola is a hyperboloïd, I am fine with the option to rotate the hyperbola, without being a real hyperboloïd after.
The cross section is important for my division method. I make big triangles in there, and if it is not a circle, there is no way to make some lengths the same as others (thats another story to explain why )

this is the first step in the division method, and all points of the triangles are lying on the surface.
Working with points exactly on the surface is important, because the form of a hyperboloïd make the ‘forces in a structure’ go nice down from the top til the bottom…

If you are willing to allow the inner hyperboloid to have elliptical cross sections, then you can get the bases to align perfectly. You accomplish this by scaling the outer structure down in one direction. The bases still coincide but the roofs never touch The panels come from the Lunchbox plugin

.

hyperboloid from circles 3.gh (18.4 KB)

1 Like

@akilli, waw, this looks really nice. I am drawing for the moment a surface of revolution, based on a curve I draw with the points of the circles I needed on. I draw it in gh, so I can change it if i would get a better curve that would be a hyperboloid function I have quiet a big grasshopper file for the moment with my division method of triangles I make, so I am afraid if I open now your file, my Rhino will crash… But I am very curious to see it and thank you very much spending time on my drawing! I am in my last step of my gh for the moment, and I am missing just 1 command to finish it… I don’t find the ‘curve-> normal to surface’ command form Rhino in gh. Do you know this?
Thanks again!

@akilli Hi, opened your file, installed the plugin… A new world is opening to me! This toolbox is great! I tried to do use the panels on a piece of a surface of an hyperboloïd I divide with my method to see the difference in division, but he is making the panels like the hyperboloïd is still complete. Can I do something to change that, that he divides only my piece into triangles?

I don’t think I can give you a reasonable reply at this point without seeing your code. Could you upload the file and also internalize whatever geometry you’re importing from Rhino? Thanks.

@akilli hi, here are both files:
test verdeling driehoek plugin.gh (29.8 KB)
test verdeling driehoek plughin.3dm (6.0 MB)

I tried to mark were the issues are and what I tried, It’s a bit messy in my rhino file as I tried things out and copies… What I was thinking if the gh maybe only want to triangulate the revolved line, and any changes on that surface he doesn’t look at. Thanks

There are some people on this forum who are absolute mind readers - and I’m not one of them. Could you please clarify some things for me? You seemed to be enthusiastic about the Lunchbox triangulation but you also use a complex cutting surface for the split. Do you intend to use one or the other? Also, your cutting surface is used once. Do you intend to use it multiple times for a different triangulation? Again, even a lousy drawing would be much appreciated.

@akilli I am sorri for being not clear. I hope this file can help… I put the steps I did of cutting in a row with the final surface I want to triangulate. uitleg hyperbool opdeling met panel.3dm (3.1 MB)
But I think the panel doesn’t want to work with a cutted object. I try out with a sphere, I cut him in half en then in half, to have a semi-dome, and it doesn’t work, but when I take a curve and revolve it to become a semi-dome it seems to work.
The reason why I am doing the last split with the surface, is because I wanted to compare the panel method with my method, and because of my triangulation, the end of the hyperboloïd is like this. Maybe I am just asking to much…

Let me know if it’s not clear… I am so into it, it is sometimes difficult to explain simpel an clear

Your model was extremely well done - and useful! I hope you’ll forgive me for returning to my comfort zone and using my code. I was unable to split using your cutter, which seemed far too elaborate for the job it was being asked to do. The problems boil down to trimmed versus untrimmed surfaces. When you split a trimmed surface, there is still trimming information out there. You cannot triangulate properly because Lunchbox is looking at the original untrimmed information. The answer, I believe, is to triangulate an untrimmed surface that already looks the way you want.

What I did was created a portion of the hyperboloid using my initial method. If you look at the isocurves however, you see they are skewed so the next step was to extract the hyperbola profile and do another sweep. This gives an untrimmed surface with nice perpendicular isocurves. I projected a curve onto this surface and performed a split. The next step was to extract the edge data. This will be the boundary of the new, triangulated surface. Using Sweep 2, a new surface is created and it is still untrimmed but looks the way you want - I think. (you’ll have to play with the curve to get the exact shape you desire) THEN apply Lunchbox to this untrimmed surface. for the final result. Since I only used one quarter of the hyperboloid, the final result had to be mirrored. It might be possible to do it all with one curve, but honestly I don’t have the time.

This has been an interesting exercise and I hope I’ve been of some help, but I do have a life of my own - such as it is…Good luck!

1 Like

@akilli Great! This is what I was looking for! Thank you very much spending time on my drawing and you have been a big help to me!