Geodesic Intersections

Hey all,

I've been doing some work with free-form grid-shells using the geodesic components from the latest releases of g-hopper and I was wondering if anyone else has run into the same hurdles. After creating the grid shell curves on the surface, I can't seem to intersect them, no matter what I try.

But, if I redraw the same curves with a closest point on surface and curve on surface, and flatten only ONE of the curve lists, they intersect just fine. Any thoughts on this?

Also, does anyone have any good ideas for a means by which to organize this collection of points? I think I've seen some solutions to this before using VB Script (which, unfortunately, I'm not too familiar with). Any help is appreciated!

Attachments:

test_srf.3dm, 77 KB
geodesic_working_04.ghx, 271 KB

Replies to This Discussion

Permalink Reply by Mårten Nettelbladt on March 10, 2010 at 2:41pm

Hi,
In this thread http://www.grasshopper3d.com/forum/topics/geodesic-dome Arthur Mamou-Mani and Jon Mirtschin describe som techniques for free form geodesic structures (scroll down a bit). Check out some interesting stuff by Jon Mirtschin at his site http://geometrygym.blogspot.com.

Permalink Reply by Patrick Delorey on March 10, 2010 at 7:27pm

Thanks Mårten. I had seen that discussion previously, and modeled the grid shell method off what Arthur was working with. Still, the problem lies specifically with the intersecting points. Interestingly, I've also heard that the problems that exist with the Rhino ShortPath command are persistent in Grasshopper.

Permalink Reply by Arthur Mamou-Mani on March 11, 2010 at 3:21am

Hi Patrick,
I saw your comment and tried applying the script to your rhino model but couldn't make it work.
The only difference with the model i was working on and yours is that you have kinks at the base and i was working with a curved base.
Here is the script attached if someone manages to make it work please let me know.

Attachments:

geodesic_working_05.ghx, 448 KB

Permalink Reply by Arthur Mamou-Mani on March 12, 2010 at 11:58am

Hi Patrick i tried applying it again with my form and it didn t work either i will need to check the script it might work differently on the new version of GH.

Permalink Reply by Patrick Delorey on March 12, 2010 at 7:38pm

Thanks Arthur. I appreciate the help. Also, could you recommend a good introduction to VB Script in Grasshopper?

Permalink Reply by Arthur Mamou-Mani on March 14, 2010 at 9:06am

Hi Patrick,

I ll be taking the Rhinoscript workshop in London in May with David Rutten so if you re planning on travelling you know where to go...
In terms of book if you completed the Grasshopper Primer then have a look at this conversation:
http://www.grasshopper3d.com/forum/topics/vbnet-book-recommendation...

Permalink Reply by Jon Mirtschin on March 13, 2010 at 2:03am

Hi Patrick

Grasshopper isn't able to find the intersections due to the tolerance of the curves created by the Geodesic function. If you relax the acceptable tolerance considered acceptable for an intersection to twice that used by the Geodesic curve calculation (which is adopted from the Rhino document tolerance setting), then the intersections are found. I did this by creating a little c# component with an intersection function relaxing the tolerance. Obviously the curve on surface works to a tighter tolerance than the Geodesic function (Which is iterative), that you were able to find the intersections using that means.

Hope this helps,

Jon

Attachments:

geodesic_working_04.ghx, 298 KB

Permalink Reply by Patrick Delorey on March 13, 2010 at 12:31pm

Jon

Thanks so much for the help, and especially the explanation! Just further proof that I really need to learn more about C# and other similar tools.

The only problem with that particular method is that it sacrifices the data tree structure. The ultimate goal with this particular definition is to develop a script capable of generating all the fabrication data for a constructable grid-shell (i.e. cnc router cut-files for plywood sheathing, etc.) I figure the biggest challenge is going to be sorting the points in some organized fashion, and I had assumed that preserving the data tree structure might make this more manageable. Any thoughts with regard to this are appreciated, but you've already been a big help in understanding the problem.

Thanks again!

RSS