geodesic

Comment

Comment by Chris Mealing on October 24, 2009 at 9:29pm

I've now completed a video tutorial (my first attempt, so go easy on it!) for the definition, uploaded here:

I hope it's helpful, and I'd be pleased to receive any feedback on the tutorial itself before I add pt. 3 regrading the framework.

Comment by Niels De Temmerman on October 22, 2009 at 2:14pm

Hey Chris,

I am still interested in part 2 'return of the geodesic'. beware of sequels..
No seriously.
@Daniel: there are different projection methods and different subdivision methods which all yield a different solution; some have more length variation than others. there is an old book (seventies) called Domebook 2 whoich has some of these solutions.
I must say: I do not inderstand you explanation concerning subdivide-project-subdivide...
Could you elaborate?

Comment by Daniel Piker on October 20, 2009 at 5:08am

Very nice. I think maybe it is possible to get less variation in strut length though. I noticed when I made a simple geodesic definition a while ago that doing all the subdivisions first, then projecting tends to bunch the triangles slightly around the vertices of the original icosahedron.
If you were to subdivide-project-subdivide-project etc then I think you would end up with a slightly more efficient structure.

Comment by Chris Mealing on October 19, 2009 at 1:15pm

Haven't written it yet... actually I was waiting to see if anyone found it useful.

Hopefully the definition file and my discussion of some of the bits (below) will stall for a day or two while I get the rest of it done. Thanks for the interest!

Comment by Dmytro Lutsak on October 19, 2009 at 6:10am

When will you post part 2 of the tutorial ?

Comment by Niels De Temmerman on October 18, 2009 at 7:43am

Hi Chris,

thanks for your response. I've read what you have suggested, but let's just say I am not enough of a programmer to do this yet.
I think (not sure) that the problem might arise from the fact that my collection of end points, (which make up the geodesic grid of points) is a bit random. I mean that it is not a nicely ordered collection of points. I suspect that your generated collection of points arises from a subdivision of a surface, which might have some kind of order to it, which comes in handy when you use the VB script to arrange them in 3 lists, later used to connect the triangles. Am I right?

And since I do not know how to order the endponts of my line segments into an ordered collection, ready to feed to your VB script, I might have to take your route of subdividing a surface. I thought my method would work...but alas.

So just to be clear, there is no method to take a random grid of points and have them connected by lines in such a way that a triangular grid arises? I'll start a topic on this

Thanks for you help

Comment by Chris Mealing on October 17, 2009 at 6:50pm

I've uploaded a replacement definition - while trying to remember how I did this (!) I noticed I was picking up some of the points in a funny order, with the result that a third of the facets were flipped inward (it'd have showed up in a render sooner or later). geodesic.ghx

The "trick" of the lines is all buried in the vb code inside the "VB (DotNET VB Script)" component I renamed "vbGeodesic". Basically the visual basic code cycles through the 15 points resulting from the divide operation, picking them up 3 at a time in the right sequence to define the 16 facets that result from the divide (ie the component takes in a 15 point list and spits out a 16*3=48 point list). Hopefully the picture below helps. Once you figure out how to select the points in correct sequence, it's pretty straight-forward to do the cull and draw lines.

A related example is on pg 106 of the Grasshopper Primer. An even better example of what I think you're after is on page 32 of the "Algorithmic Modeling" document on the tutorial page.

Cheers

Comment by Niels De Temmerman on October 17, 2009 at 6:08pm

Yeah, the PDF is useful as a format. Thanks.
I guess making video is tedious, but then again it is easier to mimic afterwards. I especially like the bit where you rotate the 3 faces to make up the 10 first faces. clever.
I am really looking forward to part 2, where you will hopefully explain in detail how you make the 3pt surface of the grid points. I've tried to copy your way of doing it, by creating 3 cull patterns, but making 3pt surfaces from the points doesn't work. How do you control which point sgo in what list? I only end up with a flat surface, not a 3D-facetted surface.

But actually, I do not need that surface, but only straight connecting lines between a triangular grid of points, to make up the geodesic grid. I am quite surprised that there is not a simple function within grasshopper that does this...or is there?

Can you have a look at my definition? I have marked where the problem is. I have a geodesic grid of points, but how to connect these by straight lines? Any suggestions would be welcome.

Nielsgeodesic_new.ghx

Comment by Chris Mealing on October 17, 2009 at 5:37pm

Thanks for the kind words! I've uploaded a pdf of my first attempt at a tutorial (still trying to figure video out!). Hope it helps some. I'd appreciate hearing whether this format is useful.
Geodesic Part 1s.pdf

Comment by Niels De Temmerman on October 17, 2009 at 5:20pm

Stunning. absolutely beautiful. It works now, thx for the tip.
I am trying to analyze what you have done to create the triangular grid. I can generate a triangular grid of geodesic points (lying on the sphere) but connecting them by lines, thus generating a geodesic grid, is beyond my skills at the moment...
How can I upload definitions in this message?

• ‹ Previous
• 1
• 2
• 3
• Next ›
• Page