Dimple Principle Paneling on Closed Surfaces

Replies to This Discussion

Permalink Reply by tomi on June 2, 2009 at 7:15pm

as i was thinking about this, this is how i would do it if i could:

1. take the object (geometry) duplicate and uniform scale it up
2. then (and this is what i can't do in grasshopper) make a grid of the surface of the duplicated form, possibly with various methods (patterns)
3. every point in the grid gets a sphere assigned
4. boolean difference all the spheres out of the original geometry object

actually seems pretty easy, but i can't do the grid stuff :/

Permalink Reply by Dirk Anderson on June 2, 2009 at 8:12pm

Tomi, try the 'Divide Surface' component to create a number of regular U & V points on the nurbs surface.

If you want to create variation in the grid, use a graph mapper or a differing counting pattern to create an xyz point grid and apply to surface using the 'Evaluate Surface' component. You need to make sure the extents of the grid correlates to the surface domain (e.g. min grid pos: (0,0), max grid pos: (10,10) / surface domain U: (0,10), surface domain V: (0,10)).

Permalink Reply by Damien Alomar on June 2, 2009 at 8:57pm

In RhinoScript 101, there is an example that does something like this on page 62-63. A while ago (on the old forum) there was someone who put together a GH version of the script. I thought I had that definition, but I can't seam to find it. You may try to see if you can use the script as a starting point and assemble this in GH. HTH

http://en.wiki.mcneel.com/default.aspx/McNeel/RhinoScript101.html

Permalink Reply by tomi on June 3, 2009 at 5:36am

thank you both very much!
i will look into your suggestions and post my results (or my failure) here :)

Permalink Reply by tomi on June 4, 2009 at 5:19am

(the reply structure in this forum is pretty confusing, should i reply to each persons post or only to the last for everyone? or to my own post? O_o)

first i tried your approach dirk, problem is i need even spacing between the grid points, which i can't achieve with the divide surface component (having as many grid points as maximum grid position in each direction is not helping either, or maybe i didn't understand you correctly) have a look at the picture i attached, even spacing is not given. i will have to combine damiens tip with yours somehow:

the algorithm proposed in the rhinoscript 101 is for evenly spacing circles on a sphere, which is pretty neat, the "problem" is i'd have to adapt it for any closed geometry, which is a 'slightly' significant step up in the complexity of that algorithm, on a sphere you can precalculate how many circles will fit on the surface, with more complicated shapes i don't see how i could go about it. :/
i have found the 'circles on sphere' grasshopper translation here
and i will try to use it, let's see what i will come up with.

any hints, tips and suggestions are welcome!

Attachments:

• dimples1.PNG, 346 KB