Grasshopper

algorithmic modeling for Rhino

Hi there,

I have created a geodesic dome using the geometry gym plugin.

When I offset the triangulation, half of the triangles on the sphere are offset negatively and the other half are offset positively.

The images I've uploaded hopefully show the problem.

Is anyone able to suggest how I make all the triangulation offset in the same direction?

Any help on this issue would be hugely appreciated.

Thanks ever so much!

aL

Views: 2032

Attachments:

Replies to This Discussion

grasshopper file

Attachments:

Hopefully the solution will help solve my loft error, I' want to achieve this attached extrusion all over the sphere eventually!

Attachments:

It's possibly something to do with the directions of the curves output by the Geometry Gym component. Perhaps the directions of those curves change from one half of the dome to the other. You could check this in Rhino - Analyze >> Direction.

I don't know if there's any way to align curve directions, you might be able to break the curves down into points and then re-create aligned, clockwise polylines so that the offsets are all the same direction.

Have you tried asking the Geometry Gym guys?

That's a great call, I've just emailed them. In the meantime I'll have a go at your suggestion. Thanks for the input.

The other way to look at this is to scale the curves about their centres and then move them along a vector between the centre of the dome and the centre of the triangles. Then loft.

I think the Area component gives you the centre of the triangle, so you should be able to get the move vectors

Weaverbird is also great for this sort of thing:

Attachments:

Hi Luca,

Is this the type of result you want?  I calculated a plane on the centre point of each polyline to control the offset.  It's also better to "join" surfaces to make a brep than calculate a boolean difference (which is much slower).

Hope it helps,

Jon

Attachments:

Thanks for all the input everyone, I tried your last suggestion Martyn, which worked and I've attached here if anyones interested.

And I'm just downloading Weaverbird to check it out, so thanks for the tip Daniel.

Thats what i was after, thanks, Jon. I didnt't know there was an alternative to boolean difference, I use boolean difference commands all the time! I've studied your definition and think I've worked out how your doing it differently, I'll try and apply that technique in future.

Attachments:

While we're all on this subject, does anyone know how to handle this with only the points as inputs, i.e.without the polyline curves that define each panel?

Triangulating a set of points when they can't be easily projected onto a surface becomes a lot harder, though there are ways of doing it - sometimes you can use Proximity3d, remove duplicate lines, then apply Weaverbird's 'Mesh From Lines', though this approach will only work when the points are fairly evenly distributed, so you can set the proximity threshold to avoid unwanted edges.

otherwise you could use Dave Reeves' ball-pivot implementation

RSS

About

Translate

Search

Photos

  • Add Photos
  • View All

Videos

  • Add Videos
  • View All

© 2024   Created by Scott Davidson.   Powered by

Badges  |  Report an Issue  |  Terms of Service