algorithmic modeling for Rhino


I did a sunlight analysis whith ladybug and i internalised values to create a delaunay mesh according to those values. My aim is to have larger triangles with less relief on sunny parts, and smaller triangles with more relief on shadow parts.

But I see two problems :

- My mesh is not clean enough, there are some stretched triangles on the borders.

- And the relief is not proportional to the size of the triangles, wich is a problem for the final design.

Can someone help me with this please. I don't know how to figure it out..

Thanks a lot

Views: 192


Reply to This

Replies to This Discussion

If I understand your issue, it's because the Delauny Mesh will will create all triangles it can (even very acute ones close to the border).

If you pull some of the closest points to the rectangular border for the region this should give you a cleaner result: 

This chooses which points to move based on proximity to the boundary by a ratio, but you could use a fixed value for the Y input of the expression.


Thanks for your quick answer.

Your example is better than mine, but there is still a square border that i would like to avoid like the script here attached. In this script, the faces are not forced to touch this square border and i think the design is more efficient in this way.


Not sure I quite follow, this example does much the same, except it populates the region and the outline, and uses these (along with the control points of the rectangle as above) to create the mesh... If you look the faces are still touching the square unless I'm missing something? The mesh in that example also has a lot less points.

If you want to use the mesh you had before I can think of two quick options:

- explode the mesh and cull by a minimum area requirement (this might cause 'holes' in your mesh though) and re-weld the mesh.

- explode the mesh and cull by distance from the perimeter curve. 

However, speaking with my engineer hat on, a planar, straight border is always easier to construct and details (even if it is more boring!)

Out of interest, tried applying these options to your script (requires Mesh+ plugin) - neither look like they work to well at first go I'm afraid! Biggest issue is the number of points and severity of change in extrusion between the points. Good luck and have a good weekend!


Thanks again for your quick answer. Unfortunately, i can't manage the size of the triangles according to different values anymore.. I tried with around 200 values instead of 1050, but i can't figure it out with your example. I think it was more efficient with the first script (i mean the triangle scaling differences). On the other hand, your is probably better to avoir "bad triangles" on the border. Maybe i have to find a way to mix both scripts..

I will try again and we'll see.

Otherwise I have another volume in my 3D, and i would like to do something similar as the picture attached. Do you know what is the best way to do this, still according to sunlight values? Values should controled the dimensions of the "hats". You can find my complete values, and the same values simplified with the geometry in the script here attached.

Thank you very much if you can have a look.

Have a good evening.

Thanks a lot


Hi Joe!

Do you know why there is no more sizes difference between triangles with high or low values with you script? I can't figure out this problem..






  • Add Photos
  • View All

© 2018   Created by Scott Davidson.   Powered by

Badges  |  Report an Issue  |  Terms of Service