Mapping regular (equal) diagrid panels for a freeform surface

Hi I'm new to this forum and still a novice in grasshopper.

I have a freeform surface which I hope to map equal sized diagrid panels onto it.

I tried creating a mesh in Rhino and use TriangulateMesh however the dimensions of the triangulations all varies and I was wondering if it is possible to minimize the number of types of triangles. I understand that given my form it would be impossible to have perfectly regular triangles but I am hoping to reduce it for ease of fabrication.

I have attached my model and pictures of the effect I hope to achieve.

As I am aware that the above regular triangulation may not be possible to achieve on my freeform model, I am also trying out another method for cladding the interior roof. The second picture shows this effect I am trying to achieve (but in terms of triangles). I am not too sure of how I can map the grid so that it is contained in the roof surface of my model.

I would really appreciate any input regarding these two effects I am trying to achieve with my model.

Thank you in advance!

Attachments:

140721 Clubhouse.3dm, 394 KB

Replies to This Discussion

Permalink Reply by Damon on July 25, 2014 at 6:45am

A few thoughts.

1. Are you sure the first image is an example of "equal sized" triangles? I doubt it. Unless you have developable surfaces, which your forms don't look like.

2. Why do you want equal sized triangles? Aesthetics? Constructability/financial? Frankly, much of the diagrid construction I've worked on and know about doesn't worry about same sizes but rather maximum and minimum sizes, and tolerances.

3. You could aim for families of sizes: understand the tolerance of the system as constructed and group similar panels by size and shape so that you could potentially have 10 of A, 8 of B, 4 of C, and 2 of D or something.

4. I highly recommend looking at Evolute Tools. It has all sorts of optimization of meshes that would allow you to control of size, shape, etc.

5. Finally, I don't think you want a simple mapping of a diagrid to your surfaces. I say that because of the way your form pinches down to a point. One thought would be to create a larger surface that has a more rectangular boundary, panelize it, then trim the panels. Then all the internal panels would be quite similar and you'd only have irregular edge panels.

Just a few thoughts.

That's my two cents!

Permalink Reply by shwok on July 28, 2014 at 7:27pm

Hi Damon,

Thank you for your reply! In response to your thoughts,

1. Indeed I doubt the first image has equal sized triangles thank you for pointing that out.

2. I want equal sized triangles mainly for constructability as it is costly to fabricate and manufacture triangles of different sizes, and there are financial limitations.

3. Having families of sizes, maximum and minimum sizes, tolerances is an excellent proposal - I was just thinking about that myself as well, except I have no idea how to get started on it (which was why my attempt to start small and simplify the problem at hand). Could I possibly take a look at your grasshopper code for the diagrid construction which you have worked on previously?

4. Thanks for your suggestion! I was looking at Panelling Tools to tackle this problem, I shall look into Evolute Tools as well (:

5. Thank you for your suggestion! I shall try my hands at producing a rectangular boundary for this - the converging end is posing many problems.

Thank you again, your suggestions have been very helpful!

Permalink Reply by Damon on July 29, 2014 at 6:42am

I'll just reiterate that equal sizes and shapes are unlikely unless you start with the right kind of surface such as a developable one. To my eye, you might be able to rebuild and approximate each surface in your form as tangent conic sections.

Maximum size is definitely important, since materials only come in certain sizes, and over-sized panels are often much more expensive. Minimum size isn't all that important, but the "pointy-ness" of a triangular panel might be: glass or precast concrete get fragile, for example. Minimizing the number of panels helps, so getting the absolute largest possible panels available and reducing the edge conditions.

Unfortunately, I cannot share with you my definition because it would require a release form, etc. However, I'm happy to continue to help with whatever you come up with.

Last thought: Have you considered quadrilateral panels? If you do use conic sections, you could get quads, which would reduce material waste. Again, to my eye this form looks like it could be divided that way.

RSS