grasshopper surface patterning

hello everyone

lately I've faced a big problem that the more i search about the answer the more i get confused... and I've searched so many sites and references but I didn't find the answer i was hoping to find...

It's the problem

Lately I've decided to design a building with grasshopper and i started to learn but no matter how many skills I've achieved I have one big gap in my knowledge that I'm desperate to fill

how can we generate a dynamic pattern on a let's say curvy surface ..

I attached a sample pic so maybe i can find the answer here because this part of the grasshopper is very important to me...

Ty for your answer

Attachments:

c8e97c81e11c135d22023d6ea1e703d2.jpg, 21 KB

Replies to This Discussion

Permalink Reply by peter fotiadis on September 13, 2016 at 5:54am

The easiest way (Tip: AVOID at any cost to offset a curve into a surface > takes 2++ years to finish):

1. Create a planar tree with the triangles (or some other polyline) - Planar due to Tip.

2. Vary the triangles (still in 2d - see Tip) via some attractor(s) - push/pull etc etc.

3. Morph the modified result into the target surface in question. That yields curves rather than polylines.

4. Split the surface (using as List of cutters the morphed curves) and get the biggest resulting Brep.

Permalink Reply by farid abbasi on September 14, 2016 at 12:20am

Ty so much Bro . Thanks so much

Permalink Reply by Daniel González Abalde on September 13, 2016 at 6:47am

Almost always the trick is to link data.
A) What generates the dynamism.
Z) What is the pattern.

Well, in your case, for example:
A = distance to a point. I want the farther away, the pattern smaller.
Z = size panels.

So to arrive from "A" to "Z", we need to create smaller steps linking data, for example:
- We need to get distances. One for each panel. How do we measure the distance from a point to a panel? For example: from the center point of the panel to the point attractor or proyecting then attractor point to geometry.
- We have distances, one for each panel, but are values useless to scale. We need to modify these values, such that are scaling factors, we can remap to a domain between 0.95 ~ 0.25 .
- We scale panels from its center with the new values.

Attachments:

PanelAttractor.gh, 13 KB

Permalink Reply by farid abbasi on September 14, 2016 at 12:20am

Ty for The file and the time u put in typing . helpful . Thanks

Permalink Reply by Hyungsoo Kim on September 13, 2016 at 11:56am

Hi.

In this case, using "Copy Trim" isn't a bad way of mapping trimming data of one brep to another.

Attachments:

TriGridPatternMapping.gh, 21 KB

Permalink Reply by peter fotiadis on September 13, 2016 at 1:39pm

Indeed but since a facade is made via some sort of modules ... I would suggest the orthodox way.

That said ... this is also highly Academic since ... well it's rather obvious I guess (requires a totally different approach: first made some "even" (or as "even" as possible") collection of facade panels [Kangaroo 1/2] and then map each planar one to the new position).

Anyway: a very challenging issue that one (in real-life).

best

Attachments:

map_v1.gh, 122 KB

Permalink Reply by Hyungsoo Kim on September 13, 2016 at 2:01pm

Indeed, and sometimes we can see more brutal parametricism by using concrete.

Permalink Reply by peter fotiadis on September 13, 2016 at 11:51pm

Deals OK VS some "trivial" paneling issues (leaks, "even" panels etc etc) ... but is a big nightmare (case: exposed concrete) with regard thermal efficiency (unless you opt for a dual skin).

On the other hand using some STO system you can "fake" the appearance and save the polar bears.

Q: Who cleans the glasses?

Permalink Reply by farid abbasi on September 14, 2016 at 12:23am

Yeap , Who cleans the Glasses then :)) . thanks btw

Permalink Reply by farid abbasi on September 14, 2016 at 12:21am

Wow i kinda feeling that it will solve the problem , Let me try then . Thank u so much

Permalink Reply by Deer on September 14, 2016 at 12:36am