Hi there folks, just sharing some architectural geometry work I did a few years ago, which has come close a couple times to being realized in an actual building. I believe it was novel at the time, and I'm curious to know if anyone has done anything similar since.
It's an approach for constructing free-form architecture using piece-wise planar tiles. When the shape is convex, the problem is very simple. When you have negative Gaussian curvature it gets tricky, and that was the problem I was trying to solve.
The shape is first segmented using any plausible scheme -- in this case, I used a relaxed voronoi tessellation. Then I optimize the vertex positions so the regions are as planar as possible, while remaining as close to the original surface as possible. I used L-BFGS for this latter step. The result in the pic below is typical.
Keen to hear anyone's thoughts on the matter. I currently work in R&D at Weta Digital, and haven't had a chance to spend any time on architectural geometry for a while, but I thought I'd dust this off and see if anyone was interested.
Cheers,
Kevin
Oct 30, 18 9:00 pm
Block this user
Are you sure you want to block this user and hide all related comments throughout the site?
Archinect
This is your first comment on Archinect. Your comment will be visible once approved.
Planar, semi-regular tilings of free-form architecture
Hi there folks, just sharing some architectural geometry work I did a few years ago, which has come close a couple times to being realized in an actual building. I believe it was novel at the time, and I'm curious to know if anyone has done anything similar since.
It's an approach for constructing free-form architecture using piece-wise planar tiles. When the shape is convex, the problem is very simple. When you have negative Gaussian curvature it gets tricky, and that was the problem I was trying to solve.
The shape is first segmented using any plausible scheme -- in this case, I used a relaxed voronoi tessellation. Then I optimize the vertex positions so the regions are as planar as possible, while remaining as close to the original surface as possible. I used L-BFGS for this latter step. The result in the pic below is typical.
This blog post has some more information on the technique:
https://methodart.blogspot.com...
Keen to hear anyone's thoughts on the matter. I currently work in R&D at Weta Digital, and haven't had a chance to spend any time on architectural geometry for a while, but I thought I'd dust this off and see if anyone was interested.
Cheers,
Kevin
Block this user
Are you sure you want to block this user and hide all related comments throughout the site?
Archinect
This is your first comment on Archinect. Your comment will be visible once approved.