My only critique is that I expect there is a lot of "cut off" when they fabricate these flat pieces from rectangular stock. (I assume the shape of the pieces are not optimized for nesting within one another when laid out on the original flat stock — picturing all kinds of odd, curved scraps.) Presumably the cutoff material is recyclable though.
EDIT: many of the examples look "sectional" where, I suppose, if you reduce the shapes down to enough sections you can make better use of sheet stock — the tradeoff of course being you need to add more joints requiring fastening and perhaps loss of strength.
>"In the end, we intend to relieve designers from worrying about fabrication so that they can focus on aesthetics and function"
I will argue that the foremost problem in the discipline of design is that many don't 'worry about fabrication' enough.
I think they agree with you, and that's why they're doing what they're doing:
> This is an important development because one of the biggest challenges in design is making sure that a shape is actually buildable.
The article is saying "we propose a solution to this problem".
A general solution to fitting flat material to 3d surfaces is available here:
This software uses dynamic programing to minimize the bending energy needed to fit a ring of 3d points.
Air ducts (mostly hidden in cavities or attics) could get some disruption from c-tubes. You'd glue the cut sections of flat polyprepylene into c-tubes. Not quite sure what the angle joiners would look like. They can't assume 90 degrees .. or can they for a subset of uses?
These are even more beautiful in real life. I don't know why they didn't show it, but the same lab has also built a structure of tubes made of bamboo strips that arch up and over a seating area into a dome. The tubes start at planters and there are vines growing up onto the structure. It's my new favorite thing in my neighborhood. Thanks EPFL!
They didn't show it because it doesn't use this C-Tube technique.
It looks like the dome could be related to the example they show of "optimized network" organization of C-tubes — if, that is, the structure that comprises the dome webbing is comprised of C-tubes.
The dome webbing is made of strips of bamboo, not any kind of tube.
This is not special or impressive. The video shows a triangle sketch following a closed curve, makes matching tangency seem like something special, then fast forwards to flattening the three faces of the resultant shape. It's like...uh... Yeah, so, possible forever, used regularly, what's the biggie?
You missed the whole point. The surfaces are developable, which means they can be cut from a flat sheet and curved into the shape, with no stretching or folding. It is special and impressive.
Does it mean the individual pieces can be forced into a curved surface? Or that the individual pieces will remain flat but can be joined with others in an angle to produce the curve?
In other words I guess my question is can the material itself be non-bendable, as long as its cuttable?
They show comparisons of the best current algorithms to their results, and theirs are better. Science moves forward incrementally.
Also, this algorithm makes a real physical difference in the world. Cool!
There are many more and interesting examples in the paper [1], which is open access.
[1] https://dl.acm.org/doi/10.1145/3730933
the paper also has a website with a longer supplemental video than the one linked to above, includes most of the examples from the paper but animated
https://www.epfl.ch/labs/gcm/research-projects/c-tubes/
My only critique is that I expect there is a lot of "cut off" when they fabricate these flat pieces from rectangular stock. (I assume the shape of the pieces are not optimized for nesting within one another when laid out on the original flat stock — picturing all kinds of odd, curved scraps.) Presumably the cutoff material is recyclable though.
EDIT: many of the examples look "sectional" where, I suppose, if you reduce the shapes down to enough sections you can make better use of sheet stock — the tradeoff of course being you need to add more joints requiring fastening and perhaps loss of strength.
>"In the end, we intend to relieve designers from worrying about fabrication so that they can focus on aesthetics and function"
I will argue that the foremost problem in the discipline of design is that many don't 'worry about fabrication' enough.
I think they agree with you, and that's why they're doing what they're doing:
> This is an important development because one of the biggest challenges in design is making sure that a shape is actually buildable.
The article is saying "we propose a solution to this problem".
A general solution to fitting flat material to 3d surfaces is available here:
https://laminadesign.com/
This software uses dynamic programing to minimize the bending energy needed to fit a ring of 3d points.
Air ducts (mostly hidden in cavities or attics) could get some disruption from c-tubes. You'd glue the cut sections of flat polyprepylene into c-tubes. Not quite sure what the angle joiners would look like. They can't assume 90 degrees .. or can they for a subset of uses?
These are even more beautiful in real life. I don't know why they didn't show it, but the same lab has also built a structure of tubes made of bamboo strips that arch up and over a seating area into a dome. The tubes start at planters and there are vines growing up onto the structure. It's my new favorite thing in my neighborhood. Thanks EPFL!
They didn't show it because it doesn't use this C-Tube technique.
It looks like the dome could be related to the example they show of "optimized network" organization of C-tubes — if, that is, the structure that comprises the dome webbing is comprised of C-tubes.
The dome webbing is made of strips of bamboo, not any kind of tube.
Can you share a picture of that arch?
I think it's this: https://bamx.epfl.ch/
Looks very cool indeed
Specifically this one:
https://bamx.epfl.ch/bamx-green-elysee/
...which is the most photogenic of them all.
This is not special or impressive. The video shows a triangle sketch following a closed curve, makes matching tangency seem like something special, then fast forwards to flattening the three faces of the resultant shape. It's like...uh... Yeah, so, possible forever, used regularly, what's the biggie?
You missed the whole point. The surfaces are developable, which means they can be cut from a flat sheet and curved into the shape, with no stretching or folding. It is special and impressive.
Does it mean the individual pieces can be forced into a curved surface? Or that the individual pieces will remain flat but can be joined with others in an angle to produce the curve?
In other words I guess my question is can the material itself be non-bendable, as long as its cuttable?
They show comparisons of the best current algorithms to their results, and theirs are better. Science moves forward incrementally.
Also, this algorithm makes a real physical difference in the world. Cool!