3-D printing has become the rallying cause for a rising generation of designers, engineers, and architects. There seems to be few limits to what the technology can do or what range of products it can spawn, from lampshades to lunar bases. Amid all the hype, however, it’s easy to neglect one key factor: Printing capabilities are directly wedded to the size of one’s printer. As home printers become more readily available, the size of their printing beds shrink. Any budding designer with a desktop 3-D printer can create an intricate scale model of the Millennium Falcon, but what about something as straightforward yet functional as a chair? It simply won’t fit inside the printing box.
MIT-based researchers and instructors Marcelo Coelho and Skylar Tibbits teamed up to tackle this very problem. Working under a grant from Ars Electronica, the pair conceived of a whole new way to do 3-D printing. Hyperform is a new strategy for designing and printing large objects irrespective of a printer’s bed size. So not only can you print out that chair at home, you can also print a table, bed frame, and everything else you need to furnish a bedroom.
The solution is breathtakingly simple. By merely folding the object you want to print, you can jig it to fit into a small-scale printer. In Tibbits and Coelho’s project, the object is rendered in 1-D--a line--and endlessly folded into a space-filling curve proportioned to the printer’s cubic dimensions. (The designers partnered with Formlabs and iterated the process using a Form 1 tabletop printer.) When the object is exhumed from the printer bed, it doesn’t at all resemble its final shape. Rather, it’s a dense cluster of thin but sturdy polymer links packaged in a three-dimensional puzzle that can be intuitively assembled.
The chains are programmed with multidirectional notches, so that they can be latched together at right angles. Assembly is quick because each chain can only bend in the way it's designed to, thus removing a large obstacle that plagues most 3-D-printing ventures. The final product, then, will look exactly as it does on your computer screen but will be structurally sound enough to stand on its own in physical space. In the process, Tibbits suggests, scale becomes virtually, if not entirely, irrelevant. (In one test, the cohorts printed a 50-foot-long chain that they proceed to hang from the roof of a laboratory building.)
“It kinda grew up out of the idea that if you wanted to take something that was very large and wanted to compress it down into a very small bed size, how do you displace that density,” he explains. The problem of displacing, or as Coelho later elaborates, “transforming” density is integral to understanding how Hyperform works and what it’s great potential really means.
Coehlo’s colorful terminology isn’t just lyrical. He earned his Ph.D. at MIT Media Lab’s Fluid Interfaces Group, where he completed a thesis on shape-changing textiles and other materials that could transform themselves upon electrical prompt. Similarly, Tibbits, an architect, designer, and programmer who holds a master’s in computation and another in computer science from MIT, has developed a series of self-assemblage prototypes he calls 4-D printing, which was a finalist in this year's Innovation By Design Awards. With 4-D, objects are encoded with assemblage information so that when they are pulled out of the printer, they begin to build themselves, without wires or electricity. “We’re both interested, of course, in folding and in programmable materials,” Coelho tells me, speaking about how the two designers came together over Hyperform. “It was a pretty obvious connection.”
As obvious as the partnership seems, so are the inspirations behind the project easy to identify. Tibbits characterizes folding as a “universal strategy,” a technique that can be applied across all scales, nano- and human alike. But he and Coelho also acknowledge that folding isn’t innovative in itself. In fact, the basis for their computational folding can be traced to recent experiments by other MIT researchers, namely, the Milli-Motein project by Neil Gershenfeld. The latter developed robotic programmable modules that could be configured with other modules and hold their position even in the absence of electrical power. Both Coehlo, a former student of Gershenfeld, and Tibbits, who collaborated with the researcher’s Bits and Atoms team, had extensive first-hand experience with 1-D folding chains.
Even so, during the course of his thesis work, Coehlo says he never considered 3-D printing as an application for his shape-changing studies. 3-D printing entered in only after the duo had taken a closer look at the technology’s supposedly revolutionary credentials. In the last few years, a whole market for consumer 3-D printers has emerged banked on the promise that users can create anything and everything with a portable printer like a MakerBot or Formlabs Form 1. That claim, however, breaks down under even cursory scrutiny.
In exploring the problem, Tibbits and Coelho identified two common strategies for most 3-D printing. The first is to print large objects in whole, while the second prints large objects piecemeal, in smaller components that have to be hand assembled. There are drawbacks to both, not least of which being the size of the printer itself. “Some people take the perspective that we should be making bigger and bigger machines to make bigger and bigger object,” Tibbits says. “But then there’s the problem that you build one skyscraper machine to build another skyscraper-sized thing.” There’s a point, he suggests, where the exercise becomes inefficient and self-defeating.
There is also the sticky question of assemblage. Since there isn’t a commercial 3-D printer large enough to spit out huge objects like, say, an architectural installation or even a car, you have to print these things in components. Though individual pieces can be numbered prior to printing, “you [still] have no idea how to put together those parts,” Coelho says. “In most cases, you have this crazy nightmare on your hands.” Tibbits describes this as “brute-force assembly,” a hellish procedure that Hyperform eliminates.
Because the printed chains are designed with joints, the user just has to latch them into place. The exact pairings are encoded in the material itself, which is to say that the chains will snap only in the direction and manner that they were engineered to. It’s entirely intuitive, not unlike snapping two LEGO blocks together. The designers developed a chandelier to illustrate the assemblage process and gesture towards the project's industrial design applications.
For now, Tibbits and Coelho have only demonstrated one path Hyperform can take. The system can be used to handle 2-D and 3-D geometries, which would involve much more testing. Where the 1-D foldable chain generates a structural shell without an envelope, additional dimensions could yield much more fractal results.
The team intends to open up the project to an online community of coders, designers, and architects. They are the key to finding further applications for Hyperform, Tibbits says. “We were conscious in taking only a first pass at this problem. But it’s a very interesting way to kickstart other people to take on this challenge and find new solutions.”