Pictures of 3D-printed houses keep popping up in the news. I've written about three such projects before for Wired, including Enrico Dini's impressive D-Shape, a gantry large enough that it can print out gazebos. There are other examples of such systems under development that work on similar principles, like Contour Crafting at the University of Southern California, or FreeForm Construction at Loughborough University. Nasa is even working on robots that may be able to construct bases on the Moon by melting moon dust into blocks and walls using lasers.
It's not a wacky gimmick, in theory. Developing societies across the world struggle to cope with the demands of urbanisation: China's urbanisation rate will reach 60 percent by 2015; a 2007 UN report found that between 2000 and 2030 sub-Saharan Africa's urban population could be expected to double, while at least 72 percent of the urban population was living in slums. A way to quickly produce homes that are more livable than slum shacks would be a valuable tool in combating poverty. (And as for moon bases, laser sintering is a much better idea that flying concrete all the way up there.)
These printers are, basically, scaled-up versions of desktop models, and they work the same way - a nozzle, on some kind of robotic arm, is programmed to follow a design which separates a three-dimensional objects into a number of two-dimensional slices. Thin layer upon thin layer of material (usually plastic on desktop, usually concrete in construction) goes down until it's built up the full object.
Anything that'll spurt out of a nozzle and set hard will work, but some materials are better than others for different uses - chocolate's great for 3D-printing food, for example. For buildings, the normal material (or "aggregate") is usually a kind of concrete. The nozzle moves back and forth, laying down material for the first layer, then it moves up (say, 5mm, or 10mm) and lays the next layer, and the next, through to the very top. Voila, a house.
The problems that researchers are having with scaling 3D printing up to the scale of houses should be pretty obvious - concrete isn't very strong on its own. A house that's 3D-printed might stay up (and some of Dini's structures are certainly impressive) but they're nothing on concrete houses built with such boring, traditional things like reinforcing rebar. Even a wooden frame is better, because then there's the possibility of having a second floor.
Nevertheless, these systems are generally said to be able to build the frame of the house in roughly two days at top speed if you're after something like a small bungalow, or maybe doing 20 or so larger buildings a year. And, because the walls and so on are done in one go, there's the potential to design certain things into the structure - like routes for ducts, or piping and wiring routes - that are fiddly, and have to be attached to a building's frame during a normal construction.
That's quite important, as the things that take up most of the time when building small homes is the fiddly, small stuff - wiring, windows, fixtures, fittings, plastering, that sort of thing. It's also quite cheap - again, in theory. There's just a problem in that it acts a replacement for the bit that's already quite cheap and quick when it comes to building homes. In rapid-build affordable housing projects the fiddly stuff could be cut back on, but that would be somewhat self-defeating.
There's quite an illustrative lesson from history that we can rely on here. Architect Wallace Neff was famous in the 1920s as a designer of mansions for the stars of Hollywood, and became instrumental in developing southern California's distinctive architectural style. However, in the later stages of career he tried to turn his hand to the problem of affordable housing. America's post-WII population boom demanded cheap housing that could be built quickly.
His solution, inspired by the resilience of bubbles of shaving foam in his bathroom sink, were "bubble houses" - concrete homes (using low-density, high-strength concrete called gunite), built by spraying a mix of water and cement at high speed over a large balloon. In 48 hours the concrete would be dry, the balloon could be deflated and dragged out of the front door, and there would be a perfectly solid and large building, ready to be used as a home.
(If you've never listened to 99% Invisible, by the way, now is a good time to start - it's one of the best podcasts available, and the episode about Neff's bubble houses is superb and detailed.)
(There's also a good article on Slate that features some wonderful pictures of some of Neff's bubble houses under construction.)
Neff loved and adored his homes, moving into one himself and dreaming of building hundreds of thousands of them. The public response to his idea was less enthusiastic, and only a few hundred were built. Only one survives in the US today, in Pasadena - the rest were demolished.
While that last bubble house in Pasadena has become something of an architectural landmark, it is a radical design even by today's standards. For all the internal space, the houses left people feeling cold, as if living inside a sculpture. You can't hang a picture on a curved wall, you see. It's hard to fit conventional furniture in without leaving dead space. Sound travels strangely, bouncing off the roof in unexpected directions. The atmospherics are off; it doesn't feel like a home, perhaps the most crucial design requirement of a house.
Developed countries have had huge problems getting people to love mass-produced, pre-fabricated, high-density concrete buildings, which went up in their thousands in the post-war period as a solution to a population crisis. The reasons for their success and failure were complex - just as much down to social and political factors, like aesthetic taste or slashed maintenance budgets, as structural quality - but
3D printing does what it does cheaply and quickly by skipping over or ignoring construction materials and designs that a lot of people like. Wooden floors with a bit of give, for instance, or brick walls, or wooden skirting boards. These are simple additions, but they are ones that require time (and the labour of actual people) to complete. The more you offer those trimmings, the more the houses move away from being cheap and quick to make, and more towards normal houses. Once you're there, you might as well just build normal houses.
These printers are also rubbish at scales normal for mass production - it's much more suited, on a desktop scale, for making prototypes, or for enabling delicate designs that might otherwise be made by hand. This applies just as much to housing as it does to trinkets, where pre-fabs are a perfectly fine method of building big buildings (even big, tall buildings) quickly, and especially in cities, where building low-density houses is often a luxury only afforded to the wealthy.
If 3D printers have a home in architecture, it will be as a tool of self-builders. Just as 3D printing is potentially great for democratising product design, it's something that - combined with open-source architectural blueprint sites like WikiHouse - would suit the architecturally ambitious just great. It's not hard to imagine Grand Designs featuring some bright-eyed young couple who have rented a 3D printer sometime in the next decade, lugging it out into the countryside where they plan to lay down some wavey-curvy mansion that looks like something out of Buck Rogers.