3D printer could build a house in 20 hours
Ron Czapala
Posts: 2,418
http://news.yahoo.com/blogs/sideshow/3d-printer-could-build-house-20-hours-224156687.html
An engineering professor, Behrokh Khoshnevis, at the University of Southern California, is really thinking big: He has figured out a way to build housing with a giant 3D printer.
Here's how it would work, according to the blog Pop Sci. The apparatus, instead of being the size of your typical laser printer, would actually be somewhat bigger than the house it would build through a concrete layering system called Contour Crafting.
An engineering professor, Behrokh Khoshnevis, at the University of Southern California, is really thinking big: He has figured out a way to build housing with a giant 3D printer.
Here's how it would work, according to the blog Pop Sci. The apparatus, instead of being the size of your typical laser printer, would actually be somewhat bigger than the house it would build through a concrete layering system called Contour Crafting.
Comments
This is going to be TOO cool!
Thanks for sharing. That was very interesting.
Bruce
We'll have to go through a bit of a realignment in what we think is a house. Very interesting.
http://d-shape.com/tecnologia.htm
I am a bit dubious about the 3D printer approach for several reasons.
The only actual film of actual machines in the TED talk shows unreinforced (or, OK, fiber reinforced) concrete being laid down. Everything else -- window and door lintels, reinforcement, let alone plumbing and electrical -- is all pie in the sky. And even in my system, with the building being made in say 5 meter segments, the number of electrical and plumbing joins needed was looking like it would become a problem. Joins are expensive and failure prone. In the computer animation they show concrete reinforcement being screwed in place a few centimeters at a time. How much does it cost to fabricate and thread all those segments, even in a factory, compared to manufacturing a 3 meter rebar rod? And how do you reconcile that with organic curved shapes such as the Iranian buildings shown?
I looked very, very hard at materials and it is a plain fact that materials are expensive. High tensile strenght accelerated flowable (small aggregate only) concrete is very expensive compared to normal concrete, which is only cheap if you can somehow use very little of it -- but normal concrete use is only practical if it's very, very massive. Fiber reinforcement doesn't start to reinforce until the concrete cracks, allowing the fibers to begin asserting their tensile strength. I looked very hard at papercrete, and even invented a compressed form superior to anything I've seen or heard about, but paper has serous issues with a wet environment no matter what you do to it. If you look at the amount of material it takes to enclose a house-sized space and form a rigid shell, even ignoring all structural considerations just acquiring and moving the material around is expensive. It's really, brutally expensive. And all these 3D printer schemes tend to use finicky, expensive printing medium in large amounts.
The gap between what the TED guy dreams of and what he actually shows might be bridged by economies of scale and mass manufacture of the segmented service components, but there is a great deal of gap there to be bridged.
I liked how he spoke about the labor impact. People will always be part of the process, just moved into different roles.
-- Gordon
I got the impression that this is more to address mass construction than the one-off home. As for rebar, I think it's just a choice he showed threaded segments. There are CAD automated wire- and tube-bending machines that also weld joints. Once you start getting those into the process most any automation is possible, if the scale warrants building the machine to do it, as opposed to having humans accomplish the task. I have a feeling things like rebar, plumbing, finishing work, and electrical will be among the last to be automated. That still leaves foundation, framing, and roofing. Humans come in and fill the gaps.
A demo showing the end-to-end process is more compelling, especially for a TED talk, though less realistic for the near term. Consider the audience the video was shown to.
-- Gordon
Despite its hollows the actual wall shown being assembled in the TED talk is going to be extremely heavy, expensive, and poor in tensile strength compared to a conventional wall built with modern engineered nails or screws. The weight makes the lack of tensile strength an extra special problem. The demonstrated technique does not provide for any of the other pie in the sky modalities promised, such as flowing curves or 3d-printed utility runs. It's much easier to build a CAD animation than an actual house. When you start asking yourself how am I going to get 500 bags of portland cement to the site, where am I going to store them, what am I going to use to mix them with the water and admixture and aggregate which I also have to stage, will I need a tank for the gasoline powered mixer because there won't be electricity, how do I get the heavy stuff on the roof, and on and on and on, yeah the 3D printer solves some of those problems but it leaves many others harder to deal with than they are now.
IOW, to justify the cost of creating automation systems, you need to spread out the cost of the development and construction of the machinery. That's best done when building houses by the thousands. Think Levittown of the Sahara.
Similarly, he shows some construction on the moon. As lunar soil is devoid of clay and mica and other materials common in earth soil, the makeup of that material will have to be vastly different than anything found here. I don't think Redicrete delivers to the moon yet!
One of the basic principles of Frank Lloyd Wright was to use the materials indigenous to the area. Here in the southwest, if we were to need such a construction technique, the better material might be adobe, not portland cement.
-- Gordon
http://www.roadsideamerica.com/story/32774
http://www.youtube.com/watch?v=36lNS1ruK4s&feature=related
Sadly it didn't last:
http://www.youtube.com/watch?v=KNl21ZX84-4&feature=fvwrel
Concrete is only half the structure. Without reinforcement you have something that is likely to collapse unless you provide stone arches and a lot more concrete.
Isn't there a guy in Australia who build concrete does in a day?
He lays out a big rubber balloon on the ground/floor, layers of rebar on top of that, dumps a load of concrete over it, then inflates the balloon...
The biggest issue seems to me to be to tie the rebar together in such a way that it ends up spread evenly throughout the dome. You also can't have plumbing or electrical conduits embedded in the walls, unless you cut into the concrete afterwards.(Most of that could be floor-mounted instead, I believe)
As for the short vertical pieces of rebar used with the 3D printer, that is probably because anything taller would interfere with the printhead.
Honestly, a robotic 'Bricklayer' would be a better design...
(Think oversized 5axis industrial robot)
You need reinforcing steel, it needs to be placed reasonably accurately and tied, and after the concrete is poured you need to keep the forms in place for about a week before it hardens to gain enough strength to remove forms.
There are fast alternatives. In newer construction, steel deck makes the bottom form - but there is still steel rebar to achieve strength over long spans.
Another alternative is a 'ferro-concrete approach' where you build the house out chicken wire or another small mesh of steel and shoot 'shot-crete' on the mesh. I suppose that mesh might all be supported by an inflated bubble.
But it all takes about a week to gain strength and one should apply water to both cool and cure the concrete during that period. After all cracks end up as a leaky roof.
The idea that today's construction methods are slow is rather absurd. I have started from bare-ground to a completely enclosed house in two weeks with a crew of three including myself. From there, you wait for the plumber and electrician to show up and do their jobs. After that, the insulation and sheet rock don't take much time, maybe two weeks. Interior painting can be done in a few days. Exterior painting was done before the siding and trim were put on the house. Doors, kitchen, bath, and flooring are the last to go in.
Does he plan to provide the plumbing, wiring, doors, kitchen, bath, flooring, and interior paint in 20 hours? Anyone can build a shack or raise a barn in 20 hours - it is the details that make a modern house a home that add cost and time.
Still, the banks don't like houses to take over 3 months to build in the USA and are much happier if you have it all done in six weeks. If you don't have a buyer for the house, putting it one the market without the flooring, kitchen, and bath done seems to actually sell quicker as the buyer can choose colors and finishes.
Using short pieces of rebar would likely double your steel costs as each rebar lap joint is supposed to be at least 18 times the diameter of the rebar.
In any event, a geodisc dome will still get a roof over your head faster than any 3D printer fantasy. And small precut steel domes are ideal for 3rd world where they are extremely strong, resist very high winds due to being round, and can be delivered in a complete precut kit to be assembled with a few basic tools. Anyone want to see the details?
Considering this idea is still in the experimental stage, I think this idea will steadily progress towards reality. Sure there will be obstacles, such as concrete reinforcement, electrical raceways, plumbing, etc..., but I believe these obstacles could be overcome.
Hmmmm..... Schedule 40 PVC ought to do the trick for wet concrete installation
Plumbing and electrical are hung from the ceiling along with HVAC and drop ceiling tiles cover all this, but maintenance and reconfiguration are much easier than before.
Personally, I'd like to have a hand-crafted home of logs, timber, and masonry. What we have now is nearer to robotic construction than to more human creations with materials from nature and the local surroundings. Factory build home with precut parts do exist, but nearly everything is on an XYZ geometry FOR the sake of economy and standard engineering.
As I mentioned above, the geodisc dome can be built extremely fast - likely out of wood or steel. Concrete has the problem of having its own weight to deal with in a transition from fluid to solid state. Slip form technologies and pre-cast yards have evolved to deal with this reality.
Having a super-sized 3D printer is a bit like dreaming of huge Lego blocks. They both attract a lot of attention, but are not practical in real life.
Schedule 40 plastic pipe? It is a sad replacement for copper water pipe, which can last far longer and withstand more hardships of freezing weather. Sure, it is cheaper and faster to install, but I'd like a house that lasted my lifetime, not just to the end of a 30 year mortgage.
Somebody has to come up with some sort of 'miracle goo' for the 3D printer to make this version of robotic house building a reality. There are thousands and thousands of contractors that are looking at every cost factor involved in present technologies.
I meant in the case of concrete domes made using the inflating balloon.
(if you're using a robot systemto build a box-shaped building, itshould be possible for the robot to install whatever is necessary. )
You would literally have to fasten the boxes to the balloon to avoid them being 'swallowed up' by the concrete or shifted around as the balloon inflates and the concrete starts moving.
And even then there's no guarrantee that the force won't pull the tubing out of the boxes.
Best bet is to either use floor-mounted tubing, or lay a duct along the wall and have all sockets and whatnot there.
Like a simple simulation video done by some college kids is really going to explain it all!
In all, this is an interesting concept that Khoshnevis himself calls a "grand challenge" with obstacles yet to overcome, not suitable for everywhere, but possibly useful in parts of the world where extreme high population density threatens hundreds of millions of lives.
-- Gordon
For them, using short pieces of REBAR probably makes sense as that can be packaged in any available space.
The Army probably want to use it to build semi-permanent camps during campaigns.
(They have two issues with the clssic way of doing it; it takes a lot of troops that could be better used elsewhere, and the locals like to take potshots at the guys building fortifications... )
NASA wants something that can build structures without requiring ANY manpower locally at all.
(Because it's not enough to send an operator, he probably want food, water and air, too. And he probably also want a return ticket, the ungrateful wretch? Doesn't he know there's a recession and that he should be grateful for having a job in the first place?)
Because the guy is a theoretician, not a home builder or even a DIYer. He has no real points of reference to base his assumptions on.
I saw his slide show and he dead wrong on conventional construction. It is not labor intensive or inefficient nor over budget. Maybe in some corrupt 3rd world arm pit like Turkey or our military, yeah it's welfare program then. But if you check out a small multistory business complex being put together today, there are very few people involved and they are busting their rear ends. Housing tracts aren't much different.
He'd benefit by spending a year in the field rather than behind a computer simulating everything. Maybe intern with KB homes and see what successful home builders do.
If there is a market for such device, it will be as Gadgetman wrote for NASA and the Army. Money doesn't mean anything to them and they can afford to build a home for a million and not bat a eye. Who cares if it looks like was designed by a Soviet public housing expert and only fit for the insane.
But take the Marines, they got smart and modified hundreds of those Chinese shipping containers and turned them into troop housing in Djibouti.
All in all the TED talk reminds me of the schemes of junior college students who think they are going to build a small floating city to achieve their Libertarian dream; they simply don't understand the logistics of working on such a large scale. After a lifetime of building things, including some very expensive industrial machines, it hit me like a wet towel to the face what a freaking big massive thing a house is compared to other hobby projects. Modern techniques like prefabricated walls and trusses, which leverage the short-term presence of a small crane on site, make it possible to build a house in a few days with only a few workers on-site. This is how you see fast food outlets appear in just a week or two on a site that starts out as raw land. In some cases they prefabricate entire sections of the building, utilities and all, and tie it all together on-site. It would be very hard for a robotic printer which is, no matter how you slice it, going to require massive amounts of a rather exotic cement mixture to compete with that.
By that measure Einstein had no business dabbling in physics!
I don't know the good doctor, but a very quick Google search shows he's director of the graduate program of manufacturing engineering at USC, and as any professor, his principle role is to guide his students who go beyond the theory. And indeed, if you look at the related literature over the years of his students, there's quite a number of real-world dissertations, and a few working in the industry.
I keep saying this is not a practical method of construction of the US, and that should be obvious. Less than half of the earth's 7 billion inhabitants live in permanent structures. Those of us in developed nations wouldn't stand for a single point of access for electrical and water in a room, but that's what he shows in the sim video. There's no wiring or plumbing in the conventional sense. We've become jaded into thinking that our style of living is how everyone lives.
There are fewer and fewer metal shipping containers to go around, as the metal is now worth more as scrap. Plus they last only about 10-15 years unless coated, I don't know about you, but I don't trust a coating to not leech or degas.
On how KB does it: If you want to build houses for a billion (or whatever) people, you can't use lumber. That's Western thinking. The idea is to conserve what we have left. Forget budgets and corruption; it's just not possible to build that many lasting structures with conventional materials and techniques. How KB or another American homebuilding company does it is not the point, nor is it a method that can apply to developing nations.
Roger, there's no reason the same 3D machines can't create tilt-up, except why? It would be like having a desktop 3D printer create sheets that you then cut out and drill. What's the point? The older fashioned method of pouring cement into a tilt-up form isn't going to solve anything, make the process safer (yeah, like cranes and joists are safe!), or use less material.
-- Gordon
A buddy of mine used to speculate that alternating layers of concrete and balsa wood might make an interesting composite building material, as one is very strong in compression, and the other one... isn't.
I'd like to quibble on a minor, unimportant point: wood is not a bad material for structures. Perhaps the way that we use it doesn't lead to lasting designs, but I believe that can be changed. In any case, wood is one of the truly sustainable resources: some managed forests are harvested as much as every 20-50 years, with trees that grow three times as fast as naturally (source).
Of course, dirt (concrete, adobe, ...) is sustainable too. Plus we could make houses out of straw bales. It's more solid and durable than you might think, when done properly.
I saw a group that made straw bale houses in Canada. These thing were AWESOME. Just straw bales stacked like bricks, rodded down to the foundation, and covered with plaster or concrete or whatever. They seem to last 100 years from what I've head, and the insulation is significantly better than wood or brick.
I wonder if straw could be mixed with the concrete, and extrude adobe houses?
(And maybe something to dig for ore, refine it and make pipes and fittings)
1. Robot grab and lays down the REBAR for floor and any tubing that is needed there.
2 Robot switches to a nozzle and sprays own a foundation of locally made concrete.
3.Robot switches to a gripper.
4 Robot grabs a brick and moves it over a 'concrete applicator' to add wet concrete on whichever surface will get in contact with existing surfaces.
5. Robot places brick in intended position.
6. Repeat 4 and 5 until a layer is laid down.
7.Robot switces to nozzle.
8.Fill hollow bricks with concrete where necessary.
9. Robot switche to grabber.
10. Robot places utility boxes (same height and width as bricks but only half as thick)
11.Robot lays down tubing between boxes, and also uses shorter pieces for vertical conduits.
(Press to fit connectors. Boxes will initially be held in place by the piece coming up from floor and through first layer of bricks)
12. Robot place the next layer of bricks. As it's a 5Axis machine it should be able to work around standing pieces of rebar and conduit, 'threading' bricks down them as necessary.
...
...
99. Humans arrive. Pulls electrical cabling and water tubing through conduits where needed.
Bonus is that the robot can be used for other tasks than just assembling buildings.
My impression of a 3d printer for houses is that it would have to be bigger than the house. Getting it to and from the site to the next job would be rather daunting.
Regarding wood...
Over half of all wood that is cut is used for paper. And in many regions, wood is in very short supply as we used it for fuel in many regions of the world before with got into coal and then oil. The USA builds a remarkable quantity of wood frame homes, but in other parts of the world - concrete, rammed earth, adobe brick, and steel seem to be far more prevalent.
While the wood producers like to claim it is a renewable and sustainable resource, the structural quality and resistance to rot have gone down with the faster grown trees of modern forestry. It is sad to look at what is standard grade now, and structural select is rarely up to what it once was. Much of the drop in lumber quality has been because the average wood home is built with stud walls for rigidity and engineering simplicity, not designed to failure strength. There is simply a lot of extra wood in a house to keep sheet rock from cracking - so lesser quality is not critical to failure.
Other choices of material...
In India, Tata Steel provides all the engineering software that one needs to design and build a steel structure. That means that after designing, it is quite easy to order all the pieces ready for assembly with very little waste.
Taiwan is a concrete jungle as termites here have rather large appetites. Actually the structure is concrete and partition walls tend to be brick. Homes are mostly common wall construction that means a whole city block is actually one big structure of buildings ties together. This is quite normal in parts of the world where people have less money than in the USA with separate homes on a plot of land. And it eliminates a lot of engineering in favor of standard solutions and standard concrete forms.
Custom homes are where the true dream houses exist. You can do anything with enough money. But it will be rare to not build out of steel, concrete, rock, brick, or wood.
3D printing is optimal for thermoplastic. There just seems to be an economic limit to the size of thermoplastic items. We can and do go a lot bigger with laminated plastic materials and carbon fiber might really be the future for robotic homes - after all Boeing is flying all plastic airliners now.
Here in Norway we have wooden churches that are nearly a millenia old... And still in use...
http://en.wikipedia.org/wiki/Stave_church#Norway
Not only is it difficult to find the quality of timber for a similar structure, now, but the sizes needed are absolutely impossible to find.
I assume the 3D printer can be dismantled into sections for transport.
But that still leaves the need for a mobile crane to erect and disassemble it, and just as importantly; the space for the crane to work.
Other than that I assume they plan to extend the rails from one plot to another and let it work its way along a street.
(How much time have they set aside for building and levelling the tracks before erecting the printer? It really sucks if oneend of a track subside while printing... )