3D printing: it's the shape of things to come
3D printing is attracting a lot of attention, but is it really feasible that it will become commonplace on live projects? CM convened a panel of experts to explore the possibilities. Elaine Knutt reports. Photographs by James Winspear.
It won’t have escaped the attention of CM readers that 3D printing is having a bit of a “moment”. In the papers or online, there’s a profusion of headlines: “NASA 3D prints rocket parts”; “New Vista print head could allow surgeons to print human organs”; “US Navy prints spare parts at sea”; “First 3D printing bureau opens on London’s Oxford Street”. Not to mention the latest acquisition at London’s V&A museum...
All of this naturally raises multiple questions about 3D printing buildings and components. To debate the issue, CM assembled a panel of experts and interested onlookers.
From left: Jeff Carter of BAM Design; lawyer May Looi ICIOB has both a professional and personal interest in the subject; design management consultant John Eynon FCIOB sees it as a logical extension of BIM and 3D design.; Dan Culling of Skanska, who with Carter believes they could be using 3D printing in the not-too-distant future; George Lee runs a modelmaking 3D print bureau.
The group is to meet Sylvain Preumont, a French entrepreneur who runs iMakr, London’s first high-street 3D print shop. Visit iMakr and you immediately grasp some of the contradictions inherent in 3D printing. Demand for the bespoke mobile phone cases and plasticky objects on display would seem to be limited – why print what can be cheaply mass produced anyway? But the very fact that iMakr exists — selling affordable entry-level 3D printers for innovators and designers to experiment with – does suggest that 3D printing is on an acceleratory phase of its development curve.
A curve that could lead us to... well, over to John Eynon: “It’s the emerging technology but it could be as ground-changing as when PCs came out. It could change the construction industry and the retail industry too. You can download what you want, pay a royalty fee and make it. Why ship materials all the way from China, when you have them on site? You take out all the travel costs and carbon, destroy the supply chain and the design industry, and create a different industry!” he finishes.
Sylvain Preumont: “market soon”
Preumont, not surprisingly, was also an evangelist. “Schools are buying printers, architects are buying them and I am selling people their second printers because they want an upgrade. People are using them to print adapters or modifications of existing equipment that just isn’t available any other way. But we’re still waiting for the ‘killer app’, the product that makes 3D printing a must.”
The others are also enthused by the potential, aware of experimental projects on university campuses that could cross into mainstream construction. Loughborough’s Freeform Construction project has already 3D printed a double curving concrete wall section with voids for service runs already in place, the deposition nozzle “printing” the concrete, layer by painstaking layer. And Looi has been reading about work at the University of California to develop new 3D printable concrete for its Contour Crafting technology that has three times the compressive strength of conventional concrete.
And because everything in 3D printing is bespoke, the technology brings the complex and costly into the same realm as off-the-shelf components. Carter, for instance, relates the potential to the challenges of building a Zaha Hadid curvilinear design, for instance. “We’ve built a couple of Zaha buildings and had to use software to transfer the design into something you could actually build. But this could extend the boundaries of what you could build,” he says.
It is clear that 3D printing lends itself to decorative elements and non load-bearing applications; Lee refers to follies and installations. But could it ever be scaled up to become a viable alternative to mainstream industrial manufacturing technologies, such as pouring concrete into shutters, extruding aluminium for cladding, manufacturing insulation materials? At iMakr, our panellists are keen to explore the technical limitations of the system and what it would take to shift it out of London boutiques and Wired magazine and on to construction sites.
First, there was a demonstration of high street 3D printing as it exists today, which is a reminder that the technology has some distance to travel. Preumont shows a printed architectural model of a group of houses, made using a fused deposition modelling (FDM) printer and a material called ABS (acrylonitrile butadiene styrene). The model looks like the sort of thing a skilled modelmaker could produce in a couple of days. As it turns out, it took the printer an equivalent 20 hours.
“So would a model twice as big be printed at the same rate?” Culling asks Preumont reasonably. But the answer is “no“. “Timing is physics, chemistry and mechanics – it relates to the power of the engine running the print head, the energy needed, the weight of the head. So twice as big is twice the length, height and thickness or 2 x 2 x 2 – that’s eight times as long.” But surely that performance will improve over time? Yes, but Preumont still points out that FDM is constrained by the basic principles of mechanical engineering. “Four times faster is four times more energy – you can’t alter that. What you can do is make the material lighter, or reduce the thickness – or use a bigger printer.”
Looi then touches on another practical issue. “Are the machines sensitive to dust and vibration? If you’re pouring concrete on site, would you have to protect the machine?” Preumont confirms that this would probably be the case, and Lee adds that today’s printers are also very sensitive to humidity. It explains why
the Dutch architects behind the most advanced 3D print-a-house project to date, the Canal House in Amsterdam, put their on-site printer in a mini-house of its own, the so-called Kamer Maker.
In fact, Lee and Preumont point out that there’s nothing very high-tech about 3D printing. Although everyone in construction thinks of it in the same breath as BIM, the technology was actually developed more than 30 years ago. “The technology is very simple – just three motors going along the x, y and z axes and software,” Lee says. It turns out that the recent upsurge of interest in 3D printing is related more to the lapsing of critical patents than any technological breakthrough, with another important patent due to expire in 2014.
But Lee sees the relative simplicity of the technology as a strength, as it lowers the barriers to entry for innovators to develop new printers, print heads or printable materials, in construction or other fields. “3D printing opens up the possibility of other players coming into the industry,” notes Eynon. And Lee has just spotted an innovation that ushers in more possibilities. “The new Vista print head can work with multiple materials and they’re selling just the print head. Previously, companies have been in full control of the machines and the materials. But if you can produce and market a print head...”
But even if we do see the birth of this kind of “open source” 3D printing, the factor that’s keeping the technology on university campuses and at the fringes of manufacturing is the range of materials, and the lack of consistent standards and certification. “The thing that’s preventing wider application in the real world is materials,” Lee argues. “Take this ABS plastic – it’s not waterproof, water goes right through it. We need companies to do research into materials resistance,
or compressive strength, to give us all the technical information about what we can achieve.”
Preumont agrees that there is a lack of interest from the major materials and chemical companies. “In the short term, there is no market for them. But soon there will be a market for the big chemical companies [such as BASF, Henkel or Dow Corning]. They have the technology and materials, they just haven’t thought about putting it in 3D printers.”
This creates considerable potential for new materials to enter the construction lexicon. “You could easily modify the concrete to make it stay in shape as soon as it meets with air, or print a conductive filament into a non-conductive material and instantly your component can light up,” Preumont says. Encouragingly, he says that the R&D team from one major multinational chemicals company has been in touch with iMakr.
So where next for 3D printing in construction? Lee feels that non-structural cladding harnesses its capabilities. “You could 3D print cladding and reduce the weight in comparison with conventional materials, which could produce savings elsewhere in the building,” he says. But BAM Design’s Carter is unsure, as extruded aluminium is fairly light and cheap to produce anyway. “Once you’ve got the die, the expense is in the die, not the fabrication itself,” he reasons.
One application where the panel agree that 3D printing is likely to emerge as a viable alternative to traditional construction methods is, ironically, historical restoration. “If you had a historic building, you could 3D print the mould from a 3D laser scan and take the plaster cast out of it. There’s already a company in the US that prints the moulds,” says Lee. Skanska’s Culling thinks it could be useful for external landscaping features – possibly because Skanska is experimenting in this area.
Culling brings up the question of where 3D printing fits in to the sustainability argument. “So many projects these days are BREEAM or LEED-led. If you’re looking at the carbon implications of the actual construction process, what is the carbon footprint of 3D printing?” Looi’s response is that “it has to be more sustainable than shipping materials from China”.
But Lee’s explanation is slightly more nuanced. “Any process that involves sintering a powder [with a UV laser] or cutting metal will involve high temperatures. It’s probably not any more energy efficient than moulding – although it depends on the temperature you need to work at. But a lot of the carbon comes down to sourcing the materials and transport, so the energy used might be less important than the transport carbon,” he agrees. “Or the energy in use,” interjects Carter. “That’s where 3D printing could be interesting – if it could print insulation.”
Researchers on Loughborough University’s Freeform Construction project are developing techniques to print concrete wall sections that incorporate voids to thread services
Rethink building design
The group agrees that the future of 3D printing does not lie in retrofitting it to today’s methodologies, but in creating new construction typologies that exploit its potential. “If, in 10 years, 99% of buildings are designed the same way as today, then no I don’t think we’ll see a 3D printing revolution,” says Lee. “We need buildings designed around its capabilities. It allows you to do things that couldn’t have been done in the past, it needs engineers with the ability to do things anew, to come up with new forms.”
Looi agrees, but points to the need for extensive research and testing. “Clients will want guarantees. Until 3D printed components have been fully trialled, I don’t see anyone signing a collateral warranty.” But she also points out that if 3D printing does become a viable option in the future, it might usher in a different approach to construction, and therefore to contracts, liabilities and warranties. “It could be that the lifespan of buildings will be shorter. If a building or a component fails, you just print some more.”
Overall the group is confident they are witnessing the awkward adolescence of a technology that will mature over time. Several times, members of the group advance a variation of the Moore’s Law argument – that computer processing power doubles every 18 months. If 3D printing is on the same technological path, we could be looking at a new method of design and manufacturing that could open up all kinds of new possibilities.
“I have a dream of a future where we build a major project and if ever you need a spare part, you would just 3D print it directly from the BIM database,” says Lee. “But maybe it’s just a dream...” He’s right to be cautious, but the fact that 3D printed cladding components were last month installed in a mainstream City of London office project (see below) certainly brings the dream a step closer.
Skanska puts 3D printing into practice on City project
Last month, Skanska achieved an industry first by using 3D printing in a live project. It turned to 3D printing bureau Quickparts (formerly CRDM before its acquisition by 3D Systems in August) to resolve the tricky technical issue of cladding tree-like columns for an ETFE roof at a City of London office development.
A 3D printing or additive manufacturing technique was used to produce cladding for the tops of steel columns for a roof garden at 6 Bevis Marks. Eight different complex interfaces between roof and column were originally envisaged as cast steel nodes, but Skanska foresaw considerable expense and difficulty. Alternative steel plate options were considered expensive, challenging or aesthetically unsatisfactory.
The ETFE roof package was delivered by specialist Vector Foiltec, and its consultant Adrian Priestman approached High Wycombe-based Quickparts, which produces bespoke parts for the aerospace, defence and medical devices industries.
One of eight different cladding modules produced by 3D printing for the tops of columns at Skanska’s 6 Bevis Marks development
Simon Hammond, regional sales manager, says the firm has been quietly innovating while others have been noisily discussing. “What people see in the news represents the consumer end of 3D printing, the machines you can buy for the home or office. But we’re a commercial bureau, we print parts for human bodies or aeroplanes or Formula One cars, and they’re not prototypes. People are adopting 3D printing because of the advantages it brings.”
Bevis Marks was an ideal candidate, because of the complexity and lack of repetition. “With eight iterations, other types of manufacturing would have needed eight moulds, but we produced eight unique shrouds. It comes into its own when a low volume production run [using conventional manufacturing] is too expensive.”
Quickparts used a selective laser sintering machine that fuses layers of powdered Nylon 12 (PA 12) to build up the complex shapes. “We took the CAD file from the architect, then ‘sliced’ it into 0.1mm layers. The laser operates above a movable platform, so it traces one slice of the layer by putting enough heat on the nylon to melt it. Then the platform moves down by 0.1mm, you put on more powder and build it up layer by layer,” Hammond explains.
The eight cladding nodes, termed “shrouds”, were printed in different sections – a process that took about three weeks – then jointed. Finally, the 600m wide, 800mm high nylon shrouds were finished and painted to resemble steel.
FDM or SLS?
There are two 3D printing techniques relevant to architecture and construction. iMakr sells entry-level printers based on fused deposition modelling (FDM), which allows objects to be made direct from 3D CAD files. The technique is probably what most people think of as 3D printing: a print head with a nozzle deposits filaments of plastic or other materials that are unwound from a coil. The nozzle is heated to melt the material, and its movements are controlled by software.
The models at iMakr are printed in ABS (acrylonitrile butadiene styrene), a plastic that is available in different proprietary formats that is often used to make car bumpers and also found in Lego. ABS components can be smoothed and sanded after printing and given additional coatings, such as paint or waterproofing. But as a pliable thermoplastic, ABS is prone to warping if it is cooled too quickly.
Ceramics or metal powder mixed with an epoxy binder can also be printed on an FDM printer, but would have to be fired in a kiln to achieve the same strength characteristics as the original.
Another issue is that manufacturers’ printers are designed to print only the materials it produces. “There are compatibility problems — filaments and printers need to match. In the future, we will probably see norms and standards for 3D printing that allow more interchange,” iMakr’s Sylvain Preumont says.
Another important limitation is that it is impossible to print above a void: if there’s an unsupported part in a design, the printer will automatically print “support”, often in another material, that then has to be discarded.
George Lee’s print bureau, Lee 3D, on the other hand, prints models using selective laser sintering (SLS), which is also a form of additive manufacturing. It involves putting layers of plastic powder on a machine bed, and using laser pulses to selectively fuse them into a solid.
The shape is built up layer by layer: after the laser completes one horizontal cross-section of the desired shape, the powder bed is lowered by one layer thickness, and a new layer of material is applied on top. Skanska’s experiment with 3D printed nylon cladding at the Bevis Marks project (see below) used an SLS machine.
SLS can also be used with metal powder, ceramic or glass. But as the technique is always based on powder, the resulting components are porous unless finished.
Lee says the technology is still maturing. “For construction, what you need is controllability, but with SLS [using plastic powder], to get the right standard you end up throwing away the sub standard prints, which can be up to 40% of what’s built. We need to get to a point where it’s right every time.