I remember the days when a laser printer was a fabulously expensive luxury item that only large businesses could afford; now anyone can buy one for a few hundred dollars. I also remember when a photo-quality color printer was far beyond the reach of the average consumer; now they’re so inexpensive that they’re often bundled with new computers at no extra cost. What’s the next generation in printing? If you have enough money—let’s say, US$25,000 and up—you can purchase a 3-D printer that will sit on your desktop and create solid plastic models of just about any shape you can throw at it. For a few thousand dollars more, you can even make full-color 3-D objects. Perhaps in a few years, these printers, too, will drop into a more interesting price range. But even now, for a certain type of user, they represent an extraordinarily quick and cost-effective alternative to older methods of generating accurate, tangible copies of three-dimensional objects.
When I was working for Kensington Technology Group, a major computer peripherals manufacturer, my job was to manage software development. But I shared an office with several people whose job was to design the hardware for new products—specifically, keyboards, mice, and trackballs. Because we needed to create products whose hardware and software features were tightly integrated, we worked as a team. I got to participate in the hardware development process, and my coworkers were also involved in shaping the features of our software. The process of designing a new mouse, say, is a lot more complicated than one might imagine. Over a period of months, we’d hold brainstorming sessions and focus groups, conduct surveys, hire artists to create concept drawings, and finally—many meetings later—decide on approximately what the mouse would look like and what its features would be.
Next, we had to get the shape just right. Usually we’d start by having a model shop create rough carvings out of a firm foam material. We’d all hold the models and choose which shape came closest to what we were looking for. Then we’d go back to the model shop and ask for modifications; sometimes, the next step would be to carve a near-final version of the shape out of wood and paint it so that it looked much like the final product. Such models were essential for ascertaining usability, designing boxes, and helping the company’s executives to understand exactly what they were investing so much money in.
Once we had the overall shape worked out, it was time to do heavy-duty 3-D modeling on a computer. Not just the overall form, but every individual plastic or metal part—all the buttons, wheels, battery doors, and so on—had to be painstakingly designed. But before we could feel confident about having expensive molds made to produce the final product, we had to be able to see, feel, and work with a plastic representation of the device—to make sure everything fit together properly (including the electronics inside) and was easy to assemble. When we got to this stage, someone would say, “I’ll send the files out to have an SLA model made.” And a few days later, a box arrived with all the plastic pieces that had appeared on the computer screen, ready to be assembled. The SLA models were made of a translucent plastic, with a somewhat odd texture—curves and angles generally had a subtle “stair step” effect. When people asked, as they frequently did, what “SLA” meant, a designer invariably replied, matter-of-factly, “stereolithography apparatus,” as though that explained everything and further probing would be inappropriate. When I left the company years later, I still had no idea how that process actually worked.
Laying It On Thin
Recently I attended a technology exposition where I saw a 3-D printer being demonstrated. Designed to produce rapid prototypes similar to those I’d seen at Kensington, it was truly fascinating to watch. Seeing this machine made me wonder once again about stereolithography and other automated methods of creating three-dimensional physical models from computer-based designs. A brief investigation turned up dozens of different 3-D printing techniques. Although they differ tremendously in the details of the technology they use, they all employ certain general principles, and are evolving rapidly to become smaller, faster, more versatile, and less expensive.
A sculpture begins with a solid block of material from which bits are carved or chiseled away to reveal a 3-D shape. And to be sure, there are any number of clever devices that can create carvings using lasers or robotic grinding tools. But the problem with such an approach is that you can only affect the outside surface of an object; hollow or deeply concave areas, or oddly shaped interior cavities are difficult or impossible to achieve. So instead, 3-D printers build up an object in layers. Special software splits a 3-D computer model of an object into a succession of very thin layers. The printer then deposits a layer of material that represents a cross-section of the very lowest (or highest) point on the object, and keeps building on that with additional layers until it reaches the other end.
Stereolithography is one of the oldest and best-known rapid-prototyping (RP) methods for small objects. You start with a platform suspended just below the surface of a vat of a special liquid polymer that hardens (or “cures”) when exposed to ultraviolet light. Using an ultraviolet laser, the machine “draws” the first layer of the object on the polymer-covered platform. When it hardens, the platform moves down slightly, allowing a fresh layer of the substance to ooze over it. Then the machine draws the second layer, which not only hardens but fuses to the first. The process repeats until, many hours later, the entire object is finished. Drain off the excess goo, and you have yourself a model.
As cool as stereolithography is, it’s not appropriate for some kinds of objects. For one thing, you’re limited by the properties of the hardened polymer (strength, flexibility, and so on). For another, overhanging structures are a problem, because if the laser draws a pattern onto an area of the polymer without a solid surface beneath it, the resulting patch, when hardened, will either float away or sink. But other 3-D printing techniques can solve both of these problems and more.
Selective Laser Sintering (SLS) replaces the liquid polymer of SLA with a powder. When the layer heats the powder in a certain spot, it melts, fuses to the layer beneath, and hardens as it cools. As with stereolithography, the platform on which the model is sitting drops slightly after each pass, and a new layer of powder is spread on top. But because every layer is supported by the unused powder beneath, overhanging or even completely independent structures can be created. There’s also a relatively new printer type that “prints” onto powder (as SLS does) but replaces the laser with a moving print head containing a number of nozzles just like those in an inkjet printer. Instead of using ink, the nozzles squirt out tiny amounts of a binding agent that causes the powder to harden almost instantly. Furthermore, different types of powder and binding agents can produce products with a variety of characteristics—and by adding dyes to the binding agents, these machines can even produce solid objects in full color.
Yet another variation on this theme is called Multi-Jet Modeling; it uses an apparatus similar to an inkjet head that deposits layers of a special polymer with a very low melting point; it hardens as soon as it leaves the nozzle. There are also nozzle-based systems that deposit liquid polymer that is then cured, layer by layer, using an ultraviolet light. These are just a few of the current 3-D printing techniques, with new variations appearing all the time.
Some 3-D printers are restricted to creating plastic objects, but these basic principles have also been adapted to many other materials, including ceramics, wax…even stainless steel, titanium, and other metals. The U.S. Department of Defense is funding development of a metal inkjet process that could be used to create replacement parts for submarines. Some high-resolution 3-D printers are being used to test jewelry designs, while others are being designed to create dense, multi-layer electronic circuits on an extremely small scale. But if manufacturers are smart, they’ll begin thinking not just about industrial and military customers, but about consumers. Imagine receiving a birthday present as a file enclosed in an email message. Send it to your desktop 3-D multi-printer, and out comes a framed photograph…or a statuette of your parents. It’s a fantasy today (except, perhaps, for the obscenely wealthy)—but tomorrow, who knows? —Joe Kissell
To learn more about 3-D printers and rapid prototyping (RP) technologies in general, see:
- Stereolithography & Rapid Prototyping by Patrick Salsbury
- How Stereolithography (3-D Layering) Works at HowStuffWorks.com
- SFF Process is a somewhat dated, but still very useful, overview of Small Form Factor rapid-prototyping technologies.
- Rapid Prototyping and Manufacturing at the Advanced Manufacturing Research and Collaboration Cluster also discusses a wide variety of rapid-prototyping methods.
- A Year Filled With Promising R&D describes a great many R&D projects underway at the end of 2002 in the field of rapid prototyping.