Every now and again, something comes along with the potential to change how we deal with the world: steam-powered machines in the Industrial Revolution; microprocessors for the Personal Computing Age; the Internet for the Information Age. We seem to be at the cusp of a breakout technology that will usher in the next shift: personal 3d printing.
Printers that can build usable objects have been around since the mid-80s. Thing is, they were R&D projects and then industrial aides — translation: rare and expensive. Now, in the 2010s, they’re breaking out into the public consciousness. President Obama touted them as a promising technology worthy of government support. More importantly, the first waves of more-or-less affordable printers have become available for the home enthusiasts. Also, for those with shallower pockets or without a desire to jump into early-adopter tech, companies such as Shapeways popped up to provide quality printing (in a variety of materials) at “good enough” prices if you design carefully.
In the 70s, I got my first microcomputer. In the early 80s, I built a working computer of my own design — not like now were you just assemble boards but by soldering and wire-wrapping at the chip and component level. It’s performance was much as you’d expect from a first effort. Anyway, my point is that I’ve been at the early evolution of a new important technology. If I were much younger, I’d be jumping in with 3d printing with great enthusiasm. But for older me…I plan to wait a bit. Why? Because, for me, it isn’t quite “there” yet.
Most consumer (i.e. sub-$2000) 3d printers output using an extrusion method. Think of them as glue guns but on a finer scale. Their material selection is usually limited to PLA or ABS. While none of this makes me want to buy them, enough of the gee-whiz-neato factor is there that I would…except for one thing: the resolution just isn’t there. These systems typically tout “100 micron” or “200 micron” resolution. It sounds spectacular — they used the word “micron”. The fact is that a resolution of 200 μm (or microns) comes out to about 127 dpi*. That’s pretty rough. When you see an object printed at this resolution, it’s riddled with stairsteps and other artifacts. There’s a reason why even cheap laser- and ink-jet printers print at least 600 dpi: because anything less wasn’t ready for prime time. (In case you’re wondering, the 600 dpi equivalent for a 3d printer would be about 42 microns, about half the thickness of a sheet of copy paper.)
I have many things I’d love to have 3d printed, a couple of which I’ve tried with Shapeways. They are in a variety of materials and typically small, less than a couple of inches on a side. But because of that smallness, they need higher resolution printing. I don’t see extrusion being the method for someone like me who wants smaller items. Sure, there’s a $5000 “pro-sumer” printer from Old World Laboratories that uses the stereolithography (lasers solidifying a liquid polymer) that boasts a resolution of 100 nm (that’s 0.1 microns…or an astounding 254,000 dpi equivalent). Needless to say, they get nice results, albeit in an acrylic material.
There are several commercial 3d printing processes that use not only plastics and acrylics, but also metals and ceramics and even a sort of synthetic wood. The thing is that many of these are somewhat messy, require other steps (e.g. kiln firing), and a fair amount of experience. What is needed is the next breakthrough that merges together all the good points.
What I’m waiting for is a machine that will effectively inkjet a layer of nanoparticles of my desired material to the target, and immediately have that melted to the substrate or welded to the previous layer with a laser** (or bonded if the material isn’t amenable to being melted). Its resolution will be no grainier than 20 microns (i.e. 1270 dpi), though 100 nm would be much better. Here’s the standard: printing without obvious or tactile planar stairstepping and minimal steps when off-axis.
Acknowledging the advantage of powder beds to support in-progress building of spanning structures, some method of constructing using an easily removed, cheap material (e.g. talc or corn starch) to act as an in-build scaffold when necessary would be ideal. Since we are sacrificing size for resolution, a build box of 10 x 10 x 10 cm (4 x 4 x 4 inches) giving a maximum volume of 1000 cm3 (1 liter) would be sufficient for most tasks — though 10 x 15 x 10 cm might add that extra bit of flexibility with some items.
And have this come in at between $800-900. I would say that it’s a lot to ask, but I remember when microcomputers were barely usable, expensive hobbyist toys that you largely had to build yourself. Now almost everyone has an incredibly powerful computing machine (and camera, and display, and networking, and media device…oh right, and also phone) in their pocket costing only a few hundred dollars. No…I’m not asking too much. I’m just asking like Veruca Salt.
* The math is pretty basic: The resolution is 200 μm (microns) per dot and there are 1000 μm to a millimeter, which gives us a conversion of 5 dots per millimeter. Multiply by 25.4 mm per inch and we get 127 dots per inch (dpi).
** I don’t think extrusion is going to be a the long-term most popular method…maybe used mostly for toys, utility boxes, and other items such as prototypes where a finished (or nearly finished) surface isn’t necessarily needed. The technology has too many inherent limitations. If I were to bet on any currently popular printer designs to build on, I’d go with a granular bed method. Despite it’s unavoidable messiness and operating space requirements, it has the most flexibility in terms of materials and still has some resolution enhancements possible. Still, for a clean, efficient machine with minimal subsequent cleanup by the user, some future nano-scale welding is really the sane way. Also, nano-scale welding will give materials a z-axis strength missing on many current builds.
UPDATE: I wanted to include an example of what I’m talking about using lasers (or as Dr. Evil would say: “lasers”). Here’s a short video from the Marshall Space Flight Center where they describe using Selective Laser Melting to form working rocket engine injectors (goto 1:07 for that bit). A closely related technique, Selective Laser Sintering is where I think practical 3d printing has to go.