Slashdot Mirror
The Modern Ease of 3D Printing
An anonymous reader writes "What will it mean when 3D fabricators become cheap and common? A NY Times article explores the ease of copying objects by scanning them with NextEngine scanner and sending them to 3d 'print shops'. The experiments were done with Legos because most of the things around his office were protected by copyright. What will happen to the economy for engineering when we can just download a pirated description of a machine and 'print' it out? 'The world is just beginning to grapple with the implications of this relatively low-cost duplicating method, often called rapid prototyping. Hearing aid companies, for instance, are producing some custom-fitted ear pieces from scanned molds of patients. Custom car companies produce new parts for classic cars or modified parts for hot rods. Consumer product makers create fully functional designs before committing themselves to big production runs.'"
"The experiments were done with Legos because most of the things around his office were protected by copyright"
Um, the Lego folks might want to have a word with him...
There are a wide variety of technologies in the marketplace and each have their advantages. Alas, I couldn't write a survey. The Z Corp models look flashy in the pictures because they're in full color, but they're probably not the strongest.
Some of the other systems from companies like Dimension or Stratasys use stronger plastics but can't produce multicolored items.
Some can produce fully working items right from the printer . They deposit two types of material: one soluable and one insoluable. After the thing is printed, you wash away the soluable stuff and the gaps open up. It's amazing. I've played with fully adjustable crescent wrenches that are built with almost the same precision as the ones from Sears. The plastic isn't as durable as metal, but you can certainly build things with the wrench. I'm told one of the cooler demonstration items is a bicycle chain that's fully assembled after the wash.
In some sense, these pre-assembled machines are better than traditional manufacturing techniques because you can build working items inside of sealed shells. There's no ship-in-a-bottle paradox because everything is built from the bottom up.
One of the reasons I wrote the piece is because things are getting pretty cheap. Not Game Boy cheap, but something that's in line with the historical cost of photography. We're not at the introductory price of a Kodak Brownie (supposedly $1 in 1900), but we're near the price of early cameras when adjusted for inflation. The NextEngine costs $2500 new and the print shops will build items for about $70-$200.
We're getting near affordability for the "prosumer" who might want a hobby. I can imagine that these devices might be very useful to model train hobbiests, artists, and others. One artist I know builds Joseph Cornell-like boxes filled with historical scenes. They're great, really.
Rep-Rap The open-source rapid prototyping system.
I've been using rapid prototyping machines of various sorts for 4-5 years now. I've been working with the NextEngine scanner since its introduction less than a year ago. Before that, I've used Coordinate Measurement Metrology (CMM) devices, calipers, datasheets, and a little artistry to reverse engineer parts and assemblies. Here are my impressions:
We are not even close to the sort of society described in Neal Stephenson's The Diamond Age , where everything is manufactured on the spot, rapidly and on-demand, from constituent atoms. I can see how we can get from the current state-of-the-art to that eventuality, but it is still a long way off.
The quality of the parts one can get from a rapid prototyper are just that: prototypes. Depending on the prototyping technology, manufacturer, and capital cost, one can get parts with minimum feature sizes of 0.025" - 0.005", with comparable dimensional tolerance, with part costs of $0.50 - $40 / cubic inch, and build times of tens of minutes to hours per cubic inch. I apologize to those who are metric-only. These kinds of parts usually require at least some post-processing, usually to improve the surface finish or strength. Some post-production machining may be necessary to firm up critical dimensions (for instance, reaming out and/or tapping holes).
The options for finished metal parts are slim. What can often happen is that the rapid prototyped parts are used in mold-making (e.g., for investment casting). There are now a few machines that can create metal parts by melting or sintering metal powders, but they are frightfully expensive, the resulting parts require post-processing on wear surfaces, and the material strength is significantly less than cast or machined metals.
With regards to the NextEngine scanner: it is a fine piece of work. It allows complex objects, particularly ones with compound curving surfaces, to be brought into a computer faster and more accurately than building a CAD model from scratch. It isn't all automatic, and a fair bit of polishing is needed to take the output of the scanner and make usable parts (either reproductions of the original part, or else parts designed around the original part). It is a lot easier than other laser scanners I've worked with, but is not within the realm of an Average Joe. Even though we aren't there yet, we are getting closer. Rapid prototyping and reverse engineering are invaluable tools that seemingly remove the boundaries between what can be designed, and what can be manufactured.
I can see how, in the very near future, rapid prototyping will become more like rapid manufacturing of one-off parts. For instance, being able to create a custom metal implant, like a skull plate, overnight. This kind of thing canbe done now, but it is far from common, and doesn't have one-day turnaround. Another for instance, mentioned by others, is "printing" out an out-of-stock part for an old car. I don't think your average mechanic will be doing that anytime soon, especially since you'd still need a machine shop to do the post-processing. But, a local or regional job-shop, with a legit (i.e., not stolen) database of part models from Delphi an others, could get your mechanic that hard-to-find part in a short time period.
These are exciting times, and the future will only provide more opportunities.
That said, there are likely a lot of things you could do with room temperature epoxies and investment casting, and cheaper modelers for those casts. Cheaper, though, still means typically $10k or more. And even then, the raw materials aren't cheap.
So, how to bring things down? How about the Fab@home project? Nothing like open source 3D fab. I think the current cost estimate to build the tool is about $2500. I'm surprised every engineering school hasn't set their undergrads to work to make something like that (as a useful project, a teaching tool, and a development platform). The material set is a bit limited at present, but really guys, they've demonstrated Chocolate!! what more could you want? :)
Check out Neal Stephenson's The Diamond Age - he goes over much of what everybody is yammering about, but quite a bit more intelligently.
Faster! Faster! Faster would be better!
The Stratasys uses the soluble supports. Having used it several times, I can say it's a pain in the ass as they take forever to dissolve even in the ultrasonic cleaner, and I've resorted to physically chipping away larger chunks to speed up the process. Additionally, 3D printers are all incredibly energy inefficient compared to an industrialized process making the same thing from the same starting materials.
"Politicians and diapers must be changed often, and for the same reason."