Low-end of 3D Printing gets a little lower

February 23, 2010 8:36 AM

Saw on Gizmodo today a DIY 3D Printer, based on an open source design, that prints ceramic structures ready for firing. 3D printing has been around for years, but the low-end of this technology fascinates me. Once these machines get more widely into the hands of non-engineers, how many Bathsheba Grossmans out there will emerge with ready-to-print designs for craftsmen around the world to tweak and innovate? Twinkling of a peer-to-peer manufacturing revolution?
posted by cross_impact (30 comments total)

See also Reprap, a cheap 3D printer open source project that's hoping to one day create a 3D printer that can print copies of itself. It currently can produce 50% of its own parts. Since that would mean you could manufacture (and redesign) the printer for the cost of the energy and materials to build it, that would mean a massive number of people would have access to a rapid prototyper, changing manufacturing and the economy as we know it.
posted by mccarty.tim at 8:56 AM on February 23, 2010 [1 favorite]

Er, this is actually based on the Reprap. Apparently, in line with the ethos of the project, someone did see fit to modify the design. Very cool!
posted by mccarty.tim at 8:58 AM on February 23, 2010

As I kid I lived in Iowa. Every small Iowa town has as a major central landmark the grain elevators. They are basically concrete tube skyscrapers. While I lived there, our town built some new ones so I got to watch. It's pretty incredible because the entire tower is a single piece of concrete.

They don't build a huge form. Instead they build a form that's maybe 10 ft tall and fill that, then just as it dries out they raise the form and fill again. They are kind of extruding the concrete. I can't remember if it took a few days or a few weeks.

I'm wondering if a 3D cement printer could do the same thing. The advantage would be a finer resolution because subsequent layers wouldn't need to be held up by thick wet clay but instead by thin(ner) hard cement.
posted by DU at 9:00 AM on February 23, 2010

Very interesting concept, but as you saw, the precision of this unit is still at the objet'art level and nowhere near precise enough for engineering. And it's, um, clay.

The immediate future of peer-to-peer manufacturing is still in designing/programming/prototyping, and maybe assembling finished products or subassemblies using off-the shelf components which are much more widely available these days. I've started buying sample quantities of electronic components and subassemblies from China, and it's very feasible for one or two person operations to produce an electronic design and have a short run of populated boards assembled overseas.

Concepts like RepRap (Tim's link above) are also promising, but it's hard to envision one in every garage. Might soon see something like this in a store like IKEA, producing cheap panels and simple components to order.

The most immediately-useful application of 3-D printers and mills is to produce physical prototypes of design ideas (eg small appliances etc) for faster evaluation of a design. Architects are also starting to use them for model-making.
posted by Artful Codger at 9:04 AM on February 23, 2010

It is super-cool. I'd have to say though it is only DIY for small values of Y. You'd need to have pretty excellent mechanical, electrical, and software skills to build something like this. Not to mention plenty of space for a well appointed workshop.

My matter compiler operating in an extruded diamond vacuum environment is one step closer. Jetpacks must be just around the corner.
posted by Babblesort at 9:04 AM on February 23, 2010

Also, Gizmodo's link isn't that helpful. Here's the blog entry.
posted by DU at 9:06 AM on February 23, 2010

I'd like to see this thing in action. It's a crime the designer didn't post it to youtube, even if it's slow and jerky and makes awful noises.
posted by mccarty.tim at 9:08 AM on February 23, 2010

bablesort:

the value of Y is larger than you think. most of the heavy lifting engineering-wise has already been done on the rep rap and derivative stuff. it's almost "insert tab a into slot b" now.
posted by Dr. Twist at 9:11 AM on February 23, 2010 [1 favorite]

You know, they have machines that cost thirty to forty thousand dollars that really do a remarkable job with rapid prototyping. I saw a short that Leno did about a machine that could 'scan' a car part with a grid of lasers, spinning it slowly to look at all the angles, and then drive a machine to 'print' a near-perfect copy out of very tough plastic. One machine I looked at could actually make simple mechanisms, by using layers of water-soluble plastic... you just rinse the finished part, and the stuff you don't want dissolves, leaving you with a working ball joint or what have you.

Man, I'd love one of those to play with. I wish they didn't cost as much as a high-quality car. :(
posted by Malor at 9:13 AM on February 23, 2010

Hi DU, I too also grew up with grain elevators (Thunder Bay ON)

Most concrete buildings (including the CN Tower) are built using slipforms as you described. A key element is that rebar is constantly fed in along with the cement, linking the whole thing into one complete structure. So some sort of rod or fiber injection would be necessary to make this concept work on a smaller scale as well.
posted by Artful Codger at 9:17 AM on February 23, 2010 [1 favorite]

Also, related: 3D Printing of Human Organs, using the same technology. There is also a TED MED talk that shows this in action.
posted by dobie at 9:27 AM on February 23, 2010

Once these things get popular, the IP lawyers are going to have a field day.
posted by ChurchHatesTucker at 10:01 AM on February 23, 2010

Joke's on them. I'll just print out a bunch of copies of open source robolawyers.
posted by mccarty.tim at 10:24 AM on February 23, 2010 [3 favorites]

There are some foundries that will print your 3D computer models in plastic for you, for a reasonable price. For example: Shapeways
posted by Nyrath at 10:26 AM on February 23, 2010

Since that would mean you could manufacture (and redesign) the printer for the cost of the energy and materials to build it...

This is actually the thing that leaves me scratching my head. These kinds of revolutions don't come for free; my gut has often been wrong, but right now it's telling me that the per-unit energy and materials cost of a part will be MUCH much higher with this approach.

Which means that a million people printing things in their craft rooms could have a much more expensive energy footprint than central manufacturing.

Which troubles me.

But, again, my gut's been wrong before. Hopefully it is this time as well. Anyone have any hard information on it?
posted by verb at 10:35 AM on February 23, 2010

Custom prototyping in metal:
EMachineShop
Printed Circuit boards:
Pad2Pad

And I should of course mention the techshop, which has all kinds of cnc and rapid prototyping tools available:
Techshop
posted by poe at 10:50 AM on February 23, 2010

Which means that a million people printing things in their craft rooms could have a much more expensive energy footprint than central manufacturing.

packaging and delivery i rest my case
posted by DU at 11:02 AM on February 23, 2010 [1 favorite]

Anyone have any hard information on it?

Nope. But
1) when folk make things with RepRaps they can repair them.
2) You can buy 1 big spool of plastic or make 100 trips to 'local plastic widget mart'.
3) I don't know the energy requirements of injection molding.
4) People like to make stuff that 'central manufacturing' either doesn't make, or makes but charges too much for. Like replacement parts.
5) I should run some numbers on this.

We don't have a RepRap plastic recycler yet, but we're working on it.

This is part of what's determining our performance envelope,
http://www.foresight.org/gadaprize.php
but it's not excluding us from bolting new features onto it.

Oh and
6) People invariably either complain that new feature X is either a) insufficiently perfect or b) completely unnecessary for self-replication. Mildly frustrating.
posted by sebastienbailard at 11:04 AM on February 23, 2010 [1 favorite]

Hmmm. Good points. Eliminating the transportation and packaging infrastructure is a HUGE win, and I can see the value in printing up replacement parts for things rather than tossing them when they break. Thanks!
posted by verb at 11:11 AM on February 23, 2010

There are also efficiencies made by the hive mind that central manufacturing does not come up with. Since this is open source and community driven, it can make strides with ideas that come from individuals doing the work, the same way that Firefox and Linux can keep up with IE and Windows. It's not hard to imagine someone using this and coming up with a more efficient way to manage the energy that a big company did not implement.

Plus, in an environment of increased competition that comes with significantly more people having their own manufacturing capabilities, the best way for the big guys to compete is to drive down cost and focus on quality.
posted by I am the Walrus at 11:22 AM on February 23, 2010

We have rapid prototypers, laser cutters and a 3-axis large-scale CNC router at the architecture school I teach at. Very cool.
I remember the frisson I felt the first time I held a physical 3d model, of something I'd drawn in the computer, in my hand. Some barrier had broken, felt like I'd stepped into Tron or something.
posted by signal at 11:28 AM on February 23, 2010

See also MakerBot.
posted by Poolio at 11:43 AM on February 23, 2010 [1 favorite]

I'd have to say though it is only DIY for small values of Y. You'd need to have pretty excellent mechanical, electrical, and software skills to build something like this.

Not as much as you probably think. Imagine a pair of saw horses made from 2x4s and 1/4 inch bolts. You'd have to have some pretty impresive skills and equipment to make them if you were starting from a tree and a big lump of iron. Nobody does that, though. The go to the hardware store and buy some 2x4s and bolts and make a few cuts and drill a few holes.

The same factors apply here. I've seen plans for a small CNC router that used MDF, All-Thread (available at Lowes and Home Despot), and in-line skate bearings for the mechanical end of things. No, you're not going to be good enough for making sintered metal replacement hip joints, but it's good enough for lots of hobbyist aplications.

Of course everyone who goes that far ends up building a second one (or third, or fourth) because they feel like they've learned so much making the first one.
posted by Kid Charlemagne at 12:07 PM on February 23, 2010

See also:
http://www.rap-man.com/index.asp
http://bitsfrombytes.com/
http://www.multistation.com/fr/spip.php?article537
http://fabathome.org
...
and probably some others I'm forgetting or haven't noticed yet.
posted by sebastienbailard at 12:13 PM on February 23, 2010

So some sort of rod or fiber injection would be necessary to make this concept work on a smaller scale as well.

One of the transit stations in Calgary used an interesting version of this idea. I found this with a quick look on the interwebs. Something about massively thin, and curved, concrete.
posted by LD Feral at 2:31 PM on February 23, 2010

"This is actually the thing that leaves me scratching my head. These kinds of revolutions don't come for free; my gut has often been wrong, but right now it's telling me that the per-unit energy and materials cost of a part will be MUCH much higher with this approach."

Could be. Thing is the cost of the materials and energy input in something, say a coffee cup, isn't reflected much in the price. A regular restaurant grade coffee cup starts at around $0.50 each when you buy a gross or two at a time. Fancier cups can easily cost an order of magnitude more. The material and energy costs are roughly the same.

Custom anything is the real power of these technologies. Maybe I want a 7oz cup that is the same width as a 10oz just 30% shorter or a cup in the shape of Florida or a coffee cup that is truncated cone shape (great for drafting board work). Can't buy it anywhere so just crank one out in the shop. Being able to make something that isn't available from an efficient, mass market producer is extremely satisfying.
posted by Mitheral at 4:39 PM on February 23, 2010

See also Shapeways, sort of a cross between 3D printing & Etsy.
posted by yoga at 5:19 PM on February 23, 2010

I suspect that there will not be a revolution of home manufacturing. The economies of scale will favor a large scale manufacturer buying arrays of devices, locating facilities near cheap power, and proximity to raw materials (thereby getting economies of scale). They will ship these devices to you from regional production centers optimized to balance transport costs vs. production costs.
I do think there will be a surge in variety and customization. When you can download your 3-d model, plug in a few simple options and get something totally custom shipped to your door in 48 hours form a 3-d printing facility that will be pretty amazing.

I think this is about 1978 right now in terms of 3-D printing compared to the PC revolution. Out somewhere right now is another Bill Gates dropping out of college to go build something on makerbot using blender.
posted by humanfont at 6:23 PM on February 23, 2010

there are additive and subtractive 3d printers. Subtractive is also called CNC, though that usally means an expensive high precision industrial gizmo
Build your own
or buy one
or a smaller one

The current issue of Make magazine (sorry, the digital edition is behind a paywall) is all about this.

Makerbot is based on RepRap but diverges from the RepRap design in that it isn't intended to reproduce itself.

There also a method of 3d printing glass with frit bonded with dextrose. I think it will sinter under heat, but I'm a little sketchy on the details.

Due to the open source hardware movement the price and accessibility of all this stuff is getting the bottom knocked out of it.
posted by warbaby at 9:42 PM on February 23, 2010

Makerbot is based on RepRap but diverges from the RepRap design in that it isn't intended to reproduce itself.

Over on the RepRap wiki we're trying to bootstrap more "Bootstrap" machines which we call RepStraps.

http://objects.reprap.org/wiki/SGBot
http://objects.reprap.org/wiki/SamBot
http://objects.reprap.org/wiki/HydraBot

People like to make RepStraps, and they're useful, and there's lots of room in the wiki. The only reason Makerbot isn't at RepRap.org is that the dev didn't want to put it there.
posted by sebastienbailard at 1:18 PM on February 24, 2010