The incredible shrinking machine
No comment
The RepRap project announced the RepRap Version III named “Huxley” a little over a month ago. 1 It’s based on the mini-Mendel by Ed Sells. Interestingly, if you compare the last few iterations of the RepRap, from the Darwin, Mendel, and mini-Mendel – you’ll notice the build volume is shrinking. As these RepRap designs have evolved, the replication time has gone down, non-printed parts have been decreased, but the usable build area keeps decreasing. 2 I suppose if self-replication is the highest goal, the machine and build volume should continue to miniaturize. 
This raises an interesting question about the optimal size of a 3D printer. How much build volume do you want or need? I tend to use an 80-20 rule-of-thumb for such things. I’m usually happy with a solution that can do everything I need 80% of the time. 3 I suppose if you refused to compromise on build area, you’d have to go with a commercial printer or design your own super-mega-9000 giant printer.
I normally print out small parts, replacement parts, MakerBot parts, or toys – so, I’ve never felt constrained by my Cupcake’s build volume. And, I’ve felt even less constrained as new printable upgrades are developed that allow you to increase your build area. I can’t say for sure, but I suspect build volume isn’t that big a deal for most other people – otherwise we’d see more printable volume-increasing upgrades and more attempts at the MakerBot longest print Hall of Fame record.
- Photo courtesy of didelco10 [↩]
- Amusingly, it’s likely Webca’s printable Cupcake, which takes a really long time, only made MakerBots even more popular. Go fig. [↩]
- That’s one of the reasons I’ve been using a netbook for the last two years. Long battery life, low price, good portability, but no optical drive – which I never used that often anyhow. [↩]
| Tagged with | build volume, huxley, reprap, self-replication | 10 comments |
10 Comments so far
Nathan Carl Lewis
Come November me and my friend are going to purchase one of the new Thing-o-matics from Makerbot. But I will admit having a much larger build area is something I would love. And I will give you some examples as to why.
With the new release of the Thing-O-Matic with its automatic centering and higher resolution the only thing left on my wish list is a larger build area.
I do a lot of designing of mechanical parts usually having to do with the automotive industry. Not to mention accessories and such. With a larger build platform I could print things like application specific products that are bigger than the 4x4x6~ish current build area. That would allow me to test fit things and prototype them, or even just polish up my printed items and sell them for next to near nothing for cars that dont have much aftermarket or restoration support. For instance, I own a 1987 Chevy Sprint Turbo, and there is next to near nothing for it. My buddy owns a 1988 Dodge Colt and he has the same problem. So that is just one of the ideas I will be working on, and a reason a bigger build platform could open up TONS of opportunity.
And I like opensource anything, free and lets me create and do what I want with out the need for deep pockets.
Just my 2c,
Nate
Erik
Actually, mendel is bigger ‘on the inside’ than Darwin. Don’t recall specs from the top of my head, though. Though I still love the ability to lift my Huxley by the index finger. And typing never gave it much strength, I’ve really tried.
About build volumes, Martijn made the Ultimaker so you can stack it (and it’s already ~21x21x21cm, or more than 9000cc):
http://picasaweb.google.com/erikdebruijn1/UltimakerWorkshopSeptOct#5519856356093443778
Wade
Yeah, even the page you linked to shows that Mendel has a bigger build volume than Darwin, despite it’s smaller footprint. 5600 cc for Mendel vs 5290 cc for Darwin – it’s pretty similar, but Mendel clearly has a larger build volume.
Huxley, of course, is smaller, but it’s not meant to be a replacement for Mendel; it’s a bootstrap machine just capable enough to print your own Mendel.
beak90
Keep in mind that a Huxley actually has a bigger build area than a Makerbot does. I’m not sure if the Huxley has the same amount of Z or not, but it is definitely larger in x and y. Also, just because its designed to be a certain size doesn’t mean you have to build it that size. Without changing the printed parts at all, you can make the build area as large as you want (theoretically). Whosawhatsis made an oversized one here: http://www.thingiverse.com/derivative:3053
colorbroken
I’d love a print bed the size of a 5×7″ photograph, something in that area. My intentions are to build things that are camera-sized, and the Cupcake platform isn’t quite that. Seeing a running Reprap at the Makerfaire sealed my intent to build one, so I can finally modify Polaroid bodies. The adventure has just begun!!
VDX
… i think ‘optimal’ building size is depending of some more parameters like accuracy/precision, fabbing speed, available materials and shrinking/warping issues …
My tripod-demonstrator ( http://builders.reprap.org/search?q=assembling+tripod ) has a much smaller building volume, but with mechanical accuracies down to 1 micron and fabbing track-width with paste below 0.1mm (and easy multi-material by changing the cartouches) it’s more meant for building hybride components, than only plastic parts.
Actually i’m nearly finished collecting all the components i need for a cartesian micro-multi-fabber:
- 3 high precise linear axes with 180mm travelling range (max. 100kg load)
- 3 2Nm-steppers+controllers with 10000 steps per rev. and 200kHz clocking speed
- 3 encoders with 4096 (x4) lpr
- 3 inductive sensors/ref-switches with 0.1 micron accuracy
- milling head, paste-dispenser, IR- and blue (445nm) diodelasers for lasersintering
- fiber-laser with 50Watt@1070nm power and 0.01mm focus size for cutting or sintering
- Arduino and software capable of milling and lasercutting
- tweaks for converting 3D-STL in path information for laser-sintering
- and some other stuff too. needed for the build-up …
I think, when it’s running, the main limits will be the calculating speed of the controller (for 200kHz or 10MHz stepping rate) and the size of the toolheads – 100microns for milling, 50 microns with paste, 10microns with the fiber-laser …
The building area (180x180x180mm) is then ‘only’ limiting the maximal size of a single part, but with adding/stacking multiple parts you can get bigger … and with free selectable materials (plastic, ceramic, glass, metal, …) i can build anything i can construct and output as fabbing information …
So it’s more the fabbing speed, accuracy and material selection what’s defining the optimal size of the building volume and footprint of the printer …
Viktor
HotKey
Hi everyone!
I just had the same thoughts last week, so I put together a parametric stage set up to evaluate what could be possible.
http://www.thingiverse.com/thing:4336
Download it and see for yourselves!
SuperMike
I’m new to this (having just “commissioned” my makerbot a couple of days ago (#1021, if I’m not mistaken). One reason there’s probably more demand on the “small side” of the build envelope spectrum is that we’re still somewhat limited by the extrusion speed. Even if it had 1-meter cube build envelope, it’d take a year to print out anything of any substance within it. I do expect to see extrusion improve, but there are a couple of useful things that could enhance larger-object building with current extrusion technology:
1. Big flat stuff: lots of big stuff is really only long or long and wide, these items actually have a dimension that’s not as big as the other two, so we’re making stuff like this thing (which I may actually print). If there were a flat, wide version, perhaps with both a laser cutting head and an extrusion head, a lot of parts could be built that are out of reach of the current bots.
2. Taller stuff: I was daydreaming the other day and realized that a polar (turntable based) bot could, without as much worry about collision, bring two printheads to bear on the same workpiece simultaneously. The g-code(s) would be freaky, and you’d probably want a way to quickly start and resume flow, but it could be done, and if two, why not three, or four?
Phil
Size matters. Think about some uses cases for makerbots…
How about game parts? So, go to a game store and survey all the boxed games. How many of the plastic parts in those games could be made on a makerbot? Not all, Legos, perhaps. But many parts range up to maybe 1 (or 1.5) feet. Can you make copy of any of the model rockets for sale today? No. Model Cars? No. Doll house dolls? No. Doll house wall parts? No. Dinosaurs? No. LGB scale train cars? No. Makerbots can’t really do games until these use cases are covered.
How about one-off classroom demonstration parts? Body parts, molecules, etc. Imagine a high-school class room. How big must the models be in a class room so they can be seen and understood by the students? Again, like plastic models, 4 inch parts are too small. What about an inventor building a demonstration model of some new discovery to show his friends? Same problems, 4 inches is to small.
How about kitchen parts? How many of the plastic parts in a kitchen can be made on a makerbot? Not even the measuring spoon in a post today could be made easily. 4 inches is still to small. x/y/z limits of 1 or 1.5 feet covers nearly every plastic object in a typical kitchen. The biggest plastic object in my kitchen is a cullender. A little over a foot across.
Think a little larger. Ikea. Sells engineered household products. Covered kitchen parts already, but the “rest of the house…” is also interesting. Of all the plastic parts sold by Ikea, even when included in a larger product, how many could be made on a makerbot? Plastic parts seem to range up to about 1 or 1.5 feet. After that other materials are used, wood and metal primarily. Here are some examples. How about clock faces? Can a makerbot make those? No. 4 inches cannot be seen when hanging on a wall. Plastic parts used in offices, inboxes? No. Bins? No. Plastic used in lamps? These are usually assembled from smaller parts, but from 4 inch parts? No. I want to be able to download a clever lamp design off thingiverse, print it, buy a socket, cord and perhaps a shade from the local hardware store and put it together… Can I do that now? No, 4 inches is to small for that.
There are other use cases too, but these are 4 big ones that makerbots currently don’t cover well. This was the primary disappointment that I had when I first saw the limited build volumes in the makerbot products.
Most plastic parts made in the world today are limited to a certain size because the physics of the plastic itself limits the overall size of a finished product.
Cover that size and you’ve hit the sweet spot for a makerbot.
A Build-O-Matic anyone? x and y of about 1 or 1.5 foot, z of about 1.5 or 2 feet and nearly all uses cases for standard plastics are covered.
Phil
tt
a suggestion: http://techreport.com/articles.x/19754