The next frontier of desktop 3D printing is multi-material and multi-color prints. Right now, you can buy a dual toolhead for a Lulzbot, and dual toolheads from other companies exist, although they are a bit rare. In the next few years, we’re going to see a lot of printers able to print dissolvable supports and full-color 3D printers.
Printing in more than one color is almost here, but that doesn’t mean we’re on the cusp of a complete revolution. Multi-material printing is lagging a little bit behind; you’ll be able to print two colors of PLA next year, but printing an object in PLA and ABS is going to be a bit tricky. Printing something in PLA and nylon will be very hard. Color mixing, likewise, will be tricky. We can do it, the tools are getting there, but think of this year as a preview of what we’ll be doing in five years.
The i3 Quad
Last week, Prusa rolled out a 4-extruder upgrade to the latest iteration of the i3 Mk 2. This weekend at Maker Faire, the world got its first glimpse of this machine in action, printing out multicolor fish.
Here’s how Prusa’s 4-color mod works. A small, SSR-based drive is wired into the controller board, allowing four different stepper drivers to be controlled independently, but not simultaneously. These steppers feed filament through a Bowden tube to a four-way Y adapter just above the hotend. To change colors, the printer backs filament out of the Y-adapter very quickly, pushes the new color in, and squirts the remainder of the previous color of filament out onto a waste tower.
When this mod was first released, there were a few questions about how wasteful this waste tower would be. It’s actually very optimized, with photographic evidence:
The ‘waste tower’ has three ‘slots’ where color changes happen. The printer only wastes filament when it needs to, and the slicer is as smart as it needs to be. There’s very little waste here.
To me, this is how we’re going to do multi-color 3D printing in the near future. Stepper motors are cheap, boards that switch motors on and off are cheap. Hotends are expensive, and dual hotends mean leveling and offset problems. That said, this setup has limitations: there is no color mixing, and printing with two materials of different temperatures (PLA and nylon, for example) is going to be hard or impossible. It’s still awesome.
Color Mixing
On the other end of the spectrum is ORD Solutions RoVa4D printer. Your desktop inkjet printer uses cyan, magenta, yellow, and black ink to print any color. The RoVa uses cyan, magenta, yellow, black, and white filament to print objects in any color. There are two more extruders on this printer designed for flexible and support material. It’s a printer with seven extruders feeding into one watercooled mixing hotend. It is a triumph of 3D printer engineering.
Right now, slicing multicolor prints is hard. ORD has a tool that allows you to take a standard .STL file and ‘paint’ an object so it can be printed on their printer. That’s amazing, but it’s closed-source. Software devs interested in 3D printing, there’s a project for you.
As far as print quality goes, the RoVa is in a weird place. Right now, nobody is doing mixing color 3D printers. Having that capability at all deserves a thumbs up. However, there are color artifacts in all the prints. Looking at the printed Einstein bust, I can easily see a streak of pink on the layers where the lips were printed. If I were awarding points, I don’t know if I could knock points off for this. It works, but it is the 3D printed version of a color bitmap print made on a dot matrix Image Writer printer. The quality will go up, but I don’t know if it’s going to happen with this machine.
This is pure speculation, but I don’t know how ORD Solutions can afford to sell this printer. Right now, the seven extruder version, with a watercooled hotend, sells for $3750 USD, with delivery in May, 2017. The general consensus is that it’s too inexpensive.
2017: Dual Multi Extrusion
3D printing is slowly moving towards dual and multi extrusion. Color is a big reason, but not the best. LulzBot’s FlexyDually extruder makes the argument for combining flexible and rigid materials for some metamaterial weirdness. Dissolvable filaments allow any dual-extruder printer to create any object without regard to geometry.
Dissolvable filament has been around for a while, from Makerbot’s water-soluble PVA filament to dissolving ABS in acetone to limonene-soluble HIPS. These filaments, with the possible exception of HIPS, never really caught on as a dissolvable support material. This might be changing. At MRRF last March, E3D announced they were working on Scaffold, a water-soluable support material that has the same base material as the stuff gel caps are made out of. To get rid of Scaffold support material, just put your part in the dishwasher. It’s coming, and soon we’ll have a real use for dual and multi extrusion.
We’re slowly creeping towards a compact, desktop device that’s able to print any object in any orientation, in any color. That’s the goal we’re all working towards, anyway. We’re not there yet, but it’s getting there, and we’re really looking forward to where this tech goes.
What about the diamond hotend from reprap.me? they being doing this for a while…
Because if you’re going to do a mixing hotend, four colors (CMYW) is the minimum needed. CMYKW would be better. The diamond only has 3.
Given that, and the lack of stuff on the software side, the diamond is ultmately meant for one-color-at-a-time printing, like the Prusa quad. The prusa is ultmately a simpler solution, making the diamond redundant.
“Because if you’re going to do a mixing hotend, four colors (CMYW) is the minimum needed. CMYKW would be better. The diamond only has 3.”
Color mixing from the diamond look like “pastel” colors.
I like this one http://www.ordsolutions.com/rova4d-full-color-blender-3d-printer-pre-order/
A bit high for my budget :(
People have been dual-printing for a while now. I just started last week after upgrading one of my printers.
Way too expensive for sure, but I have to wonder: how difficult would it be to “just” come up with a 5-to-1 assembly functionally comparable to the Rova4d, but compact, cheap and capable of being fitted onto some mainstream machines like a Prusa i3 or so, the way the Diamond is?
I don’t mean so much the work that would go into actually manufacturing such an assembly, but rather the design and engineering that precedes it.
The diamond hotend can make black by mixing equal parts of C, M, and Y. What the hotend can’t make is white. Any print made can be no lighter than any primary, and mixed will only make it darker.
Nice bust of Grandpa Walton…. ;-)
The development of home 3D printing reminds me of the early days of home colour printers. Well home printing in general.
Dot matrix 9 pin, 24pin etc
Thermal colour printers.
I remember seeing the demo prints from some NEW colour dot matrix at the time and they were amazing!
3D printing appears to be traveling along just as fast.
I still have nightmares about color dot matrix printing, the sort of ones where given application has it’s own shitty print drivers and screws up your $$$ ribbon in one muddy awful print.
Yeah, i have been bracing myself for the 3D equivalent of the “DTP Revolution”, when every Joe and his mongrel was suddenly a design expert because the had gotten a free publishing software off the cover of a magazine.
I am happy to see that the 3D printing revolution is finally starting to output some useful stuff and not only scruffy dashboard figurines.
Saw the Ord color 3D printer at Makerfaire on Saturday. It was tempting. Also saw the Wasp printers with clay adapter kit (still need a kiln) and some laser resin printers as well.
The temptation to buy any of these printers was so great that they all ‘canceled’ each other out. After reading this article I will once again ‘wait it out’ until next year before buying my 6th printer.
I’m the same way with mid engine supercars. ;-)
I think that the real hard part with these multi-extruder printers that have a single hot end is going to be the filament. Even with the same type of plastic all from the same manufacturer, different colors of filament can require different temperatures to work properly. I have a black PLA that needs 10-15 degrees more heat than the same brand of purple PLA.
That’s a little counter intuitive, you’d think the black would soak up the heat like anything and it’d drip out unless you backed it off. I guess it could be that carbon black in the pigment is making it more thermally conductive and sucking heat out back up the filament to atmosphere.
Although should it be a market demand, that all colors print at the same temperature, one supplier might make it happen.
The color of the filament within the dark innards of a print head is unimportant. Color is all about light reflection and where there is no light, there is no difference. Filament is a lot like so many women out there. Some are prettier than others, but they all “look” more or less equal in the dark. :-) (Differences in behavior and odor are still possible, however. :-) )
The problem is related to the coloring and has been an issue in plastics for a very long time. Just different colour of the same material can have significant variation in their expansion and contraction when curring and also their rigidity.
I’m looking into getting a 3d printer are there any people would recommend. It’s seems people are very happy with the higher end ones and the budget models are very batch dependant.
Are there any diamonds in the rough? Prusa i3 on ebay is going cheap at £145 some assembly required (however I am put off by the reviews saying the base parts are often mis-assembled or even broken).
The version with metal parts (i3 pro B geeetech) and 2 X-bearings needs heavy modifications to work flawlessly, because it has X-axis bearings with play in astronomical units. Other than that it can print a benchy just fine with only little overhang. Modifications needed: new bearings, 4-bearing mod or tensioner. then PSU fan (overheat), reposition PCB fan, bendo-straightenerr plate to get 90° for X-Y, underplate to get stiffnes to not bedlevel like crazy, fume extraction box, nozzle cooler duct, LED lights…
I bought it cheap to play with it. If you need perfect printer then get prusa from Prusa :-D
This would be for some rapid prototyping, so I am thinking of it in contractor-hours. Pursia i3 was 145 but if it takes more than 2 hours to get up to speed with something like a CTC duplicator (£265) it’s not really cheaper,,,
Check out the Wilson II printer. It has very clear English assembly videos, priced complete around $500. The Wilson II also has a pretty decent help forum.
Amazingly smooth transition from never having printed anything to printing stuff daily.
I see a 3D printed steak. The future is now!
If the entire feed system right up to the point of extrusion could be refrigerated, ‘printed’ hamburger patties could be a thing. Squirt out and parcook the ground beef through a hot nozzle onto a clamshell grill then close that to finish cooking and mash the meat so it holds together.
No need to stock different size frozen patties and the various seasonings could be applied as the extruder spirals out the patty. Could do the same with pretty much any ground meat.
Check out the Wilson II printer. It has very clear English assembly videos, priced complete around $500. The Wilson II also has a pretty decent help forum.
Amazingly smooth transition from never having printed anything to printing stuff daily.
This isn’t 100% the same thing, but a couple years back, I wrote an open-source color slicer for 3D files. I was slicing them into images for printing, so there’s no GCode generation, and the output is rastered images rather than vectors. Here it is (https://github.com/enjrolas/glasscutter). If anyone is working on color 3D printing/GCode generation, get in touch — I’d be stoked to talk/help.
I worry about the chances of the filament, especially suboptimal filament, breaking during retraction and then blocking up the whole works with a broken strand of filament in the head, unable to retract, and blocking the path for the next filament. The big question is can the software detect that? (I think not) and What will be the consequences. (probably air-printing) Still this opens the door to finishing up one roll of a color and following it up with another of the same color coming down another channel. I doubt the software can do that today, but there is potential. This kind of system really needs more sensors to indicate actual filament travel and position, but small steps are best. One at a time.
I would like to see something similar to Stratasys’s method of automatically switching filaments on their Dimension FDM printers but patent issues could be a problem.