Meet pwdr, the open source 3D printer that is a complete departure from the RepRaps and Makerbots we’ve come to love.
Instead of squirting plastic onto a build surface, pwdr operates just like the very, very expensive powder printers used in industrial settings. Pwdr uses gypsum, ceramics, and concrete for its raw stock and binds these powder granules together with water deposited from an inkjet cartridge.
Inside pwdr there are two bins, one for storing the raw material and another for building the part. The part to be printed is built one layer at a time, just like your regular desktop printer. After each layer is finished, a counter-rotating drum scrapes the raw material over the build area and another layer is printed.
There are a lot of advantages to pwdr versus the melted plastic method of printing used in the Makerbot; because each build is self-supporting, it’s possible to print objects that just couldn’t be made with an extruder-based printer. Pwdr also supports laser sintering, meaning it’s possible for pwdr to make objects out of ABS, Nylon, and even metal.
Right now, pwdr is still in the very early stages of development, but you can build your own powder printer from the files up on Thingiverse. Check out the video of pwdr printing after the break.
26 thoughts on “Pwdr, The Open Source Powder Printer”
slow as hell
Slow compared to what?
It reminds me of the CandyFab machine from a few years ago, similar idea but using sugar as the medium, though inkjet does give you a bit more precision than a hot air gun does.
At one minute per layer the latest model I printed out in a plastic squirter would’ve taken about fourteen hours instead of one and a half. And, it would be brittle instead of strong. The diy versions of this type of printer still has a bit of catching up to do but hey, at least it’s a start.
Such potential, should prove so much easier to dye this multiple colours, printing speed is slow but again potential it should be pretty easy speeding that up but detail will always take time especial printing at layers that thin. brittle? im sure that can be solved ether by changing the compounds in use or a combination of changing them and the liquid used to solidify
on the other hand its damd messy
Commercial printers using this method (I used to use a ZCorp one) already use four cartridges for CYMK printing of binder into the white powder medium, which would be a pretty obvious development.
ZCorp prints are also extremely brittle when first removed from the printer – to toughen them up you have a choice of an expensive resin-infusing process or a big bottle of cyanoacrylate. Once you’ve brushed the loose powder off the surface the superglue soaks right into the porous structure – the final result is solid enough to survive fairly rough handling, and easily capable of being used as a master to cast further parts from. A two-step process is the most sensible way to go.
Someone is already using this method as a business. http://www.myrobotnation.com
It could probably be used to produce molds for metal casting, it’s nice to have alternatives to plastic deposit systems.
Mold making is what I was thinking as well! Time to finish our furnace so we can try this!
or you could use wax in an extruder type or mill styrene with it set up as a router.
wax in powder form, then sinter into a solid
All I saw was some printing on a white surface.
Actually, I think this would make much more sense with a laser. Thous the head would be extremely lightweight and could be moved fast. The laser could bake anything that melts.
This water thing will never reach a useful precision and robustness.
Considering one of the most populer 3d printers on the market does this exact thing i think you might be wrong there…
ok Well it uses a binder but thats semantics you could just as easly squirt a binder as water.
britle yes the raw item is britle. but there easy to stirdy up.
just dip them in a binder bath or cure them.
Speed wise… yes this is a bit slow… looks like they whent for a CNC mill type gantry wich is a bit on the slow side as it was designed more for power over speed which isnt realy needed here…
I’d swich to belts and opto feedback then you can push the speed more. alot more
alltho i’m not sure about the liquid output needed fore each layer.
i couldn’t tell in the video but i’d probly go for a offhead resevware as well.
The only drawback to this is the open air mixing method. For most materials, ink jet printing (aka electrostatic deposition methods) have a very tiny niche – and in terms of volume per minute (or hour) is unlikely to ever be a strong competitor.
But that doesn’t mean such materials can’t, don’t or won’t exist – it’s just that there’s work to be done, and you still have to figure out how to mix the water with the base…
However, this is a good first step towards the continuous pizza extruder.
That printhead is moving slooooow.
And simply replacing the multi-nozzle printhead with a single-point laser will make this slow mechanism even slower. Plus they’re talking about SLS of metals as a possibility, which as far as I know doesn’t work without vacuum or inert gas to prevent oxidation.
Might be worth exploring Selective Inhibition Sintering as a possible upgrade instead. It still uses a printhead, but deposits a liquid that inhibits sintering in the chosen powder medium. Plastics can then be selectively sintered in-place, layer by layer, with the addition of a simpler and cheaper nichrome heater. Metals can be bulk sintered outside of the printer once the entire part is built up. It’s sure easier to enclose a part in vacuum or inert gas, than a whole printer.
A single point laser mounted to the same head would be slow, sure. Figure a 2-axis laser mounted above the surface and it’d be significantly faster, just point it where it needs to go.
For that matter, mounting a point laser on a 1 axis mount, moving that along horizontally would be much faster as well. This print system looks like it could do with a lot of optimization. As a first run, though, I like the concept.
You’d use a galvo system, just bouncing the laser off a couple of mirrors.
That’s how laser etching systems work (for cutting you usually move the head).
I don’t think the v0.1 pwdr could produce metal parts, a heated chamber is probably necessary to avoid deformations.
Also I read somewhere (probably the MetallicaRap thread on reprap forums, don’t remember exactly where though) that you need a vacuum chamber, something about controlling the porosity of the object… Maybe we’ll see that on v2!
On the other hand printing stuff using concrete is totally awesome, thinking about art, garden stuff like large plant pots, indestructible outdoor furniture…
Those concrete examples you mention would require a very much scaled up model and it would take a long long time.
I think you’d be better off just chiseling it by hand or even making a clay mold and pour some concrete into that.
Or simply hire a guy – possibly from a 3rd world country – with skills to do that for you.
Looks very similar to the machine i built 15 years ago. But mine was much faster.
US patent: US6403002
Exactly and your patent (and all the others) is expiring.
That is why the hobbyists are now free to actually use this technology instead of it being locked up with assholes like Zcorp.
correction: my website is
Wow Aad your patent is referenced by the zcorp patents. Do you make any money off of that?
With the people saying the print would be fragile it this thing can print with clay then wouldn’t it possible to simple fire the final print in a kiln to turn it into a ceramic?
i though the zcorp printing printing the working wrench was pretty cool because the moving parts works.
I saw this on tv but here’s the youtube video
Please be kind and respectful to help make the comments section excellent. (Comment Policy)