3D printers such as the RepRap and Makerbot turn spools of plastic filament into just about any object imaginable. There’s a problem though: this filament costs about $40 a kilogram, and raw plastic pellets cost about 1/10th of that. Obviously, there’s a lot of room for improvement. The folks at Inventables are throwing $40,000 at the problem in a contest to build a machine that takes plastic pellets and turns it into usable plastic filament.
The object is simple: build a device that takes ABS or PLA pellets and turns them into a 1.75mm filament. The machine has to cost less than $250, be able to add colorant to the plastic, and be usable in a 3D printer. The winner gets $40,000, a laser cutter, a 3D printer, and a CNC milling machine courtesy of Inventables. Sign up on the official contest website and don’t be shy about sending your progress into the Hackaday tip line
If you’d like to get started, here’s a great page that goes over the basics of plastic extrusion, and a few attempts (1, 2) from [Adrian Bowyer] and [Forrest Higgs] that show exactly how hard this is. There’s also the Filabot that had a successful Kickstarter, but there’s apparently been no (or very limited) progress in the four months since the Kickstarter. I’ve even given this idea a go, but am currently stuck at manufacturing a proper auger. To put this in perspective, this is the moonshot of the current crop of 3D printers; a simple device to lower the barrier of entry to 3D printing is desperately needed, and we’ve got to give props to the Inventables crew for putting this contest together.
47 thoughts on “Win $40,000 For Squirting Plastic Out Of A Nozzle”
How about a stake in the device and credit in the patents that will come out of it? I don’t see anything about the results remaining open source.
G’Day, If you read the full media release it states
“The first team/person to build an open source filament extruder”
Which makes this so much more honourable :-)
There’s what I was looking for. Thank you! Guess I know where my long weekend is going.
turning pellets into a plastic filament is very easy, the problem is getting it to stay as a straight filament while it cools so that it can be wound onto a spool.
It would be much easier and probably cheaper (1 machine rather than 2) to modify an existing 3d printer to take pellets.
A hopper can feed a melting pot with an auger to force the fresh plastic straight out the nozzle, similar to how the 3d printer that can do colour mixing we saw a few days ago works
I agree. It seems very inefficient to melt pellets into a strait filament just to be melted again when coming out of the hot tip.
I agree, making filament is retarded, we should be making 3D printers that simply take pellets, AND can color as needed with say… an open source color printerhead
G’Day, Can you please add a link/url to the Colour Mixing article you mention in your comment :)
Was thinking the same exact thing. Or at least if you are going to make your own filament, that you fix your printer to accept variable width filament so you don’t end up solving another unnecessary problem.
Also +1 on the request for the color mixing bot link.
Actually, come to think of it, with pellets, it seems like color mixing with a single head (via controlling where the pellets come from) would be so much easier…
$40,000? Are you kidding me? If I designed this thing, and it worked, I’d go into business!
All you people saying that you’d go into business with it don’t realize there are already plenty of commercial machines that do this. You can’t patent it so open source is the only route.
You can patent improvement on existing technologies (like say alternative methodologies that use lower cost equipment) and I wouldn’t completely discount someone pioneering a new technique for filament creation in the course of this contest either. The maker community is full of a lot of extremely bright people. I really like the immediate jump to eschewing a filament maker altogether made in the comments here.
Do they have a statement of participant’s rights on the contest? I didn’t see one. They may. Never make the “it’s been done” argument when it comes to new technology. That’s how businesses lose millions on patents.
Filament is better than pellets in 3d printing because you can adjust the flow rates with some degree of reliability. It is hard, even with flow controlled augers and such as the pellets/air gap variances cannot be controlled. Also not being able to heat the pellets evenly makes for bad prints.
You can’t have a single machine that extrude the pellets into a filament and then use it for 3D printing? as a bonus you would not have to cool it down/warm it up between the 2 processes.
The point is you *do* need to cool it to a point where you can control the flow for extrusion.
The trouble with making filament is not that it’s hard, it’s that it is hard to do consistently. You want filament that’s round and doesn’t vary more then 0.05mm in diameter.
That’s why you currently pay $40 per kilo. There are only a few locations that can do this accurate. There are china based sellers that go as low as $25 per kilo, but the quality is china crap.
(I’m a 3D printer owner, I know my filament)
>this filament costs about $40 a kilogram, and raw plastic pellets cost about 1/10th of that.
Let’s say I wanted to scare up a few pounds of ABS pellets in short notice.
Granted, it was a few months ago, but I couldn’t find anybody on-line selling these mythical cheap plastic pellets. I’m sure the big guys selling to injection molding companies deal in the TEUs and railroad car quantities. They might see $4/kg prices
There were some sellers on feebay, but the prices weren’t all that great. ($10 for two pounds, which is a saving, but more like 4 to 1 rather than an order-of-magnitude difference)
If the sponsors want to spur some innovation, then maybe they could, you know, add “plastic pellets” to the inventory over at inventables.com , pricing the stuff at cost+shipping+handling or something. It seems surprising that no one had thought of this yet.
I looked on Ebay and found 20-lbs for $35 (1.75/lb or $3.85/kg)
That looks like a off spec material that is left over. You have different grades of ABS with diferent MFI’s it is not just like throwing anything at it. Also extruded
The contest site says “This is between 5-10 times the cost of the raw resin pellets” and Hack-a-day just took the 10, but you can buy pellets from here for about 9$ per kg (1/4 ish) plus shipping and tax. I hope that such a device would cause a large increase in demand for the pellets, which should make them easier to find and cheaper.
Well keep in mind that there already exists a large market for these things. They’re used the world over to make plastic objects in injection molding machines. It’s just been hard to buy them cheaply in small quantity until recently.
18 months ago, all I could find were flea-bay sellers selling the stuff at par with 3mm filament.
Both are listed now as “sign up for email alert when we start to sell em”. No word on price yet.
Im not gonna say here but i know of an extremely cheap and effective way to do what they want. Nice prizes too. could use ALL of em. $40G’s, thats a new smart car and a segway to me.
My 2 cents:
Making a machine that can recycle thrown-out plastic into filament would be awesome, for us, our wallets AND the environment
At work I see pounds of plastics being thrown away on a daily basis because it’s such a hassle to put them in a recycle bin.
Here’s a video of some crazy guy doing it; though at this point quality seems the major issue:
Actually your value hypothesis is off, currently spools of plastics from companies and a variety of manufacturers are around $25 per pound, very reasonable, and only will drop further with larger market adoption. The filament has to be ultra high quality, that means, in part, there has to be a consistent width of plastic, issues that would result from the use of low quality plastics could include jamming issues to outright failure and destruction of the machine. Additionally most companies specialize in calibrating and testing plastic spools for the best resolution, something that would not be possible if individuals “created” their own. This says nothing of the implication that “double heating” the pla (once from pellet, and a second time as filament through the printer) might also require additional polymer binding material that could be hazardous, and much more expensive than purchasing filament spools. Unless you own stock in pellet plastic companies, I could easily think of 100 more worthwhile projects to improve 3d printing with the $40k. In closing, the doubleheating issue may make this infeasible and putting through low quality filament spools can have devestating effects on these expensive machines.
Gee I wonder how many pounds there are in a kilo?
Get with the metric system [man!]
100? name 3.
lemme guess, you sell filament?
A: 2.205… ish :P
I’ve heard the ‘double heating’ argument before, and I’ve got to ask this:
Isn’t all 3D printer filament ‘double heated’? As in, once to make it into a filament, then again to make it into 3D prints?
Not neccesarily, if it is virgin material you’re getting the result of polymerized resins that have had no heating applied to them. At least not anything substantial enough to drive off volitile and neccesary constituents that keep the plastic from being brittle.
No – you could make filament straight out of the compounder (the machine that makes pellets).
/me grabs some milk jugs and the blender and sneaks outside…
Has anyone found any good specifications for the ABS plastic and ideal working properties? Specifically, preferred melting temperature and pressure, as well as the volume of 1kg of liquid plastic.
I’m an industrial mechanic and have a few ideas…. Maybe I missed the information in one of the articles.
Your best bet is data sheets from pellet distributors/manfucaturers. Also consider the reprap/makerbot forums since these machines melt those plastics. Some rough numbers I’ve found for melting points are 219F(ABS) and 175F(PLA).
Though not too much ;)
You also need information such as the flow rate, elongation at break etc. There is alot more than just the temp.
Ok, For the company I worked for I built A protype Extruder that would extruded a 220mm filament.I don’t think changing the nozzle would be that hard. Most designs I’ve seen uses air to cool the plastic. we use water misting system. the cost was around $500 plus a ton of machining time.We feed off 3 different hoppers to give us the varies of color that we were looking for.
If anyone wants to team up let me know,Neonmerlin17@gmail.com.
Looks like these guys did it earlier unless I am missing something.
The point is to make it under $250. The un-assembled kit for filabot is $350.
That thing will never work anyway. It is alot more complex than what they think it is to re-compound material. That will be full of air bubbles and un-melted material. As well as that if you don’t know what your doing and mixing materials some can give of harmful gases and some can be explosive so just saying you can recycle this and that without knowing what it is isn’t right. I can list a huge amount of other things why they will never produce that!
Don’t own a 3D printer, but from what I read the filament is actually pulled back out of the hot end just a little to stop the flow of resin when moving from one point to the other without printing. Need a hard filament to do that.
I also worry about melting the pellets then remelting in the hot end. If filaments are drawn directly from fresh resin i.e. you make either pellets, filaments, rods, or sheets. Not pellets then filaments.
This is not that complicated, don’t use mechanical push with a screw driver, use preassure push like a tube of toothpaste.
Put pellets in hopper, apply pump to remove air.
Heat, till melting, where lack of air removes carbonation.
Stir to make sure all is liquid inside hopper
Apply pressure with a piston in the hopper pushing down the entire side of the inside of the hopper.
Push it down a cone into an exit hole.
That exit hole has back preassure by winding a few times after cooling stage, that will not allow air bubbles into the mixture if any are left.
It is not that complicated, and can be built really cheap, you can make a stable vacuum with a backward bicycle pump, and a hopper like a piston with a high temp gasget attached to the push item, weight can even be a large screw press above the piston, if you want the press to be really cheap. Think of one of those old Presses for printing where news paper ink was pressed into parchment.
The heating till melting can be on the bottom part cone where a stir stick turns, and can be any torch on the side, then a wire filament can be the final heat to make sure it is liquid. After passing into 3mm out tube, a cooling fan blowing on a heat sink, cools the filament while back pressure keeps it solid without air bubbles.
Liquid should not be moved by a screw, it should be moved by pressure. use the screw press to create the pressure that pushes out the liquid into filament form
Hopper is a sheet of aluminum made into a cylinder, with straps around outside for extra strength.
piston is a thick piece of metal with a fat gasget around outside so plastic is not pushed up out around the piston, and a screw push from a hand press mechanism from the top pushes piston down into the hopper.
under the hopper a sheet of high temp plastic as a separator
under that a copper cone the stir stick could be complicated, might not even be needed if you vibrate the hopper from the outside. That is the part that is heated where all the plastic in that is liquid.
Then an exit hole to a tube with added heat to make sure it is always liquid there. 3mm
Then a spacer of some other metal. 3mm
Then a winding tube for output and back pressure 3mm with a heat sink and fan blowing on it to cool at that stage. Any form of back pressure could be used but forcing cold filament to wind in a circle would work.
And you have a cheap pellet to filament system.
JB Weld could probably be used to smooth inside of hopper seam.
Although best to put a plastic box around the whole thing so if it breaks liquid ABS does not fly all over the place, for safety.
also a relief valve could be added if pressure gets to high as a secondary exit of liquid for someone that over cranked, but someone would not do that if they had a pressure meter on that safety valve.
Sounds cheap and easy to make.
It wasn’t easy but I did it, see my video:
It is now in validation at Inventables.
Good speed on that also, that looks great. converts to filament at a pretty fast output speed.
And great music there, one of the best songs.
How does that (the extrusion part) compare to what these guys are doing? (They’re building a 3d printer with built-in extrusion.)
Is this contest still open? Haven’t seen a winner announced yet.
I have no idea why everyone thinks this is something new . I have been using a Nordson Hot melt system that is developed for glue, I use it to melt a resin. It is supplied with a melting pot temperature control, positive displacement pneumatic cycliner, heated hose of any length and nozzles of any size. The system can be controlled manually, foot pedel or through a Plc. ITS NOT NEW GUYS
Please be kind and respectful to help make the comments section excellent. (Comment Policy)