PetBot: Turn PET Bottles Into Filament

Recycling plastic into filament normally involves chopping it into tiny pieces and pushing it through a screw extruder. [JRT3D] is taking a different approach with PetBot, which cuts PET bottles into tape and then turns it into filament. See the videos after the break.

Cutting the tape and extrusion happens in two completely separated processes on the same machine. A PET bottle is prepared by cutting off the bottom, and the open rim is pushed between a pair of bearings, where a cutter slices the bottle into one long strip, as a driven spool rolls it up. The spool of tape is then moved to the second stage of the machine, which pulls the tape through a hot end very similar to that on a 3D printer. While most conventional extruders push the plastic through a nozzle with a screw, the PetBot only heats up the tape to slightly above its glass transition temperature, which allows the driven spool to slowly pull it through the nozzle without breaking. A fan cools the filament just before it goes onto the spool. The same stepper motor is used for both stages of the process.

We like the simplicity of this machine compared to a conventional screw extruder, but it’s not without trade-offs. Firstly is the limitation of the filament length by the material in a single bottle. Getting longer lengths would involve fusing the tape after cutting, or the filament after extrusion, which is not as simple as it might seem. The process would likely be limited to large soda bottle with smooth exterior surfaces to allow the thickness and width of the tape to be as consistent as possible. We are curious to see the consistency of the filaments shape and diameter, and how sensitive it is to variations in the thickness and width of the tape. That being said, as long as you understand the limitations of the machine, we do not doubt that it can be useful.

It doesn’t look like PetBot is open source at the moment, but if you don’t want to figure out how to build it yourself, you can buy one for $400. [JRT3D] also says there is a crowdfunding campaign in the works.

We’ve seen several DIY filament extrusion machines over the years, but  if your primary goal is to save on filament costs they might not be worth the trouble. A 3D printer that can use plastic pellets is likely a better way to do that.



44 thoughts on “PetBot: Turn PET Bottles Into Filament

    1. Hi, I’ve got a PETBOT, myself. The bottles need to be clear of glue and other debris. A quick soak in some hot water, some oil to loosen up the glue and a dip in some soapy water is all it takes to prepare them. Takes about 2 minute/bottle!

      1. Is the length of the filament restricted to what you get out of one bottle or can you chain multiple bottles together to make a big roll? this is absolutely interesting as I spend most of my time in a country that just throws these out and it makes me absolutely sick… also how long is the overall process per bottle?

      1. Unfortunately while it’s technically “biodegradable” PLA doesn’t actually break down naturally in the environment, it needs a purpose built facility to give it the right conditions (>60C, proper microbial cultures, etc).

  1. I Love the idea, I’ve been recycling PET myself, but it’s quite brittle when used like filment unfortunatly.
    Also as rope or fiber it also breaks easily.
    It would be great to find a way to add plasticicer to softten it, like commercial PET-G

  2. In Germany a PET bottle is 15 cent deposit returned at almost any shop. Austria: 29 cent, Denmark 13 cent, Norway 29 cent, Finland 20 cent, Netherlands is 25 cent. (Warning: might contain traces of errors.)

    With 15 cent a bottle of 18 grams it makes around 8 € in deposits per 1 kg PET. 1 kg PETG/175 goes at its lowest for around 13 € incl shipping.

    That is a tight business model.

    1. Some of what you’re paying for with filament is tight tolerances and particular desirable mechanical properties.

      I’ve bought dodgy seconds PETG on eBay for €10 / kg, but it was real crap — bubbles and kinks and stuff like that. Still good for non-critical pieces, so worth it. (The good stuff, when it matters, I pay €18 / kg.)

      Still, I’m surprised by how close this is, at least in the 15-cent regimes.

      1. The good stuff compared to the home brewed bottles stripes. Like Laphroaig and Jim Beam.

        I find it difficult to see differences in quality and relate them to anything. My printing process has way to many parameters that might change, material parameters that change etc etc. Non-dried and new-out-of-the-box-factory-dry PETG result in something I find no difference in… I have expensive PLA that prints like horse-puke and some cheap stuff with great results. Both handled equally, at least I think so (next problem I have).

        I’m completely lost in that.

    2. There are between 35-40 countries that have deposit schemes. JRT3D is US based and there are only 10 states that have a similar deposit systems. The US has a very strong curb side general recycling program and most regions generate more rPET (recycled PET) than existing demand. Somewhat ironically, in some locations it is illegal to put this overflow in a landfill, so it is shipped to places like India and China. Some of it then makes its way into rivers, then into the ocean, then into the floating trash islands that everyone in the west is trying to stop from happening by recycling in the first place.

      Reduce and Reuse are by far the most effective R’s in the Reduce-Reuse-Recycle triad, but they are the hardest and usually first forgotten.

      1. Our local transfer station finally took the 3R logo off the wall. It used to really piss me off because they would not let you grab stuff out of the various dumpsters to re use. I guess they did not want people leaving with more crap than they went in with? The new unspoken rule there is you can stack stuff you came in with next to your car if you want to sort it, and if you are discrete you can do car to car transfers but if someone brings something in and it touches the floor, they consider it theirs.

        1. Also, RCRA (waste management law) strictly controls waste going into and out of facilities, meant to prevent illegal diversion of waste. So, once it enters the dumpster, it’s waste and the transfer station is required to track it until it reaches final licensed disposal. “Losing” waste could hide illegal dumping or improper disposal of materials.

      2. Most of the stuff that ends up in the ocean is because of poorly secured loads on the way to Asia from the US, that is one reason there is so much more plastic in the Pacific than the Atlantic. Almost all of the overflow that ends up in Vietnam, Burma, etc is burnt in open pits.

        Western politicians pass laws that everything must be recycled, even things that are not economically viable. China agrees to buy the junk so they can have access to the market for the plastic that is worth recycling (it is cheaper for them to buy it all than to not use the recycled plastic in manufacturing). They don’t want the trash in China so they pay to have it shipped to countries who agree to ‘dispose’ of it, most often by burning it in big open pits – which is not ecologically ideal as you can imagine. And there are kickbacks at every step of the way so people can ‘feel good’ about recycling. I have multiple classmates who are merchant mariners who have shipped this stuff and the trades run articles on it. It is all a big racket.

    3. I’m from NY but live in Texas now. In NY it’s $0.05 per bottle of soda or water, juice bottles, milk and other bottles get nothing. In Texas there is no bottle deposit, everyone just tosses them in the trash. There is a recycling collection in the city I’m in but most don’t care. It’s really sad. Their big sloging here is “Don’t mess with Texas” and they are proud of being Texan but what good is it that they all trash it.

    1. Since it’s being pulled out instead of pushed, the engineering challenges of such a printer would be insane… I just watched tenet, so maybe you’re talking about inverted filament? Lol

      1. Don’t think the idea is that much of a challenge, fundamentally its the same as a round filament in all aspects but how you drive it – as being a ribbon it would need to be driven in accordance with it’s geometry, but that isn’t actually hard, or all that different – still squeeze it between some rollers of known dimensions, just have to be accomodating to the wide but thin shape… Does mean the filament will be harder to move against its thicker faces, probably not that much as it should twist pretty freely. But still ideally the printer design would only move the head in the way it does bend very freely – making it a bed slinger of some sort.

        The shape will also change a few details that were once very fussy in round filament FDM – how hot and large the melt chamber, how much cooling does it need to only melt inside that chamber so it doesn’t jam etc. But none of that is really any different to a round filament, its just we have done round and found the right sweetspot for good reliable and affordable hot ends.

        1. But then you’re not pulling it but pushing it again. No different than regular filament beyond what you described. I see now that the original question was more about printing with strips that with his hotend. A logical leap on my part. Plus I just watched tenet as I said, so the idea of a printed part sucking into the extruder and becoming a strip of pet was amusing to me. Yes it would be a bit more efficient to extrude from the strip directly if the electricity used for heating is the only measure of efficiency.

        2. Check this out. Someone already tried. Not sure what happened to the project. I know this is an old thread but I just had to respond, LOL.

          I tried to find one several months ago and couldn’t. Someone should bring it back if the patent ran out. Could prove very useful.

          In case my copied link doesn’t work, it was called the Renegade 3d pen. I just googled 3d pen recycled soda bottles.

  3. Big question for me is can you on the fly feed in bottle number 2,3,4’s ribbon to get whatever length of filament you want reliably, as the tiny amount you can get out of one bottle is going to be very limiting to the usefulness of the result. It looks like you probably can, might futz with the output result a little but feeding the fresh end in by hand at the tail of the first does look possible.

    Still interesting, inspirational etc – I really like that is used the same motor for both parts, nice to keep the part count low and functionality isn’t hurt by it.

  4. Clever! Would be so easy to add PID thickness control.

    What about simple water PET bottles? There are plenty of those
    In Hungary they don’t pay you for PET bottles.

  5. An issue with this is that these bottles are not pure PET. The preforms are made with several layers. Food safety regulations require a layer of virgin PET on the inside. The outer layer is usually the thickest, made of recycled PET. In the middle there’s an oxygen barrier layer and who knows what else in other layers.

    When the preform is heated and blow molded, all those layers stretch out really thin.

    1. I’ve never heard of this? I’ve seen many companies in the last 2-3 years really heavily market testing food safe 100% rPET bottles, and I’ve seen rPET used a lot in non-food containers, but I’ve never seen a double layer/barrier bottle.

  6. First of all: This idea is not new!
    Correct me , if I´m wrong.

    I remember a couple of russian guys got it first. They invented it, developed and discussed it in a blog and published the blueprints and also showed the mechanism in youtube videos.

    I hope the “inventor” gives them the credits they deserve before starting a crowdfunding campaign.
    I have to say, that the author did very poor research in that case…

    And yes, there are 2 or more-layer bottles around.
    Mostly the brownish orange-juice ones. In the middle there is a thin layer of LDPE.
    This is the ulitmate horror for unmixed recycling and in my eyes the total wrong way.

    just my two cents

  7. The idea is great. Some work went into this machine but I must say the build design is awful and it looks like a pure functional prototype to me. No way this is worth 400 bucks. It’s actually a mixture of home depot and 3d printing stuff. It’s missing the final touch in product design to explain this price point IMO

  8. The idea is great. Some work went into this machine but I must say the build design is awful and it looks like a pure functional prototype to me. No way this is worth 400 bucks. It’s actually a mixture of home depot and 3d printing stuff. It’s missing the final touch in product design to explain this price point IMO

  9. Actually this is Russian developed method and name is not PETBOT but PETPULL Already more than 2 years we are doing this grate filament and have a lot of experience. Right now we are busy with fully robust filament welder.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.