Regrind: A Route to Closed-Loop Printing

Experiments, Massive Dimension, MDPH2, Pellet Head Extruder -

Regrind: A Route to Closed-Loop Printing

One of our particular ambitions at Massive Dimension is to make 3D printing a more sustainable practice. Plastic waste is an enormous problem, and the myriad solutions presented thus far have helped to stem the tides but are just a beginning. Problems of new production aside, there is already a vast enough quantity of plastic present on our planet that creative recycling efforts must be introduced to find useful applications for what we have otherwise deemed garbage with an extremely long shelf-life. We now have clothing made from recycled plastic, plastic aggregate substituted into concrete, and in several countries around the world there are now roads made in part from plastic waste. None of these processes make the plastic disappear, but they do give the materials a second cycle of use. In a similar vein, we are enamored by the idea of being able to use exclusively recycled plastic as a feedstock for pellet-driven extruders. 

Printing with recycled waste material is not itself a new frontier- the entire Filabot product line was created with the intent of giving users the ability to reclaim failed prints and turn them back into filament- and it has been successful in that project. However, the more elusive achievement has been to reliably run large format prints on the Massive Dimension extruders directly from regrind material- skipping over the “in-between” steps of compounding the regrind into new filament and pelletizing it to make feedstock similar in consistency and composition to virgin polymer pellets such as you would receive from a manufacturer. While we haven’t quite achieved the level of reliability we’d like to see, we’ve run a lot of trials lately on the MDPH2 and would like to share our findings thus far with anyone who might be on a similar endeavor. You might want to print directly from regrind because you are also ecologically-minded, or perhaps you are just attracted to the lower price of recycled materials, which are readily available on the internet. Furthermore, you can cut down on labor if you can successfully skip over the steps of compounding/pelletizing. Whatever your reasons, we applaud your entry into the frontier of regrind printing! The three largest problems you are likely to face can be broken down as follows: purity of regrind, melt flow, and feedstock starvation.

Example of underextrusion - unmelt

Purity of regrind is the most straightforward problem to discuss and potentially the most difficult to solve. By purity we mean simply that it is a single polymer, not contaminated by any other polymers with different melt temperatures, though of course other contaminants such as dirt, metal, etc. could theoretically exist depending on your source. If you are printing with PLA, such as in the case of many of our recent trials (ubiquity of PLA lends itself to being an ideal candidate for creating sufficient supply of regrind) all it takes is a sprinkling of a higher-temp polymer such as ABS mixed in with your regrind to cause drop-outs in extrusion. We have run trials both from regrind sourced “out-of-house” from other recycling companies as well as regrind produced from our own failed prints in the lab. If you do not have the capability to create your own regrind, the best you can do is shop around for a source that can deliver on consistency. If you find a source that works, stick with it- and look for an opportunity to order a smaller quantity to use as a test benchmark before committing to purchasing in bulk. We have had mixed results with outsourced regrind with evidence of polymer contamination, and are striving to maintain more stringent processing standards to ensure purity in our own regrind. Unfortunately a single misplaced failed print can compromise an entire batch of regrind. We have multiple grinders, each dedicated to a specific polymer to ensure no cross contamination within the machines.

Coarse granules created by pulverizing hooks

The second and third problems of melt flow and feedstock starvation are interconnected in so far as they are both dependent on particle geometry. Again, if you are without your own grinders then you will have to shop around to find a regrind source that meets your specific needs. Within the broad category of regrind you could have dramatically different sized granules, and particles might be well sorted or else intermixed- all depending on the blade configuration of the grinders and the screens used during processing. For example, we have one particular grinder that has two large hooks on a drum that breaks down waste mostly by pulverizing the plastic, with auxiliary smaller blades that further work the crushed material. That grinder creates a fairly coarse flakelike regrind. On the other hand, we have a machine that has a more uniform assortment of shorter, sharper blades on its drum. This machine creates a much finer grind, closer to what you might end up with if you pelletized filament. Both of the aforementioned machines offer the ability to insert screens of varying mesh above the collection pan, which could be used to further sift and sort regrind into different categories if desired. To be honest, this is a technique we need to do further research on, as it offers a potential solution to some of the problems of printing with the resultant regrind. 

Fine grind from shorter blades

Consistent particle size is desirable to achieve a controllable melt flow. If you have a large disparity between particle size you cannot arrive at ideal barrel/nozzle temperatures because the temperature setting for the smallest particles might not be high enough to completely melt the largest particles, and if you turn it up to account for the largest, the smallest particles can begin to melt too early in the screw and cause blockages, halting extrusion. Therefore, best results will be achieved without either of these outliers. This “Goldilocks” particle geometry has additional requirements to prevent other extrusion dropouts. Broadly speaking, regrind has a more difficult time moving through the feed throat and screw and successfully arriving at the nozzle. Virgin pellets are almost always bead-like geometry, often spherical and sometimes cylindrical- but either shape encourages them to drop and roll through the feed throat and into the heat zones of the barrel. Regrind particles have a greater number of hard edges and sharp points, geometries that lend themselves to bridging and jamming in the feed throat. 

Medium-fine grind from exterior source

Granules above a certain size threshold are certainly more difficult to print with, though smaller isn’t always better. Vibratory devices can be employed on the feed throat to break up jams and assist gravity in delivering the particles to the screw, though in our experience if the particles are excessively fine the vibratory device can encourage them to pack too densely into the screw and overtorque the motor. A pulse signal to the vibratory device with an adjustable period can be utilized to fine tune & match the grain size and density of the regrind. We’ve had the best success with particles sized less than or equal to an average virgin pellet. 

Regrind and virgin LX-175 blend

While we still have plenty of work to do to achieve consistent results with printing direct from 100% regrind, we’re happy to be making progress in that direction. Our pure regrind prints have failed at various points in the course of their extrusion, often running for hours before dropping out entirely, and we hope to refine that process to have greater success. On the other hand, we’ve found that printing with a mixture of regrind and virgin pellets greatly increases the reliability of extrusion- 50/50 blends or even 75/25 regrind to virgin polymer makes a huge difference. While the goal continues to be to be able to print entirely from recycled sources, we still consider this an important step in that direction. If you are struggling with printing regrind, we recommend you start by blending in this manner and see how your printer handles different blends. An interesting side effect of blending the regrind with clear polymer is that it allows the individual colors of the regrind (assuming your regrind is a mixture of different colors) to remain distinct in the print, whereas with 100% regrind prints they tend to all mix together into a single color, often grey or brown. As always, we would love to hear your experiences and see pictures if you are trying similar experiments at home!


Below are some videos from the experiments:

1. MDPH2 large format printer on SCARA movement system with regrind colorant - June 3 - part 1 of 2 


2. MDPH2 large format printer on SCARA movement system with regrind colorant - June 3 - part 2 of 2 


3. "Fly's Eye" Geodesic Dome - June 15 - part 1 of 2


4. "Fly's Eye" Geodesic Dome - June 15 - part 2 of 2


Below are some more photos from the experiments:

Extrusion direct from pure regrind

Extrusion from 50_50 blend

Finished print from 50_50 blend

In process print 75_25 blend




Leave a comment

Please note, comments must be approved before they are published