Is the main problem behind recycling 3D prints one of sortation and contaminants? I’ve been researching this for a bit - any input would be appreciated.
I looked up the spectroscopy machines used to identify different polymers. They seemed expensive but also a bit overkill for the job. No need to be accurate enough to separate every polymer out there - only “good enough” to categorize a few of them from a controlled source.
Most makerspaces I spoke to are willing to share their filament order history to facilitate recycling. If we know the exact plastics we start with (including their visual color and IR profiles, brand, additives, etc.), could cheap hobby-grade sensors be accurate enough to tell the difference for small-scale sorting?
The easiest example would be one club I found that carries just 3 kinds of PLA: red, black, and white by Elegoo. If I see a black plastic in their junk bin, it’s probably black Elegoo PLA and I know to put it in a unique bin just for other black Elegoo PLA samples.
Another club prints black carbon fiber filament, so the color test would fail there, but both clubs’ plastics would be in two separate boxes and I would obviously not mix them. (That said, I think carbon fiber absorbs radio waves, so maybe there’s another test I can devise to tell them apart anyways.)
Yes, this is a significant problem. If the contaminates have a higher melting point then they will remain solid inside of the screw extruder and plug the exit dye/nozzle. For a highly contaminated waste stream you will need a screen inside of the extruder that can remove these particles from the melt. As this screen becomes plugged it needs to be cleaned or rotated out, which is complicated and expensive to implement at the hobbyist scale. Contaminates with lower melting points will either burn (with this residue making it into the final output) or are likely immiscible with the desired plastic. Both scenarios will result in an output from the extruder that is likely brittle with poor mechanical properties.
In large recycling facilities they still rely on identifying the plastic by the number on the bottom of the bottle/container. They will also sort the plastic by floating it (plastics have different densities).
Likely not. The other issue is that you would have to scan and analyze each plastic piece. This could be automated, but it would still take time.
This is exactly the solution that I have come up with. Instead of the recycling number, I am color coding the material based on plastic type. Just FYI, I have found that this does disappoint some users of my makerspace as many want to print aesthetic models of different colors and I only stock white PLA.
Yes, sensing and analysis would have to be automated. I’ll look into IR spectroscopy as it seems to be the fastest and most mature option.
From my limited understanding of spectroscopy techniques, it seems their high cost stems from trying to achieve high precision. But if we start with a discrete number of known plastics, do we even need precise sensors? For X unique plastics, we only need to be precise to +/- half the difference in color, IR spectra, or whatever variable we’re measuring that tells them apart.
Actually, the bar is lower than that. As long as it can pick out at least one plastic in a bin of X many plastics, we can separate it from the others. That ought to be an improvement over sending everything to landfill.