According to this, the problem with grinding materials and re-melting them is that the shear stress from the grinding process makes the plastic unusable, but wouldn’t re-melting undo any damage from shear stress? I suppose you might break some polymer chains or something, but I would think that you would have to grind extremely finely for that to be a problem.
After all, isn’t most filament (good filament, at least) made by extruding the filament once, then cutting it up (presumably with a shearing cutter of some sort) to turn it back into pellets, to be extruded again? My understanding is that this is standard practice, at least for color filaments, which won’t get sufficient mixing in a single extrusion pass. The pellets they show while talking about the shearing problem were clearly made this way… https://www.youtube.com/watch?v=_iFfehELSkU
I would think that the absence of any special modifiers which may be added to the master batch, to aid initial extrusion, which aren’t present in recycled pellets is more of a problem. I can’t see how you prepare the material mechanically for recycling would affect the melting process unless you add contaminants at the time. I’m happy to be proved wrong.
Regrind/remelt is very common in the injection molding industry, but uneven extrusion is less of a problem because they are filling a closed die. Also the process is typically used for high volume production so any flawed parts can just be rejected and sent back for regrind again. Speaking to friends who used to be in the industry they had some components with cosmetic requirements when almost half of their production was scrapped and remelted.
As far as removing the shear stress it would probably depend on the processing parameters, if they are only softening the material enough to extrude then their could be residual stresses, if they are fully melting it to a true liquid state there shouldn’t be any residual stresses.
I vaguely recall hearing something about remelting the plastic too many times making it more crystalline. Perhaps being forced out of an extruder is enough rearrangement of the plastic to avoid that. I do not know.
Usually only up to about 15% of regrind (runners/sprues or recyclables) gets mixed in with virgin material, at least at a reputable US-based molder. Every time you regrind the material, you are mechanically shearing/cutting the polymer chains. Additionally, every time you run them through a screw, die, mold, etc, the shear stress from mixing and flow also shortens them (it is not that shear stress is somehow stored in the polymers). Thermal aging also becomes a problem, whether it is causing chain scission or promoting the loss of plasticizers and other additives. Eventually, the molecular weight gets low and therefore the polymer’s ability to transfer stress from one chain to another is greatly degraded. The loss of additives impacts the material’s toughness, etc.
Virgin pellets are manufactured by compounding virgin polymer powders and additives. This is the longest (greatest molecular weight) the polymers will ever be. And the point at which the ideal ratios of additives are present. Yes, there is some loss due to the extrusion/pelletizing process, but subsequent processing starts to add up exponentially so it does not take many regrinds to start to greatly impact properties.
@Ulrich_Baer shear thinning refers to a fluid whose viscosity decreases when under shear strain. This is not something that is “stored” in the filament we buy.
During the manufacturing process of pellets and filament, the rate of shear strain in an extruder is high enough that some materials may exhibit this behavior. All it really means is that the filament can be extruded exponentially faster as the shear rate is increased in the extruder. Whether or not a filament manufacturer achieves this has no bearing on the printing properties of a filament used in a 3d printer.
I just came to G+ to post the very same thing, with similar questions. Mainly, what are they doing instead of regrinding? Are they simply compactifying plastic objects into bricks that fit into the extruder? I seem to remember a few printer startups that ran on filament sticks instead of coils. I seem to remember they weren’t well liked, but not why. Maybe the concept is good enough for this.
Since polymer shortening can’t be eliminated, perhaps they will try for the longest possible chains in virgin material, in order to maximize the number of times it can be recycled. Won’t they need a way to assess and manage material to control the extrusion? Can the polymers be lengthened without taking the whole mass back to short hydrocarbons and starting over from scratch?
@Dale_Dunn I don’t think controlling extrusion is such a problem in the 3D printer. Generally speaking, the extruders in 3D printers are positive displacement. That’s why we calibrate E steps and then find boundaries (maximum flow rates) that we can print up to for a material. In general, a virgin material will have a higher viscosity than a recycled material (not accounting for any flow promoter/additive loss) so I would think that as a material is recycled, its maximum printing flow rate would increase. It is entirely possible that they are using extrusion methods that are not based on positive or fixed displacement, in which this would be a concern.
Increasing the length of polymer chains without bringing them back to chemical constituents is kind of the pie in the sky of recycling isn’t it? To my knowledge, there have not been any cost-effective breakthroughs in this.
I can not help but think that it would be best if they could grow the plastic or similar material on the ship. Early records were made from the shells…or was it the shit…of some sort of bug. Either they could have such bugs to help make filament or they could engineer the ability into something else or they could find something else that they could use. maybe some items could be made from something like clay. There are clay extruders too.
My ignorance of plastics has been decreased a little bit. I had not thought about this deeply, and I had not even considered the polymer chains could be forever broken due to mechanical shearing. I thought as long as the plastic reached a liquid state that it was fully reconstituted and that this could be done virtually infinitely. I now know that is not the case. Thanks for this post! I’m really going to miss G+ when they shut it down.
I’m fairly sure the issue is going to be repeated melt processing, not so much the mechanical grinding. Yeah, regrinding will damage the polymer a little, but re-extrusion causes polymer shortening via multiple mechanisms (heat, shear, chemical attack, etc) AND ALSO out-gassing of important additives like plasticizers.
It will depend significantly on the polymer type and processing conditions though. For example, virgin PET and to a lesser degree PETG is damaged by moisture – water aggressively cleaves the chains in half via hydrolysis at melt temps – so it has to be incredibly scrupulously dried prior to remelting. With PET, one melt cycle with moisture contamination will destroy it. With PETG, each melt cycle with moisture causes progressive loss of material properties (eg weakening and brittleness). Is proper drying going to be practical in space? Maybe; heat and vacuum are readily available… except you’re outgassing additives with every drying cycle.
I am positive that it’s POSSIBLE to meaningfully recycle materials in space, but to achieve more than ~2 cycles of material reuse I think you would need a non-negligible amount of “infrastructure” in place to manage long-term degradation of material properties. For example, some combination of fresh feed blending, regrind drying, and additive replenishment.
What’s the underlying goal here though? Give the ISS a way to make widgets between resupply missions? If that’s all you want, you might as well just run everything single-pass and ship fresh filament as needed. If you’re trying to help a Mars mission or moon colony be more self-sufficient, I think you’d be looking to get some significant reusability. It’s a different application at that point. You’d be willing to invest R&D and infrastructure into optimizing it.
@NathanielStenzel it should be possible to make PLA and/or PHA in space from some sort of fermentation or photosynthesis pathway. But it’ll be fairly complex equipment with a lot of heating/cooling/power/space/weight requirements. And zero-G chemical plants aren’t a thing yet… tons of R&D required there. Not real practical for the ISS, but if you’re building a moon or mars colony, maybe you’d want to go that way.
Mechanical stress breaking polymers seems like a specious argument. The amount of molecules on the surface of a reground part is tiny compared to the number in the core of a granule.
The video doesn’t say anything about how the machine processes the plastic into a new part.
I’ll have to talk to my local plastics engineer and see what he says. I had the impression that the number of melt cycles was a big factor.
Well, it’s obvious that it’s possible to re-melt plastic without first breaking it into small pieces. Being in large pieces of non-uniform shapes with air trapped inside just makes it more difficult to apply heat evenly.
I’m sure there is stuff you can do with atmospheric pressure to help the process (it seems like it would be particularly easy if you could reliably trap a vacuum inside your parts (and keep it a vacuum as the part is used), which would make it want to collapse back down to a bubbleless mass when heated. Of course, you would be applying heat to the outside of the model, and the infill would still heat more slowly, so it would brace the structure and the softened skin would stretch inward and break, allowing air to enter.
Now, if you could generate heat from inside the part, that might solve the problem. You could get partway there by mixing metal particles into the polymer and then using microwaves or induction heating for the re-melt. Actually, if you did that in a vacuum (which is free in space), you could quickly burst the part, causing all of the air to escape, then polymer tension should cause the part to settle to a bubbleless mass.
Mixing metal particles into the plastic would explain the color of the prints shown in the video. It’s otherwise silly to add pigments that would look like that to the plastics for this purpose.
At MRRF 2018 and MRRF 2017 there was a booth with someone printing a paste with metal in it. It needed to be cooked in a kiln to finish the part. I believe it was a clay base instead of a plastic/polymer base. I imagine that plastic/polymer might work too. Light hardening polymers might also work. I believe there were posts about a light hardening polymer printing delta printer 3-5 years ago that would squirt out the gel and then the LEDs on the effector would harden the plastic.
@NathanielStenzel Markforged and Desktop Metal are both currently developing (and suing one another over) metal printers that do FDM with metal particles in a polymer matrix that are designed to have the plastic remove and the metal sintered. You can also buy filament from The Virtual Foundry that can supposedly be used to do this on printers that use open filament. That’s not the case here.
In all of those cases, they’re still extruding through a heated nozzle, but what I’m suggesting is that the secret sauce here might be using metal particles embedded in the plastic to allow parts to be heated electromagnetically from within, at least during the re-melt phase, to get around the problems involved in re-melting a print without first grinding it into pieces small enough to effectively conduct that heat into.
@NathanielStenzel microwaves can heat carbon fibers internally too; induction can heat metal. CF is a lot lighter and more likely to improve the mechanical properties so that might be a better way to go. Although I’d be worried about CF filament throwing conductive dust over time with use/abrasion so metal microparticles might be better.
