Originally Posted by un-midas touch
Picturing a tank of feedstock connected to a hose with a robot at the end of it small enough to climb up the structures it has already built (Waiting for Zero4549 to tell me how I'm crazy
). Or some other way to avoid constraining the size of the product to the size of the printer without having to jump through ridiculous engineering hoops to make a functional puzzle out of everything you build.
On a tangent, has anyone seen a commercial model that uses epoxy+hardener type mediums instead of high temps?
The materials used in those is insanely expensive.
Last I recall, the biggest current hurdle is being able to use plastic pellets instead of the extruded filament. I think someone finally did manage to build a working extruder that works with plastic pellets and makes a properly sized filament.
The problem is, if you want to do it at any significant rate of speed, you need some very expensive hardware.
I have suggested to people on other forums who do this stuff to look into repurposing an extruder from an old 10-15 ton injection molding press. Small barrel, capable of extruding basically any type of plastic (including polysulfone, which I work with daily), and you would simply need to assemble a nozzle with a carefully designed cooling structure and a means of pulling the extrusion from the cooling nozzle/die at a slightly lower speed than what the extruder screw is capable of pushing it out at. This way it comes out unbroken. The significant issue here is that it would likely require active
cooling in order to achieve a useful feed rate that can then be spooled or immediately fed into a machine. The spooling idea would work better though, letting you extrude faster than the machine consumes so you dont have 2 kw of heater bands running all the time. I imagine, with some cash, one could theoretically use a large water cooling loop. It doesn't even have to be PC components, go buy a properly sized car radiator, a 120v pump, some high temp hose, and a big box fan. This will not bring the die/nozzle below maybe 150 degrees or so, depending on the plastic you are running, but it would be more than enough to solidify the melt enough for a gripper to pull it.
The problem is that people don't ever listen to me, and I keep seeing all these attempts at doing EXACTLY what I am suggesting, but using other, cobbled together parts, melt chambers with pistons that force the stuff through a die... etc. You can buy a whole injection molding press for only a few grand, buying the extruder assembly is significantly cheaper, especially if you get something off a press from the 70s that uses an electric motor instead of a hydraulic one.
I work with plastic 50-60 hours a week, so someone needs to pay attention