One of the goals is a solar cell production plant design that MetalicaRap will be able to print, that will utilize MetalicaRap's vacuum chamber and beam for the solar cell manufacturing processes. For a typical family home electrical system we may bring the solar cell cost down from 10,000 euro to 400 euro by self printing. (solar cell installation, inverter, and other costs would obviously be on top of this price).
It should be possible to use a process that creates thin film solar cells at a vacuum of 10-4 Torr by directly co-evaporating copper, gallium, indium and selenium onto a heated substrate[3]. Other precursor choices should also be possible[4]. The CIGS manufacturing process will consist of electron beam physical vapor deposition[5] (EBPVD summary by material [6]) and DC sputtering in the vacuum chamber(DC sputter animation [7]).
If self-replication really is the holy grail it seems like it'd be easier to have the system set up as a (self-)assembly line than a jack-of-all-trades: Print photoresist, etch, drill, pick & place, wave-solder, cut, etc. And doing anything "assembly" will also require intervention without also doing significant robotics work.
It might be straightforward to optimize the number of needed stations to "print" a reasonably-sized PCB that's fully populated and ready to use, though. Particularly if some stations could be combined due to similar hardware and precision needs: drilling vias and mounting holes could be done with almost the same hardware as pick & place. But full self-assembly will be orders of magnitude more difficult because of the need to move through several stations, potentially with manipulation in between. There you're getting into essentially small-scale factory design.
It'll be awesome when it happens though: An assembler that can autonomously create, "from scratch" (only basic components), another fully-functional assembler. That can create, "from scratch," yet another fully-functional assembler.
Just, uh, don't go combining all that with mobile resource foraging and extraction and an imperative to keep making copies.
I think a more likely scenario is cities having little 'auto-factory' locations that have an automated machine shop create things for the neighborhood. They'll get shipments of raw materials, electronics and energy pipelines. These shops can assemble the robots for new shops and so on.
Or if you've got $50K to spare, checkout the Dimatix material deposition printers. I spent 2 years on one in a printed electronic lab at Stanford. You'll need access to a chem lab to get research-grade printable & conductive polymers :)
In 1963 Australians Bolto, DE Weiss, and coworkers reported iodine-doped oxidized polypyrrole blacks with resistivities as low as 1 ohm·cm. This Australian group eventually claimed to reach resistivities as low as 0.03 ohm·cm with other conductive organic polymers. This resistivity is roughly equivalent to present-day efforts.
not very conductive. You can used various PeDOT polymers inks but you're best best is usually to go with metallic nanoparticles. Those aren't too great because they limit the number of substrates you can use due to required post-deposition processing (annealing mostly, to make them conductive). We also managed to print nanowires and carbon nanotubes. Those are extremely conductive and require little post-processing. In fact, we did most of our work on paper substrates.
12 comments
[ 1.8 ms ] story [ 41.9 ms ] threadOne of the goals is a solar cell production plant design that MetalicaRap will be able to print, that will utilize MetalicaRap's vacuum chamber and beam for the solar cell manufacturing processes. For a typical family home electrical system we may bring the solar cell cost down from 10,000 euro to 400 euro by self printing. (solar cell installation, inverter, and other costs would obviously be on top of this price).
It should be possible to use a process that creates thin film solar cells at a vacuum of 10-4 Torr by directly co-evaporating copper, gallium, indium and selenium onto a heated substrate[3]. Other precursor choices should also be possible[4]. The CIGS manufacturing process will consist of electron beam physical vapor deposition[5] (EBPVD summary by material [6]) and DC sputtering in the vacuum chamber(DC sputter animation [7]).
http://reprap.org/wiki/MetalicaRap
But I can dream, right?
Sure we are.. http://code.google.com/p/homecmos and irc.freenode.net ##hplusroadmap says hi.
It might be straightforward to optimize the number of needed stations to "print" a reasonably-sized PCB that's fully populated and ready to use, though. Particularly if some stations could be combined due to similar hardware and precision needs: drilling vias and mounting holes could be done with almost the same hardware as pick & place. But full self-assembly will be orders of magnitude more difficult because of the need to move through several stations, potentially with manipulation in between. There you're getting into essentially small-scale factory design.
It'll be awesome when it happens though: An assembler that can autonomously create, "from scratch" (only basic components), another fully-functional assembler. That can create, "from scratch," yet another fully-functional assembler.
Just, uh, don't go combining all that with mobile resource foraging and extraction and an imperative to keep making copies.
In 1963 Australians Bolto, DE Weiss, and coworkers reported iodine-doped oxidized polypyrrole blacks with resistivities as low as 1 ohm·cm. This Australian group eventually claimed to reach resistivities as low as 0.03 ohm·cm with other conductive organic polymers. This resistivity is roughly equivalent to present-day efforts.