5 ms·

Quantitative improvement, not qualitative. They don't need to separate the resin from the base plate after every layer, which means vastly improved speed and pr

by saulrh 11y ago

Quantitative improvement, not qualitative. They don't need to separate the resin from the base plate after every layer, which means vastly improved speed and precision and much less complexity. It's the difference between a car that can go 5 miles per hour and a car that can go 50 miles per hour - only quantitative improvement, but now you have a reason to use it every day.

mcphage 11y ago
Of course, in manufacturing they're competing against injection molding, which can go a metaphorical 100,000 miles per hour.
- saulrh 11y ago
  But has its significant limitations, notably that it takes a colossal machine and days to weeks to design, optimize, and machine a mold. Injection molding works great if you want to get to Pluto, but it'll cost you two years, a partnership with NASA, and a thousand tons of LH2 and LOX, and that cost remains constant even you just want to hit the supermarket. Current 3d printers can get you to the grocery store pretty well, but you're out of luck for anything bigger. Carbon3d offers a new middle ground - fast enough and good enough for small production runs, but without the titanic constant-time overhead demanded by injection molding. This isn't going to be used for parts that you need ten million of. It'll be used for parts that you need a few hundred of.
  - Retric 11y ago
    Seems like there is a huge opportunity for 3d printing injection molds. It's probably more a tool chain question than an actual 3d printing question. But, by reducing the cost gap from mid-scale 3d printing to full scale production seems like a large potential net win. AKA, if the costs are similar at 10k and drop after that.
    - minthd 11y ago
      3d printing of injection molds is already being used for some low volume medical products.
    - Abraln 11y ago
      Injection molds need to be ridiculously precise, as in 1/1000 of an inch of can render it useless. Also, I have doubts that the layered result of 3-D printing can hold up to the extreme abuse these things get put through after tens or even hundreds of thousands of cycles. We are talking about foot thick plates of steel bending and warping from fatigue damage after a while. Even if you could 3-D print the bulk of it, the job of polishing and detail work would still need to be done manually. Edit: after Reading the comment below, I admit I may be somewhat biased from working with more complex and detailed parts with very right tolerances, so there may be some applications where it is feasible. All the molds I've seen/worked with were made the old fashioned way.
      - Retric 11y ago
        tens or even hundreds of thousands of cycles If it's reasonably cheap to print, then only lasting say 2,000 cycles would not prevent a useful mid-scale production runs. Also, the high end of 3d printing is a lot tougher and has tighter tolerances than you might think. http://www.gizmag.com/ge-fires-up-all-3d-printed-jet-einge/37448/ http://www.gizmag.com/ge-fires-up-all-3d-printed-jet-einge/3...

by saulrh 11y ago

mcphage 11y ago

Of course, in manufacturing they're competing against injection molding, which can go a metaphorical 100,000 miles per hour.

saulrh 11y ago
But has its significant limitations, notably that it takes a colossal machine and days to weeks to design, optimize, and machine a mold. Injection molding works great if you want to get to Pluto, but it'll cost you two years, a partnership with NASA, and a thousand tons of LH2 and LOX, and that cost remains constant even you just want to hit the supermarket. Current 3d printers can get you to the grocery store pretty well, but you're out of luck for anything bigger. Carbon3d offers a new middle ground - fast enough and good enough for small production runs, but without the titanic constant-time overhead demanded by injection molding. This isn't going to be used for parts that you need ten million of. It'll be used for parts that you need a few hundred of.
- Retric 11y ago
  Seems like there is a huge opportunity for 3d printing injection molds. It's probably more a tool chain question than an actual 3d printing question. But, by reducing the cost gap from mid-scale 3d printing to full scale production seems like a large potential net win. AKA, if the costs are similar at 10k and drop after that.
  - minthd 11y ago
    3d printing of injection molds is already being used for some low volume medical products.
  - Abraln 11y ago
    Injection molds need to be ridiculously precise, as in 1/1000 of an inch of can render it useless. Also, I have doubts that the layered result of 3-D printing can hold up to the extreme abuse these things get put through after tens or even hundreds of thousands of cycles. We are talking about foot thick plates of steel bending and warping from fatigue damage after a while. Even if you could 3-D print the bulk of it, the job of polishing and detail work would still need to be done manually. Edit: after Reading the comment below, I admit I may be somewhat biased from working with more complex and detailed parts with very right tolerances, so there may be some applications where it is feasible. All the molds I've seen/worked with were made the old fashioned way.
    - Retric 11y ago
      tens or even hundreds of thousands of cycles If it's reasonably cheap to print, then only lasting say 2,000 cycles would not prevent a useful mid-scale production runs. Also, the high end of 3d printing is a lot tougher and has tighter tolerances than you might think. http://www.gizmag.com/ge-fires-up-all-3d-printed-jet-einge/37448/ http://www.gizmag.com/ge-fires-up-all-3d-printed-jet-einge/3...