28 thoughts on “The New Industrial Revolution”

  2. I’d like one DC-3 and one F-86, please. In titanium. To go. [Hmmm; I wonder if the IP for these two aircraft is ‘public domain’?]

  4. About five years ago, I ran across a project to build houses with similar technology. I think it was at UCLA.

  5. Of course, now the govt. will have to step in. If they can print titanium why not weapons grade uranium and the explosive trigger as well? The plans for a rifle type nuke are public domain.

  6. I think there is already a serial number embedded by printer manufacturers, to prevent copying and printing of paper money. I’d expect something similar might be required for the fabbers eventually.

    There are a lot of material processing issues to be worked out. Even if I had, say, a ratchet fabbed out of plastic (or even steel), it wouldn’t have the strength to turn a rusted bolt, it requires forging to get a quality product. Eventually I suspect it might be worked out, but not in the next decade or so.

  7. I think we were the first people to ever fire a 3D printed Rocket Motor.
    First attempt was about 6 weeks ago, first real success was last Saturday. The problems in between had nothing to do with the 3D printing, just other misc problems like contaminated propellant and some cat pack issues.

  8. it wouldn’t have the strength to turn a rusted bolt, it requires forging to get a quality product

    But then I keep reading about stuff like nanotube reinforced materials being super strong and stuff.

  10. A fascinating and informative article.

    I’m hearing Neal Stephenson chuckle, however. This is The Diamond Age coming true. Where’s my Feed?

  11. At the trade show in Chicago last year, the largest company in the metal-additive process was printing a small turbine blade assembly in 20 micron titanium powder with laser sintering. Print time: ~41 hours. Same time to machine on a 5-axis machining center with automatic tool changing: a little more than 2 hours. There is still a lot to done to make 3D printing competitive with these types of parts.

  12. I look forward to Open Source Engineering. Download the Linux of rockets, fill up the tank, and away we go!

  13. “….a 5-axis machining center with automatic tool changing….”

    Sounds like a rather large fixed asset to me. A 3d-printer could offer greater portability which could mean that the printers can be closer to the end point where the parts are desired; better, faster, cheaper — pick two.

  14. Not Diamond Age quality yet – not when the tolerances (voxellation?) are still in the hundreds of thousands of nanometers.

    But IIRC the most interesting technological change in the plot of Diamond Age was from the Feed (where you own a hookup that lets you make whatever you have permission for) to the Seed (where you own a self-contained fabricator that lets you make whatever you have a design for). The most interesting 3D printer from that point of view might be the RepRap.

  15. 41 hours vs 2 hours…

    Doesn’t sound too bad all things considered.

    1) We aren’t talking about replacing existing technologies with the current SOTA in additive manufacturing (which is nowhere near mature yet). Where will we be in 5-10 years? That might be a much more interesting comparison

    2) There are niche markets now where a 20-fold disadvantage in time of manufacturing isn’t a real problem. Low volume, high quality items might be a good example of this, though I leave that as an exercise for the creative student

    3) The ability to manufacture with much higher efficiencies in raw material consumption, the ability to create a remote template for manufacturing, etc. might be said to (somewhat) offset the disadvantage in manufacturing time. Probably not enough (yet), but the technology is young

    4) At some point, it is likely that instead of simply using this technology to mimic existing capabilities, we will find new kinds of manufactured products that we didn’t previously make. The linked article refers to different part shapes and efficiencies, for instance, but I am sure that there will be others

    All in all, much to be excited about

  16. What if you wanted than turbine blade to have conformal cooling channels? Not doable on a 5axis machine at ANY cost.
    That is the real power…..

  17. 3D printers are great. But they do have two major disadvantages (for now):
    1. Speed – as mentioned above printing takes a long time, especially at high resolution. For some additive technologies, you also have to do a lot of post-processing. For example, the ZCorp printers produce very fragile parts that must them be very carefully blow clear of extra powder and hardened with what amounts to superglue. A Stratasys printer requires support material removal in a bath of hot caustic chemical soup. An EOS metal powder process has the additional steps of baking and second metal osmotic infiltration.

    2. Money – even a low-end Stratasys or ZCorp printer at a maker lab I go to costs $20 per cu. in. of build material consumed. That adds up in a hurry!

    Having said that, I’d love to have one of those machine in my workshop 🙂

  18. AI being depicted in the novel as having failed

    This is something rare in SF. Enough that I’d like to get this book.

    Printers are always going to have their limitations. I don’t believe any are able to completely replicate their own parts yet. Even if they could, those parts might not be the quality they need to be for a non-fragile high precision machine. Of course, I’d still like to have one.

    Design becomes a much bigger issue. Because of so called IP, we can expect lawyers to clog the works. I think 3D printer will tend to continue to be for small lots and prototypes for quite a while. The really good ones are still quite expensive.

    A good materials engineer might get more out of it than your average person assuming they like to tinker.

  19. Goodbye Wal-Mart! Goodbye low-wage workers in China=unrest, eventually = a Chinese Cairo, really more successful Tiananmen repeat?

  21. I’m still trying to wrap my mind around the concept of 3-D “printing”.

    I work in a printing shop and I’m more accustomed to the 2-D kind. And I still have trouble with that sometimes.

    Maybe the extra time involved won’t matter so much if the printing could be distributed among a variety of manufacturers, each of whom would focus on getting their own parts exactly right.

  23. @ Paul Breed – Some larger turbine blades are still designed to use conformal cooling channels. But more commonly these days, small turbine blades require controlled growth from seed crystals, after determining the ideal alignment and composition of the material at a molecular level. What you have then is a blade that has been designed from the inside out for optimal strength retention and heat dissipation, and perfectly built to match. From what The Economist reported, it doesn’t look like the product of 3D printers are being manipulated at that level. (Which I’m sure their designers and backers would qualify by adding, “Yet.”)

    That said, even though you can’t really build a jet engine from metal dust with one of these right now, you probably could build some of the larger blades in batches without any trouble. The whole lost wax casting process could become a thing of the past, depending on how many emitter heads you can use at any given time, or on how quickly the 3D printing process can be made to work using just one.

  25. @ Paul Breed: Dammit, you’re changing the meaning of the word “unreasonable”.

    ——————

    There is a huge range of possibilities for 3D printing. I would love to eliminate the toner transfer and etching steps from my printed circuit board prototyping process, and simply print traces on… well, pretty much any insulator would do. Multilayer boards could be prototyped in a hurry if you can print both conductor and insulator.

    I can’t remember where I read it but apparently some semiconductor manufacturers are turning to a sort of 3D printing with ion beams (IIRC) for the next generation of chips.

    The Raprap idea mentioned above is one possible way around the IP issue. The goal of Reprap is to build a machine that can make (among other things) a copy of itself, and everything is Open Source. The first generation could only make copies of its own plastic parts (gears and so forth) but the next iterations will be trying to make their own circuit boards. The beauty of it is that once a buddy has a Reprap, he can print another one for you, and you can print another one for someone else, and so on.

    Right now, 3D printing is somewhat coarse-grained. However, that’ll change over time and get finer and finer scale, eventually approaching the nanometer range. When that happens, things like Star Trek-style replicators become possible. Imagine starting out with a machine, dirt, energy, and data, and ending up with bacon that had never been a pig.

  26. The encroachment of 3-D printing tech into manufacturing will continue to be incremental and based on increases in speed, accuracy and achievable surface quality of the 3-D fabricators combined with decreases in system and raw material cost. The extremely fine metal powders needed by laser sintering machines, for example, cost many times per pound what billet stock of the same assay does. This price differential offsets a lot of the gross raw material mass savings of pure additive vs. pure subtractive manufacturing processes. This price differential for powdered vs. solid raw stock may be subject to future reduction, but I don’t see it going away.

    Broadly speaking, there are four classes of 3-D printers:

    1) those using a tank of liquid raw material that is rendered into successive layers of solid polymer by laser (usually UV) as the object a-building is lowered, incrementally, into the material on each layer fabrication pass.

    2) those using powdered raw materials and some kind of chemical densifying or laser sintering process to achieve solidity of their output. The laser sinterers are the only systems at present that seem able to make metal parts.

    3) those using melted polymer raw materials which “print” their outputs using printheads that are, effectively, miniature hot glue guns.

    4) those using thin film raw materials that are built up by being laser cut and glued atop previous cutouts to produce a net shape. It’s possible, in principle, to build such a unit that uses metal foil or thin sheet stock as raw material to make metal parts, but I’m unaware of any extant example. Sheet material is much cheaper, per unit of mass, than powdered material, but the waste involved is, overall, similar to what results from conventional subtractive fabrication.

    As noted elsewhere, the powder and “glue gun” machines use very expensive raw materials and also need to use some type of support material to deal with overhangs in a part’s design. The support material needs to be straightforwardly removable when fabrication is complete. The support material is typically close to the part material in cost per unit mass and, for many “there’s no way to make this by conventional means” parts, a great deal more support material may be needed than part material to produce the finished part.

    Bottom line: 3-D printing has its own problems and limits just as do conventional subtractive fabrication techniques such as machining. I suspect we will see an eventual fusion of both kinds of technology in individual machines once the tribal barriers between the old school metal benders and the new kids on the block break down a bit.

    And when we get to where true Drexlerian hard nanofabrication becomes possible, the 3-D printer will join the Bridgeport mill in the museum of obsolete manufacturing technology.

  28. I think we were the first people to ever fire a 3D printed Rocket Motor.

    I recall Aerojet made some thrust chambers by a printing-like process: they scribed holes into thin metal sheets, stacked up the sheets, and bonded them together by solid diffusion at elevated temperature. The holes were designed to line up to form cooling channels and outlets for transpiration cooling.

Comments are closed.