I spent way too much time today fantasizing about metal 3D printers. I understand they typically work by using lasers to melt a fine layer of metal powder into solid metal, and then they add another layer of powder and melt more of it, and repeat, then drain away the unmelted powder. Currently they cost several hundred thousand dollars and can build stuff in something like 20cm x 20cm x 20cm volume. Well, here’s my fantasy design for an industrial-scale metal 3D printer, that would probably be orders of magnitude better in every way, and thus hopefully orders of magnitude cheaper. Warning/disclaimer: I don’t have any engineering or physics background and haven’t bothered to check several things so probably I’m making stupid mistakes somewhere.
1. The facility consists of a big strong concrete-and-steel dome, with big airlocks in the sides. The interior is evacuated, i.e. made a vacuum.
2. The ground floor of the facility is flat concrete, mostly clear. On it roll big robot carts, carrying big trays of metal powder. They look like minecarts because they have high sides and need to be able to support a lot of weight. They have a telescoping platform on the interior so that the surface of the powder is always near the top of the cart, whether the powder is a meter deep or a millimeter. The carts heat the powder via induction coils, so that it is as close as possible to the melting point without actually being in danger of melting.
3. The ceiling of the facility (not literally the ceiling, more like a rack that hangs from it) is a dense array of laser turrets. Rows and rows and rows of them. They don’t have to be particularly high-powered lasers because the metal they are heating up is already warm and there isn’t any atmosphere to diffuse the beam or dissipate heat from the target. When laser turrets break they can be removed and replaced easily, perhaps by robots crawling above them.
4. Off in one corner of the facility is the powder-recycling unit. It has a bunch of robot arms to lift out the finished pieces while the carts dump their contents, the unused powder being swept up and fed to powder-sprinkler-bots.
5. There are a bunch of powder-sprinkler-bots that perhaps look like the letter n. They roll over carts and deposit a fresh layer of powder on them. The powder is slightly cooler than the powder in the carts, to counteract the extra heat being added by the lasers.
6. The overall flow is: Carts arrange themselves like a parking lot on the factory floor; powder-sprinklers pair up with them and go back and forth depositing powder. The laser arrays above do their thing. When carts are finished or powder-sprinklers need more powder they use the lanes to go to the powder-recycling unit. Said unit also passes finished product through an airlock to the outside world, and takes in fresh shipments of powder through another airlock.
7. Pipes of oil or water manage heat in the facility, ensuring that various components (e.g. the lasers) stay cool, transporting the waste heat outside.
8. When machines break down, as they no doubt will, (a) teleoperated fixer robots might be able to troubleshoot the problem, and if not, (b) they can just wheel the faulty machine through an airlock to be repaired by humans outside.
9. I think current powder printer designs involve a second step in which the product is baked for a while to finish it? If this is still required, then the facility can have a station that does that. There are probably massive economies of scale for this sort of thing.
Overall this facility would allow almost-arbitrarily-large 3D objects to be printed; you are only limited by the size of your carts (and you can have one or two gigantic carts in there). Very little time would be wasted, because laser turrets will almost always have at least one chunk of cart within their line of fire that isn’t blocked by something. There are relatively few kinds of machine (carts, turrets, sprinklers) but lots of each kind, resulting in economies of scale in manufacturing and also they are interchangeable so that several of them can be breaking down at any given time and the overall flow won’t be interrupted. And the vacuum environment probably (I don’t know for sure, and have no way of quantifying) helps a lot in many ways—lasers can be cheaper, less worry about dust particles messing things up, hotter powder melts quicker and better, idk. I’d love to hear an actual physicist’s opinion. Does the vacuum help? Do open cart beds with ceiling lasers work, or is e.g. flying molten metal from neighboring carts a problem, or lasers not being able to focus at that range?
Outside the vacuum dome, printed parts would be finished and packaged into boxes and delivered to whoever ordered them.
If one of these facilities set up shop in your city, why would you ever need a metal 3D printer of your own? You could submit your order online and immediately some unused corner of some powder bed would start getting lasered, and the finished product would be available for pickup or delivery, just like a pizza.
If we ever start doing fancier things with our 3D printers, like using different types of metal or embedding other substances into them, the facility could be easily retrofitted to do this—just add another type of sprinkler-robot. (I guess the heated powder could be a problem? But some carts could be non-heated, and rely on extra-duration laser fire instead, or multiple redundant lasers.)
Cool sci-fi-ish idea, but my impression has been that 3D printing is viable for smaller and/or specific objects for which there is not enough demand to set up a separate production line. If economies of scale start to play a role then setting up a more specifically optimized process wins over general purpose 3D plant.
This sort of thing would bring the cost down a lot, I think. Orders of magnitude, maybe. With costs that low, many more components and products would be profitable to 3D print instead of manufacture the regular way. Currently, Tesla uses 3D printing for some components of their cars I believe; this proves that for some components (tricky, complex ones typically) it’s already cost-effective. When the price drops by orders of magnitude, this will be true for many more components. I’m pretty sure there would be sufficient demand, therefore, to keep at least a few facilities like this running 24⁄7.
I spent way too much time today fantasizing about metal 3D printers. I understand they typically work by using lasers to melt a fine layer of metal powder into solid metal, and then they add another layer of powder and melt more of it, and repeat, then drain away the unmelted powder. Currently they cost several hundred thousand dollars and can build stuff in something like 20cm x 20cm x 20cm volume. Well, here’s my fantasy design for an industrial-scale metal 3D printer, that would probably be orders of magnitude better in every way, and thus hopefully orders of magnitude cheaper. Warning/disclaimer: I don’t have any engineering or physics background and haven’t bothered to check several things so probably I’m making stupid mistakes somewhere.
1. The facility consists of a big strong concrete-and-steel dome, with big airlocks in the sides. The interior is evacuated, i.e. made a vacuum.
2. The ground floor of the facility is flat concrete, mostly clear. On it roll big robot carts, carrying big trays of metal powder. They look like minecarts because they have high sides and need to be able to support a lot of weight. They have a telescoping platform on the interior so that the surface of the powder is always near the top of the cart, whether the powder is a meter deep or a millimeter. The carts heat the powder via induction coils, so that it is as close as possible to the melting point without actually being in danger of melting.
3. The ceiling of the facility (not literally the ceiling, more like a rack that hangs from it) is a dense array of laser turrets. Rows and rows and rows of them. They don’t have to be particularly high-powered lasers because the metal they are heating up is already warm and there isn’t any atmosphere to diffuse the beam or dissipate heat from the target. When laser turrets break they can be removed and replaced easily, perhaps by robots crawling above them.
4. Off in one corner of the facility is the powder-recycling unit. It has a bunch of robot arms to lift out the finished pieces while the carts dump their contents, the unused powder being swept up and fed to powder-sprinkler-bots.
5. There are a bunch of powder-sprinkler-bots that perhaps look like the letter n. They roll over carts and deposit a fresh layer of powder on them. The powder is slightly cooler than the powder in the carts, to counteract the extra heat being added by the lasers.
6. The overall flow is: Carts arrange themselves like a parking lot on the factory floor; powder-sprinklers pair up with them and go back and forth depositing powder. The laser arrays above do their thing. When carts are finished or powder-sprinklers need more powder they use the lanes to go to the powder-recycling unit. Said unit also passes finished product through an airlock to the outside world, and takes in fresh shipments of powder through another airlock.
7. Pipes of oil or water manage heat in the facility, ensuring that various components (e.g. the lasers) stay cool, transporting the waste heat outside.
8. When machines break down, as they no doubt will, (a) teleoperated fixer robots might be able to troubleshoot the problem, and if not, (b) they can just wheel the faulty machine through an airlock to be repaired by humans outside.
9. I think current powder printer designs involve a second step in which the product is baked for a while to finish it? If this is still required, then the facility can have a station that does that. There are probably massive economies of scale for this sort of thing.
Overall this facility would allow almost-arbitrarily-large 3D objects to be printed; you are only limited by the size of your carts (and you can have one or two gigantic carts in there). Very little time would be wasted, because laser turrets will almost always have at least one chunk of cart within their line of fire that isn’t blocked by something. There are relatively few kinds of machine (carts, turrets, sprinklers) but lots of each kind, resulting in economies of scale in manufacturing and also they are interchangeable so that several of them can be breaking down at any given time and the overall flow won’t be interrupted. And the vacuum environment probably (I don’t know for sure, and have no way of quantifying) helps a lot in many ways—lasers can be cheaper, less worry about dust particles messing things up, hotter powder melts quicker and better, idk. I’d love to hear an actual physicist’s opinion. Does the vacuum help? Do open cart beds with ceiling lasers work, or is e.g. flying molten metal from neighboring carts a problem, or lasers not being able to focus at that range?
Outside the vacuum dome, printed parts would be finished and packaged into boxes and delivered to whoever ordered them.
If one of these facilities set up shop in your city, why would you ever need a metal 3D printer of your own? You could submit your order online and immediately some unused corner of some powder bed would start getting lasered, and the finished product would be available for pickup or delivery, just like a pizza.
If we ever start doing fancier things with our 3D printers, like using different types of metal or embedding other substances into them, the facility could be easily retrofitted to do this—just add another type of sprinkler-robot. (I guess the heated powder could be a problem? But some carts could be non-heated, and rely on extra-duration laser fire instead, or multiple redundant lasers.)
Cool sci-fi-ish idea, but my impression has been that 3D printing is viable for smaller and/or specific objects for which there is not enough demand to set up a separate production line. If economies of scale start to play a role then setting up a more specifically optimized process wins over general purpose 3D plant.
This sort of thing would bring the cost down a lot, I think. Orders of magnitude, maybe. With costs that low, many more components and products would be profitable to 3D print instead of manufacture the regular way. Currently, Tesla uses 3D printing for some components of their cars I believe; this proves that for some components (tricky, complex ones typically) it’s already cost-effective. When the price drops by orders of magnitude, this will be true for many more components. I’m pretty sure there would be sufficient demand, therefore, to keep at least a few facilities like this running 24⁄7.