NSA Just Emailed A Wench To The International Space Station



  • Story here: http://www.iflscience.com/space/how-nasa-emailed-wrench-space

    Not as sexy as 24th-century Star Trek replicators, but given what we currently have available, it's pretty cool.


  • Grade A Premium Asshole

    @redwizard said:

    it's pretty cool.

    Yes it is. I so want a 3d printer, even though I have no idea what I would use it for.


  • BINNED

    @Intercourse said:

    I so want a 3d printer

    I you had. I would email you the blue prints of these


  • BINNED

    @Luhmann said:

    I would email you the blue prints of these

    Thinking of it ... that would start a hole new era of trolling. :trollface:


  • Grade A Premium Asshole

    @Luhmann said:

    I you had. I would email you the blue prints of these

    Evil idea: MakerBot add-on that monitors an email for appropriate files and automatically prints them...without any authentication. Someone should do that as a social experiment.


  • 🚽 Regular

    @Intercourse said:

    Yes it is. I so want a 3d printer, even though I have no idea what I would use it for.
    Around these parts there's a town where they used a couple of them to print a relatively large nativity scene.

    They were asking visitors whether they'd like to have their face scanned for next year's scene.



  • @redwizard said:

    Not as sexy as 24th-century Star Trek replicators, but given what we currently have available, it's pretty cool.

    I can see this being really handy for one-off fixtures/jigs and such -- the kind of tooling that you don't need very often, and you don't need over and over again, but boy, when you need it, it sure comes in handy!



  • It's indeed a perfect fit for the ISS, where sending a tool can cost thousands of dollars... assuming it can even wait until the next supply ship.



  • @tarunik said:

    I can see this being really handy for one-off fixtures/jigs and such -- the kind of tooling that you don't need very often, and you don't need over and over again, but boy, when you need it, it sure comes in handy!

    $1k or so just so that you're able to get that Philips screwdriver when you can't find one? Well...


  • Grade A Premium Asshole

    @Maciejasjmj said:

    $1k or so

    The prices have come down. You can pick them up now for ~$400. No clue what the quality is like on those, but there are less expensive options now.

    That would still be a damned expensive screwdriver.



  • Today I went to the local Media Markt (if you don't know it, it's the world's second largest retailer of consumer electronics, after Best Buy) and they had 3D printers for sale. So they have officially gone mainstream.



  • @tarunik said:

    I can see this being really handy for one-off fixtures/jigs and such -- the kind of tooling that you don't need very often, and you don't need over and over again, but boy, when you need it, it sure comes in handy!

    If you're looking for that sort of stuff, a small CNC mill is much more useful and has been available for $1000 for a very long time. 3D printing allows more shapes to be printed, but the product is much less durable and not that great looking.

    It's the right choice for the ISS because they don't have the space for two tools, but most hobbiests looking to make tools or jigs should go with a mill. If you are looking to make chess pieces - go with the 3D printer.


  • Java Dev

    3d printers, or at least most of the plastic-based ones, can reuse materials. So if your 3d-printed chess piece breaks you can melt it down and make a new one from the same materials. This is not that relevant when materials are cheap, but can be key on the ISS.



  • @Jaime said:

    3D printing allows more shapes to be printed, but the product is much less durable and not that great looking.

    Much of it depends on what you're printing in, not on 3d printed vs. machined by itself, I suspect...you are right that small mills are fairly inexpensive any longer, but that's not something apropos to put on the ISS all the same -- floating mill swarf in the air circulation is a bad idea.


  • Discourse touched me in a no-no place

    @Jaime said:

    3D printing allows more shapes to be printed, but the product is much less durable and not that great looking.

    What material are you printing with? If you can use a printer that uses laser metal sintering (!!!! 😃 !!!!) then the part can be made extremely durable. I happen to know of a very large US corporation that is working on using this for an very production component, and if I said what the component was, you'd probably be able to guess the name of the corporation. I'm not entirely sure whether I was told in confidence or not, so I'll stick with vague hints…

    On a more normal-workshop scale, nylon's pretty durable by comparison with ABS. The main downside is that it is a bit trickier to work with, as it needs a higher temperature and better control.



  • @dkf said:

    What material are you printing with? If you can use a printer that uses laser metal sintering (!!!! !!!!) then the part can be made extremely durable.

    I'm aware of laser sintering. The best you can do with laser sintering is equal to the typical material used on a mill. A $600 Sherline mill can cut tool steel. No 3D printer that costs under $100,000 can come close to that.



  • @anonymous234 said:

    So they have officially gone mainstream.

    My son and I were looking at one in a Microsoft Store just before Christmas. We even got to print a couple of demo items they had pre-loaded into the printer's memory. So I'd call them mainstream, too.

    @Jaime said:

    I'm aware of laser sintering. The best you can do with laser sintering is equal to the typical material used on a mill. A $600 Sherline mill can cut tool steel. No 3D printer that costs under $100,000 can come close to that.
    The PLA plastic used in 3D printers can be used for lost "wax" casting, which would be my main application for the printer.



  • @Jaime said:

    I'm aware of laser sintering. The best you can do with laser sintering is equal to the typical material used on a mill. A $600 Sherline mill can cut tool steel. No 3D printer that costs under $100,000 can come close to that.

    Shall we compare material-on-material then? (Say, an injection molded nylon or ABS part vs something 3D printed from the same material?)



  • How far away are we from being able to print off touchscreens, motherboards and graphics cards, that kind of stuff? I guess if it's feasible to print CPUs/GPUs in the longterm, then having a machine which spits out new computers cost effectively would be pretty compelling... (especially if you can feed it the blueprints for Computer Printer 2.0 somewhere along the way...)

    I heard you like computer printers, so I put a computer printer in your computer printer so you can print computers while you print computers...


  • Discourse touched me in a no-no place

    @tar said:

    How far away are we from being able to print off touchscreens, motherboards and graphics cards, that kind of stuff?

    A long way, unless you count the way that they're manufactured already as “printing” (which some of it sort of is). The problem is that they require lots of very different materials (glass, pure silicon, copper, etc.) some of which require very toxic chemicals to handle, and all of which requires being much cleaner than a normal home or workshop.

    Or at least my home isn't about to get a hydrofluoric acid supply. Not even a very little one.



  • @dkf said:

    my home isn't about to get a hydrofluoric acid supply. Not even a very little one.

    How much HCl are we talking about here? If it's a few hundred millilitres, then I'm already storing some as it is...

    To ask a slightly different question, what level of technology could I expect to print in the next 5-10 years? Could I print a transistor, or failing that, a vacuum tube?

    D.I.Y. E.N.I.A.C.



  • Hydrofluoric. HF. It's scary. Really scary. It eats through glass.



  • @aliceif said:

    Hydrofluoric

    Oh. In that case, less interested in that then. Maybe time to sleep until I can distinguish [i]C[/i]s from [i]F[/i]s again...


  • Discourse touched me in a no-no place

    @aliceif said:

    Hydrofluoric. HF. It's scary. Really scary. It eats through glass.

    In fact, that “eats through glass” is why it is used in the manufacture of silicon chips. One of the types of processing steps involves converting (part of) the surface of the silicon wafer to glass and then using HF to etch away some of it.

    Mind you, HF isn't nearly as “exciting” as F2 or some of the fluorides of oxygen…


  • FoxDev

    @dkf said:

    Mind you, HF isn't nearly as “exciting” as F2 or some of the fluorides of oxygen…

    do you want a scary acid?

    try H2SbF7

    https://www.youtube.com/watch?v=ckSoDW2-wrc#t=515



  • @tarunik said:

    Shall we compare material-on-material then? (Say, an injection molded nylon or ABS part vs something 3D printed from the same material?)

    I'm not sure why we would, but I'll play. Injection molding has stupid high first-part costs, but give an excellent surface finish. 3D printing is much cheaper if you want less than 1000 pieces (or if you want one piece today), but results in an inferior surface finish. 3D printing can make parts with hollow centers and/or complex concave sides that would be difficult or impossible to de-mold. Laser sintering or UV curing 3D printers can even make parts within parts.

    I have nothing against 3D printing. I was simply responding to your suggestion that 3D printing is a good way to make jigs, fixtures, and tools. These things are usually much better when made of metal, and a mill is a cheaper way to shape metal. I bought my CNC mill fifteen years ago for $1200.

    3D printers get a lot more buzz than mills because they require less expertise to operate. You just feed it a model and press go. Sure, there are some techniques to learn, but not that many. Milling requires a lot more skill - figuring out how to mount a part, zeroing after every tool change, zeroing after remounting the part to machine a different surface, etc. Because of this, CNC mills will never be a good choice for the general public.



  • @Jaime said:

    I was simply responding to your suggestion that 3D printing is a good way to make jigs, fixtures, and tools. These things are usually much better when made of metal, and a mill is a cheaper way to shape metal. I bought my CNC mill fifteen years ago for $1200.

    For reusable tools -- I agree with you that metal is a better choice of material than any plastic. I was speaking to single-use fixtures or jigs though...(which would be the likely usecase on the ISS).

    @Jaime said:

    Milling requires a lot more skill - figuring out how to mount a part, zeroing after every tool change, zeroing after remounting the part to machine a different surface, etc. Because of this, CNC mills will never be a good choice for the general public.

    There's also a CAM step in there, which varies from "annoying" to "GAAAAH" depending on how sophisticated the software driving the mill is.

    (I've used the PCB-making endmill at school a couple of times.)



  • @tar said:

    How far away are we from being able to print off touchscreens, motherboards and graphics cards, that kind of stuff? I guess if it's feasible to print CPUs/GPUs in the longterm, then having a machine which spits out new computers cost effectively would be pretty compelling... (especially if you can feed it the blueprints for Computer Printer 2.0 somewhere along the way...)

    I don't think these printers are capable of 10-nanometer-resolution yet.



  • @tarunik said:

    There's also a CAM step in there, which varies from "annoying" to "GAAAAH" depending on how sophisticated the software driving the mill is.

    CAM software depends on sophisticated algorithms and fast mills to work well. If you are trying to keep a million dollar mill busy, this works fine. For hobby stuff, I've found that a change in methodology is useful. Instead of building a positive model and asking software to figure out how to mill it, I wrote software that allows me to design in a subtractive manner. These types of models are trivial to turn into tool paths.



  • @Jaime said:

    For hobby stuff, I've found that a change in methodology is useful. Instead of building a positive model and asking software to figure out how to mill it, I wrote software that allows me to design in a subtractive manner. These types of models are trivial to turn into tool paths.

    We basically had to "trace over" our PCB artwork by hand, IIRC, to set the tool paths up for our mill.



  • Hydrofluoric acid is what they used in Breaking Bad to get rid of [spoiler]bodies[/spoiler]. Although Breaking Bad takes place in a world where chemistry behaves differently to real life.

    Also, what this guy said:

    @tar said:

    Could I [3D] print a transistor, or failing that, a vacuum tube?


  • Discourse touched me in a no-no place

    @tar said:

    Could I [3D] print a transistor, or failing that, a vacuum tube?

    Printing a transistor is close to how they're made now. You need pretty exotic materials to do it (control of the dopants is critical) and you need hyper-pure silicon to start with. The chemistry is… well, not easy stuff to do in the average home.

    You could probably print the components of a valve (I don't know how to print glass, but it may be possible or they may be a plastic that is a suitable alternative) but you'll have problems with the assembly, since a thermionic valve needs a high vacuum inside it to work right.

    It's probably cheaper to just buy ready-made valves of much high quality than anything you'll make yourself. It's definitely cheaper to buy ready-made transistors of tremendous quality.



  • @dkf said:

    You need pretty exotic materials to do it (control of the dopants is critical) and you need hyper-pure silicon to start with. The chemistry is… well, not easy stuff to do in the average home.

    I have been speaking with someone on IRC for quite a long time now who has been off-and-on tinkering with CMOS and MEMS processing in his apartment (he's a grad student who teaches IC reverse engineering). He didn't have too much trouble finding a supply of silicon or workable dopants (what he uses, IIRC, is in liquid form that's spin-coated onto the surface of the wafer), from what I recall -- etching wasn't an issue either (KOH happily wet-etches silicon).



  • So... fuck, next you'll be telling me I can't even build a difference engine with a 3D printer. I may as well just buy gadgets off Amazon like some kind of plebian then. At least until nanotech gets commoditized and then I can think about engineering a self-replicating gray goo scenario. Who's laughing [hypothetically] then, eh? They said I was mad...

    Fiked under: Mad?! ...I was furious!


  • FoxDev

    @tar said:

    So... fuck, next you'll be telling me I can't even build a difference engine with a 3D printer

    in one pass, probably not.

    but if you'r patient and willing to do a bit of assembly, yeah you can print that today.


  • Discourse touched me in a no-no place

    @accalia said:

    but if you'r patient and willing to do a bit of assembly, yeah you can print that today.

    A PCB will probably be printable quite easily for a single-layer board (as you wouldn't need to screw around with printing the substrate) and it is going to be easier than making the board by hand. That'll hold doubly true for multi-layer boards, which I will never make by hand again; getting the layer registration right was awful.

    I'm not sure what you'd print with. Probably not copper, not in small-scale work…



  • @dkf said:

    A PCB will probably be printable quite easily for a single-layer board

    Uh, what they were talking about is mechanical (at least the famous parts anyway).


  • Discourse touched me in a no-no place

    @dkf said:

    That'll hold doubly true for multi-layer boards,

    Can you even DO that as a hobbyist? I always wanted to mess around with making a PCB for a little project of some kind or other but I can't imagine how you'd make a multilayer thing without gluing multiple boards together or something.



  • @FrostCat said:

    Can you even DO that as a hobbyist? I always wanted to mess around with making a PCB for a little project of some kind or other but I can't imagine how you'd make a multilayer thing without gluing multiple boards together or something.

    Multi-layer boards really are several double-layer boards (copper clad prepreg) stacked up with unclad prepreg (a woven glass fiber mat that has yet-to-be-activated epoxy resins in it) sheets between them, more or less. The main issue is registration, though: you have to be very precise in how the layers match to avoid through-plated via barrels shorting to things they shouldn't.

    Fortunately, services such as OSHPark exist, and are far nicer/cheaper than DIY, even for double-sided (as PTH, soldermask, and silk are all very good things to have) -- just keep in mind that you're looking at a two-week turn time if you order PCBs this way.


  • Discourse touched me in a no-no place

    @tarunik said:

    The main issue is registration, though: you have to be very precise in how the layers match to avoid through-plated via barrels shorting to things they shouldn't.

    Sure, that's fairly obvious if you think about it, although since I've never actually done it I didn't know the term "registration" until now.

    @tarunik said:

    just keep in mind that you're looking at a two-week turn time if you order PCBs this way.

    Without clicking the link, I assume these guys either take one-off designs and batch a bunch of them from different people onto a big board and then cut 'em apart, or are capable of doing small-batch runs. I am on the mailing list of a company like that that I found via sparkfun, but never have had the time to mess around with it.



  • @FrostCat said:

    Without clicking the link, I assume these guys either take one-off designs and batch a bunch of them from different people onto a big board and then cut 'em apart, or are capable of doing small-batch runs. I am on the mailing list of a company like that that I found via sparkfun, but never have had the time to mess around with it.

    They do the former -- that's the way proto/short-run orders are handled in the PCB world (chip fabs do the same thing under a different name, too, see "multi-project wafers").



  • @dkf said:

    A PCB will probably be printable quite easily for a single-layer board...

    I'm not sure what you'd print with. Probably not copper, not in small-scale work…

    Maybe they can use multiple material tanks like they do today for Magenta/Yellow/Cyan/Black color printer tanks: Instead, use some combination Plastic/Solder/Copper/Epoxy?


  • Discourse touched me in a no-no place

    @redwizard said:

    Maybe they can use multiple material tanks like they do today for Magenta/Yellow/Cyan/Black color printer tanks: Instead, use some combination Plastic/Solder/Copper/Epoxy?

    I'm just trying to imagine what sort of environment could print those wildly different materials in quick succession. I've got this feeling that getting plastic and copper in the same process would involve chemical magic, as one doesn't melt at all until massively above the temperature where the other chemically decomposes, even without oxygen present. You'd probably be better off with a conductive plastic of some sort (as well as a more normal insulating plastic) since then you'd be able to keep the temperature range more similar, and you also wouldn't need the epoxy or solder (those are features of the limitations of the traditional process) but even so…

    If you could find a suitable compatible pair of materials, you could do it I guess.



  • The mistake you're making is buttuming that the raw materials used for this process would be the same as the raw materials currently used.



  • @dkf said:

    I'm just trying to imagine what sort of environment could print those wildly different materials in quick succession. I've got this feeling that getting plastic and copper in the same process would involve chemical magic, as one doesn't melt at all until massively above the temperature where the other chemically decomposes, even without oxygen present. You'd probably be better off with a conductive plastic of some sort (as well as a more normal insulating plastic) since then you'd be able to keep the temperature range more similar, and you also wouldn't need the epoxy or solder (those are features of the limitations of the traditional process) but even so…

    There's also sequence of printing, with perhaps a queue tray. The material that requires higher temperatures gets printed in advance and cooled, then the lower-temp material is added surrounding/layering it so it melts around the higher-temp material.

    That's just one way I could imagine working around the problem.


  • Discourse touched me in a no-no place

    @redwizard said:

    There's also sequence of printing, with perhaps a queue tray. The material that requires higher temperatures gets printed in advance and cooled, then the lower-temp material is added surrounding/layering it so it melts around the higher-temp material.

    That's just one way I could imagine working around the problem.

    Given that we were talking about fairly large objects with multiple layers (PCBs or something very much like it) then you'd need either a separate stage to stick the individual pieces together, as now (difficult!) or you'd be using some sort of more complex 3D structure, but then you'd need a much more complex printer since you would no longer be producing a simple layer at a time. Which would be an interesting challenge in control systems I guess, but not really for the average tinkerer.

    Finding a conductive/insulating pair of materials with relatively similar plastic and thermal properties would probably be easier. Then you could put the complexity in the print head and stick to simple printing.



  • @dkf said:

    I've got this feeling that getting plastic and copper in the same process would involve chemical magic, as one doesn't melt at all until massively above the temperature where the other chemically decomposes, even without oxygen present.

    Still catching up on the backlog of unread articles from the holidays.

    Copper melts at 1084.6°C (almost 2000°F) and oxidizes readily in air when it's hot. 3D printing it in the way that low-melting plastics are currently printed does not seem to me like it would ever be practical, at least in a consumer printer. Laser sintering might be able to make almost solid copper traces, but I'm not sure that would be practical in a home DIY printer, either.

    OTOH, printing fine particles of copper in a binder, like the copper paint used to repair PC boards, does seem like a viable option, although the resistance of the traces would be higher than solid metallic copper traces.

    The substrate could be printed using any suitable plastic, probably even epoxy. Epoxy is a thermosetting resin (plastic), rather than a thermoplastic, but some kind of curing agent could probably be printed at the same time as the resin, or maybe a liquid epoxy could be used in a UV curing printer. (I don't know what kind of resins are currently used in them; maybe they already can use epoxy.) Printing an epoxy/fiberglass composite, like typical PC boards, doesn't seem too practical, so the resulting boards might be a little more fragile. Thinking about it further, is seems like epoxy or some other thermosetting plastic would be required for the substrate; soldering would tend to melt a thermoplastic board. (It can even damage or destroy an epoxy/glass board, if you don't know what you're doing.)

    Through-hole vias pose another problem. It seems to me that they would pretty much require printing the substrate and copper together; it would be rather difficult to get the copper printed into the small-diameter holes if the substrate were completely printed to its full thickness first. Printing a thermoplastic substrate and copper powder in a thermoplastic binder could probably be done together (but see above for disadvantage of using thermoplastic).

    If both were thermosetting, that could work, but to my limited knowledge, all 3D printers that use UV cured (thermosetting or similar) plastics print by curing a layer of resin at the surface of a container of liquid or powdered resin; this doesn't seem easily adaptable to printing two materials at the same time. However, my knowledge is clearly incomplete; I just found this, which, if a version with two print heads were to be made, could do the job.

    TL;DR: It could probably be done with printers not too different from some available today1 and suitable materials (particularly, a powered copper/resin mix with good conductivity).

    1 At least if the DeltaBots project linked above got their Kickstarter funding.


  • Grade A Premium Asshole

    @HardwareGeek said:

    Epoxy is a thermosetting resin (plastic), rather than a thermoplastic, but some kind of curing agent could probably be printed at the same time as the resin, or maybe a liquid epoxy could be used in a UV curing printer.

    Mix the catalyst in to the epoxy in the print head. We sometimes use an epoxy (or polyurethane adhesive?) that has a dispensing head that mixes the two parts in a helix fashion as it traverses through the nozzle. Not sure exactly what it is, we just call it by the trade name, "202".


  • Grade A Premium Asshole

    @dkf said:

    You'd probably be better off with a conductive plastic of some sort (as well as a more normal insulating plastic) since then you'd be able to keep the temperature range more similar, and you also wouldn't need the epoxy or solder (those are features of the limitations of the traditional process) but even so…

    Also, something similar to Gallium could provide the conductivity, with a suitably low melting point. Some manner of conductor like silver suspended in a plastic or epoxy matrix would likely be a better choice. Provided that a change in manufacturing processes could get the resistance down to a suitable level.

    I have used various conductive epoxies to make repairs. I would imagine they would have very high resistance over long runs though and be unsuitable for such purposes as they are now.



  • @Polygeekery said:

    Also, something similar to Gallium could provide the conductivity, with a suitably low melting point.

    Hmm, interesting idea. Gallium itself melts at too low a temperature; you don't want your PC board melting when you touch it. Most alloys melt at an even lower temperature — that's the main use of metallic gallium1 — but I'm sure that there must be some ratio of, say, gallium and copper that would have a melting point similar to commonly printed thermoplastics. It would probably need a modified print head, as it would be a lot less viscous — a real liquid, rather than the very soft goo that results from heating a thermoplastic. The printer would also have to purge the head before being turned off; like ice, gallium expands when it freezes, which would almost certainly damage the print head.

    One thing that might prevent this from working is gallium's tendency to attack other metals and make them brittle. This might or might not be a problem if the gallium is already alloyed with copper or whatever.

    Surprisingly, at least to me, is that it is fairly safe to handle. I get the impression that it's probably less toxic than the epoxy catalyst. Gallium (III) acts very much like iron (III) in the body, and is even used in some pharmaceuticals because of its similarity.

    1 The main use of gallium in general is in semiconductors for microwave and LED applications.


Log in to reply