3D printing

[Jonfliesgoats]

@DBowman

Cool link, Dnowman!

[Shpaget]

1 hour ago, DBowman said:

Granted, but 'large production runs' are not the 'long space mission' use case - take 5000 parts vs mass of 5 parts of gunk. 

That's not how 3D printing works. You can't just have some all purpose goo that turns into whatever material you want. Not only such a goo doesn't exists, but different materials require different 3D printing processes, and most meterials used in our products can't be 3D printed at all.

Take any man made item around you and count the different materials in it. Furthermore, making a product usually does not involve just the shaping. More often than not there are additional steps that need to be done. For example, some steel parts need to be quenched and then tempered. Others have to be forged. Those processes were developed because there is a genuine need for the properties they introduce into the material.

Glass products require their own production processes.

Electronic components, such as microcontrollers and CPUs require huge production facilities.

You can't just walk up to a 3D printer and ask it to be a Star Trek replicator.

[Nuke]

multi filament extruders are going to be the next big thing. not only printing color but also printing different materials, conductive/insulative filament (for 3d printed circuitry), and support material. the support material allows you to print machines pre assembled for example using a dissolvable scaffold, just soak your part in water (or other chemicals) and all thats left is a working mechanism. cool stuff. the other side of it is its gonna be expensive and more parts to maintain. i think half of the work i do with my printer is maintenance.

that said the 3d printer isnt the end all solution to all your manufacturing needs. think of it as another tool in the shop sure i have a printer, but i also have a drill press, a cordless drill, and my most recent acquisition, a scroll saw. i suspect future builds i have in mind will use all of them.

Edited January 1, 2017 by Nuke

[DBowman]

17 hours ago, Shpaget said:

That's not how 3D printing works. You can't just have some all purpose goo that turns into whatever material you want. ... You can't just walk up to a 3D printer and ask it to be a Star Trek replicator.

No kidding. For example I cannot in the near term envisage 'desktop chip fabs'. For sure they will have to take some particular spares for un manufacturable parts. I think they will design for parts minimisation / multi usability and manufacturability, e.g design it to take a non optimally big bolt to avoid needing another gauge of bolt, use a suboptimal material because you are taking some anyway so 'its free' on the spares gunk mass budget. My 5000 part vs 5 parts of 'gunk' is just an illustrative example of the principle. I could imagine seals & gaskets being significant and maybe you could cover lots of your parts with a few materials, similarly all sorts of support structure for seals, gaskets, tubing etc might be able to be covered by a few materials. Choosing a sub optimal material that requires a little more mass but is 're usable' could make sense if using broken parts as feedstock for repairs was possible.

My main points are

1. being really sure means taking a lot of spares, if you can manufacture them it can be a big mass win. If you have to take 5 * 20 parts worth of 20 kinds of gunk it's still less mass than 5000 parts. If you cannot manufacture some item like high precision high temp turbo blades then take them but plan to be able to manufacture the housing or support structure. 
2. due to #1 the switchover point between prototype / once off items vs mass produced items is going to fall in a different place than in an earth-side supply chain with a huge market. Something you buy in the store is optimised for cheap manufacturability, short lifetime, and zero repairability ( at least it seems so ). Something on a 15 year Mars base mission is going to be optimised completely differently - for few materials, robust long life, and repairability. For example poly ethylene will be pretty easy to manufacture out of Mars atmosphere so probably we would use it for all sorts of 'inappropriate' tasks since we'd have heaps of it.

[wumpus]

On 12/31/2016 at 6:19 PM, Kerbart said:

Polishing afterwards does nullify it. What's the point of printing everything constructed together if you have to take it apart, polish it, and assemble it again? It's easier to print the parts separate, in that case. But than you cannot create "impossible" assemblies that 3D printing makes possible.

Not only that, for something like a turbine they are likely already using subtraction methods on a CNC machine.  Put a block of forged alumninum in one end, get a forged turbine (shaft or housing) out the other end.

12 hours ago, DBowman said:

No kidding. For example I cannot in the near term envisage 'desktop chip fabs'.

Desktop chip fabs make even less sense than most 3d printing dreams.  Wafer creation, silicon purity control, the incredibly nasty chemicals used for those processes pretty much the naive approach silly for the desktop.  Not to mention that FPGA/CPLDs exist that would effectively replace much of what you would use a "desktop fab" to do in the first place (while I don't think their analog counterparts took off [other than multi op-amp chips], I suspect they would become a thing before "desktop fabs" ever did).

Besides, what you would really want is the "desktop PCB manufacture/Pick&Place".  The pick and place is getting there, but I'm less sure about PCB manufacture (yes, I'm old enough to have etched single layer PCBs with acid and pens).

[Nuke]

automatic desktop pcb printers do exist but they leave a lot to be desired. most use conductive ink and require cold soldering of components with conductive epoxy. this is fine if you want a board with a short life span. etching copper clad at home is cheap and easy if you use the toner transfer method, uv photo transfer is also popular. with the chinese pcb mills offering boards at a few bucks per square inch (and for professionally etched, drilled and silk screened boards at that) its kind of a hard sell to have a desktop machine that does all that for you.

also the whole bitcoin mining fiasco showed that you could crowd fund an asic if you really wanted to. when the quest for bigger racks of servers (and lots of gpus as well) revealed diminishing returns, some of the smarter miners moved on to fpga solutions. eventually enough bitcoin miners were using fpgas to warrant an even faster system, so they managed to crowdfund the million or two bucks needed to have do a run of asics to really accelerate mining. doing chip fab at home is probibly not going to happen, and an fpga is fast enough for anything you might want to do, and is faster than anything you could build. if what you are building is in high enough demand, then crowdfunding becomes approachable.

Edited January 3, 2017 by Nuke

[Jonfliesgoats]

Interesting!

[DBowman]

aviation week's take on what NASA is up to re in orbit assembly in the near term. Mainly how to build things that are unlaunchable like very big solar arrays or next gen telescope. Robotic arms and 3D printing mainly. Made In Space Inc put the 3D printer on ISS (made some replacement connectors for some irrigation thing) and has; a vacuum printer tech (spars in space), 'redo' machine to recycle printed things, and some other stuff. Early days obviously ...

[Bill Phil]

On ‎12‎/‎31‎/‎2016 at 5:19 PM, Kerbart said:

Polishing afterwards does nullify it. What's the point of printing everything constructed together if you have to take it apart, polish it, and assemble it again? It's easier to print the parts separate, in that case. But than you cannot create "impossible" assemblies that 3D printing makes possible.

You can polish the outside so it looks better to the unaided eye. Doesn't nullify the internals being 3d printed without any polishing.

[Kerbart]

3 hours ago, Bill Phil said:

You can polish the outside so it looks better to the unaided eye. Doesn't nullify the internals being 3d printed without any polishing.

Good luck driving around in your 3d printed car, beautifully polished on the outside. Where it matters most, because that's what we can see, after all!

 

[Jonfliesgoats]

Well, there are advantages to printing which yield improved structural integrity.  For example we can use material gradients.  Again, I refer to the GE GENX engine with its printed nozzles that are subject to high heat and stress.

To be sure initial printing of materials today can't compete with forged and milled parts.  I don't use printed cylinders on my personal airplane, for example.  While the GENX engines at work use printed nozzles, they still use milled turbine blades for the parts under the most combined physical and thermal stress.

Technology advances, though.  Perhaps we need to think where these additive manufacturing techniques will be in thirty years relative to milled parts?

[Bill Phil]

On ‎1‎/‎6‎/‎2017 at 6:39 PM, Kerbart said:

Good luck driving around in your 3d printed car, beautifully polished on the outside. Where it matters most, because that's what we can see, after all!

 

The only part of the car that is 3d printed is the outside and part of the frame. The frame doesn't need polishing. Polishing the outside makes it look no different from a normal car. Not to mention that it would weigh less than a conventional car using the materials ORNL used. Higher performance from the same powertrain, which is electric, if I recall. And not 3d printed.

And, in any case, you misunderstood the intentions of my original comment. I was pointing out that high quality productions can still have been 3d printed. It was, after all, in response to a comment on the quality of 3d printed objects. Fresh out of the printer (after cool-down and setting, as it is required... if it is required) it won't look good. Not unless you managed to get a very high resolution print, however that would take a very long time to print. But that doesn't prevent you from going in afterwards (as you would for a good number of things not 3d printed) and smoothing out the outside.

In the case of moving parts, you can't really print them as they need to be on a desktop printer, as far as I am aware. Some of the much more advanced printers, sure. But those are expensive as heck. And if you want any good performance out of said parts, you'll need to be on a very high resolution. Basically, 3d printing said moving parts isn't going to be very helpful, since we'll need the better performance you get after the print is done.

Think about it like this: rapid prototyping on an even greater scale. Functional prototypes, to boot.

Edited January 8, 2017 by Bill Phil

[DBowman]

These guys are 3D printing solid fuel for hybrid motors. A hybrid motor is like a solid fuel rocket but the solid propellant wont 'burn' by itself, another liquid or gas propellant is passed through the solid fuel - so you can throttle / start / stop the motor. It's not quite 'printing high explosives', but nearly.

Controlling the time evolution of the 'burn area' controls the maximum thrust. For example in an SRB with a cylindrical void the 'burn area' increases as the propellant burns, leading to increasing thrust as mass decreases. Some SRBs use 'star prism' voids so the burn area decreases as the mass increases - leading to more constant thrust. These guys are doing 'something smart' re 'printing the voids' - I've not looked at the patent so I'm not sure what.

I imagine these hybrids start to 'waste' the liquid/gas propellant as the void gets large and it stops contacting the burn surface. hmm dual helical fuel ports in 3D printed motor