Extrapolating ~10-20 years from our current technology, what are some limitations with 3-D printing? Is the ability to print any type of matter in reach, or are there some fundamental issues that might limit this technology?
So I've been 3D printing for nearly 2 decades so I feel I can extrapolate fairly well here. A lot has already been said but I'll just add:
Dirt cheap printers requiring branded filaments:
The cost of a machine will keep coming down (My first entry-level machine in 2005 was 2 weeks x Australian minimum wage. My most recent entry-level machine (2020) was 2 days x Australian minimum wage.), and, as people compare machines on price, you'll end up with machines sold at a loss but requiring overpriced filament. Similar to razor and blades / 2d printer and inks. So a limitation of home / hobbyist machines in the future will be they need manufacturer approved filament - with the microchip in the roll to validate it.
You can of course pay a premium to side-step this with an expensive printer, get hacked filament, or build a printer yourself from parts, but how many people today build their own 2D printer to get around the overpriced ink?
Positional accuracy imperfections.
Subtle amounts of slack in the belts, morphing of the guide rails, twisting of the build plate, etc limit how accurately the head can be positioned. My best printer can position itself to 0.002mm electronically but realistically it's off by about 0.05mm minimum (0.5mm if I haven't calibrated it in the last few days). 20 years in the future a home machine may knock another zero off that but it's still not accurate enough for printing say, a computer chip, which needs ~0.0001mm accuracy or better.
You're going to have to service it still.
They will still be susceptible to jams in some form (a common one is material expanding in the tube from heat and blocking it). Every few months minimum you'll need to clear a clog of some form. It'll be an improvement from today (my 5 machines are running 247 and every week I need to service at least one).
This may become a service done by a profesional, like a car mechanic. Expect a counter on the machine saying "Oh you've printed 200kg of plastic - due for a service!"
Tiny layer imperfections wont go away
It will still have tiny layer variations, mostly due to subtle differences in material flow rate. A common cause now is temperature variation and back pressure causing skips (the extruder cant push the material fast enough so there's less material in an area than expected). These may be mitigated by sensors either detecting the abnormal flow or scanning the model after each layer and retouching, but there will be some blemishes from the skips and slips of the extruder and tiny variations in filament width.
As already mentioned, you wont be printing a transparent plexiglass plane anytime soon.
Will still need some post processing
Prints will still suffer stringing (where tiny trails of molten material leave the model footprint like molten cheese from a pizza) and pimples (where tiny extra bits of material accumulate on the head and merge with the surface later). These will require post processing.
Upper end machines will be able to assist with this but no 3d printing a ventilator and connecting it straight to a patient without tiny bits of plastic hairs getting into their lungs.
In the early stages of covid19 I was working with other home hobbyists printing health supplies as needed, and most of the home printing tech could only do PPE (eg face shield mounting bands). Only one of my printers Resin SLA tech) was able to do ventilator parts due to concerns about stringing tiny microfibres.
Fault lines and non-uniform strength.
Prints will still have fault lines. Even with a multi-axis rotating base plate allowing the z axis to pivot in multiple directions to mitigate this, there still will still be weakness between the layers, I think we can get to the point that so it isn't a series of straight lines in the z plane, but the fault lines between layers will still be there as a weak point.
I can already 3D print things that can hold my own body weight (handles / straps / steps / etc), but I'm not going to be 3D printing my own carabiner and climbing a mountain with it.
As someone who 3d prints and fire dances I will never be able to 3d print my own fire dancing toys. If we need something to become stringy when hot to print at desktop temperatures, that will not be able to resist flame. Even the high temp resins and thermoset resins available in sla lose strength at like 110 degrees. In a few decades this will go up with new material tech, but we wont be able to achieve strength under fire yet melt when required for forming in a machine on my desk.
Government restrictions on materials
Greenwashing efforts will apply some legal limitations. With concerns about the ocean becoming plasticised and jurisdictions banning plastics straws and cutlery (instead of banning the more substantial fishing nets and plastic packaging), expect the "home novelty plastic figurine creation" industry to take some legal hits. Expect some government bans on useful filaments in a misguided effort to cut down on pollution.
Expect laws regarding plastic labeling on home prints.
In addition to government banning useful filaments, expect government legislation on labeling prints. PLA filament can be trivially recycled (or composted), but it shouldn't be mixed with other plastics in the recycling plant in any decent volume as it can degrade the final product, and if output plastic is inferior, it can't be sold, making recycling unviable.
Expect to print the little chasing arrows with a number on the base of all prints, from cute figurines to machine parts. It may even be done automatically by the firmware.
Expect 'Taggants' in high strength filaments to target 3D printable guns.
How do you trace a 3D printed gun that's been used in a crime? The same way we trace explosives back from a bomb blast crime scene back to manufacturer / batch / invoice number. It's called Taggants, and has been used successfully in prosecution of bomb cases. Tiny microparticles encode bits of information allowing forensic investigators to identify the filament batch number by looking at the part under a microscope.
So expect the FBI / etc to be able to trace 3D printed parts back to the sale of a batch number of a filament wholesale to at least the final retail outlet (where they case seize security footage / credit card receipts), perhaps even to the individual roll.
I'd expect this to apply on any high-strength, high temp, filaments. Anything good enough you could use to make an automatic rifle out of minimum.
People will start selling their creations online as 3d-printable as an alternative to shipping it. However to limit the propagation of copies, expect that design to be licensed to your printer only, and only available to print for a limited window.
So of course when a printer dies, or a manufacturer goes bust, or a hard drive dies, or you forget the password, or you just have an off-brand printer, you won't be able to print something even if you have the design locally and paid for it.
When a part breaks after the warranty period, DRM will prevent you from reprinting it, you'll have to buy the rights to print another one.
Aside - expect society to change tastes to geometry that can be 3D printed.
Rather than a technical limitation stopping people from printing their dreams, I expect society will (in some cases) change their dreams to the technical limitations.
Which furniture company do you expect to remain in business in 20 years?
- The one making furniture in some 3rd world country on the cheap and shipping it to markets in shipping containers.
- The one that has a big 3D printer out back and prints stock as it sells? Each retail store has one in stock of each design that, when it sells, a replacement is printed.
Rather than keep warehouses of stock from cheap overseas manufacturers (and pay interest on all that finance), expect things to be printed just-in-time where possible. Because there are geometry requirements for 3D printing things (as detailed in the other answers so I wont repeat), the available products will change in line with that printing requirements, and tastes will change to follow.
Every trendy apartment will be decorated with things that were 3D printed, and that distinct style will become common due to its economy.
A table made from a big peice of wood will look clunky and obsolete in 20 years when compared to the modern, 3D printed on demand furniture.
(Or even cheaper idea: A furniture store only has display stock. When you order, they drive a trailer to your house and plop it in your driveway for a few hours / overnight. The new chest of drawers you ordered is printed on your property, they come back, unload it for you, and take the trailer to the next customer.).
Predicting 20 years into the future is incredibly hard
Let me begin with a disclaimer. 99.99% of the technology we enjoy today was invented in the last 150 years. Twenty years ago (2,000ish), the era of Palm Pilots, the idea of fully-graphical hand-held computers was still very much the domain of Star Trek and the flat screen monitors that are a part of our everyday life were only just hitting the commercial market — most of us were still using CRT monitors and televisions.
Twenty years is technologically forever.
Therefore, the only limitations I can reasonably predict are (a) geometric and (b) mechanical
3D printing will always have problems with geometry. For example, 3D printing a can for spray paint will have trouble with the marble inside the can, which must be entirely detached to operate. The same could be said for bearing assemblies. And those are just the easy-to-grasp examples. 3D printing describes the result, not the process. The process is very much 2D in nature.
3D printing will also be limited by the number of materials that can be brought to bear. Printing an entire car as we understand them today would mean printing multiple kinds of metals, plastics, ceramics, textiles... This is primarily an issue of practicality as, perhaps, one could imagine bringing different print heads to bear at different times to build things... but that brings us back to the issue of geometry. The window in the car door has a bolt through it. How do you keep the glass in place while you shift back-and-forth between print heads to lay down the metal of the bolt? How do you keep it all in place in the first place? But I digress...
Another limitation is anything having multiple pressures. Let's go back to that spray paint can. What if we try to 3D print the entire, finished can? That means we're trying to lay down pressurized paint at the same time we print unpressurized metal. Theoretically, in this case, we could pressurize the entire print space ... but what if you needed two propellants at different pressures?
The size of the print space will always be an issue. This might be more an issue of economics than anything else, but it seems that the ability to print a microchip in the space needed to build a dump truck would rarely make sense. It would be faster to have one printer to print the control electronics and another to print larger components then assemble. But, you will always be limited by the size of the print space. The support bars needed to hold the print heads and transport the printing medium would need to be fantastically strong to accurately print a 747 airplane. (And we're back to that first bullet... good luck printing the tires and wheel assemblies in-situ.) Support bar sag will probably always be a significant limitation (limiting the size of what can be printed).
Finally, 3D printing will always be limited by what can be made a liquid within acceptable conditions. 3D printing with molten steel will likely never make sense. It's why most 3D printing today is done with, basically, glue. It anneals to what's already there and hardens almost instantaneously. Few materials can be worked with in that way. In fact, anything that requires tempering (like tempered steel) would intrinsically limit what can be printed at one time (if at all).
It's worth noting that while it's tempting to see 3D printing as morphing into something akin to Star Trek replicators, they won't actually do that. To me, 3D printing feels like a "cul-de-sac technology." A cul-de-sac technology is one that is either necessary or, at least, obvious in hindsight to the development of technology as a whole — but in and of itself, the branch of technology doesn't go anywhere. It could be said that all technologies are intrinsically cul-de-sac technologies (telephone land lines are a good example), but some are very short-lived. Palm Pilots, for example.... My gut tells me today's 3D printing will morph into something on an industrial scale that will be a hybrid of technologies — and I wish I knew what... I'd make billions....
JBH does a great job covering the limitations of geometry so I will just add a few things.
Printing metal is severely hampered by the fact iron based metal end up brittle since they can't be properly forged or tempered. so don't expect 3d printing of steel for anything structural.
Anything that must be altered after the fact will be out, so no vulcanized rubber, this applies doubly so for anything reactive like a battery.
The same goes for things that need to be uniformly transparent like phone screens. 3D printing can't lay down a uniform sheet of material without being made only for that.
Don't expect 1 unit to do everything. you will need different printers for drastically different materials, the machine that makes leather will not be making dishwasher parts.
Expect big improvements of 3D printing of biological materials like food or tissue, They are already being developed, it needs a lot of things worked out but there is no hard barriers. expect 3D printed organs and transplant tissue. They may not be perfect but they are much better than the alternative, especially if you make them from your own cells, don't expect this at home however, this will be something that gets done in specialized hospitals.
Expect 3D printed foods. Do not expect things that are largely unprocessed like steak or fruit, although vat grown steak may become common and this may start as 3D printed precursor. Some things will still have to be cooked afterwards, you can 3D print a cake but you have to bake it afterwards. Chemistry might even let us custom make more complex biological molecules from simple ones although 20 years might be too soon for that.
Expect 3D printed leather, its already in the works, printed and woven cloth would not surprise me either. So you could have a home machine that makes custom clothes. Although the machine will have to do more than just print.
Expect 3D printed home made circuitry. they may not be quite as good as normal circuits but the on demand customization ability will more than make up for it. Home made electronics will explode in popularity. buy a bespoke laptop.
Expect integration with other technology, micro-shops are combination 3D printing and CNC machines that can do things neither one by itself can do. Medical tissue printers may include gene sequencer, protein sequencers, and incubators.
Lastly expect a term to preplace "3D printed", something simpler or shorter. The more common a technology becomes the shorter its name tends to get.
3D printing will become a daily tool for construction specialists/handymen.
It is inevitable that 3d printing will:
- Produce prints faster
- Print stronger materials
- Become easier to use
Once the technology reaches the level where a fist-sized part can be printed in under an hour, it will revolutionize the way that various "handymen" or construction specialists do their work. For example, right now, a plumber needs to have the back of their van full of every conceivable pipe fitting so that they can minimize shopping runs and order delays. If, for example, they arrive at a customer and that customer has an oddball pipe fitting somewhere, there is no way for the plumber to simply improvise a fit--they have to measure, order the part, and then come back to install it. With a fast 3d printer, the plumber could simply say, "I need a 7/8ths to 1 3/8ths male to male adapter with a 90 degree bend" and half an hour later, the part is printed and ready to install.
Similarly, repair technicians of all flavors could do their jobs much quicker if they could just punch in the serial number for a specific device--say a washing machine--and then print the exact part needed to repair it. Of course, the manufacturer would probably extract a licensing fee for the service, but it's better than ordering and waiting forever to get a small plastic do-dad which only takes a couple minutes to print.