This is the 2nd in a semi-regular series I’ve started doing “why things change and why they stay the same”, in which I will focus on a particular topic or theme and explain what technology can a achieve, where’s its going and why certain things will always remain pipe dreams.
There’s been a considerable talk of the emerging technology of 3D printing, with some even going so far as to predict that it will replace conventional manufacturing methods where anything we want, rather than going out and buying it in a shop, you’ll just download it off the internet and print it out. Now while certainly I agree 3D printing has an important role to play, both now and in future, I suspect such talk merely demonstrates how little many know about modern manufacturing processes.Certainly, it has to be said that 3D printing techniques, often referred to within industry as “rapid prototyping” has grown in leaps and bounds in recent years. When I started my career about 15 years ago, a decent rapid prototype system cost hundreds of thousands of dollars, took a good half a day to finish something and was fairly limited in what materials it could support. Now you can pick one up for just a few thousand pounds and it can support a wider selection of materials.
Although at present commercially available ones will only work with various forms of plastic. However, experiments are continuing on rapid prototypers that can utilise metals. Indeed NASA has been experimenting with them for the manufacturing of rocket engine parts.However, if this is the technology that’s going to replace conventional manufacturing techniques it has a long way to go. A typical “run” from a rapid prototype machine can still take hours, whereas a skilled machinists can run off a complex part at a rate of multiple parts an hour. And if you don’t want to pay for the machinist (as is often the case these days), CNC technology means this process can be automated.
And if that doesn’t sound fast enough, there’s a host of other techniques, such as pressing and stamping (often used to make flat sheet metal parts, such as bits of car bodies), extrusion, injection moulding and vacuum forming (often used to make hollow plastic parts) or casting (pretty much any shape you want really!). All of these techniques can manufacture parts at the rate of parts per minute, with hundreds or even thousands a day flying off a single production line. Even the slowest of these methods, casting, while it does take time to set up moulds and cool casting down again, often this is compensated for by batch production (e.g. making ten or twenty parts at a time) and low labour and material costs.As Carl Bass of Wired magazine points out 3D printing represents a sort of reverse “Moore’s law”:
“…So if we want something twice as big, it will cost 8 times as much and take 8 times as long to print. If we want something three times as big, it will cost about 27 times more and takes 27 times longer to print. And so on…”
Furthermore, there is the issue of material costs. 3D printing uses relatively expensive feedstock compared to most of the other techniques I’ve mentioned. Which generally use cheap materials, such as Steel, Aluminum or plastics such as Polyethylene or Acrylic (while 3D printers can also use similar materials, such as HDPE or ABS, this often has to be of a higher level of purity).
This is also important when we consider the issue of reliability. The fact is 3D printing isn’t the most reliable of manufacturing techniques, as this set of epic 3D printing fails emphasises. This is of course important when you consider that you’re often making multiple parts that need to be fitted together accurately. And in a modern factory, you’re making multiple parts by the thousands that need to fit together!Case study – a Kettle
Perhaps we can illustrate these points by considering a few examples of stuff you might want to build with a 3D printer at home.
So let’s start with say, a kettle (first thing I spotted in my kitchen when I went to make a cup of tea a minute ago). A kettle consists of two basic building blocks, the casing of the kettle (purpose, contain water, isolate user from electricity and the boiling hot water) and the electrical element (purpose, heat up the water!).
Making the body of the kettle would be complicated using 3D printing, largely because of the types of plastic used in them (mine appears to be made of polypropylene) isn’t currently compatible with 3D printing and the vessel itself is relatively large and thin walled (not impossible to make with a domestic 3D printer, just difficult).
Furthermore we’re dealing with more than one component, you’ll have the main vessel, a lid, a clear window of some kind (to see water level), perhaps a separate handle and base part and a couple of smaller parts and dowel pins to fit everything together. Manufacturing all of these with a 3D printer is inevitably going to be harder and more time consuming than the injection moulding technique often used today.
Then there’s the electrical side of things. Currently no commercially available 3D printer is able to make this system, as it would need to be able to process material with good electrical conductive properties, which is non-toxic and soluble in water. Plus you’ve got the controller (e.g. the bimetallic switch or pressure transducer that turns the kettle off when the water boils) plus the cable, fuses, plug and all the associated wiring.
Indeed as an aside, its worth considering that practically everything these days from your toaster to the keys of your car has some level of semi-conductor based electronics within them. And that’s not going to be easy to make reliably with a home 3D printer. In short take a quick look around your kitchen or living room and you’ll suddenly realise few are the sorts of thing you want to make DIY style.Then of course there’s the matter of putting all those bits and pieces together. Firstly, costs. Time is money and if it takes me several hours to make the parts for a kettle and then assemble it, I have to balance that with the fact that going down to my local store and buying one off the shelf will take but half an hour and cost say £30 for a decent kettle (that’s what the last one I bought cost), which is probably a lot less than the depreciation price on a 3D printer and the material costs, never mind the cost of my time (and that’s alot more than £30 for half a day!).
Secondly, health and safety. Do we really want people wiring up kettles or other appliances from base parts at home and plugging them into the national grid? How many people will fry themselves (or their kids) if that starts happening? And if someone does get fried who is to blame? Inevitably the lawyers will chase the person who produced the drawings for the kettle online. And anyone who knows anything about product liability (particularly for any domestic appliances sold in the West) will know that they are going to struggle to mount a defense against such a case. This is why most manufacturers these days seal anything dangerous in there products up as tightly as possible, often making products impossible to dismantle without breaking them.
This is one of the reasons why the developers of a 3D printer compatible gun were forced to take their design down, as lawyers pointed out they could be criminally liable if anyone committed a crime with such their design.
Case study two – the liberator….or Jihadiator?
Indeed this 3D printed gun, which sparked a lot of this speculation is an interesting case study in itself. Firstly, as an article in “the register” highlights, it ain’t a particularly effective weapon. Its basically a crude musket that would be of little practical use in a real gun fight. If the US government ever did turn “tyrannical” as many Tea Party types fear I don’t think you’re going to be wanting to take on US marines armed with M-16’s…or indeed helicopter gunships and main battle tanks, with a one of these little plastic muskets.
Indeed I doubt you’d have much luck stopping a burglar, unless he was kind enough to stand still…and isn’t armed (which in most countries means you’ll be done for assault with a deadly weapon…or murder!)….or indeed you convince him to take the gun and shoot at you with it, as recent tests of such weapons by police suggest its as much of a danger to the user than whoever he’s shooting at!And more “advanced” versions of the weapon haven’t performed much better, which is hardly surprising when you realise there are good engineering reasons why gun barrels are made of metal alloy’s. While I don’t know a lot about guns, I suspect an old 1890’s Colt Frontier or Winchester rifle would be a far better choice of weapon.
Indeed just about the only use for such a weapon is someone, such as a criminal or terrorist, who wants to kill or threaten someone but doesn’t want a weapon that can be easily traced or that can be picked up by metal detectors. Leading me to conclude they should have named it the “Jihadiator” or “Muggers friend”. Fortunately, as the key parts of the gun, the bullet (which includes the casing and lets not forget the gun powder) and firing pin, are both made of metal (and can be traced via forensics), this does limit the potential for harm from this weapon….of course a harmless weapon is sort of a contradiction in terms!Case Study three – Frying pan
Still not convinced? let’s look at something even simpler, a frying pan or some other kitchen utensil. These are generally made of either cast iron or stainless steel. In either case some form of “heat treatment” would probably be applied post-manufacture.
One of the reasons why iron and steel are still the top engineering materials (aside from it being cheap!) is the fact that we can easily apply heat treatment. Correct application of heat treatment can yield a material that is either soft and ductile (important if it’s going to have to bend a lot and absorb a lot of punishment, e.g. the suspension on your car for example or a small sharp cutting knife) or you want something that’s hard and strong (such as the head of a hammer or axe, or a heavy meat cleaver) or even a bit of both! And of course stainless steel has the additional advantage of being non-toxic, corrosion resistant and easy to clean (although care must be taken when heat treating stainless steel).Also its worth mentioning surface finishes. In the case of a frying pan you’ll probably want a nice smooth surface inside (achieved by grinding), to which you then apply the all-important non-stick coating. While on the bottom you want a good flat surface (to ensure good efficient heat transfer) but not too smooth, else the pan is liable to easily slide off the hob potentially burning somebody. Again, this is one of the advantages of making stuff out of steel, it’s easy to machine after forming.
Making such components out of any available 3D printing technique is impossible right now, largely due to the fact they cannot output suitable materials and its unlikely to be possible in the future, and I can’t see your average householder performing heat treatment using a stove and sink (again health and safety!). And again, even a knife or frying pan needs some assembly which takes time and building a machine to do all that (and assemble hundreds of them an hour) is going to be a lot more cost effective and safer!
Environmental issues – Rapid prototyping or Rapid garbage generation?
And of course we haven’t considered the environmental issues. Plastic based waste, as I’ve discussed before, is a very potent environmental issue and the last thing we need to start doing is create a machine on every desk that spits out more cheap plastic “crap”.
I’ve done a trawl through the literature online (sciencedirect, etc.) and I can’t seem to find a good comprehensive analysis of the Life Cycle energy and environmental impacts of 3D printing. However what I have been able to identify is far from encouraging. It is suggested that a 3D printer can consume 50 to 100 times the energy as an injection moulding machine to produce the same part as well as generating potentially toxic fumes and wastes in higher quantities.Other studies tell a similar tale, much higher rates of energy use by 3D printers and more toxic materials used. The supporters of 3D printing often counter by trying to argue it is better for the environment as it produces less waste and considering the energy costs of the supply and distribution network (while cheap plastic stuff made in China and sold in an air conditioned Wal-Mart has a large carbon footprint here, a wood or metal component made by a company down the road from me and ordered online doesn’t).
However this often involves a flawed methodology of ignoring the energy costs of the supply chain for 3D printer (that high purity feed stock comes from somewhere!), assuming little to no waste material due to mishaps (and as I’ve shown 3D printers make more mistakes than conventional manufacturing), no trial and error experimentation by people printing stuff out unnecessarily (manufacturing tends to only produce the finished product in volume after the design has been finalised and tested, amateurs inevitably won’t be that efficient) and that 3D printing can manufacture complex devices (such as those I’ve demonstrated above that it would be impossible or impractical to make with 3D printing) and ignore economies of scale (i.e. the LCA of a manufacturing system in constant use is always going to be superior to one used occasionally).
Also such studies often ignore the fact that much of the “waste” from milling or moulding processes is often collected and either recycled or reused (either for sound financial reasons or enforced by legislation). By contrast lots of decentralised 3D printers would produce waste that was harder to reuse/recycle and there is thus a higher probability of it ending up in landfills.
And again one of the many reasons for engineers preferring metals such as aluminium and steel or plastics such as polyethylene is that they are relatively easy to recycle. Although polyethylene, like many plastics, is a little more tricky to recycle than metals, this can often be got around by reusing it for lower grade products (i.e. not 3D printing).There is also research ongoing into replacing plastics with natural materials such as hemp or starch, readily compatible with existing manufacturing methods, but not so easy with 3D printing.
The future of 3D printing? Much like the present!
In short it is difficult to avoid the conclusion that existing manufacturing techniques will remain dominant in for some time to come. Indeed none of the examples I used above represented particularly challenging parts. Inevitably we start trying to make things such as gas turbine blades or the engine block of a car, or indeed anything with electronics in it, and suddenly 3D printing is struggling.
As things currently stand, existing techniques will dominate for good sound engineering reasons. Not least of those being the complex material and surface finish requirements such components demand, which often require the use of specific alloy’s, heat treatments and machining methods.
So what role will there be for 3D printing?
Well I suspect an expansion on what they are currently used for! Often 3D printing is used in support of design (by allowing one to print out a 3D mockup of what a part should look like) or in support of manufacturing. Ironically, one of the things rapid prototyping is most used for in industry is to make the moulds that are then used to cast parts for mass production. Again this is a role that can only expand further as technology improves.
I can also see a sort of hobbyist industry building where people print out lots of odd shapes, jewellery or artworks for fun or home use, much like some use existing hobbyist scale CNC machines (notably millers) today. There’s also a host of small household items that are often charged for in shops at an exorbitant rate (e.g. accessories for your I-Phone, etc.) simply because the manufacturers know that they can charge than and you’ll pay. So there could be a market for such machines to fill in this niche.
Specialists parts, such as the NASA engine parts I mentioned, might well begin to be manufactured using this technique, particularly if they can be developed to produce metal parts at a reasonable costs. But recall we are talking about parts made in extremely small batches v’s modern industrial manufacturing which is often trying to hammer out hundreds or thousands of parts a day, as cheaply as possible and of a consistent size and quality.
In short, I see 3D printers doing pretty much what they are doing now, just a more expanded role, in a wider variety of industries and in larger volumes and possibly with the use of metals in future. I see them more supporting existing manufacturing, rather than replacing it.
All watched over by machines of love and grace
Of course part of the problem here is that many of those driving this speculation about 3D printing tend to be computer science geeks and others technocrats and quite a few libertarians. Both groups have ideological reasons for wanting to see a 3D printing future, as it would bypass existing the manufacturing industry with its standing army of tens of millions of skilled workers and all the costs and excess baggage that comes with these industries. Notably the need for either governments or large corporations to support and finance such operations.
However, I would argue that they are, much like those who believe nano-technology will also replace conventional manufacturing (good critique of that here), perhaps letting their ideology cloud their understanding of engineering.