That's a misleading title if there ever was one, the output from the 3D printer is used to make molds, not car parts.
It's a tool in the chain, not a one-step scan-print have a usable car part process.
They could have sent out the original to make a mold of or they could have made a precise casting of it and then reversed that for a positive.
This looks like a great excuse to use a cool tool. Any competent machinist could turn out that nut in 1/2 the time it would take to scan/print it. And if you have access to a CNC rig (and most machine shops have 4 or 5 axis workcenters) you could make as many of them as you wanted once you had the shape programmed.
For a one-off a mold is a crazy process anyway, the only reason molds are used is if it is impossible to machine something (rarely) or to save cost in higher volumes (the most common case).
The reason "People say, “Why not just give the part to your machinist to make?” Well, if the machinist makes it wrong, you still have to pay for it." is absolute nuts, if you give a machinist (a competent one) a part to duplicate you'll be hard pressed to tell it from the original if you haven't marked it in some way.
My EcoJet supercar needed air-conditioning ducts. We used plastic parts we designed, right out of the 3D copier. We didn’t have to make these scoops out of aluminum, plastic is what they use in a real car.
Yes the title doesn't mention that molds are made as well but I would say the title is correct.
If you make ducts by milling you're doing something terribly wrong or you are in the aerospace business.
Leno has this thing about bashing 'machinists', I've heard him rant about that before.
At the end of the day a machinist can only make what you spec, if you spec it wrong you end up paying and you'll have to do it again. Measure twice, cut once.
Aerospace CNC work is amazing stuff, it's not rare to see a block of aluminum be reduced to less than 5% of its starting weight as a finished part, the rest is scrap.
Yes, absolutely they do. But it's not as straightforward as it should be because the aluminum swarf is contaminated with cutting oils and other nasty materials.
All the scrap gets recycled, a typical price for a pound of 'dirty' scrap is $0.09. Contrast that with $.70 for a pound for stock (and more expensive if it is a special grade or shape).
RTFA- He gave an example of plastic AC fittings made directly from his printer. He also uses an automated metal milling machine to make parts directly from 3D files.
Printing is an additive process, milling is a subtractive process.
If the 'printer' would be used to directly make a functional car part (I don't care if it is made of some high grade plastic, as long as it is used directly in a load bearing situation) that would be revolutionary, this is just a rich guy playing with his toys bashing a profession that he hasn't a clue about.
Machinists the world over would cringe if someone said 'if they make the wrong part you still have to pay for it', that should read 'if you spec the wrong part you end up paying for it', there is a huge difference there.
He mentions in the article that some of the 3D Printers have the ability to make metal parts using thin steel wires.
Now, of course, Jay Leno has buckets of money to throw around to hire professionals and get professional level equipment, but he's got a point that this stuff is coming down in price all the time and that it is disruptive.
No disrespect to machinists, but getting all the measurements right is non-trivial. Letting people take 3D models, either scanned or CAD'ed and then use sub-millimeter accurate automated tools to create the parts is huge.
Yes, those metal parts will then have to be sintered. This works but leaves a fairly brittle piece, nothing you'd want to entrust your life to.
The parts that are being replaced were engineered to a certain spec, you can't just come along with your 3D printer and 'print' a copy of the original shape in a different material than the original and expect it to work long or safe.
I agree that getting all the measurements is non-trivial, but the way to use that data is to then send it to a milling machine to recreate an exact copy (not just shape, also material grade).
Sub millimeter accuracy will not do it for real parts, you'd have to go down to about 0.01 mm for anything beyond the trivial.
Machinists and their gear routinely achieve such accuracies in an enormous variety of materials.
Rapid prototyping, fine. Checking shape for fit, also fine. But to suggest that you can 'print' working parts with the same characteristics as machined parts from a 3D printer after scanning a machined original right now is simply not true, and to suggest that such a one-off is cost effective is really nonsense.
But it quite possibly will be true at some point in the future.
Sintered parts are everywhere, especially gears. In your car. Perhaps not in the tranny but window motors, wipers etc. They are used when plastics are not strong or temp resistant enough.
Emachineshop is a giant step in the right direction and they have been around for a long time, so they get points for being early, but they are only a part of the picture.
Check out http://www.cnczone.com thousands of people making Automation and sharing manufacturing knowledge.
A couple of comments. First, I wonder how the copyright will shape up for this kind of things. When they become ubiquitous, will a Mercedes still be a Mercedes part if it's downloaded? And will you have to pay a licencing fee to use it? Will we see a ground-up reimplemented Open-Mercedes?
Before you think this is science fiction, think about the current market in spare parts. Far too often you have to pay 100+ for a piece of plastic or metal of a certain shape. The moment the price for such parts gets under 100-200, you will actually be able to re-build your car radiator in the service cheaper then ordering it from the producer. Customized and upgraded too, if you want it.
Which will cut into a huge source of income for producers - and open up one for enterprising repair shops. Both industries who care about money and don't shy to show it. It will be interesting.
Yeah, I was wondering the exact same thing about copyright. Given the current state of copyright law, it seems like it might be of dubious legality to take a carefully engineered and manufactured part and duplicate it at the press of a button. In much the same way as the advent of digital distribution of music (and video, books, etc.) made the music industry crack down on file sharing, are we going to see producers of engineered products attempt to shut down technology such as this when it becomes more widely available and cheap?
Not even for premium parts, but just for everyday items. Coat hangers, hooks, cups, stands, all those random little plastic connectors that are sold at huge margins. Heck, even plastic army men, or miniature figures for various board games. It will be really interesting how this plays out...
It would be awesome to be super rich like that and be able to say "oh yeah, I make car parts directly from my 3D printer as a hobby" when you really mean "I tell my full-time 3D modelling team that works on my cars what parts I want made as a hobby."
(At least that's what I assume is the truth- I didn't get the impression his hands actually ever touch a computer, cleaning up errant polygons from a 3D scan :-)
My dad has been an antique (primarily 1920's and earlier, authentic) car geek for a long time, and he and my mom bumped into Jay Leno at a car meet years ago.
While there, he was just one of the guys, just there for the cars, and he definitely knows his stuff.
I think of him as a car guy who got a job as a late-night host so that he could pay for his true love.
Looks like you could buy a basic setup for under $20,000. certainly not cheap but about the price of an average mid-size car. No doubt within reach of anyone that wanted one bad enough.
Once trickle down makes these more cost-effective and once they can create things in metals as well as plastics, we are really going to start seeing some amazing changes in the way we operate as a society.
IKEA will have to adapt or be demolished. They could actually leverage this kind of tech.
I can picture all sorts of flat-pack furniture made by specialized machines to automatically rout and drill holes, with specialized parts made by 3D printers.
The whole back catalog would always be available. Designs are already modularized and extendable. I could also picture DIY one of a kind "custom closet" kits designed on a computer, with the parts bagged for you, and custom directions printed just for you.
3D printing vs injection moulding are 10000:1 cost wise right now, in the future this will come down but I doubt there will ever be parity.
The reason companies like IKEA standardize connectors, handles and all kinds of other bits and pieces is to maximize the use of these hard to design parts. Mold making is an art and molds are expensive, even for relatively simple parts. But once you have a mold the parts become absolutely dirt cheap.
Specialized woodworking machines exist, such as CNC rigs to create roof trusses straight from a customized CAD package.
One other nasty little gotcha about 'printed' 3d objects is that they are really 2 1/2 D printed objects, in other words the design is deposited layer by layer, this usually means that in at least one direct (the deposition direction) the material is not very strong.
Slightly off the topic of the original article, but imagine an Ikea store with a "print behind the counter" making semi-custom furniture by CNC milling/routing the MDF/OSB components and using existing tech to make the camlocks, screws, pins, etc.
That doesn't require any new tech at all, just for the capital equipment to become cheap enough (or the utilization high enough) to justify doing the manufacture at the store instead of at a centralized plant.
It wouldn't need that much $ at all, a fair sized CNC rig that could do this would run you about $20,000 including software and tooling. I built my own once (for plasmacutting sheetmetal) for about 2/3 of that, and that's already quite a while ago, I'd expect it to be cheaper today.
This is actually a really good idea you've got there, you should do something with that!
Just select your furniture from a computer and the automated ikea design system will automagically build your furniture infront of you. You could even make slight modifications on the spot.
Thanks for that link, that's a lot closer to the real thing than the machine the article talks about. Fully dense metal and a reasonable range of materials (including tool steel, which is a must).
I wonder what the cost per part would be compared to 'traditional' methods (abrasive / subtractive), if they manage to get that down far enough it will be quit the miracle. Metal powder + energy in -> parts out.
If you make the machines mobile you could have them extrude your steel beam construction in one go :)
I could see these having an impact in places like Africa and South Asia, as mobile phones are doing. It's depressing to hear about wells being dug but then going out of service because there's a part missing.
sure, it would be a lot cheaper to just ship bulk parts around than set up 3d printers in poor towns, but the rub is that you don't know what metal/plastic things people will need in advance.
I think that's focusing in the wrong direction. The costs involved set a high-enough barrier for the first world, let alone impoverished countries.
The focus should be on simplifying the technology these communities need and making it more robust, as well as showing people in the community how to repair and maintain it.
Not to mention that this technology would represent another point of failure -- sure, you could print out parts for the well, but what happens when the roof over the printer fails and everything gets soaked by a monsoon.
Thing is, the 'printers' are expensive but getting much cheaper, and the materials are very cheap. It's not a panacea by any means, but a versatile low-volume manufacturing facility might have greater value than a specialized high volume one in places with mediocre industrial and commercial infrastructure.
I see your point 100%, but these are basically the same arguments people made about cellphones and they turned out to be better option than trying to deploy the simple and robust technology of landlines. I suggest ad-hoc communication and manufacturing technology may be the best thing in an ad-hoc economy.
>A versatile low-volume manufacturing facility might have greater value than a specialized high volume one in places with mediocre industrial and commercial infrastructure.
I would imagine you are correct. However, conventional manufacturing practices are not solely high volume operations, and can be implemented with fewer resources.
>[T]hese are basically the same arguments people made about
cellphones and they turned out to be better option than trying to deploy the simple and robust technology of landlines.
Cellphones don't cost $15k per village, nor do they require training to use special computer programs, or a constant supply of raw materials.
If they did, you can bet that they wouldn't turn out to be a better option ;)
>I suggest ad-hoc communication and manufacturing technology may be the best thing in an ad-hoc economy.
I agree, provided the technology becomes sufficiently inexpensive and robust.
But for today, use that cash to leave the community with an extra pallet of parts, instruction on how to maintain the well, and the rest to educate the people.
Ultimately, I share your optimistic view. The technology will be there someday. We just can't force it in the meantime.
> Cellphones don't cost $15k per village, nor do they require training to use special computer programs, or a constant supply of raw materials.
Really? How much does a cell tower, a reliable power supply for same, etc cost? For widely spaced villages, I'd certainly think this cost would be well over $15K per village. Hell, I can't put a 50K sq ft office building size, cell phone repeater system in for $15K.
(I'd expect widely spaced villages would be the common case in areas where cell deployment is cheaper than copper lines.)
The cost and simplicity argument is a valid one. On the other hand the reality is already a different one: People in Africa use mobile phones much more than we in the West. Even banking is managed via cell phones. So 3d printers might allow the to leap frog the industrial society stage. It's the same with alternative energy: They don't need nuclear power plants down there, solar and wind is often enough.
I know little about materials science, but I know that metal grades vary significantly in performance (and cost). Instead of saying, "Any part can be made via a 3d printer," shouldn't it be said that, "Something in the same shape and size of any part can be made using such a printer."? It's not like we can print out a new camshaft and drive the car for 5K miles using it, right?
As other people have said, he's mostly using this printer to print molds of the parts. These molds are then used cast the parts out of whatever grade of metal is up to the task.
Having once cast an aluminum cornbread pan in high school shop class, I can understand how using such a printer for mold making could be extremely useful. And, I can see how it could have niche uses in various industries. But, the people saying that the 3d printer will change the world as we know it (or whatever other grand predictions) haven't explained to me how it will be anything more than a great proto-typing/mold-making tool for casting/machining operations or customers of those businesses. Am I missing something?
That all depends on the speed of the machine, the price of the 'input' material and the versatility with which you can apply it (and the cost of the energy required to run it).
If it is done 'right' (price point at or below casting, fast to the point where making a complex part would take a couple of minutes at the most and create any solid shape, even multiple parts in one go assembled and ready to go) then it would change the world as we know it in ways that I can only describe as science fiction.
Drive the price down even further (to say below what it costs to mass produce castings) and you're looking at something where even the imagination will fail.
Make it precise enough that you can create nano machinery and you get yet another level of technology that gets unlocked. It's hard to predict what any of that would be like, but I can see all kinds of good and bad stuff happening.
Suddenly the blueprint is the machine, if you can think of it you can make it.
Right now the barrier to entry to manufacturing is comparable to lets say big iron computing in the 70's.
Think of this thing as the PC of manufacturing, suddenly everybody can quickly and easily make just about anything.
Mold making is an art and terribly expensive, so for now prototyping is handwork or at best one-off CNC runs. To replace that by a process that would cost a very small fraction would upset quite a few applecarts.
If it stays slow and expensive then it will be a niche technology, but the speed at which the prices are coming down and process speed goes up is very impressive.
You can't cast a working crankshaft either; the surfaces won't be smooth enough. You can cast a "crankshaft" and then machine the surfaces down to the micron tolerances you need.
I can see how "soluble supports" enable the moving parts to be printed - but how can a scanner see which parts are connected/unconnected, when viewing from one direction? How could a scanner see that an axle is a freely moving rod snug within a cylinder, when that is hidden? An extreme eg: put the part inside a blackbox, and scan the box.
This looks impossible to me (that you'd need to touch up the model manually), but the article said that's just what "old school" guys think...
Amazing technology when you think about it. Sure it's clunky now, and material is limited and expensive, but it wont be long before we can print copies of the machine itself.
I collect old Transformer toys and already I have begun to see some 3rd parties sell off "unpainted" transformers. I would not be surprised if in 10 years it would be very common for their to be printed addons and or all new toys.
There would be a huge market for things like this. My father collects multiple copies of war toys (things ranging from Micro Machines all the way up to planes with three feet wingspans). He's said that good ones are hard to come by. With a tool like this, he'd probably churn out several copies and spend considerable time detailing each one.
This is good stuff, but the important part is just briefly mentioned: It's the Fadal CNC machine. The other machines scan and build plastic prototypes or occasionally functional parts if plastic is sufficiently tough for the job.
But once the plastic prototypes are tested and proven, the Fadal CNC machine then makes them out of metal. I would still like to know what kind of metals the Fadal can machine?
And then there's the bet that this technology would continue to get cheaper. This is certainly true of semiconductors and software but will it be true for this, for CNC machines?
And lastly, manufacturing "works" on a large scale. You buy raw materials in bulk, manufacture and then sell the now transformed materials in small units. Buy in bulk, sell in small is where the profit is made.
But this is very long tail stuff. No one will need a large volume of a special custom widget.
Would individuals buy this tech to use over a life time, like a good expensive tool box? Or would a small business buy this and then offer to ship custom parts in a small area where shipping is cheap?
I dream of the day where I could potentially have my own audio hardware startup for under a couple grand. Being able to print custom parts would make it so much more viable to make boutique audio equipment, instruments, etc.
Are you familiar with the work being done at Stanford on music? I saw a special on TV (I think it was on Nova Science Now) about the group. They are doing some interesting things in terms of making new instruments out of things they find sitting around. They also have their own "computer" symphony. Here's a link: http://ccrma.stanford.edu/overview/Overview.html
Slightly off-topic, but, was I the only person slightly disturbed about the rant about the 3D printers being American made machines?
Maybe it's because I'm British (living in SF) but I still find that kind of patriotism smacks more of jingoism. I'm all for working with local craftsmen and supporting local business, I think that's great, but I still don't get how this kind of transparent nationalism is seen as a good thing in the US.
I would much rather go with the machinist I've met quite a few veteran machinists and they are very good at what they do, especially the ones who have learned to use cnc in conjunction with traditional work.
Leno however probably has enough money that he employs multiple people who know very much about engines and cars that they are the ones worrying about the molds.
Zach Smith, one of the founders/makers is in both projects. Also, we (I'm RepRap developer) collaborate on designs to print and reuse each others technologies and ideas. Because both designs are completely open source, the can reuse whatever is practical and also differentiate themselves to be used in different 'market segments'. RepRap is meant to be more adjustable and modular, and meant to be able to replicate most of its own parts. The Makerbot doesn't have this requirement and can mostly be made on a laser cutter. Although several parts of the Makerbots are printed on Makerbots and/or RepRaps.
Yea, there's actually a pretty wide array of scanners. The Konica is like $80k with high speed and quality. The 3D3Solutions.com machine is low quality but fast and pretty cheap ~$5k. The 3Rivers3D.com machine is $15k but can scan fast at very good quality. The NEXT engine is good for scanning a few things at okay quality, but there are less hyped machines that have really impressive results for high value.
It's a tool in the chain, not a one-step scan-print have a usable car part process.
They could have sent out the original to make a mold of or they could have made a precise casting of it and then reversed that for a positive.
This looks like a great excuse to use a cool tool. Any competent machinist could turn out that nut in 1/2 the time it would take to scan/print it. And if you have access to a CNC rig (and most machine shops have 4 or 5 axis workcenters) you could make as many of them as you wanted once you had the shape programmed.
For a one-off a mold is a crazy process anyway, the only reason molds are used is if it is impossible to machine something (rarely) or to save cost in higher volumes (the most common case).
The reason "People say, “Why not just give the part to your machinist to make?” Well, if the machinist makes it wrong, you still have to pay for it." is absolute nuts, if you give a machinist (a competent one) a part to duplicate you'll be hard pressed to tell it from the original if you haven't marked it in some way.