3-D Printing with Molten Steel (Video)
Steve Delaire is making a 3-D printer that uses steel instead of plastic. Specifically, he's using TIG welding to build up layers of steel, just as most 3-D printers build up layers of plastic. He says he's "still working it out," but eventually hopes to use 3-D welding to make larger than life art pieces that are strong enough to be placed safely in public areas such as parks, where children are likely to climb on them. Steve's blog is called Molten3D, and it's a diary of his work, including the problems he encounters and how he overcomes them. He's not the only one doing metal 3-D printing; a Texas company has even made a printed metal gun. So there's plenty of people working in the field of what we really should call "additive manufacturing" instead of "3-D printing." But whatever you call it, every year we see this kind of process being used to make stronger and more complicated shapes, using an ever-increased variety of materials in ways that have been developed since this seminal paper, Liquid Metal Jetting for Printing Metal Parts, was written in 1997. (Alternate Video Link)
Some Meta
The good:
This isn't some naval gazing blog post by someone whos opinion we don't care about babbling about how 3D printing will change everything. In fact, I didn't see anything about future implications of this technology. This is just a blog by a guy building something really cool in his garage, and I like this.
Maybe it won't pan out. Maybe it will be impractical. Maybe he'll hit some insurmountable wall. It doesn't matter, he's actually doing something! He has actually got a physical thing in his garage that he's tinkering with, and that's cool in my books.
The bad:
Bold comic-sans esq font. I'm not usually a font snob, but this is really hard to read. I actually copy+pasted the contents into a text editor to read it.
Good luck.
I was a welder for quite a while. Molten steel behaves more like water than plastic. Steel "beads up" as you heat it. Being precise with welding is very hard. Plastic behaves more like frosting when molten so it's easier to use in a printer scenario.
I think his big mistake is taking the ID of 3D printing and applying it to a martial for which it wasn't intended. I'd think he'd have better luck if he instead looked at Powdered metals: http://en.wikipedia.org/wiki/P...
In industry, they design a part, make a mold for it, press a mixture of powdered metals into the mold and then sinter it (basically the same as firing ceramics)
ooo... hey look, I tried looking it up and that's exactly how they 3D print metal:
http://en.wikipedia.org/wiki/S...
anyways, yea, that's the direction he should go. I'd use electro static charge to hold the shape and then use something like an Xray laser to melt/fuse/sinter it.
Has anyone seriously suggested that 3D printing will lower the cost of mass production? The cost of prototyping and small runs maybe, but for anything where even a few tens of thousands of units are produced, modern manufacturing techniques promise to be cheaper while delivering a superior product for the foreseeable future.
--- Most topics have many sides worth arguing, allow me to take one opposite you.
Well, besides confused microcontrollers? From his blog, he's using a 3D plastic printer to prototype the parts for this printer. Cool. Some test pieces in the blog photos, but let's see the sparkenmaken!
"Win treats sysadmins better than users. Mac treats users better than sysadmins. Linux treats everyone like sysadmins."
You can already run your programs on a mainframe - just as fast as this microprocessor BS would be, if not more so.
I saw a demo, around 2003, of a sintering machine the military used to build prepare parts in the field. Rather than shipping a part they could produce and machine it as needed remotely; all they needed was the appropriate instruction set and they were good to go. When I asked abut the strength and durability of the parts they said it was as good or better than normal spares.
I'm a consultant - I convert gibberish into cash-flow.
Well take the bait out of your mouth and abate your breath, but please not for long. It's not that mass-produced things will be cheaper by the piece to the consumer - it's hard to compete with lost-wax, styro molding, sand casting, injection molding, etc and the cheap labor to do them. But if you need that first one made, you can get access to 100 micron precision on shoebox-sized builds for the cost of a week or two of a decent salary. Which can save you time and money compared subbing it out. And since time is money...
"Win treats sysadmins better than users. Mac treats users better than sysadmins. Linux treats everyone like sysadmins."
Interesting article, however, I suspect the editors are a bit mistaken. I strongly suspect that Mr Delaire is NOT using TIG welding in his machine, but instead is using MIG welding. Also I have to wonder if Mr Delaire is aware of http://hardware.slashdot.org/s...
If not, he may be able to save a bit of effort and time by building upon the work someone else has already done.
Even with the "2.5D" limitations of a three-axes CNC machine, I still prefer subtractive manufacturing.
Get free satoshi (Bitcoin) and Dogecoins
https://www.youtube.com/watch?...
Sig?
This guy wanted to be able to weld cool stuff without having to sit there holding a welder for hours at a time, plus he wants to have the welds be of uniform quality. What makes this a "3D printer" is that it uses an X-Y drive rather than an arm with elbow/wrist joints like most robotic welding machines.
dom
But that's the thing. If the environment isn't also controlled, the weld will not be of uniform quality. Welding is an art. You have to pay close attention to ambient conditions, the material being used and the settings on your welder. You need more or less Gas pressure, to adjust the gas mix, higher or lower welding speed, and need to change the voltage. You can also reverse the polarity of the weld depending on if you want the puddle to push or pull based on the direction you're welding. When I welded for a living, storms coming would mean I had to re-setup my machine if I was on a big job. I don't know if it was the barometric pressure, temperature or humidity that did it, but it was definitely something we had to deal with. As soon as I'd lift my helmet to make the change (back then there were no auto-dimming helmets) I'd see everyone else in the shop doing the same thing. Moving to someone elses welder meant re-setting it up to, so personal style must have had an effect as well.
On assembly lines they do just that... they control the environment. It's at a relatively constant temperature, humidity and the welding arm is doing the same exact weld over and over again so they operator can program exactly what it's supposed to do. But it's by no means as simple as feeding in a cad file and pressing "weld" You can do that for sure... crappy welds are good for several metric tonnes of force and can hold on a bumper as well as a good weld... but you're not going to get consistent quality or a "Pretty weld"
Wonder if it works like this one, http://www.mtu.edu/news/storie...
Passionately Indifferent
Does this mean I can finally download a car?
Why would there be any size limitations to laser sintering? I don't see any reason why it can't be scaled to any size required. SpaceX is building rocket engines using the process, for example. The rocket engines in question aren't exactly huge, but they still put out more than 16 thousand pounds of thrust.
Yes, there are people who really believe we're at the dawn of Star Trek because a hot glue gun on a stepper motor made a Yoda coffee cup.
Cool, sounds like a job for sensors and algorithms.
"Anything you can do I can do better..." sung the old computer...
http://www.masturbateforpeace.com/
When outlawing ownership is too difficult, oppressive lawmakers limit availability of certain items — like guns or drug paraphernalia or alcohol-making equipment — by banning their sales or transport (withing or across State-lines).
If the end-users are suddenly able to make them themselves, some of these laws would not apply making us the people a little bit more free.
In Soviet Washington the swamp drains you.
SpaceX is using laser melting, also called selective laser melting, explained: https://www.additively.com/en/...
Laser sintering is commonly used the term when plastic parts are produced: https://www.additively.com/en/...
Video about steel-welding-3D-printer without actually showing it in action. *facepalm*
Assorted stuff I do sometimes: Lemuria.org
Why would there be any size limitations to laser sintering?
Because the bed holding the part and the powder that's going to be sintered needs to be moved with very good precision. It also needs to be at least as deep as the part you're making. It's generally very difficult to get large things to move small distances.
If God forks the Universe every time you roll a die, he'd better have a damned good memory.
Laser sintering is commonly used the term when plastic parts are produced: https://www.additively.com/en/...
I'm not sure what you're saying here. If you mean that the word "sintering" is used mainly when plastics are involved then I'm afraid you're mistaken. It's used when referring to sintering as opposed to melting; melting and sintering are quite different and produce very different finished parts. The question of whether plastics or metals are normally sintered or melted is unrelated.
If God forks the Universe every time you roll a die, he'd better have a damned good memory.
Welding identical parts with predictable properties in the same places is one thing, and robots are ideal for it*. So long as there's a need for custom fabrication and patching damaged parts there will be work for human welders.
*Probably why many earlier robots were used for just that, although spot-welding rather than arc.
If God forks the Universe every time you roll a die, he'd better have a damned good memory.
I really wish they'd stuck with calling it "rapid prototyping".
If God forks the Universe every time you roll a die, he'd better have a damned good memory.
The company that makes the 3D printers SpaceX use call it Direct Metal Laser Sintering. Not selective laser melting.
http://www.eos.info/04ea46b971...
But that's the thing. If the environment isn't also controlled, the weld will not be of uniform quality.
Depends - I recall welding kits a couple decades ago that used preheat torches and submerged the arc under a pile of powdered flux, which made the environment pretty uniform and automated (at least good enough for x-ray inspection - the parts being welded were steel bridge-beams). The welding kit was a 1/4-ton monster that ran along sections of track, which in turn attached to the steel by way of strong-assed magnets. The operator only had to set the machine up for proper amperage, make sure the flux hopper stayed full, that there was always enough wire on the spools, that the preheat torches had enough gas, knock off the semi-hardened flux once it cooled down, and stop the thing before it reached the end of the track. Mind you, the bead was about an inch thick and two inches wide, but it was incredibly uniform, and you could do multiple passes over the same seam without any grinding in the interim.
The only real hard part was setting up the right amperage and prepping/grinding the seam surfaces properly before a run.
Quo usque tandem abutere, Nimbus, patientia nostra?
Everyone makes awful, sputtered welds on their first few tries. Being able to lay down a good bead is not a skill that comes easily, not least of all when you're trying to build a machine that can do it from scratch.
If God forks the Universe every time you roll a die, he'd better have a damned good memory.
There was a project going on, first at Carnegie Mellon University and then at Stanford's Rapid Prototyping Laboratory, where a very similar but more sophisticated process (Shape Deposition Manufacturing) was investigated to make precise metal parts with full strength (unlike the sintering methods mentioned here and elsewhere). A number of methods were used or tried to melt the metal, including TIG welding, laser fusing and induction heating.
The biggest hurdle to success was the huge internal stress that built up in the process. Remember that one puts layer after layer of molten metal on top of the previous one. The new metal layer solidifies and shrinks, creating lots of compressive on the previous layer. Put down enough of these layers and the part will crack.
And no, doing this process with Invar won't help, because Invar doesn't have this beautiful near zero thermal expansion close to its melting point.
Lets hope the blogger reads what has been published about this process before he commits more effort and resources to his project ...
Some really intricate parts are cheaper to 3D-print than to try and cast/forge/machine traditionally. Think hollow structures with stiffening ribs or cooling channels inside, or other similarly complicated shapes. The SuperDraco rocket engine falls into this category (though of course also is a small production run).
Seriously, it's like we we've been having a conversation about home-built plastic ultralight aircraft, and then somebody says, "Hey, there's this company out in Seattle that makes aircraft out of metal."
Doing it with molten metal is a new one on me, but people have been laser sintering metal powder for thirty years now, and I bet most of the dollar volume of 3D printers shipped today are of this type. They've been coming down in price too. I have a friend who's a research machinist who has one in his lab, and he tells me that the strength gap between cast metal parts and laser sintered parts has closed significantly over the years.
Post may contain irony: discontinue use if experiencing mood swings, nausea or elevated blood pressure.
If you think you can't train a computer vision system to do it... Sure, maybe it's 20 years before it's cost effective(although for underwater welding...), but it's coming.
If you can boil it down to an algorithm, however complicated, you can get a computer to do it.
Eventually the computer does it better, because it has more sensors than you, thinks faster, has finer muscle movements, and can execute more complex algorithms than you can.
We're about to see this with driving.
http://www.masturbateforpeace.com/
"Why would there be any size limitations to laser sintering? "
Exactly. And the 'small' ones to print guns, cost less than 2D Laser Printers 30 years ago, dollar for dollar, even without considering inflation.
Except the metal is not actually melted, it's sintered.
If the metal is melted it's either going to bead up or flow across the surface it's sitting on due to surface tension.
Well, if you visit any of the big 3d-printing websites out there, you can already print & order your 3D model in a variety of metals including stainless steel, gold, silver, aluminum etc. 3d printing is not only in plastic anymore. They shoot lasers at metal powder to melt it (much like a laser printer).
It works fine without Flash if your machine pretends to be an iPad.
If it is large enough, why not move the laser instead?
You may even be able to do this by reflection without moving more than a very small surface which reflects the correct wavelength. This does not seem like an problem which cannot be solved.
My first thought as to what would stop it is the amount of energy required, as lasers are not very efficient, and melting (or sintering) large objects may take enough energy that the losses involved in transferring energy through a laser may make it inefficient compared to direct heating.
I'm a signature virus. Please copy me to your signature so I can replicate.
We should start worrying about 3D printed guns when it becomes easier/cheaper to print a gun than it does to build one out of scrap in a basic machine shop. That day is a ways off yet.
Welding is an art.
True, but that's because humans do it by hand, and are not necessarily aware of all the changing variables, let alone be in a position to do anything about them. In theory a machine that does have the appropriate sensors would always do a better job as a result.
So, there is a six-minute video of a guy talking about 3D printing. Am I out of line for expecting a video that shows the 3D printing? I'm not even interested in the first word the guy spoke much less six minutes of him talking. If you promise me "3D Printing with Molten Steel (Video)" is it unreasonable to assume that the Video is of 3D Printing with Molten Steel?
If it is large enough, why not move the laser instead?
In an SLM machine you have a cavity that is filled with a fine metal powder, the laser selectively melts/sinters parts of that, the base of the cavity moves down slightly and more powder is added to the top. Rinse and repeat until you have a solid part in the cavity surrounded by the powder that wasn't touched by the laser. I'm not going to cop-out and give a link to wikipedia, but if you go there and only look at the block diagram you'll see why the bed needs to be moved with precision at least equal to the minimum feature size that you're aiming for.
What this project aims to achieve is akin to 3D printed plastics where the raw material is deposited right where it needs to be instead of selectively converting parts of a much larger amount of feedstock. I suppose you could move the laser if you want, but it on the face of it I think it would be simpler to just use a few mirrors attached to servos to direct the beam where it needs to go.
If God forks the Universe every time you roll a die, he'd better have a damned good memory.