3-D Printing with Molten Steel (Video)

Video no longer available.

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)

Tim: Steve, here at Maker Faire you have on display a machine that’s a little bit unlike most of the 3D printing equipment that I’ve seen here—because it’s a welding machine. Talk about the reasons for that.

Steve: Okay. Well, what I’m trying to do with this is copycat a typical plastic printer and use a welder instead and metal materials. So it is a TIG welding process, which is very common, and I am feeding cold wire feeding material into the weld puddle. To create a metal object instead of plastic. The end goal of what I am trying to do is to make public sized art with welded steel. As it turns out most public art has to be shown to be structurally sound. And so by using the welded process I’ll be able to show the engineering behind the product and it will be approved by the cities to be out in the public. A very important part, the key, in the process is being approved. So that’s the end goal. This is the bench test model—I am still working it out. The challenge is using open source slicing software in my application—it doesn’t quite exactly blend over perfectly. So I have to learn how to change the software for my needs. The welding equipment confuses the motion drive. So there are some lessons to learn there on how to solve that problem. But by and large it looks like it’s going to work. We’re getting some welding samples that are starting to look promising.

Tim: Can you show us the machine itself?

Steve: Sure. So the machine itself really copycats what’s called the gantry style printing machine. I’ve done it in steel because of the heat requirements. Being in a welded environment I can’t have a wooden structure—it’s going to burn. And I’ve eliminated as much plastic as I could for the same reason. So it’s mostly metal construction. In my drive system I’m using a cable to make the movement work because it will scale up. If I use the standard gear and rack system or pulleys and belts it doesn’t scale as well. With this I can get any length cable I want and it scales easily.

Tim: And the feedstock is pretty standard stuff?

Steve: Yes, absolutely. The feedstock is just standard MIG wire that you can get at Home Depot or Lowe’s or any welding supplier. It’s easily obtainable, and it’s not expensive. If you just use a steel wire for 2 lbs it is $12, $13; 2 lbs in plastic actually is considerably more money. I think it will match in the actual volume. My slicing heights for this particular machine is about 1 mm whereas in plastic it’s very common to see 0.2 mm. So my stack height is five times bigger per pass, so that 2 lbs of material will actually go quite a way.

Tim: How about the power requirement? I think you’re hooked up to some pretty hefty looking batteries back here.

Steve: Well, the battery, all of them here those particular batteries are recycled from an electric car that we have. It’s actually a 24 volt input for the motion control. And what I’m doing is, I’m separating out the power requirements of the motion control versus the power requirements in the welder. I want them to come from two different sources, to try and isolate some of the crossover electronic problems that this machine currently has.

Tim: Could you talk briefly about those problems you mentioned before?

Steve: The microcontroller has been quite robust but when the welding process starts up it creates a high frequency radio wave that then passes across that board and freaks it out. So my fix at this point is to start the welding and then start the motion. Separate the two things and it’s working. Ultimately, I need to solve this problem.

Being here at the Maker Faire I’ve had enough interest from people coming to see what I’m doing that I think I found somebody that knows what to do to solve this problem. And that’s been really good—I’ve made some good contacts for solving my software problems. So that’s been the real asset for coming here is to find people to help support this project. Ultimately, I hope this to be an open source thing. It’s called Molten3D and I’ve created a BlogSpot to share my ideas. I’m sure by me sharing other people will chime back in with “You could solve that problem by doing this”

Tim: To be clear, I want to make sure people realize that the size we’re standing next to is not the final size what you want to do, you want to scale this up, 2, 3, or 4 times.

Steve: Yes, absolutely. I want to be able to make public art pieces that are 8’, 10’ tall all welded sculpture type thing....

Interesting but not new

[Registered Coward v2]

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.

Re:Wow

[Charliemopps]

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"