Low-Cost, Open-Source 3-D Metal Printer Could Bring Revolutionary Technology To Millions
http://www.isciencetimes.com/articles/6452/20131203/3d-printer-metal-open-source-1500-low-cost-michigan-tech.htm?fs=69b92
No pics of the machine’s output. Resolution is probably a few mm. MIG is good for welding bulk parts together, so current (common) implementations are not well suited to laying down small amounts of material precisely. Parts from this machine probably need significant secondary subtractive processing to be useful for anything.
@Dale_Dunn it’s a concept I’ve considered previously - in theory, you can dial down a mig so it deposits a fairly thin layer; that’s not practical for handheld welding as you can’t keep the tip steady enough or moving in sync with the electrodes feed rate to avoid material build up and subsequently melting the tip. Did you spot that they’ve inverted a Delta and are moving the work piece with the end efector while keeping the MIG head stationary. Would have thought a Cartesian arrangement would have been more applicable…
@Tim_Rastall , I don’t doubt that with refinement from the current material joining process it could become something useful. I guess I could have said so, right? What I see described so far looks like a first try or maybe proof of concept.
Keeping the feed rate right for small deposits does seem like a problem, especially with a MIG. Maybe a pulsed wire feed with a TIG torch could produce some interesting results, or even a TIG torch fusing metal powder. Not that I don’t want to be wrong about this. It wouldn’t be the first time. I once thought bowden tube extruders were a passing fancy and an insurmountable problem. If they keep going with this, they may very well refine it into something more interesting. I hope they do keep going, now that they’ve taken the first step.
My feelings exactly. I’d sort of discarded the idea as I couldn’t get my head around how you’d remove the component from the build surface -assuming you needed to weld to a metal surface for the first layer without a cutting tool - and how you’d manage any kind of overhang. Perhaps some very aggressive cooling or secondary ceramic support structures could be used…
Anyway, I did some googling and another article explains things more clearly: ‘The printer works by laying down layers of steel that are sealed together with a commercially available welder that is integrated into the printer’
So, it sounds like it uses a bunch of precut steel sheets and the MIG just welds them together. Who knows how the steel is cut or placed in position.
What’s clear from the googling I did is that someone has done a press release with very limited information that has then sparked a bunch of lazy reporting :(. (as is depressingly usual at the moment)
Huh - and after lots of further googling - here’s a wiki on it, basically tells you how to build an inverted delta printer and point a MIG welder at the the end effector :/.
So no metal plates, just molten wire.
http://www.appropedia.org/Open-source_metal_3-D_printer
The good news is the developer is a prof at Michigan tech. So development may continue. Hopefully he’s not like a couple of the profs at my school who wrote about some undergrads’ projects to meet minimum publication requirements.
A british university has been doing this in partnership with boeing maybe? There was a video on 3DPrinting.com i think.
Seems like it’d be hard to build up parts with any accuracy/precision.
I do know another group has worked out how to do this with electron beam welding. I seem to remember an aerospace company involved. That process welded directly to a metal substrate that had to be machined off. With precision that looked like a mm or so, it also required secondary machining. They described it as a “near net shape” alternative to casting or forging. Cheaper than making tooling for complex short run parts typical of aerospace applications, it still wasn’t cheap. Perhaps this group using the MIG welder is trying to create a cheaper version of that. Electron beam welding equipment isn’t cheap.