I agree with @Joshua_D_Johnson , being a dental technician, casting a plastic or wax printed piece would look a lot nicer.
@Camerin_hahn I meant how small can you get the head and wirestock and therefore how small you could get a bead of molten material.
Tesla, i love when you guys discuss
@Tim_Rastall that would make more sense…
@Darin_McDonald I believe there is a project working with metal clay. You print the fine metal clay then fire it to make small parts, that is expensive and a very different project. There have been lost pla casts in aluminium, it seems to work well, I do not have the equipment or experience to do so however.
@Tim_Rastall I can’t imagine any mechanism that would put a lower bound on the size of MIG wire. However, to make it bond well you need to be using enough power to heat the target object some too. Wire melting onto cold metal doesn’t stick worth beans.
That particular barrier could probably be lowered a lot by heating the chamber real hot.
@Nick_Parker to make a “really hot chamber” would both drive the cost of energy and setup through the roof.
Viscosity is a problem as well, metal is very different then thermoplastic, you don’t get the nice semi-solid form. I think the the only viable methods will be sls (although very costly) or bonding a powder with flux/glue then casting it. The mig/tig solutions are neat but not useful.
@Camerin_hahn SLS has even stricter chamber requirements. Heating normal air is far easier than filling a chamber with inert air, particularly for DIY.
As for viscosity I don’t know much, but from my small welding experience the bead begins to act as a solid very quickly. Far more quickly than extrusions on my RepRap. And that was with heavy gauge wire for learning, not fine stuff that we would presumably use for printing.
Serious, the whole thing wasn’t more than a nice exercice to start with. The process of MIG welding itself isn’t reproducable enough to be used in this way. The variations thru the process itself result in structures 5 to 10 fold the size of the wire used +/-5. Also the resulting pice is layered in a quite unfortunate way, making parts prone to break easy. Thats why you’re supposed to never weld a broken welding again without removing all prior added material. The process shown does exatly the same, rewelding existing welds.
In contrast laser sinter process works on a way smaler scale thus creating a way different crystal struyture. Not to speak of a lower energy intake 
believe this man, he knows his way around welding ^^
@Camerin_hahn I’ve seen lost-PLA casting in bronze, aluminum, glass, and pewter. Metals like gold and silver should work too. I find the pewter process particularly interesting as the melting point is low enough that you can use room-temperature-vulcanizing silicone for the mold.
As for overhangs, what if you extend the wire and touch the side of it to the existing part. I know you’re not supposed to weld that way, but a computer-controlled welder might be able to do it in a way that would allow it to fuse material to the side of existing material rather than just the top, making overhangs possible. Theoretically, it could to completely horizontal overhangs of virtually unlimited size without drooping due to the speed with which the metal solidifies.
@Whosa_whatsis , are you saying simply tack welding filler rod at the ends off the bridge? If there is no metal below your welder, there’s no way to strike an arc, so you can’t heat the feed. So the"filament" would have to remain cold. At least this is my understanding. I’ve only done minimal TIG welding… and I’m awful at it
@Eric_Moy , what @Whosa_whatsis is describing is “out of position” welding. Basically, there are welding techniques that don’t require the weld puddle to be on top (in a gravitational reference). I know those techniques exist, but I’d have to refer you to YouTube to find out how they’re done.
@Tim_Rastall , could you share the link to the paper? Thank you.
@Whosa_whatsis I believe that for the mean time Lost pla casting is the way to go.this does limit the shapes that can be made (some shapes simply are too difficult to get the air bubbles out of), but it is the cheapest and easiest solution to making metal parts in a hurry.
Also the problems that the method described in the posts inherently faces are more easily solved in lost pla casting,
Maybe some kind of artist might like the result. I don’t. That stuff won’t fit anything… It’ll need a long developement in case to rapid prototype…
It’s an idea that I have thought about as I am a fabricator. As far as starting the piece on a copper sheet would work to remove after. Heat is always a problem when starting but once your welding, too much heat is a problem as your puddle will drop. I think a small wire and an argon chamber would allow for better detail. Nozzle positioning would have to be an option for welding over a void.
I believe strength would be an issue on highly stressed parts but general use items not an issue. High detail maybe achieved by using a carbon plug preshaped to negative of detail desired. Super fine detail such as a internal spline would still be a challenge as sharp points are not a spherical puddles preferred shape. However you could push a puddle with compressed gas.
These are just suggestions. It’s always good to push the possibilities. Cheers to the folks that are pushing the edge.
Carbon plug, huh? What if you had a carbon cylinder mounted next to the wire and allowed the head to rotate around its axis? You could form the edge of the unhardened metal similar to the way some of the concrete printers work. You would have to either be able to retract the former or always use toolpaths that leave a space for it (printing inside-to-outside).
Of course, you could do the same with any process of extruding material, but it really only makes sense for those with large extrusion areas like this and the concrete machines.
Carbon is brittle and its use as a forming tool would be subject to wear. But other non conductive heat resistant material could work as long as it could be kept cool.
Also chambers are not expensive due to fact argon is heavier than air and any box would work as pressure is not an issue. Just run a hose in a box and let it push out air.
it certainly needs some work but if they can perfect it while maintaining low costs, then it will be revolutionary.
Expensive 3d printers for metal already exist what people want are, cheap and high quality 3d printers that can give expensive ones a run for their money. not expensive ones.
What I want to see these things make are things like a car engine. manufacturing the parts yourself and putting them together.
grinders that can re-use metal parts and then you putting those into the 3d printer then out pops another item.
maybe some kind of home smelting machine/molder should be in the works with the same concept of 3d printing