Two ideas: for your welding shield gas be it CO2 or Argon, if you do your welding in an open-top "tank" then once you have filled the tank your gas will sit in there as it's denser than air. Looks like you are almost there with your build.

Second if you can spot-weld how about adding tetris parts with a robot arm and building up the shape with solid metal? Yes the voxel volume will be rather course bit that might not matter if you can grind the assembly afterwards.

Hi, I think I will relocate the cooling system radiator to a place inside the printer... maybe underneat the Z axis assembly and add a polycarbonate front plate, so that the air stays inside the printer.

That sounds interesting and would likely be a very unique machining method :)

As you're the new king of DIY metal 3d experimentation, and have talked about possible sealing of the chamber, I wondered in a mad moment how the reliability and lack of sticking might actually be improved underwater, salty water, as you'll really be able to stab that wire into the plate without risk of it sticking... I've heard you really have to apply force or drag it around to maintain a spark when welding underwater, but it was a random welding video ( https://www.youtube.com/watch?v=xigAr1aqZzY ) using stick with a protective outer coating, still I think the same would be said using wire and mega-current, but safely inside a box...I'm curious and wondered if you might be too.

Keep up the great work either way!

Hehe thank you :)

Should I try printing under water?

The problem I see would be, if the water is conductive I think the current would no longer flow across the wire or at least less current would flow across it and maybe it would no longer melt.

But I haven't tested it yet.

EDM machines use either distilled water, or mineral oil to get around the conduction problem

Yes, but he said salty water :)

A few weeks ago I saw a video of someone building his own edm drilling machine for which he also used distilled water. 

https://www.youtube.com/watch?v=rpHYBz7ToII&feature=youtu.be

I will try to print my own shielding gas shroud with the wire printer. This way I could stay with the contact tip nozzle and keep it simple.

Oh your absolutely right it will go through the wire and leech into the water too, but I believe that adds an interesting characteristic whereby the sticking issue is reduced and a more predictable behaviour is exhibited. 

Just curious, what's the highest current you've used so far?

I testet once at which point the breaker trips which protects the 50mm² cable from overheating with 7A input and 375A output.

https://www.youtube.com/watch?v=cDjd9VCVpoI&feature=youtu.be

I'm printing at around 1,5A to 2,5A input which should be 80A to 130A output what goes through 0.8mm welding wire.

By using more intensity there is a lot of arcing until the wire get stuck in the nozzle or the nozzle stops conducting due to soot on the nozzle.

Seems like there would be difficulties; if the results were great then a lot of welding would be done in tanks of water.

I'd watch for:

1.  Electrolysis:  generation of hydrogen and oxygen gas --> corrosion one side.  Also a combustion / explosion hazard.

2.  Cold joints.

3.  Increased shock hazard.

Yes, think I will continue working on a shielding gas option.

Hi,
well regarding your argon/co2 endeavours:
When you use CO2 in your shielding gas, it will increase the carbon content in the liquid metal. Not sure which effect that will have, but I'd recommend going with argon 4.6 which should be sufficient.
Also take a look at common mig gun supplies for using as a shielding gas outlet.
Argon is quite expensive, so If you don't want to waste as much, I'd recommend you using a construction as close to a common MIG gun as possible.

In TIG welding I love to use whats called a gas lens, which makes argon flow laminar going out of the nozzle, which makes it an absolute joy to work with. Maybe you can find something like this for a MIG gun? not sure if it makes a difference but I can imagine everything which takes variations and chaos out of the equation will make your results much more consistent and predictable.

Love the project :)

Hi thank you for your advice :)

I think I will likely design a new printhead with a mig gun head or a tig gas lens with a contact tip in the middle to have the possibility to use shielding gas and keep it as an experimental feature. I have absolutely no welding experience and still never used a welding machine, so it's impossible for me to give a recommendation on shielding gas use.

I have seen DIY wire benders use welding wire feed rollers for the wire feeder. Might be a good place to start if you are considering adding one to your machine. See 3 minutes 20 seconds into this video: https://www.youtube.com/watch?v=WENEK-LJA6E

Hi John,

thank you for the link. Will try to build something like this.

I have not used my 3D printer in a long time

your spool looks like the one on my 3D printer

Consider running the metal filament through a bowden tube or similar to reduce the chances of it unspooling and causing problems? 

Hi Wayne, yes it will run through a bowden tube.

How do you prevent oxidation of the liquid metal? Do you use shielding gas and/or flux-core wire? 

Hi Klaus,

I thought about it, but couldn't test how oxidation affects the print quality until now. I think in the next few days the build should be ready for a test run. Actually I want to prevent using shielding gas to keep it cheap and simple, so the only option would be flux core wire, if oxidation is a problem. Maybe it could be a benefit that the wire heats up / melts from the inside because of the high current flowing through it, but thats only speculation.

Hello. I have some experience with wire welding and i think you really should do something to prevent oxidation, otherwise the printed object quality will be greatly degraded both in terms of physical and visual properties.

Also i need to strongly advise you to stay away from flux core. I leaves slag, which needs to be constantly removed before next layer can be welded on top of it.

Use CO2 gas! It's very cheap and easy to do. You can either provide constant flow of CO2 directed to your nozzle (in similar way plastic FDM printers use part cooling fans) or you can enclose whole 3D printer in watertight vessel and fill it with CO2 before printing. CO2 is heavier than air, so you can still have some openings at top, given you don't blow it too much. Only issue might be caused by pressure regulator freezing, if you use too much flow. There's no need for high flow if you use CO2 filled enclosure. If you need to have high flow for a long time, some people use warming devices. I saw people using high power resistors to heat up the pressure regulator. Quite easy. Also you can put electricaly controlled valve at low-pressure side(!) of the regulator so your 3D printer will be able to automaticaly stop the flow once the printing is done.

Go to any shop with welding equipment and they will tell you everything you need to know, it's really not that hard. Basicaly you need CO2 bottle, pressure regulator and some hoses.  Also you can get these things in shops who sell restaurant equipment, because the very same stuff is used to pour a tap beer from the keg.

I know it might sound bit uncomfortable at first, but i believe you will save yourself great ammount of trouble if you go the CO2 way rather than flux core.

Just remember to close the valve before you go to bed ;-)

Hi Tomas,

In terms of flux core wire you are 100% right. It behaves like you said and is not suitable for printing. In terms of CO2 I just added a inlet for CO2 for further testing. Do you think a hole next to the nozzle is enough or do I need a proper shielding gas nozzle to benefit from shielding gas?

The idea of sealing the printer and filling it with CO2 to the top is also good, but would require a relocation of the watercooling radiator.

The point is that the molten metal should be always covered in CO2, so it does not come in contact with the air. I am not really sure if single nozzle next to the wire nozzle is enough, but i think you can easily try it, if you somehow manage to inject smoke into the CO2 line (3d printed venturi and vape-pen?), so you will see, how the CO2 moves through the space.

You can take a look at commercial TIG torches and how they are made. I think it will give you some idea. It's basicaly big nozzle for gas and the smaller wire nozzle is in middle of it. But i think it might not fit nicely to your machine.

What if you take some copper pipe (brake line), seal one end (crimp with plyers and solder it). Then you can bend this pipe to "?" question mark shape. So you will have plugged end forming the ring around your wire nozzle and straight end providing inlet of CO2. You can then use tiny drill bits to make series of holes in the ring all around the printing nozzle. Dunno. Just idea.

Or just take a look at ducting part of cooling fan assemblies of FDM printers and make similar ducting from something heat resistant.

BTW if you have some watercooling system, you might run the hoses to pressure regulator of CO2 tank, because this gets really cool as CO2 gets to atmospheric pressure and warm water can help to prevent it from freezing, if you figure out how to exchange the heat between water and regulator. (Might still need to cool the water afterwards if it gets too hot)

Hi, I've ordered a tig nozzle which already has the holes for shielding gas and a thread for a shroud. Your ideas are also good. When I have the new toolhead ready I will test it with fog to see whether it covers the pool of melted material.

How do you insure that the wire disconnects from the welded spot? It seems like the opportunity to just weld the head in place is very high.

Hi,

It should be as followed:

As long as the electrode is made of copper it should not weld to the workpiece. The most heat is generated at the point of the greatest resistance, so the copper electrode stays cooler because of its low resistance and greater diameter and should not melt.

I tested it many times and it never welded in place.

I meant the welding wire not separating from the welded spot

Yes that could be a concern, thanks for the hint. I will test it and if the head welds to the welded spot I will maybe need another nozzle with an orifice shifted inwards or something like that.

I remember a project, a friend did in the late 1990ies. In that time they called it "rapid prototyping" and not so much 3D printing. They tried to melt metal like iron with an induction coil or some welding torch to build up an object layer by layer.
But that led to massive thermal distortion. AFAIK they tried even grinding each layer with a good (6 axis) CNC mill, but in the end they could not design a reliable 3D printing process.

Hi Martin,

I think that's the thing with metal 3D printing. Many people have worked on it, but only a few companies were successful with complicated and costly methods. It seems like it's extremely hard to get usable results at low cost.

The objects which I want to print/weld will likely be more partially together welded wire parts than solid metal parts, but I think it could be a cheap way to create 3D printed/welded parts at all. Maybe later the process can be somehow improved to create solid metal parts.

Post-print cleaning and electroplating?

Maybe when it works more reliable.

Very cool, have wanted to see a 3D printer operate like this. 

Thank you,

there are still some things to figure out but I think it should work.

I really like this idea. I can see it starting out really simple, with wires tacked to the build plate to make "low poly" shapes. Then you increase the number of spot welds to improve the resolution, and then build up layers. Really cool idea.

One place you might have trouble is the duty cycle of your spot welder. You might have to build a pretty beefy welder to keep it more or less continually operating.

Good luck, and keep us posted.

Thank you,

I hope I can build a capacitor based spot welder that can work for hours, like a 3D printer. Maybe with multiple capacitors to switch between them for reducing time between two welds and for using double pulse spot welding if needed.

What would you be spot welding?  I understand spot welding as joining two or three objects pinched together between (usually large) electrodes that run a heavy current through them.

Are you thinking of capacitive discharge welding, maybe wire-feed?

Hi, I started this project as a cnc spot welder,

https://hackaday.io/project/169412-wire-3d-printer/log/172893-mot-spot-welder

but the machine is still capable of spot welding 

https://hackaday.io/project/169412-wire-3d-printer/log/175294-welding.