I was wondering if anyone tried / researched the idea of 3d printing useing a welding machine with a wire feed.
Basically, is it possible to fit a [www.hobartwelders.com] end tip to a RepRap and try printing layers of welded aluminium? Can it be stacked on layers? Will it be able to produce anything usuable out of metal?

I'm thinking about to try this, but have no clue or experience with welding.
Any opinions are welcome

Edited 1 time(s). Last edit at 05/21/2013 09:54AM by rapcraft.

NASA has done something like this, but with electron beam welding. From what i understand feature size can't be made very small because of the size of the melt pool, the thermal conductivity of the metal, and the surface tension of the molten metal. But you can create blanks that you can then CNC. printed parts have to be heat treated to remove stresses, but when finished are close in strength to cast pieces.

search for EBF3 or check out the video below
[youtu.be]

Edited 1 time(s). Last edit at 05/21/2013 12:37PM by Buback.

No, using a MIG welder as a PSU won't work. That particular welding process relies on the filler material, in this case the wire (electrode) from the welder, striking an arc due to potential difference between the work piece and the electrode. Without the arc there is no puddle and nothing melts. My guess is that sooner rather than later someone will perfect a commercially viable additive manufacturing process for metals on a larger scale.

See also:

Metal FDM thread
Metal-print RepRap
3D Metal Printer Projects

Tig with a tungsten for the arc and a filler wire feeder might work. I know a welder who repairs injection molds working under a microscope with a welder that goes down to 2 amps! I also worked on a machine to make microwave coax cable tig welding 0.020 copper at 400 ft per minute. I have rebuilt lots of aluminum parts just building up the missing pieces with weld. So this could work.

... some years ago I've made tests and developments with micro-plasma-welders and lasers for welding kovar and stainless steel with adding wires to gaps or on the surface - the welder, even with smallest possible power and plasma-torch and 0.1mm-wire, created a melted track of maybe 1mm width (wire completely blended with the solid material) and overheated (coloured) maybe 3mm of the surrounded surface.

The laser in comparison had a spot of 0.1mm and only 'coloured' a spot of 0.2mm while welding/fusing wires from 0.2mm down to 0.02mm diameter to the center of the spot ... so I could 'draw' solid lines with the wires diameter and only 'overheating' a zone of 0.2mm width surrounding the wire.

Now I have IR-lasers with spot diameters of 0.02mm to 0.01mm capable of melting or evaporating metal, so have plans to repeat this testing/developing in smaller dimensions and higher accuracies

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

GTAW (TIG) is not a good process for this application either. The tungsten and the potential difference between the work piece is what generates the arc. TIG down to 2 amps is pretty common for high performance welders but as with MIG, it's not the application for this. Additionally the high frequency start for the arc will likely disrupt any electronics nearby like the electronics that control the printer. This is common also in CNC air plasma cutters that use a high frequency torch. It's going to need to be a different process than traditional arc welding. That's not to say some of the technology won't be used, but in the current form and design off the shelf TIG, MIG or stick welders won't make a good choice for printing metals.

Frame This

Already done daily at IBM
see the man made of atoms
play with a ball made with an atom... includes quantum aura

Using a standard welder to melt metal for deposition will produce very low resolution prints that will not be very useful for anything, the wire is already at 0.8mm. So for this to be viable you need to go small!

An interesting idea is to mount an induction heater to the printer head and then feed the wire through it.

[www.youtube.com]

But as Viktor mentioned, lasers can be focused to very tight resolutions which allows us to print parts that are very useful. This process is know as 'selective laser sintering' and has benefits over any 'deposition' method because you do not need support structures (usually).

---

3D Metal Printing and more - visit [www.metalbot.org] !

Well, with laser sintering, you sometimes do need support structures, the printed object can become heavy on one side and sag down into the powder. But I agree, Laser sintering is a much more useful way of printing metal..

Here are some thoughts I posted on the Facebook group and add here as some of it may give the DIT crowd some ideas before they get patented.

I was reading the Metalicrap topics here and Metalbot forums. Lots of fun stuff about laser and electron beam melting of powders.

I have seen video clips of ultrasonic wire bonding machines used for IC chip lead bonding. These use very thin gold, copper or aluminium wire to connect from a metallised pad on a silicon IC to the metal lead frame. They do this fast and with no heat. They literally shake the metal wire tot he substrate and it welds. They have bond strength tests that I seem to recall cause the wire to break not at the weld so the bond is pretty good.

What if one were to keep bonding the wire and moving the head in the outline of a part and then raise the head a little. The wire bonding machines already have the 3 axis and wire feed all built in so it could be tested just for a little bit of software work. The wire diameters max out at about 0.01 inches or 0.25mm so they are smaller than a plastic nozzle. I expect such fine wire would be expensive but perhaps still affordable for small objects. Perhaps EDM spark erosion wire might be usable made from brass.

This is a public domain disclosure if not already patented but perhaps one could use plated wire that might be steel or brass with a brazing or soldering alloy plated on top. The wire would be bonded only in spots for higher speed and the when the model is complete it could be sweated (bulk brazed/soldered) in an inert atmosphere and all the wire bonded together (have to be careful it does not all fall apart and tangle).

(this section is discussed in earlier posts here)
But in keeping with laser melting and RepRap there are production laser systems (expensive I expect) that are used to build up damaged parts of hardened tool steel injection mould cavities. They are manually controlled to fix the chip or pit or ding by melting microscopic spots of metal from a fine feed wire onto the workpiece. This could surely be automated so one feeds out fine wire and pulls it along ahead of a pulsed laser that follows it. If the manual production is able to make repairs to high pressure dies then there should be a way to make things strong enough using automation.

So perhaps for metal printing we need to look at wire feeding and outside source of energy to bond.

Kalle
--
Johannesburg, South Africa

Hi KalleP, some ideas on metal wire deposition (using spot welding, cold wedling, vacuum welding, diffusion bonding, ultrasonic welding, etc) are discussed here. Check out the link to the Ultrasonic Consolidation video.

As many people have indicated above this (additive manufacturing, free form fabrication, laser sintering, etc) is being used/experimented with now by a lot of companies and organizations including our own Government. Whether its laser -or- EB Welding, using wire feed -or- metal powder and be it aluminum or any other metal... the real bottom line is that the companies that will experience the GREATEST benefit from this technology are typically aerospace/defense manufacturers who have parts that typically would have to start from cast and then they subseuqntly machine the hell out of to meet there end spec. As people have mentioned, most parts being developed this way currently are on the small side. With EB Welding, the process is generally performed in a vacuum chamber,so that fact alone can restrict the types of parts that they would REALLLY like to be manufactuing utilizing this technique. However when it comes down to it, the process is not far enough along right now that it is will be able to meet the stringent industry standards. And I am unsure if you are aware of the costs of buying/running/maintaining EB or Laser Welder equipment but these are not inexpensive processes. Things ARE coming along but for this to be successful it's all in the metallurgy; meaining reproduceability is key. Some of the other joining processes mentioned above just cannot deliver those results, and with EB & laser.. present day, it is just not at a point where it is commecially viable for the industries that would get any real benefit from using it.