Low-cost, open-source 3D metal printer in development

by Mark Tyson on 10 December 2013, 12:00

Quick Link: HEXUS.net/qab565

Add to My Vault: x

Most of the news in 3D printing technology has been centred on various kinds of plastic outputting devices until recently. While the plastic producing 3D printers are offering refinements such as full colour at keener and keener prices, for some purposes only metal printed parts are useful. So a Texas firm 3D printing metal parts via a sintering process was pretty interesting to hear about last month. However that kind of printer isn’t as available or accessibly priced like the ‘revolutionary’ plastic fantastic versions. Now it looks like that problem is to be addressed by a low-cost, open-source 3D metal printer in development at Michigan Technological University.

Joshua Pearce and his team at Michigan Tech admit that the 3D metal printer that they have developed is a work in process and raw. However it is ‘low-cost’ with a parts list price totalling under $1,500 worth of materials. Compare that to many 3D plastic printers costing much more or even commercial metal printers costing around half a million dollars.

Associate professor Pierce, said that “Similar to the incredible churn in innovation witnessed with open-sourcing of the first RepRap plastic 3D printers, I anticipate rapid progress when the maker community gets their hands on it”. Thanks to its low materials cost and open-source nature he added “Within a month, somebody will make one that’s better than ours, I guarantee it”.

The printer’s key components include “a small commercial MIG welder and an open-source microcontroller,” we are told. As the welder can be dangerous equipment to use it’s recommended that only people with relevant experience construct and work with this device.

The Michigan Tech 3D metal printer lays down thin layers of steel to make objects but, at this in-development stage, the team hasn’t printed anything more elaborate than a sprocket. Hopefully, as Pearce says, the impetus gained through the open source community can help refine this 3D printer design quickly. A full academic paper describing the project is available here.



HEXUS Forums :: 4 Comments

Login with Forum Account

Don't have an account? Register today!
Interesting idea, basically FDM but with the plastic extrusion nozzle replaced by a welder.

I seriously doubt this is ever going to amount to anything useful, the fact that molten metal isn't capable of supporting itself combined with the huge amount of energy involved here, means that no real dimensional control will ever be possible. Small improvements on the sprocket may be possible, but even the best efforts will still be lumpy and unusable.

The best method of making an affordable metal printer would be to use the same concept of a metal wire as the feedstock instead of a powder, then replacing the welder heatsource with a stationary laser. It'd eliminate the two biggest obstacles to cheap metal printing: The need to use a powder, and the need for a system to direct a laser beam.
Cog
the fact that molten metal isn't capable of supporting itself
I wonder if this is totally insurmountable. With sufficient point-cooling you could use free-form-fabrication style continuous contact to form self-supporting structures (no built-in moving parts though). Or if a support material were used. A quick-drying ceramic output by a linked head might work (maybe use the welder to supply heat for quick curing?), or if the moving bed setup were replaced with a moving-head setup, sand could be used to fill out each layer by brushing to a level after each weld pass (may have some issues with cavities, unless you dusted the sand over the top then wiped down to level, but this may have sand inclusion issues in the metal).
Cooling isn't the issue, it's controlling the heat input. Mig welders use an electric arc which is difficult to control with any precision, and the welding process involves a large amount of heat input which melts more of the material you're welding onto than the metal rod you're applying. On top of this the electric current has to travel through the model you're making which will add a lot of extra latent heat which will be difficult to predict and control.
Using a laser solves all this as you simply control the energy output to melt just as much metal as you're feeding in.

A ceramic support material is feasible, but problematic. It takes a lot of heat to harden a ceramic, and in the process you're driving off a lot of steam which can damage the metal you're printing as well as the machine itself. Sand bound with fast setting resins may work as well, but it's going to be burning away as you're applying the metal, and take it from someone who's worked in a foundry, that's not something you want to be happening in a confined space.
But of course, if you're designed a printer that will be producing ceramic or resin-bound sand as a support structure, you can print a ceramic mould then cast metal into it afterwards.

Loose sand may help a little but is unlikely to provide much support.

The problem with laser sintering machines that use a powder is that the current set-ups need the laser to move, which means expensive and intricate lenses to direct the beam. When you get rid of the powder a moving bed becomes much simpler meaning a stationary laser can be used instead.
I think I'll wait 'til they create a graphite printer and print myself a Space Elevator.