Aluminum filament 3D printer

Hello,

after having seen a lot of video of induction heating on Youtube, and a lot of easy DIY circuits, it comes to my mind that it would be relatively simple to assemble a 3D printer that would use aluminum or brass filament instead of PLA/ABS.

Basically it would be made by a classic reprap replacing the brass nozzle with a ceramic one, having around the coils of the induction heater. The extruder would push the metal filament, the induction would "instantly" melt while the pushing force would deposit as it happens with the plastic.

Apart the toxic smokes of the melted metal that would advice to not to print in your bedroom, I think that with this printer we could create more robust objects.
The print would NOT be as a cast melted object, but it would be more fragile, still having the classic reprap rowed structure, but however interesting.

- What do you think of the feasibility ?

Best Regards

Antimix

You could make filament from a mix of nylon and aluminum powder, which is Shapeways' Alumide powder. You can use leftover PA2200 powder from an SLS printer to make filament with a Filastruder. Tim asked them about using leftover Alumide powder, and they said it wouldn't be suitable for 3D printer filament because they find the aluminum particles wear the nozzles on the melt flow index test machines. If you could make a nozzle that will withstand the abrasion, then that might be a possibility.

An interesting proposition but there might be quite a challenge to overcome. Unlike plastics, most metals (and especially aluminium) transition directly from solid to a runny liquid. There is no gradual transition into a paste like consistency like plastics do at the temperatures we extrude them. The metal softens OK as the melting point is approached but will then suddenly become a liquid and run off. This will make it very difficult (maybe even impossible) to get the molten metal to stay where you put it.

Perhaps instead of melting with an induction coil one could use a MIG welding head to deposit the aluminium (or indeed many other metals). It would just be a matter of putting the welding head instead of the extruder. It might need a more hefty carriage though as the weight of the electrical welding cables will probably be too much for the steppers of a normal 3D printer.

Actually on thinking about it I am guesssing there might already exist industrial scale machines of this sort, but I don't see any insurmountable difficulties. All the same firmware and hardware of a 3D printer could be used with the exception of having more powerful steppers and sturdier rails and linear bearings.

The MIG welder could be modified to have the feed driven by a hefty stepper motor which would become the E axis, effectively giving you a Bowden extruder, except that the hot end is a MIG welding head rather than an extruder.

Quote
lister6520
Actually on thinking about it I am guesssing there might already exist industrial scale machines of this sort

Yeah, those machines are called welding robots.

See these threads for extensive discussion of metal printing ideas:

3D Metal Printer Projects
Metal FDM thread
Metal-print Reprap
3D Printing aluminum with a welder
Induction Heated Nozzle

It's one of those things that raise a lot of questions.

Would surface tension would hold molten metal in the nozzle, or would it always just drip out for any practical size of nozzle?
What sort of power is needed for heating?
What bed would you use?
Does extruded metal warp?

Seems like a good research project!

Commercially, laser sintering seems to be the most commonly used method for 3d printing metal. I'm guessing that's for good reason. I think any approach to print metal would probably be easier if it followed this already beaten path.

Quote
Twitchy
I think any approach to print metal would probably be easier if it followed this already beaten path.

Metal printing via FDM-like extrusion has been demonstrated too.

See this paper by Mireles et. al.: Fused Deposition Modeling of Metals
and this RepRap blog post: a new approach to printing metals
and also this thread: Printing Tin???

Quote
lister6520
Unlike plastics, most metals (and especially aluminium) transition directly from solid to a runny liquid. There is no gradual transition into a paste like consistency like plastics do at the temperatures we extrude them.

Apparently a big part of the trick to getting FDM to work with metals is to design a non eutectic alloy that has a "slushy phase" or "mushy zone", analogous to the "rubber plateau" seen in plastics during the glass transition. One could probably come up with an aluminum alloy that was suitable, but like bobc says, it would be a research project.

Check out some of these further relevant threads...

This guy has a similar idea... induction printing

This is an interesting article on 'droplet printing', which apparently provided quite impressive results!

In any case there would need to be some heavy duty modifications to any existing FDM printer, so much so that it it is more practical to design a printer for this specific purpose from the ground up.

---

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

Wouldn't the bonding between the metal droplets/extrusion onto already solidified metal be pretty crap?

A lot of interesting observations.

I would add also another fact I have observed, and that can be applied: I have seen that under opportune frequencies in the induction heater spiral, the dropped piece of aluminium stays in the middle "without gravity" (e.g. does not drop on the floor) and become red , white, and melts, staying still melted in the middle of the spiral until no signal is applied.

This feature could be used to maintain the melted aluminium in the ceramic nozzle, and control, based on the frequency of the electrical signal, when and how much quantity of aluminium extrude.

Still I imagine that this kind of printer would have a very thin layer height due to liquid capillarity when the nozzle spreads the liquid metal. The capillarity (e.g. using distance between nozzle and previous layer) could be used to drive the liquid metal shape avoiding the excessive circular expansion and maintaining the layer linear shape).

This metal printer is going to take a lot of energy to run

Instead of melting a metal just dispense the paste and then sinter/bake the object in an oven in a box filled with sand.

Of course we are not melting plastic but metal, so we should expect more power consumption.

On the schemes I have found, considering that we are speaking of a coil not greater than 2cm in diameter, I have seen a power supply used as source around 24V 15A DC PSU.

So the matters that should be handled seem:

* Aluminium filament suddenly melts.
- Have an alloy of Aluminium and something else that helps the extruding / deposition
- Use magnetic forces generated inside the nozzle and the liquid capillarity between the layer and nozzle to drop/stop the fluid and keep the liquid metal on its track.

* Metals scratch the nozzle
- The nozzle will be in smooth ceramic, with some tolerance since the drop will be probably driven by magnetic force more than push/pull effect of the extruder.
- we could have a sort of cut straight channel under the hole of the nozzle to better drive the metal capillarity.
- a ceramic extrudes would solve the issue (a link to a vendor was posted in this thread)

* Metals solve the nozzle
- from some experiments it seems that the nozzle dissolve after very few metal usage, so the nozzle should be anodised or the nozzle should be in ceramic.

* Strong power supply

* Metal bending warp effect
- to be further understood

* Bed attachment
- liquid metal will need something on the bed enough strong to resist to the heat, but also to grip on.

* Heat transmission on the nozzle
- Even if is it only the wire to melt itself inside the ceramic nozzle, it is a question of time but then also the nozzle will reach almost the temperature of the melting metal, acting itself how a furnace, and affecting the hot/cold control on the filament handled by the coil and the board logic.
- some logic will need to be implemented on the firmware to control the frequencies and in order to be sure the metal has the correct temperature
- Infra-red sensor for getting so high temperature of the filament / nozzle ?
- very good hot insulation from the other mechanic element of the printer. We don't want to melt electrical cables, insulations, near the hot-end etc.
- we may need water cooling some parts

I am continuing to look at this new printer possibility. Someone has already tested it, but there were no updated news.

Regards

* Aluminium run off issues
How about a dual extruder system. The first extruder lays down a channel which surrounds where the aluminium should go. The second extruder places the liquid aluminium in the channel where it can't run out and will harden in the correct shape. You could basically print a circuit board by printing the non metal parts first in plastic then filling the cavities in the plastic with the aluminium. The heat might cause some damage to the plastic but it will be very minor if done correctly and at a very thin layer and it might work.
Good luck guys
Mike

Instead of melting a metal just dispense the paste and then sinter/bake

If you want to extrude metal directly without major changes to your setup, you could try a Gallium wire. It melts around 22°C, is non toxic, but may cause problems in a metal nozzle. Then again your enviroment has to be quite cool as well. You can increase the melting point if you alloy Gallium with other metals such as Tin.

---

Blogs:
Meine 3D Druck Abenteuer
[3dptb.blogspot.de]
FLSUN Delta Drucker für Deutschland
[flsun-deutschland.blogspot.com]
Books on 3D patents:
[goo.gl] (english)
[www.amazon.de] (deutsch)

PeterGrant: Your posts are obviously thinly-veiled advertisements. Please indicate that you are associated with a company by putting a link in your signature, and please post all future advertisements in one of the "For Sale" subforums.

Quote
forum rules
If you run a company (selling parts or services, etc) please place a link to your company website in your signature.

I'm not familiar much with aluminum, but when I've forge welded steel together in the past, it required flux between the layers to adhere. Molten metal won't simply fuse to solid metal (at least not 100% of the time); both the deposited metal and the metal being deposited on, need to both be a little melty, and the conditions need to be right. Even if you could get it to adhere, Metal contracts when it goes from molten hot to hard. It contracts a lot, and it doesn't have the give that plastics do.i think you would have a lot of layer separation issues due to the metal cooling, contracting, and pulling away from the insufficient layer bond, if you did get it work.

Regarding the mig welder idea that was brought up: I've had the idea to replace the extruder with a mig welder too in the past; same basic principle - filament is shoved down a tube and melted at the very end, depositing onto (into in the case of the wekder) the work piece. However, having given it some thought, and having some experience with a mig welder, I'm not too optimistic. If you've got a mig welder, go lay a bead down in a line and then immediately lay another on top, and another, and another. Let me know how many layers you can make before you have a deformed orange/red glowing wad of metal that sits there spitting sparks and crackles for 15 minutes. You would have to pause for an unacceptable amount of time between each layer, not only to let your print cool, but also your welder; these things usually have a very low duty cycle, like 20%-40%. Nit to mention all the crap that will flying out from the print bed. Your lead screws (hoping you don't envision this thing using belts) and your linear rails would need flexible bellows covers. And I'm pretty sure it would look like crap.

Edited 1 time(s). Last edit at 02/18/2014 11:18PM by strantor.

Michigan Tech has made a RepRap 3d metal printer using a Mig welder, and a delta bot.

DIY $1500 Metal Printer
[forums.reprap.org]

Idea: Immerse the print in water. With a buildplate moving in the Z-axis and a head in the other two, you could support it using only waterproof materials. Then have the head just barely above the surface of the water. As the aluminum hits the water, the water boils and cools the aluminum, keeping it from running.