Metal-print Reprap

in order to do this, there has to be a stable part directly below to where you are spraying, would it be possible to do 6-axis with this method?

I suppose you could but personally I think it would just add unnecessary complexity over the ordinary layer system. A support material would be essential. What that would be I don't know yet and there is so much work to do before worrying about that.

I do have to make it work first before getting too carried away.

---

Make your Mendel twice as accurate.
[www.thingiverse.com]

Some people have managed to do "lost pla casting". I think that's the farthest the accesible 3d printing world has come to metal free forming. It looks pretty easy, and with a metal 3d printer you'd just eliminate the oven step, which you actually wouldn't if it's done with laser sintering.

By seeing some metal casting videos, it seems aluminum would be the most maneagable material (and documenting a procedure to recycle everyday aluminum objects would be cool, although I've heard soda cans aren't good for anything). Support material could be green sand (?), liquid aluminum may be layered and cooled so the next layer doesn't drip and the nozzle could be milled conrete. (I'm just babbling what first comes to my mind, it sounds too simple).

Edited 1 time(s). Last edit at 12/15/2012 08:11PM by Pixbae.

... 'made today some experiments with metal powder from conventional sintering systems and different IR-lasers - you need >2W to burn the powder in the spot, above 4W it starts to sinter, but for further tests I'll need a housing and inert gas support.

---

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

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

Pixbae Wrote:
-------------------------------------------------------
> Some people have managed to do "lost pla casting".
> I think that's the farthest the accesible 3d
> printing world has come to metal free forming. It
> looks pretty easy, and with a metal 3d printer
> you'd just eliminate the oven step, which you
> actually wouldn't if it's done with laser
> sintering.
>
> By seeing some metal casting videos, it seems
> aluminum would be the most maneagable material
> (and documenting a procedure to recycle everyday
> aluminum objects would be cool, although I've
> heard soda cans aren't good for anything). Support
> material could be green sand (?), liquid aluminum
> may be layered and cooled so the next layer
> doesn't drip and the nozzle could be milled
> conrete. (I'm just babbling what first comes to my
> mind, it sounds too simple).

Yes, yes...and yes. Why ever not? This all seems pretty intuitive to me. I will try and let you know if it works in the coming months. For Alumn can, I know hobbyists successfully get real metal out of it.

My biggest concern now has shifted to the evolution of heat. Aluminum is the worst material other than brass and copper because it conducts heat SO fast. That means there is potentially little chance that a 1200 degF layer of aluminum, say 1 mm thick, would be able to melt and truly bond to the layer below it. Just touching the air could cool it so fast that it goes through phase change. So I wonder how that problem would be overcome, and how you would know whether or not bonding fails due to oxides preventing layer adhesion, or raid cooling preventing layer adhesion? They could both look the same...

What metal did you use VDX?

... it's fine iron powder with some percents of copper and lead as 'lubricant' for sintering ...

---

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

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

I don't think extruding aluminum and layering on top of each other is going to work. Aluminum and all metals are not amorphous and it will never act like ABS/PLA when heated to it's liquid state. Once aluminum reaches melting temperature, it will simply flow like water and you will get a big mess. Amorphous materials can go from solid to a rubber like state hence the glass transition and this is the main reason why FDM works. Unless we can locate an amorphous metal, i doubt this method will ever work.

Laser sintering seems to be the most practical approach for a metal-print reprap. Perhaps using a low temperature metal like pewter in powder form could be used instead it has a melting temperature of 170–230 °C. This would require less energy for the laser to melt powder pewter (if it exist).

Edited 1 time(s). Last edit at 12/24/2012 03:06AM by gyronictonic.

gyronictonic Wrote:
-------------------------------------------------------
> I don't think extruding aluminum and layering on
> top of each other is going to work. Aluminum and
> all metals are not amorphous and it will never act
> like ABS/PLA when heated to it's liquid state.
> Once aluminum reaches melting temperature, it will
> simply flow like water and you will get a big
> mess. Amorphous materials can go from solid to a
> rubber like state hence the glass transition and
> this is the main reason why FDM works. Unless we
> can locate an amorphous metal, i doubt this method
> will ever work.
>
> Laser sintering seems to be the most practical
> approach for a metal-print reprap. Perhaps using a
> low temperature metal like pewter in powder form
> could be used instead it has a melting temperature
> of 170–230 °C. This would require less energy
> for the laser to melt powder pewter (if it exist).



Good points, and as it turn out there are MUCH greater challenges ahead.

I was able to verify that at the melting point of aluminum, viscosity is nearly the same (.89) as water. I dont think this is a problem though, since i could certainly hold water in a 1 mm nozzle still due to surface tension. Its a problem at large diameters.

The BIG CHALLENGE to printing metal is the extreme heat conduction of aluminum. Accoridng to my simulations, it is PHYSICALLY IMPOSSIBLE to melt aluminum at the surface before it wicks away all the heat. Even if the aluminum reached melting point, there would be very little temperature differential, and heat of fusion would keep the material below from melting.

It MAY be possible to melt aluminum at the interface (800C) using aluminum molten to about 1200, even or higher...

After this, I found the next enourmous issue is warp! Aluminum changes density from 2700 to 2300 when heated. There is no print bed in the world that will keep the part from shrinking. Are the any solutions besides heating? I'm not sure.

For both reasons stated about, aluminum will need to be kept a few degrees below it's heat of fusion point (<10 C below melting temp). Based on the naive thermal simulation (convection, conduction, no radiation or heat of fusion) aluminum starting at 650C heated with a nozzle at 1200 C would melt at the surface within a second, rising to 960 degrees with a 3 mm nozzle size. In theory, the heat of fusion of aluminum is 400 kJ/kg, or roughly equivalent to a 450 degree C temperature rise (based on heat capacity of 890). Therefore, if the simulation shows heating from 650 (melting) to about 1100, the phase change energy would be overcome; this is extremely naive, and the temperature differential would be faster in practice since the temperature would never truly exceed much past 650. So I consider it reaching 960 C in the simulation a successful phase change melt based on the derivative of the curve (rate of heat transfer) at 650C versus 960.

Now that that is potentially solved, there comes the HUGE issue of melting aluminum to itself. I'm *very happy* to report having melted aluminum to itself, again naively. I used bare nichrome wire to melt 2 mm wide strips of aluminum to itself. It curled around the nichrome and coated itself. The resulting aluminum was slightly more brittle than regular aluminum foil, but it was truly stuck to itself. So the surface oxide issue may be LESS than once thought. Again, I suspect the oxide is present between layers as a large percentage gap (60-90%?) preventing true bonding, but that few dozen percent is still enough to make a strong part if it is random. Also, a shielding gas is not a big deal if needed.

In any case, owing to thermal losses of radiation, the issues of heat conduction, safety issues, etc. I can't believe how big a challenge this is. Big challenges include heating the aluminum block-to-be-printed without loosing all that heat. Its like a toaster oven from hell (650C = 1200F) and I'm not sure how you create such a part, unbounded, without monitoring the part, and insulating it from all three thermal loss types properly. Luckily, the thermal conductivity does allow the whole part to be about the same temperature unless there are tight or small junctions. Printing extrusions would be no problem.

Still, compared to welding with 100 amps, I'm surprising there isn't a thermoplastic-welding-rod style welding machine for metals-- or is there - like brazing but out of a nozzle?? Like the last user said, maybe the problem is the lack of being able to "push" metal onto itself without it flowing away.

Overall, I think such a device can be achieved in the next few years and will beat out SLS from metal powders in terms of cost dramatically.

Edited 1 time(s). Last edit at 12/24/2012 04:07PM by Simba.

... look at the 'real' videos here (the virtual ones shows/advertises the technology): [www.beam-machines.com]

This DLM (='Direct Laser Melting') makes better surfaces than SLS, but are best performed with 5-axis-machines

---

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

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

Is there a possible reprap SLS 3d printer?

There is a guy on the R2C2 group who is hoping to deposit metallic ink with an inkjet printer head and then use a laser to sinter/fuse the metallic particles.

I don't how practical that is, but it seems novel.

VDX Wrote:
-------------------------------------------------------
> ... look at the 'real' videos here (the virtual
> ones shows/advertises the technology):
> [www.beam-machines.com]
> =en_US
>
> This DLM (='Direct Laser Melting') makes better
> surfaces than SLS, but are best performed with
> 5-axis-machines


Yes! This. $1M dollars right? $2? And not useful except for uniquely expensive shapes and prototypes?

My goal is a straightforwards non-laser, non-high current metal melting solution that is like what reprap did with FDM except for printing metal at home (under 5K). I don't like the laser because of the extreme cost, high power requirement, low power output, safety issues, etc.

Hi Simba,

> Yes! This. $1M dollars right? $2? And not
> useful except for uniquely expensive shapes and
> prototypes?

... with DIY the price will be mostly defined by the 5-axis-machine (this was the cause for the ( )-smiley) - could be something below 2K with home-brewing ...


> My goal is a straightforwards non-laser, non-high
> current metal melting solution that is like what
> reprap did with FDM except for printing metal at
> home (under 5K). I don't like the laser because
> of the extreme cost, high power requirement, low
> power output, safety issues, etc.

... IR-diodes with 9Watts@9Amps CW, capable of melting most absorbing powders, are comercially available around 300 Euros - I have some salvaged fibercoupled diodes with 5W and 9W and made some tests with different powders like plastic, lignin, glass, ceramic and metal ... you'll need at least 4Watts to start sintering or below 2Watts for MID (laser-activating conducting/metallized tracks in plastic)

---

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

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

What about a powder/laser reprap?

CPS Wrote:
-------------------------------------------------------
> What about a powder/laser reprap?

... would be a real interesting fabber with different available materials - essentially all meltable powdered materials when running with inert gas ... otherwise something limited to materials not oxidizing when melted

---

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

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

CPS Wrote:
-------------------------------------------------------
> What about a powder/laser reprap?

There is a very nice writeup on a low cost powder/laser printer on the RepRap wiki here: SLS Wax Printer

Sorry meant metal powder.

Aren't you able to do lost cast with plastic (ABS or PLA can't remember)?

Since we can print with plastic like ABS,PLA,and also wood.The idea to print with metal is also very interesting,though the metal have a rather high melting temperature.Perhaps we can print with metal power.

... for the FDM-process a metal-powder filled filament could be used - similar to the wood-dust filled polymer-filament actually used for 'wood-printing'.

Another possibility are metal filled pastes and a paste-dispenser ... or eutectic alloys, that melts first time at lower temp - had some gold-tin alloy (80% gold, 20% tin) in 0.3mm big spheres, mixed to a paste with Dexpanthenol as base,´that melted first time below 400degC, and then needs over 700degC for remelting again ...

---

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

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

There is also a carbon based filament too = electronics =)

Yay everyone buy 80% gold, only going to cost about $1million for a decent sized print, we all have $1m+ laying around lol

I love the discussion here but you guys are trying to reinvent the wheel. There is already a FDM system for metal, MIG welding. You just have to make your table so that the part can be removed at the end. There are metals which steel won't go to stick to which could be used. You could also try using a thin sheet of steel over that metal as a base plate for the first layer to stick to. There will be some limitations due to how the metal/system works but it could be done.

I suggested a mig or modified tig with mig feed system ages ago. The problem comes the resolution would be low and you need a constant feed of some specialised compressed gas to keep the air out. Unless you built a chamber and filled it with argon and did not open it till it was finished.

Anyway my goal was to make a system that will do a large variety of matirials as possible with only one depositing head which is why I have changed tack to everyone else.

Got the nozzle system to make the collimated beam (Well see if it works.) all finished and working on the compressed air heating chamber. Powder feed and container after that.

---

Make your Mendel twice as accurate.
[www.thingiverse.com]

+1 to jzatopa. MIG and TIG welding are well developed, and work. Aluminum is notoriously difficult to weld with, because of gas issues, only TIG welding works.

Some other general comments:

Eutectic alloys typically have the lowest melting temperatures. Just off eutectic alloys often have two phases intermingled, and exhibit complex melting behaviors which can produce amorphous mixes that might be more conducive to FDM techniques. There are dozens of metals that alloy with aluminum, with different effects on strength, melting point, morphology, etc. The little I know about metallurgy isn't enough to determine whether this will work or not.

Shielding gases shouldn't be a huge problem - a nitrogen filled box should work fine. All one needs to do is remove the oxygen, which can be done pretty well by burning a candle in an enclosed space. This won't completely do it, but is a good start. CO2 would be another possibility, but if temps get hot enough, that might release oxygen, which would defeat the purpose.

In terms of the high thermal conductivity of aluminum, the main issue with FDM is the energy density. If a laser, electrical, or other heat/energy source could be focused finely enough, not only could the aluminum melt, but the aluminum oxide might be able to be reduced back to pure aluminum and oxygen gas that would mostly diffuse away. Further, using an arc would not only heat the filament, but also heat the contact point, so the molten aluminum is added to other molten aluminum, not laid down on a solid surface. This may or may not deform the previous work significantly - it seems likely to me that the working area would have to be very small, and the print speed would have to be very slow for this to not leave a molten slag.

Nozzles - I am certain there are metals that could make a nozzle for molten aluminum. Go back to the printing press. A letter (in negative) was cut from a pure metal, usually copper. A standardized block of copper was put around the cut mold, and then a non-mixing alloy was poured in. This enabled one cut letter in negative to make lots of letter molds in positive. Cutting 52 letters (UC and LC) enabled the making of enough letters to print a full book page. And this was 1400's technology. Surely we can do better in 2013. (Also note the DIY nature of making the press - Gutenberg and his pals didn't make it proprietary, helping it to take off. This tech was used pretty much until the mid 20th century for printing.)

All in all an interesting thread, but difficult to test these ideas out.

Regards,
aeronaut

Well I'm intending to test my idea out. If its successful I can streamline the whole assembly in to one neat nozzle heater powder injector. Its all seperate bits at the moment to allow easy swaping of parts for testing. Obtained some tungsten wire for the heating element to try and achieve the maximum temperature I can.

I'm not sure you could make things propitiatory in the 1400's. Just keep them secret.

---

Make your Mendel twice as accurate.
[www.thingiverse.com]

Madkite Wrote:
-------------------------------------------------------
> Well I'm intending to test my idea out. If its
> successful I can streamline the whole assembly in
> to one neat nozzle heater powder injector. Its all
> seperate bits at the moment to allow easy swaping
> of parts for testing. Obtained some tungsten wire
> for the heating element to try and achieve the
> maximum temperature I can.

Let us know how it works out. Good luck.

> I'm not sure you could make things propitiatory in
> the 1400's. Just keep them secret.

Actually, you could. Information dissemination was much slower then. Further, there were guilds of various trades, one of the functions of which was to keep information secret. People were actually murdered for exposing guild secrets from time to time. Masons, tanners and dyers often had secrets they guarded very carefully. Even recently, the Zildjian family keeps they way they make cymbols secret - in the era where the USSR and USA were prime adversaries, the Moscow Symphony Orchestra had to order their cymbols from this company in the Ozarks.

But in the sense of having patent rights, or being able to initiate legal action at anyone profiting off of your intellectual property, no, you are entirely correct, there was little facility for that in the middle ages.

Regards,
aeronaut

Actually this raises an interesting question.

Lets say one of us comes up with a novel patentable thing for Reprap that might be worth some money if patented.

Apart from patenting it yourself, what could you do to keep it free for people to use? Can you prevent somebody patenting it and stopping its free usage without patenting it? Otherwise if we did invent anything new and potentially valuable whats to stop a company patenting the idea and preventing the inventor or us proliferating it as we wish?

It would make all that persons work pointless and be of no benefit to the community.

---

Make your Mendel twice as accurate.
[www.thingiverse.com]

That's a good question. In theory, publishing an idea is sufficient to prevent a patent due to prior art. "Publish" in patent sense means to make the idea well known to the "trade" that there is no excuse for not having known about it. Writing about it on your blog may not be sufficient.

It is also quite easy for a patent lawyer to create novel additions to a public idea which essentially wrap the public idea in a patented version. In practice, patent offices seem to be quite generous with applications when it comes to prior art and novelty. If they grant the patent incorrectly, there a lengthy process to have it reviewed and invalidated. In the meantime, the patent holder is open to sue any infringement.

The IP system is really geared up to protecting people who claim rights to an idea, it does not really have provision for people to make ideas explicitly public/free. I guess the only thing one can do is try to disseminate an idea as widely and in as much detail as possble, and hope that the patent examiners find it.

As I thought. Nothing to help an open source project.

It would be very hard for any individual to afford all that patenting unless they were quite wealthy. And a waist of crowdfunding.

The system does favour big business or the wealthy.

If there was some kind of open source license for this kind of thing then they at least could not claim prior art. Though as you point out it would probably be easy to get round with enough money.

Hmmmmm.

---

Make your Mendel twice as accurate.
[www.thingiverse.com]

... I've tried some times with patenting or registered designs / utility models in Germany to 'hold my back free' when developing new ideas ... but then a patent attourney told me, it should be enough to publish the concept somewhere, so yo can block any claims someone would have against you, when his registration was made later.

So I'm posting/spreading my ideas and concepts in magazines or forums ... or sometimes holding them back too for 'later use' if they are too extraordinary/prematurely for now

Edited 1 time(s). Last edit at 01/28/2013 03:15AM by VDX.

---

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

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

Check out these:

[www.youtube.com]

[www.youtube.com]