Metal-print Reprap

Why hasn't this been done yet??? Other than printing ultra low temp weak soft metal (solder) that dissolves the brass nozzles eventually, why hasn't anyone tried to ram a metal wire into a really hot hot-end ? This seem like it would be very very easy compared to all the great work that has been up to now, with the only exception that ultra-high temp nozzle designs would need to be made? I refuse to believe sintering powder is simpler and more efficient than laying down molten wire i.e. very similar to welding rod work?

"really hot hot-end?" Hava a look at this list: [www.engineeringtoolbox.com]

I'd love to see your design for a hot-end that can be safely and reliably heated to at least 660°C to melt Aluminium, which is one of the lowest melting points or useful metals in that list.

I'm sure the metal working industry, including those toiling in smelting plants, furnaces and such, would also love to see it!

"seem like it would be very very easy "

Really? Don't you think that if it was easy it would already be in use? Working with metals, and considering the energy needed, is not a trivial problem I imagine.

---

------------------------------------------
garyhodgson.com/reprap | reprap.development-tracker.info | thingtracker.net

Quote

This seem like it would be very very easy

If it is that easy, do it!

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

I will (make a hot end for aluminum wire), if users think it is worth it. I personally feel it is somewhat of a holy grail for 3D printing, since you could do strong metal parts, heat dissipation, and high conductivity electronic wires. However, is the reason there was limited discussion in the past a lack of interest or just a feeling that it was too far out?

Also to rehash what I said earlier about it being easy. Given all the incredible work in making repraps work (RAMPS, stepper control, custom hot ends, custom filament rolls, custom software, etc..) it would seem that if the only limitation to getting started with aluminum 3D printing is just a hotter hot end, then it would seem easy because it is only one step to solve, rather than a total reinvention of the system.

I will be attempting to printing aluminum 5356 and 1100. These are somewhat readily available as "filament" for welding and shaping. I will be using a hot end made entirely from high temperature ceramics. The *easy* part is that seems to have been done. It doesn't make sense to use a metal hot end since these would obviously be dissolved by the aluminum and have a short lifespan.

On each individual matter (like heat generation) I could go into detail about why it would work. For example, the latent heat capacity of aluminum is similar to melting ice - its not like it is unusually hard. So getting to 1200F is only going to be 3 times the energy + the energy from additional radiative thermal losses (which is scaling with T^4). Though the fact that a toaster oven reaches 2500 F and tungten 4500 F (daily household items) doesn't motivate anyone to think the energy is possible surprises me.

Likewise I could argue how the other parts are possible, but people will still doubt, and even more simply will not believe it. So lets jump to the end result. If a 3D printer was in existence today, that makes aluminum parts that weld-together nicely, look good, but have ~1.5 mm nozzle resolution, is that useful and a good start?

I do have a few dumb questions:

Would aluminum Curl?
Will aluminum stick toitself when heated enough, or will the ozone resist?
Would aluminum be able to dissolve a ceramic nozzle (I know silicon is heavily used in alloying at small %)
Why are ceramic insulators Sooo expensive? (i.e., mcmaster 9360K3, 87675K24, 93615K18 )
Would this be worth it since a company would have to mark up reprap to ($3-5K?) to compensate for all the expensive heater parts.

The 3D printer would look like this, but obviously less industrial, and without argon gas.
[www.youtube.com]
And less like this: [www.youtube.com]

Edited 1 time(s). Last edit at 12/01/2012 12:11PM by Simba.

It just occurred to me that aluminum flow through a ceramic conduit would not ever see any oxygen. For this reason, only the aluminum extruded at the very tip would be oxidized. The rest could stick! So the real question is, how fast do oxides form, and how thick do they have to be that they prohibitively prevent layer adhesion (more than say 50% for a reprap). Obviously, a higher-end version would use argon or nitrogen and have near 100% adhesion.

Simba Wrote:
-------------------------------------------------------
> scaling with T^4). Though the fact that a toaster
> oven reaches 2500 F and tungten 4500 F (daily
> household items) doesn't motivate anyone to think
> the energy is possible surprises me.

2500 F for a toaster oven?? That seems a factor of 10 out.

Heating elements are not quite the same thing as melting metals. Incandescent light bulbs are quite specialised devices.

To be honest, I can't see printing molten metal other than low temp soldering alloys ever being practical. Certainly far from easy.

A toaster reaches 450 F inside because it is 100 times small an area than the element. The element is nichrome. It is metal or ceramic coated metal glowing red. Its 2500 F. Okay, so, more like 2000 F : D, case in point, its HOT compared to what i'm going to do.


C F Color
400 752 Red heat, visible in the dark
474 885 Red heat, visible in the twilight
525 975 Red heat, visible in the daylight
581 1077 Red heat, visible in the sunlight
700 1292 Dark red
800 1472 Dull cherry-red
900 1652 Cherry-red
1000 1832 Bright cherry-red
1100 2012 Orange-red
[www.hearth.com]

There are lots of problems to using another material for printing besides "can I melt it?".

You need to ask - When I melt it, what is it's viscosity? Will it maintain the extruded shape during deposition? How quickly will it solidify? Will the heat of new material melt or weaken existing layers? Will layers bond effectively? Can I feed the unmelted material easily enough? and so on...

In many ways, the advancements in 3D printing will be made by materials scientists, rather than the engineers that design the extruders.

It just occurred to me. How come laser sintering doesn't have problems with oxides preventing melting of the material together?? And they start with the oxide of the material. What am I missing...

... for not precious metall powder you have to flood the building area with an inert gas as argon or pure nitrogene.

Or you have some complex methodology, where the powder particles has a lowtemp melting hull, what's needed for fusing, and then the printed object is sintered in an oven with inert gas and/or soaked with the metall or finishing chemistry.

There are many different methodes involved in SLS, not only melting the powder with a laserspot ...

---

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

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

That makes sense, but surely there isn't enough air in between the particles that the molten metal couldn't deplete it? Also, this doesn't address how inert gas solves the issue of particles already having a huge surface area of oxide?

... when you melt a particle, then usually the oxidized surface 'shatters' in small pieces that are mixed with the inner fluid material.

If this runs in air, every fresh 'opened' crack is instantly covered with oxide, so you didn't get a chance to connect the pure metal of adjacent particles ... and the oxides usually don't have enough adhesion to give a strong connection.

When done in an inert atmosphere, the 'inner' pure metal can mix and fuse ... and the remains of the oxidized surface (often an ultrathin layer of only some atoms thickness) are intermixed and not visible any more ...

---

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

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

Hey Simba (and everyone else),

I checked the other metal FDM thread and some of the metal-depostion pages on the wiki for mention of this and couldn't find it, so figured I'd just throw it out here.

Aluminum is hot extruded in industry all the time, so a metal FDM might be more feasible than many think.

Here are some videos:
Animation of aluminum extrusion process
Porthole die extrusion

It is possible to extrude fairly small features (below 1mm, see for example this company) so the scale of the process might be acceptable for a reprap-sized machine.

Would the extruded metal stick to itself? There is some indication it could. In the words of wikipedia: "... dies incorporate the mandrel in the die and have 'legs' that hold the mandrel in place. During extrusion the metal divides and flows around the legs, leaving weld lines in the final product." So, at least in the die, the soft aluminum can be separated and re-welded together. But out in the open is another question. The ambient temperature might have to be very high.

The dies for aluminum extrusion are commonly made from AISI H13 "hot work tool steel". Fancy ceramics not necessary. So, you might be able to make a prototype hot end out of machined H13 steel, much like one of the all-metal hot end designs that are already out there.

According to wikipedia, extrusion pressures range from 4,400 to 100,000 psi (30 to 700 MPa). That sounds like a lot, but a force of 50 lbf (222 N) applied to a 3mm plunger (or filament...) generates a pressure of 4,500 psi.

So, building a prototype aluminum extruder might not be that difficult. But, of course, that is just the first problem to be solved. I'm sure reprap enthusiasts are not the first to consider this idea. There is probably a reason that most metal 3D fab systems use some kind of laser sintering or welder-based methods.

TL;DR Look up Aluminum Extrusion, make the hot end out of AISI H13 steel.

Even if you can get these temperatures and pressures with kitchen equipment, extruding alu doesn't weld it onto the layer below.

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

There are some printers that print plastic coated metal powder by laser sintering, followed by evaporation of plastic and metal sintering in an oven.

This concept should work with a RepRap too. However, someone would have to develop a well working metal loaded filament. And a steel nozzle would be needed i think, otherwise it would be grinded by the metal particles. The object is then printed as usual and put into a ceramic oven for metal sintering.

For precision parts, there could be some stl preprocessing, adjusting for the shrinkage and distortion of the sintering process.

Quote
Traumflug
Even if you can get these temperatures and pressures with kitchen equipment, extruding alu doesn't weld it onto the layer below.

Yes, Traumflug, I understand that
That is why I qualified my comment with statements like "But out in the open is another question" and "of course, that is just the first problem to be solved".

This is my point:
We know that a welding process takes place within the die - the aluminum separates into multiple streams as it flows around the legs, and then the streams weld together as they flow past the legs.

Wouldn't you agree that it is at least plausible that the same kind of welding process could take place a short distance outside of the die, provided that the layer was kept sufficiently hot?

... there were some results with placing solder-tracks on plastic or in grooves ... 've got different solder types melting between 180degC to 350degC. And I'm melting "Roses Metall" (less toxic than Fields metal, which melts at 71degC) at 98degC without problems, so for first experiments you can use this types of 'low-temp' eutectic alloys.

Next step could be melting the tip of a placed wire with an IR-diodelaser (have some ten 9Watt-diodes with 0.1mm fiber attached) or try SLS with powders ... I've made successfull tests with powders from dark plastic, lignin, ceramics and even metal-dust

---

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

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

Thanks guys!! I'm really pleased with the way this conversation is progressing. Thanks for all your insights and positive comments. Surely this will lead to great developments.

Again, as my experience has always been with new technologies, the "state of the art" is always due to the accepted wisdom that something better "can never work." Its only when we put aside assumptions, sometimes just doing the damn experiment, that we find out how wrong we are.

The evidence that aluminum flows can separate and rejoin, and that sintered metal (with an extremely high surface area with oxides) can bond together suggests that the brittle oxide layer breaks apart and creates a small fracture region inside the aluminum, but does not prohibit reasonable aluminum bonding. I for one am certain that aluminum can be welded under a mildly shielding environment, perhaps so that inert gasses are necessary to make a somewhat strong, if not slightly-weaker-than aluminum part. I will keep you posted as I'm trying as I develop ceramic-nozzle in the next few months to augment my multiple-material-3d-printers.

Best,
Mike

Edited 1 time(s). Last edit at 12/10/2012 03:07PM by Simba.

Hi Mike,

I have some small brass-nozzles with ruby orifices with bores of 0.1 / 0.15 / 0.2 / 0.3 mm - and some other stuff for micro-tooling, laser-brazing or such.

Could be interesting for some high-temp experiments, when removing the ruby's and inserting in more heat resistant material ... but I didn't have the time to do this in the next time ...

---

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

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

Hey,

So one of the assertions here I keep hearing is that you could use a steel nozzle for melt alumium. I would assume as solder rapidy eroded the brass nozzles in prior tests, and gallium degrades aluminum at room temp, certainly molten aluminum with destroy steel nozzles. It may even slowly dissolve some ceramic. So I think a ceramic inner coating is the only option.

VDX - can you send me info or links to products where I could buy such nozzles? I want to make sure they aren't obnoxiously expensive as I've heard they can be for waterjets. I will gladly try any starting at .1mm and up.

Thanks

Just some thoughts:

Extruding straight rods in 1 direction would be much easier than the current setup for reprap (vertically above and sticking to the layer below). Maybe you can come up with a design to extrude from the y direction instead of from the z direction (flipping reprap 90 degrees). The pressure you would need would be huge though.
Probably could use very very very thin aluminum rods and i'm not sure if oxidation would be a problem at that size(surface area of the extruded aluminum not final product), so you might need an inert atmosphere.

Hi Mike,

Simba Wrote:
-------------------------------------------------------
> VDX - can you send me info or links to products
> where I could buy such nozzles? I want to make
> sure they aren't obnoxiously expensive as I've
> heard they can be for waterjets. I will gladly
> try any starting at .1mm and up.

... I've got them without charge for some tests from here: [www.drwiesner.de]

They're labeled as "Precision fixed resistors, ruby type" and AFAIK the brass+ruby types were something around 3 Euros.

The ruby's are available with bores from to 0.03mm to 1mm diameter, and you can to ask, if they can give you some spare for testing or set in another material than brass.

For use in a waterjet you have to fix them extra or the pressure will separte them from the brass nozzle ...

---

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

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

Been working on a system to deposit metal and plastic for some time.
It works by accelerating the dust particles of the desired material to hypersonic velocity (800 m/s+ target for my attempt.) causing them to bond on collision. Its normally used for coatings as it is a spray with commercial equipment but I am trying to build a system that creates a collimated beam of particles you could build an object with. Mine will use compressed air and probably a lot of power but no fancy gasses. The dust should also never melt at any point except when it deforms on collision. The mechanism the particles join by is theorised to be similar to that in explosive welding but its net totally certain how it works.

Currently I'm machining (From workshop scrap.) the test of concept model to see if I can get it to work and get the desired result. The only problem I have is lack of time and money right now.

But come next year things may be better.

When its working I will post some videos of the results. Though I'm 50/50 on it being a total disaster and back to the drawing board affair.

Note such systems to join metal will be getting in to the KW of power range not the few hundred watts the current machines use. High temperature and strength means high power. Whether you cast it or join with a laser it still all needed melting with equal energy.

---

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

... we used sputtering to 'draw' metallic conducting traces in ceramics - it's working as expected on atomic+molecular level, but when the size of the 'projectiles' overgoes a specific size_vs_energy relation, then the high kinetic energy needed for implanting/fusing the particles in the material will more 'engrave' it than add some volume.

DirectMelting uses dust, blown in a spot of a laser, so the dust will start to melt short before hitting the surface and form a blob ... but only with moderate speeds, or the kinetic energy will blow a crater in the molten material inside the spot.

So my guess is, that with ultrasonic speeds you'll need really small projectiles (molekular dimensions) and really gigantic building times to fab parts as big as some cubic millimetres in volume ...

---

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

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

[en.wikipedia.org]

All you are doing is substituting heat energy for velocity. Quite a lot of material can be laid down as the particles are 1 to 50 micrometers in diameter for this technique. Research shows smaller is better but that does not mean the volume has to be less.

---

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

... hmm, pretty old article, some of the interesting links are dead ...

How is the process changing when 'spraying' not on the solid base surface, but on already fabbed structures?

---

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

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

Of what I have read it is fine to build up multiple layers. The sucess or failure of adherence can be very material specific when joining dissimilar matirials but not when joining to the same material. It just welds on to the lower layer. It is being used in the aircraft industry for repair of parts where multiple layers have to be built up to replace lost material or create strong reinforcements. It is just they are all sprays not jets.

A problem that could occur is if you wanted to spray a material which needed a high velocity (Like metal.) on to a soft base low velocity material (Like plastic.). The high velocity particals would cut straight through in to the other material. I hope to experiment if I can get it to work with laying down a single thin layer of metal at low velocity on to a soft material so boding it to it. The the next high velocity metal layer will hopefully be stopped by the lower layer of the same material and bond to it without damaging the plastic. But that is getting more than a little ahead of myself.

Reliable powder feed and as well as the flow dynamics are all things that might need endless tinkering before I get this far.

---

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

Madkite Wrote:
-------------------------------------------------------
>
> Note such systems to join metal will be getting in
> to the KW of power range not the few hundred watts
> the current machines use. High temperature and
> strength means high power. Whether you cast it or
> join with a laser it still all needed melting with
> equal energy.

According to my math, a reprap-style aluminum filament melting system (At reprap printspeeds) only requires about 200 Watt at the nozzle.