How does a 3D printer head work?

I may have chosen the wrong place to ask this question, but this seemed the closest match.

I am interested in adapting the 3D printer technology to building things out of low melting point metals. Before I put any money into this, I would like to understand just how a 3D print head functions. From my internet meanderings, I realize that a normal 3D printer melts a plastic fiber in the hot end part of the print head, and the feed process is due to the extruder. So when plastic is needed, i assume the extruder just steps the fiber slightly forward. I can't find any description of a "valve" or any similar device controlling flow at the end of the nozzle.

I'd appreciate very much if someone would either verify this, or set me straight.

Thanks in advance for any help.

Basically, yes - the hydraulic action of the filament pressure controls the extruding. Other factors that come into play are of course the amount of heat applied to keep the plastic in a liquid state (without vaporizing or burning); this is controlled via a feedback loop using a thermistor or thermocouple (generally the former since it is much easier to interface - the latter requires extra circuitry).

Retraction of the filament lowers the pressure in the extruder head, and "sucks up" the remaining filament - this is done automagically by the software prior to a move of the head that doesn't lay down any filament. It is done to reduce or eliminate filament "fibers" stretching all around (which have to be cleaned up afterward - and may impede the print as it progresses).

I am not sure if low-temperature melting metals (like Wood's metal or the like) would need to have this ability or not, or if the melt pot could just be kept under a light vacuum or something. Certainly sounds like an interesting project.

I have seen a YouTube video of a conventional 3D printer using solder. It looked very bad. The thermal conductivity of the metal means that it is very difficult to lay down fresh metal without completely melting the layer underneath.

Powder-based sintering looks like the most promising method. There is no reason why that can't work with low temperature metals. I don't know of any home printers using this.

You would probably find more useful info on the Reprap forum. I suspect there are people there who are interested in extruding all sorts of things.

...R

MorganS: Powder-based sintering looks like the most promising method. There is no reason why that can't work with low temperature metals. I don't know of any home printers using this.

That is because the machines use a different technology and are very expensive . The powder is very expensive too. Typically titanium is used but there are others.

There is an annual exhibition at the NEC which is worth a visit, all the various technologies are displayed.

This explains diy plastic sintering. http://reprap.org/wiki/DIY_Selective_Laser_Sintering_FAQ

Having seen metal sintering machines i suspect you may be able to build one for about 1/2 cost of commercial machines but it will still cost 100k or so even then. Just too much engineering involved.big heavy stuff.

The laser alone will cost 30 k or so. In theory any alloy can be used but only a handful have been developed so far. There is considerable work involved in developing a new alloy to a workable use.

Plastic filament doesn't melt suddenly like a metal, if softens progressively, and 3D printheads completely rely on this to avoid molten material squirting out of the back of the head, as the softened plastic blocks this path. With a molten metal you would have the issue of forming a good seal somewhere, so I suspect its more difficult. You'll also need to find a medium and extruder material that doesn't result in the molten metal disolving the extruder, many molten metals dissolve other metals or diffuse into them.

Solder (ie tin) dissolves copper very rapidly, but also no effect on iron, which is why soldering iron bits are iron plated copper.

You have not defined what you mean by "low melting point".

Have you considered wax printing and lost wax casting ?

Lost wax can be used with PLA too...

Regarding the solder melting comment, that's a strong argument against using a eutectic for this purpose, since they have sharp melting points, whereas other alloy combinations can melt progressively. Well, I think so...

Very careful control of termperature is everything with this sort of technology.

I wonder how hard it would be to do extrusion at subzero temperatures with water ice?

Could you have a reservoir of low temperature liquid metal and pour it like water from a tap ?

...R

Metal and water also share a characteristic of crystallization. Even with the liquid cooled below its melting point, it doesn’t solidify until there’s a crystal initiation event. Search YouTube for “supercooled water” experiments - you can do some really weird things like drinking from a beer bottle, tapping it and having it turn to ice in the bottle, becoming undrinkable instantly.

The reason why this is a problem for 3D printers is once the crystallization starts on the build plate, it can go right back up the extruder and crystallize the melt chamber and it can do this at almost the speed of sound.

I have seen that happen in rechargeable handwarmers. Its like dot punching a windscreen whole thing crazes in an instant.

I guess you need good temperature measurement and control as close to the nozzle as possible then, keep that just above melting point despite the varying conduction losses.

MarkT: I guess you need good temperature measurement and control as close to the nozzle as possible then, keep that just above melting point despite the varying conduction losses.

That's how a regular 3D printer works. It doesn't seem to be applicable to metal.

If it was both technically and economically feasible we would know about them already.

I'd consider it more a materials science project ...

MorganS: That's how a regular 3D printer works. It doesn't seem to be applicable to metal.

Or in theory everything could run faster with metal because the heat flows faster... Just scale the time and the physics might be the same (except for the momentum of the extruded metal!!) As I mentioned before I suspect eutectics and pure metals aren't the best choice, use a system with a range of melting.

Maybe in the future diamond will become cheap enough - its by far the best conductor of heat available - to allow better thermal control right at the nozzle.

Anyone who's ever seen an old linotype machine working will know there are ways to rapidly create stuff from molten metal.