Staged heating extruder

Hello,

Away from 3d printing community for awhile, but recently started a job where it would be real nice to make big parts with limited detail fast. I've been researching what the speed limits are, and to my understanding the primary limit to higher speed is control of the extrusion process, rather than accurate positioning of the extrusion head.

With too high a speed, it appears difficult to ensure the right amount of plastic melts at the right place, and not anywhere else, and to turn on/off the flow quickly and consistently. So it seems to me the speed improvement approaches focus on reducing the melt area as much as possible. Induction heating, and concentrating some nichrome right at the tip have been investigated and discussed on these forums, but getting the needed power in the intended small melt area remains a challenge.

Have there been any investigations into a staged heating system to reduce the melt volume? I remember doing some crude experiments trying to anneal the surface of pla parts by putting them in the oven, but it seemed the material has a fairly sharp solid to liquid transition. It seems like this could be exploited to reduce the heating power in the melt zone, by preheating filament just prior to the melt zone. That is, upstream of the nozzle there is a fairly long, high thermal mass heating block controlled to a temperature sufficiently below melting to avoid ooze. At the nozzle tip is a lower power 'melt' heater that heats the plastic over the melt point.

From the pla heat capacity data on the wiki I got ~320 kj/kg latent heat input to get PLA from room to melt temperature. I found here, for the crystallinity in a real PLA product from Natureworks, a heat of melting of only 29.1 kj/kg:

[ac.els-cdn.com]

Obviously there has to be some margin for control, but it suggests a relatively small amount of heat is required for the melt compared to the preheat, so a physically smaller heater, with consequently lower thermal mass might be capable of controlling whether or not there is a melt zone.

It seems like preheating could work as a lever that allows melt to be turned on/off with smaller changes in heat flows. Of course, if the melt zone heater has good thermal contact with the preheater, your control probably goes away, or gets worse, because the time constant for the melt section stays the same as without preheat but the temperature difference is smaller. So there has to be some thermal insulator in between. With a lower thermal mass for the melt zone hardware, fast start/stop seems more feasible, maybe it can even be modulated by pressing/retracting an air-cooled, large thermal mass block against it with a solenoid.

I didn't have much luck searching, but I think there's likely a good reason this doesn't work based on obviousness. Will the filament buckle too much at any reasonable preheat temp? Is the filament too thermally conductive for effective thermal isolation of the melt and pre-heat hardware? Another issue? Would really like to know about any obvious issues prior to trying to prototype.

If your goal is simply to melt plastic as fast as possible, limiting the melt zone is counterproductive. Plastic is a poor heat conductor and the limit to extrusion flow rate is almost always getting heat to penetrate to the center of the filament without burning the surface. MIT likely has the fastest printer at the moment, and they heat their filament by passing it through a laser in a mirrored chamber. The filament is semi-transparent to the laser, which bounces through the filament and simultaneously heats its entire volume. [hackaday.com]

Print quality is less affected by melt-zone (start and stops) than you might expect, since at high enough XY speeds and accelerations the extrusion rate is more or less constant. Extrusion only needs to stop for travel moves, and if you take the travels fast enough you will at worst get some stringing no matter how sloppy the extruder is. Bowden extruders are extremely sloppy but have only limited print quality impact.

I've found that the limiting factor for printing PLA is cooling it fast enough. Corners start to pull inwards, and as you become more reliant on extreme amounts of cooling the print becomes more sensitive to orientation and geometry. Shapes that deflect or channel airflow will affect how quickly the subsequent layers set which can make your surface finish very inconsistent.

Edited 1 time(s). Last edit at 06/09/2018 09:05PM by 691175002.

Someone here made an experiment with a Diamont hotend. He could print way faster, when three filament path's were used simultanously.
Now imagine what a printhead with a few thin filament path's could do. ( 0.8mm filament is a good candidate IMHO )

... I'm "playing" with the idea, not to melt+extrude the filament, but using thinner plastic wires and "laser-weld" them onto the surface (what's common with metal wires in industrial laser-welding, but with much thicker wires and bigger weld beads too).

Found some cheap and avaliable sources for 0,3mm filament (used for tooth-brushes) and some manufacturers, which have 0.1mm and 0.05mm thin plastic wires too.

Could be, I'll have a head for this until end of the year ...

---

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

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

0.3mm filament would be nice too and transporting it to the laser head would be much easier, than trying to push it through the melt zone of a regular extruder.
I'm just concerned about the laser stray-light.

... you'll need shielding - but this is common for all SLS systems too ...

---

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 have an enclosure around my printer as well as the filament roll. The temperature within this enclosure sits around 35c degrees which basically preheats the complete roll of filament.
I found that I got better filament flow and my prints were far better with this arrangement.

The other thing I also get better prints in the winter when the air temperature was low.
A more complete temperature controlled environment which includes the filament is got to be better..

RepRot

Quote
o_lampe
Someone here made an experiment with a Diamont hotend. He could print way faster, when three filament path's were used simultanously.
.........................................

That would be me then (https://somei3deas.wordpress.com/2017/06/22/exploration-of-print-speeds-with-a-diamond-hot-end/). Now that I have 5 colour mixing hot end, seeing what melt rate I could get out that is on my list of things to try..............

---

[somei3deas.wordpress.com]
[www.youtube.com]