3DPrinting tips: cooling, fans, and PLA temperatures

This is hugely important to 3D printing, and generally: no-one talks about it online (at least, not with any authority: more “blind leading the blind”). Which is silly: the engineering behind this is obvious (and it’s often obvious from looking at a 3D printer that it’s been under-engineered in this critical area).

NB: I don’t know which parts of “correct, obvious” cooling are blocked by stupid patents that should never have been granted. That might excuse the poor design of most printers today. Might.

My experience so far is that “more cooling of printed object is better”. But that’s an over-simplification…

What should a 3D printer do, ideally?

Ideal as filament moves through printer:

  1. Filament: room temp
  2. In Extruder: room temp
  3. In nozzle: > 180 C / liquid temp
  4. Exiting nozzle: > 180C
  5. Touching object/build-plate: 150 C / high glass temp
  6. “just printed”, last few centimeters: 100C / mid glass temp
  7. “recently printed”, previous layer: 80C / low glass temp


Glass Temperature vs. Melting Temperature

Long story, short: most things melt/solidify quickly … TOO quickly for us to use in a 3D printer.

We need:

  1. Something we can melt safely in the home/office (i.e. fairly low melting temperature – something closer to “a kitchen kettle” than “an industrial blast-furnace”)
  2. Something that goes “sticky” and STAYS sticky when we push it into shape
    • NB: a 3D printer draws a whole layer at once, then moves up, and does the next layer
    • Side effect: if the object has a large footprint, the layer takes a long time to print
    • …so, by the time it gets to next layer, the start of the first layer may have been cooling for many seconds – even a minute or more
    • …if it cools too quickly, or has no “glass” temperature, the first layer will now be solid and NOT sticky; the new layer won’t “stick” to it, and our print is ruined
  3. Something that has minimal warping (i.e. bending, twisting, changing shape) when it goes solid

3D printer filaments are generally (simplification) chosen to have a broad range for the glass-to-melting temperatures. As they cool, they remain “sticky” for a long time, giving us plenty of time to put a new layer on top.

Heated build-plates

Absolutely required for ABS, but even with PLA: The obivous huge advantage is that they keep the object at its minimum glass temperature. You could achieve the same thing many other – much smarter, more effective, more intelligent – ways, but this is the “super easy and cheap” way to achieve it.

I see lots of talk about “enclosed” printers “solving” this problem because they can regulate temperature much better. This is half-true: they can easily guarantee a minimum temperatuer of the whole model (rather than just the bottom portion). BUT … as noted above, other parts of the model need to be kept at different temperatures, so they only half-solve the problems.


Ideally, you’d have a couple of thermal zones where “all plastic HERE is THIS temperature, all plastic THERE is THAT temperature”, etc.

I’m not a hardware engineer; I’m sure there are awesome things you can do with modern technology to achieve this.

But my degree did include some electrical engineering, and I spent many years experimenting with heat/cooling systems for computers. You can get a pretty good approximation with a simple set of directed fans. And this is where many 3D printers fail spectacularly: they simply don’t bother.

How much difference does it make?

Here’s a clear example, using a simple improvement on the existing (supposedly refined! this is from a $3000 printer!) fan duct of a Replicator2:

(source: A radial fan duct)

And here’s another, this time on a MendelMax 2, comparing “two cooling fans pointed at object” to “ducted cooling fan”. Both objects are upside-down in photo – it was towards the top of print where the unducted print goes all wobbly and crap:

(source:A Better Cooling Fan for PLA)

Fan duct: Replicator2

For instance, here’s the Makerbot approach. They have a fan for cooling the outside, and a separate one for cooling the printed-object. The 2nd fan is ducted down and underneath the print-head, and channeled carefully to target specific bits of “recently printed” versus “printing right now” filament:

NB: I believe the Rep2 uses a centrifugal fan (c.f. notes on Prusa duct below) – it’s a different shape to the main fan, with different shaped blades and a special case on the outside.

(source: approximate replacement for Rep2)

Fan duct: Replicator2 replacement

Here’s an interesting variation – not official! – from a user. It’s hard to see, but the air comes down through the big square hole at left, and gets squirted out through very thin, wide holes on the three square edges on the inner-edge of the shiny areas:

(source: custom fan duct)

It’s worth reading the description:

“fan would sometimes push the filament to the right causing it miss the layer below”

…i.e. the AIR CURRENT from the ducted fan would “blow” the liquid filament to the side, for the very short distance between the print-head/nozzle and the main printed-object. That would distort it enough to ruin the print.

(mostly applies with filament that’s too liquidy, not solid enough – in this case, NinjaFlex – but it will have a small effect on any filament).

Ultimate rep2 fan duct?

I haven’t tried it, but this one is same idea, but instead of blowing sideways, blows more down:

(source: Downward Blower Duct)

Fan duct: Prusa / RepRap

IMHO this is an even better solution: a pure radial duct (like a doughnut, with a small gap on inner edge, it blows air from all sides, rapidly dropping the temperature of the filament without blowing it off-course).

Note the author’s comments: this requires a different kind of fan, one that produces airflow centrifugally.

(source: Cooling Duct)