Part Cooling Fans/Ducts



  • Morning. Just wondering if anyone has any good reads on part cooling fans and ducts.

    Interested to know about how focused the airflow should be and whether it should be aimed at the base of the nozzle or spread a little wider.

    I've gone from axial fans to the radial blowers and have learned many tid-bits with my own work, most of which are fairly intuative such as; avoid over restriction, aim away from heat block (even a socked heat block), cool symmetrically about nozzle.

    Previous attempts have failed due to issues such as the heater cartridge wires being too hot and deforming it, and the whole thing being too close to the heat bed and it getting warm and bendy.

    My current favoured design philosophy is design chunky and modular for ease of testing and service and then revise later to speed up build times etc so bare that in mind before reviewing any pictures I post later! 🙂

    Final question slightly related: I had wanted to put a connector in line near the hotend for ease of service and facilitate the later addition of a cable chain/protector. How far back along the wires do you need to go before the temperature of the core is safe for your typical 80C tollerant connectors? I'd like to process PA and PC soon but need enclosure first.



  • Can't find a link right now but Michael Hackneys "tusk" design comes to mind. One thing I would say is that Michael claims very little air flow is needed and I'd agree for "normal" printing. But if you want to use a big nozzle with big layer heights and widths, then in my experience you need masses of cooling air.

    In general, it's the age old problem of trying to keep the air flow away from the nozzle itself. Easy enough to do when the part has gained a reasonable height but it's almost impossible to prevent air flow being deflected back from the build plate for the first "n" layers.



  • Over the years I've tried all sorts of fan setups but now this mini-Kossel (ancient T3P3 kit, much modified, still working hard!) is equipped with the simplest possible solution, just a blower pointed at the spot below the nozzle. It works fine for all the prints I have tried it on. Not claiming it would be fine for everything. Here's a bad picture of the setup...

    0_1560501119507_IMG_20190614_092744424.jpg



  • @burtoogle That's not far off what I have running on my most standard Ormerod, but that does suffer when a 45 degree from vertical is behind the part relative to the fan. That said, mine is restricted with a small mount/duct.



  • @deckingman Tusk sounds like it is similar to what I have come up with.

    I'm trying to address issues with poor surface finish on surfaces 45 degrees from the vertical. Perhaps if it is not normally need then this may be more down to where the slicer is dumping the start/stop of the extrusion. More checking & tuning neeed! 😄



  • Ok here's the stick man cooling duct. Accidental likeness.

    0_1560506509917_DSC_3939_red.JPG

    0_1560506522509_DSC_3936_red.JPG

    The mount is more chunky than necessary as it is designed to be able to take a direct mounted filament drive or, as shown here in a trial fit bowden.

    The slots are 17 * 2mm aimed to hit the build surface 5mm either side of where the nozzle touches. I'm just re-building with 3mm wide slots and the 'feet' dropped about 3mm to clear the silicone sock better. So far this one hasn't gone bendy...

    Edit: Trying to keep the path lengths to each side as even as possible, inlet to the fan as unrestricted as possible, and the inlet not sucking straight off the hot part bed.

    At 3mm the total area of the slots is still about around 50% of the outlet hole on the fan so I feel I still need to tweak a bit further...



  • These are my previous iterations:

    First mounting of radial fan on standard reprap nozzle. Still going! Not great finish on 45 from vertical when the side is sheltered from the fan.
    0_1560513151227_01_OrmerodNozzle.JPG

    First attempt at cooling on both sides of part around a V6 mounted on an Ormerod: The power cable from the heater was resting on the duct, was too hot, and closed the vent off on one side.
    0_1560513194610_02_symetric.JPG

    Vikinghat style cooler. Alas to restrictive.
    0_1560513258451_03_vikinghat.JPG

    Final iteration before current was this one. It was an attempt to be less restrictive. It that last less than one build before curling down and colliding with the print.
    0_1560513316505_04_HiFlowSymetric.JPG

    Finally, yes I really should try the pumped systems soon but thought I should have been able to get a reasonable job with a radial blower.



  • Also on my todo list is to make something similar to a Berd air but I'd likely use something like a Koi pond air pump because I've heard the Berd ones are noisy. (Koi like deep water so a suitable air pump would have good pressure as well as flow). I already have a micro pipe bender so forming the tube should be easy enough.



  • This one has become popular on the RailCore II coreXY printers: https://www.thingiverse.com/thing:3367622. One fellow did mod his further by adding a small hole in the center, between the two primary outlets. This design is actually similar to MHackney's tusk technique.



  • @plasticmetal Thanks for the link. I make the area of the exhaust to be about 45% of the outlet of the fan. If that's working well then I've no need to try and get better than my last attempt which is about 75%!

    Second iteration of the stickman has dropped the feet by 3mm and opened the slot out to 3mm wide. I've yet to test this. Got to go pickup my son now from the inlaws.

    The third iteration is just about to finish printing which has the feet 10mm longer which will cool the IR sensor a little but also cool a little better around the part. When I stuck my my finger under the first iteration with the full on full chat the airflow feels like it is going around the back of the heater block more than the front, so hopefully this should balance it a bit more.



  • Try placing a small dish of water on the bed, under the nozzle, to help visualize the air flow.



  • @plasticmetal said in Part Cooling Fans/Ducts:

    Try placing a small dish of water on the bed, under the nozzle, to help visualize the air flow.

    That's a very good tip 👍



  • @plasticmetal I'll second that. Neat trick.



  • I read it elsewhere. I'm not that ingenious.



  • @plasticmetal fair enough, although kudos for not running with the credit! It is refreshing within this community that people actively credit work that has inspired or educated themselves.



  • @plasticmetal said in Part Cooling Fans/Ducts:

    Try placing a small dish of water on the bed, under the nozzle, to help visualize the air flow.

    What should be the ideal pattern? How do you decide if design A is better than design B?

    [I am using the hypercube evo design, strong flow from one side, and a silicon suck on the heat block]



  • I haven't tried this yet myself, but I would think you'd want to see an even depression in the water, centered around the nozzle. You don't want air blowing on the nozzle, so the depression would look like a doughnut/torus, with the water level under the nozzle not depressed.

    Getting a fan duct to do this, is another story. Obviously circular ducts, like Berd Air, can do it more easily than others. OTOH, it doesn't have to be perfect - just better than whatever you're using now ☺



  • I just adapted a bracket to fit the classic Dii cooler to my setup. I have yet to find a better duct.

    alt text


Log in to reply