This is not so much as i need help but want some data

[unborn owl]

i want people to print the part attach to this thread in PLA , you need to set your settings so that it prints the part at 13^m/s for everything except the first layer , you may need a brim added to.
Please take photo of the bottom and a side view of the top so i can see bridging and the benchy chimney .

then i need the make and model of the cooling fan that you have used and the number of fans used.

[unborn owl]

Dont be surprised if the print fails its not meant to be easy to print !

the print uses 3.25M of filament, thank you for printing it .

[subtle tangle]

When you say "13^m/s," what exactly does that mean?

[quartz echo]

Volumetric flow rate. Just set your cap to that in your filament settings, but I am assuming here that he also wants it printed at exactly that speed, not just up to that speed so you may need to raise your wall speed up to hit that flow rate.

I'm running a 0.5mm nozzle on my main printer, so that will be about 150mm/s

On my delta running a 0.8mm nozzle it's going to be about 50mm/s

Awesome, needs to be slower than my first layer to keep it around 13mm^3/s 😄

[quartz echo]

The nice thing about that flow rate is nearly every printer sold in the last couple of years should be able to hit that flow rate. Not sure if that is why he chose it or not, but does work out well.

[subtle tangle]

Going to be about 182mm/s for me (0.4 nozzle).

[quartz echo]

@unborn owl Did you want the flow rate capped at 13mmmmms or maintained. As in was this meant to be a torture test like the benchy smokestack where we don't slow down for the top chimney, or should we stick to usual minimum layer times at the top?

First one overcooled and had layer separation with fan at 60%, I'll drop it to 40%. Not a flow rate I usually print at.

[unborn owl]

maintained through out the print including the small features

[quartz echo]

Got it, I'll disable slowing down for min layer times then.

Oh shoot, I might have forgot to set bridging and overhangs to that speed as well. I'll throw this on a Bambu A1 Mini once it's done printing current job as well.

Damn, as I suspected, I missed the bridge speed. Easy fix.

Or maybe not an easy fix. How do I make the bridge not print at 50mm/s here? I set everything to 150 with a fixed wall width to maintain 13mmmmms yet it seems to just ignore that for bridge and choose 50 for some reason I can't find.

[unborn owl]

Find it 😆

[quartz echo]

It's over my head today. I would suspect a bug but am told confusion over bridging and overhangs is a feature. Also, share one of your sliced 3mf files and pics and I can just use that to get as close to your settings and picture angles as possible.

And I see you already included the pictures, I just forgot 5 minutes after reading it. Getting older sucks some days.

[quartz echo]

@unborn owl I know silk pla doesn't have the best thermal characteristics but I had it loaded up and don't like it so was trying to use it up. If you need I can redo with plain pla. Dual Delta FFB03612EHN 3628 server fans. I ran it with everything at 150mm/s on my 0.5 nozzle which maintained 13mmmmms, was slow but didn't really slow down as I ran it with 100scv for silliness and to stick to as constant a flow as possible. Filament temp was 230 which I usually run at 50-60mmmmms so I really should have dropped the temp here but just forgot, you can see the high temp in the chimney.

[unborn owl]

thanks for the test

just on the limit for the top of the chimney

Dual Delta FFB03612EHN 3628
Current: 0.75A
Power: 9w
Speed: 16500RPM±10%
Wind speed: 10.5m/s
Air volume: 65CFM
Wind pressure: 0.66mmH2O

[quartz echo]

Hope you're collecting weight info on those as well. The sites that sell them list them at 32g but I think you said you weighed one and it was heavier right?

[unborn owl]

32g to 40g depending which data sheet you read

[quartz echo]

Kicking it off on my wife's A1 mini now. I'll need to do a nozzle swap on my delta before running the test. Sort of useless info with a 0.8 nozzle.

[unborn owl]

thats why i did cubic mm rate so the nozel size had less impact

[quartz echo]

Less, but 0.8 can't overhang with any amount of fan. The filament strand becomes too heavy.

I did a test for fun and it printed bridges and chimney perfectly, but actual overhangs on top of the cone literally fell right off.

But from weight, not cooling, so I didn't want you to waste time putting that in your dataset.

[quartz echo]

Bambu Lab A1 Mini. Fan, a gentle breeze, wafting elegantly over your print. It just works™

[unborn owl]

was the good bit pointing at the nozzel

[unborn owl]

what type of fan is it can you see any identification

[quartz echo]

It's tiny. Also their extruder motor is tiny. Let me see.

Oh, not as small as I thought. I can't easily get to it, but Bambu parts shows it. I did a quick search and multiple people say it's a custom 5015 blower, and that it has been revised over time so I likely have the older one, I guess the newer one has higher output in the same form factor. https://us.store.bambulab.com/products/part-cooling-fan-a1-series?srsltid=AfmBOoqF3PtsePCMF2niicW4UFN5NkLxZNpf5aUygWvYegVNqyrdUnYV

Here are some specs on them. https://forum.bambulab.com/t/different-5015-part-cooling-fans/95773/2

[quartz echo]

I will say it normally has stunning print quality for a $200 USD printer, but it doesn't normally do overhangs at that flow.

[unborn owl]

thata an odd looking 5015

[quartz echo]

Yeah, custom shell it looks like.

I reprinted on my 350 for fun and to burn some more of that filament I don't like at more realistic settings, dropped the temp to 210 which is what I would normally use printing at that flow rate, and dropped the SCV down to 10, all other settings the same. Quite happy with it, might be able to dial the seam in a touch better but overall no complaints.

[unborn owl]

thats good cooling but you do have up to 20cfm of air flow 🙂

[quartz echo]

Yeah, it's nice having the fans right there with hardly and ducts to speak of.

[unborn owl]

look like you need a little defeltion towards the nozzle

[quartz echo]

20cfm each according to the specs.

[unborn owl]

65cfm peek 🤪

[quartz echo]

Eh....it doesn't normally print at this speed, and as you print faster you are in each place less time so I dont find pointing right at the nozzle to be good, better to spread the air around a little more. Plus it struggles to keep heat up as is. I'm about to overvolt the heater, it's 120w but I can easily drag the temp down printing fast with full fan.

[unborn owl]

fit 2 heaters per heater block

[quartz echo]

Interesting. Might actually be able to do that. I'm running a regular volcano length heater but it has room for a super volcano heater, which might also be enough for two. Interesting idea.

[unborn owl]

just get one machined by jlcpcb

[quartz echo]

They sell a super volcano length 140w and I figured with just a few extra volts at most it would be plenty for me though.

But if I run out of heat, I am going to keep that in mind.

Or I guess mill my own depending on how ambitious I'm feeling.

[unborn owl]

@quartz echo i have one of these 39g 12CFM

interesting values of CFM change depending on wich data sheet

static preasure 12mm of water

[subtle tangle]

Considering how the blades of these radial fans work, I wouldn't be surprised if this fan/duct design is putting out significantly more airflow on one side than on the other.

[subtle tangle]

Dual GDSTime 4010 blowers, 24V, dual ball bearing, 9600rpm version

[quartz echo]

My V0 is broken at the moment or I'd test but I have the same fans I think, dual 4010 which I have found do around the same as a 5015 but sometimes make packaging easier on the toolhead.

This test is ignoring duct design which makes such a huge difference. Hard to control for that though.

[torn timber]

Overhang and bridges ran at 33% fan, chimney maxed out at 45%.
Had to drop my beloved scarf joint seam to reach constant flow..
0.4 nozzle, 0.2mm layer height, cpap fan - 37k to 40k max rpm, can't find more precise info about it..
also this is abs, because I don't own play x)

[unborn owl]

Thank's very much for the test will look at them closer in a while .

[unborn owl]

was it just starting to struggle with cooling in this area,? where was this in the relation ship to your duct nozzel?

[unborn owl]

are you sure your fans are 3cfm now 😆...did you do your bin bag test?

[subtle tangle]

Not yet, but I am not optimistic.

[torn timber]

I manually put the seam vertically on the square of the lower section. So your marked area is on the left side of the print. More to the front to be precise. The extrusion starts there. I think the marked part looks the worst because of the seam, while the last overhang stage doesn't look great over all.
The airducts of my printer are on the left and right as you can see in the pic

[unborn owl]

interesting they are not aligned with the nozzle ..

its hard to see on the photo but is the top angle just starting to struggle with cooling or is that the photo making it look worse than it iis?

[torn timber]

I'd say it looks worse in real than the photos x)
I'll do another one with higher fan speeds

[unborn owl]

at some point the bridges should be blown away but the air stream

[torn timber]

exactly! that's why I keep the fan speed as low as possible. 33% works fine most of the time

[torn timber]

I think it looks even worse with 66% fanspeed. But I also just realised that my wall printing order is set to Outer/Inner.. so I'll do another one with Inner/Outer and 66%.

[unborn owl]

interesting the ripples by the seem i have this on acoupe of my prints too im wondering if its to do with PA values

[torn timber]

possible. I also get ripples on the opposing side. My front looks nearly flawless (the right sides in the photo) and the rear is much worse (left sides in the photo), which is a flaw in my stock duct

[unborn owl]

@torn timber on the picture of your tool head which is the front?

[torn timber]

bottom is front

[unborn owl]

iteresting so the air flow is going from rear to front

[torn timber]

yep, and front looks much better than back

Here 66% fan speed with Outer/Inner walls printing order. Not much of an improvement. I think I'm mainly limited by my duct design :
But Outer/Inner got rid of the ripples at the seam

[unborn owl]

that why i set the speed at 13^3mm/s it seems to be the point at which most fans seem to struggle, thanks for printed files, there is deferentially room for duct improvement on your tool head, ill publish all the findings in a month or so , there are other forums printing too

that looks like bambu lab PLA 🙂

[torn timber]

oh, interesting! Did you also test a bigger overhang model? This one is relatively small and with all the curves those printers probably are also limited by their accel (to decel) and scv and won't reach higher flows, even if the slicer says so. Just a thought though.
Looking forward to your findings!

It's the cheapest abs I was able to get my hands on. Around 10€ a role. It's from Jayo if I remember correctly

[unborn owl]

i did a 66 degree overhang but it was very difficult to print for most

[subtle tangle]

Even with accel/decel off, mine barely reached 13 by the end, and the overhangs were already junk long before that.

[torn timber]

I got the idea because I just sliced the model with my normal settings and got similar printtimes. So pretty similar to your finding. While my hotend can push 55mm³/s (messured by weight with 4,5% tolerance)

[subtle tangle]

Considering when the overhangs failed, flow and accel aren't really the limiting factors anyway, haha. I might try it again with accel and SCV cranked to 30k/30, just for the fun of it.

[torn timber]

Agreed! I'm kind of wondering how speedkills is able to print 60° (or 30° depending how you look at it) overhangs. The outer wall has so little contact to the previous layer. Is it really just his insane cooling solution?

[subtle tangle]

Probably. It makes a huge difference.

[torn timber]

well, then I got a new project. Designing a better duct.

[unborn owl]

@torn timber nozzels pointing in at 40deg seem to work best so far

and 50deg tilt

[torn timber]

oh thank you so much for the starting point! :)

[unborn owl]

your need to find the sweet spot front to back of the nozzel im still doing these tests

[torn timber]

now I'm really curious for your upcoming conclusions!

[unborn owl]

so am i 😆

[quartz echo]

I think the design of the fans helps a lot too. My 350 doesn't have any better cooling than My T1 Pro, but the duct design is really good. Not too focused on the nozzle, which I think hurts you at higher speeds. On my T1 I just switched to Cuttlefish from the previous Sweeper fan as I like the slightly broader area of flow vs the sweeper duct.

[torn timber]

hey @quartz echo , are you able to share your fan ducts? I went through a couple iterations without any big differences between the results

[quartz echo]

Sure @torn timber , they aren't anything I designed, it's an open source design. https://github.com/Kizime123/Takeoff-Toolhead

[torn timber]

Thank you! Totally different from what I imagined!

[quartz echo]

Very little in the way of the fan to the part, and not too focused so if you have a fast printer it can be over the hot area for a longer period of time.

[torn timber]

how does it behave with warping materials since it's cooling the whole print?

[unborn owl]

thats a lot of air pitty its so bulky

[quartz echo]

I don't lose any space on my bed so it works for me, but yeah, space is tricky packaging wise. No issues with warping but I print with high chamber temps which pretty much negate warping. I just did a 12 hour PC print that was half a kg without any warping, but I had both the chamber and build plate at 90C. Same for ABS at 80C chamber and plate. You can crank the fans as high as you want and it doesn't seem to cause problems.

[unborn owl]

nicley done , i found you can get ABS to warp even in a hot camber if the feature are long and thin

im trying to work out haw to get a 3628 fan on my tool head for a test ,

[quartz echo]

How hot, and what bed temp? I printed a corner to corner part in abs, long and thin (a one piece voron skirt) with no warping. Can't think of a much harder test print than that off of the top of my head.

[unborn owl]

ive noticed with one of my ABS prints it has a feature 100mm long 10mm wide if i hit it with too much cooling it starts to lift , even with some ears on the corners

[quartz echo]

What chamber temp? I mean I know your printer goes as hot as you could want, but what actual chamber and bed plate do you use for abs?

Like, I have had that problem with chamber at 70C and bed at 100C, but woth chamber and bed both at 85C the warping just goes away for me.

Plus slow cooldown at the end of the print.

[unborn owl]

80C ~85C , bed 110C , and yes leave it to cool down for a while, it may also be the ABS i was using

[torn timber]

90°C chamber temp?! You build a freakin oven! And I thought colder outside air would benefit me. So I'll also have to change that

[unborn owl]

keep my coffee warm 😁

[quartz echo]

Try dropping the bed. When I get the chamber that hot and drop the bed temp, it doesn't try to warp so I don't need as much bed temp. My working theory is that if bed and chamber are the same temp then cooling is similar rate and no warping. I also try to keep it near the softening temp of the filament. .

@torn timber @unborn owl build the actual oven, that's what Valkyrie H(igh)T(emp) is.

@unborn owl You can go to like 110C right?

[quartz echo]

Cool air warps things like crazy, so with your close the door and if you could make something to just recirc the air at least that would be much better. But with abs and cooler air you definitely need to reduce cooling, else one half cools faster, then warps towards the cooler side as it shrinks.

[unborn owl]

we are just doing the final steps and testing to go to 110C

[torn timber]

most consumer high temp printers I saw went up to like 60C. So 90 is already insanely hot for my knowledge

[quartz echo]

When I do 95C my acrylic door warps so hard I have to put a chair in front of it to keep it closed. I definitely need to deal with the door before I try to go any hotter.

It gets tricky above 80C, abs parts don't hold up, cameras fail, smart toolhead boards that are usb or canbus shutdown from overheating. But 80C really gets you pretty much everything you want, more is just for fun for me.

[torn timber]

how do you guys keep your mechanics cool?

[unborn owl]

dont have that issue have a small double glazed polycarbonate window .3

[quartz echo]

You don't really. Use high temp epdm belts, make all of your parts out of metal or high temp materials like pc, water cool if you have to.

[unborn owl]

everything is ok to 135c ish